Reference For Chapter 3

Chapter One
Introduction
Welcome to the World of
Photovoltaics
Thank you for your interest in photovoltaic technology and system design, and
welcome to the world of solar electricity! Siemens Solar is proud to be recognized
as the world-wide leader in solar electric power generation, and we wish to support
you in your efforts to apply photovoltaic technology to solve electrical power
problems, whether you are an individual designing your own system, or a system
designer working on large applications.
Purpose of This Manual

This course material is designed to be a self-teaching and technical resource for
people interested in learning about how to design stand-alone photovoltaic power
systems. Although the focus is on professionals that will be designing systems for
clients, individuals interested in designing their own systems will also benefit from
this program. This program is used by professionals from a wide variety of fields
including industrial technology, business, finance, military, construction, real estate,
and education. The topics are presented in small steps, with exercises throughout.
The main desired outcome is an ability to successfully deal with the many aspects of
photovoltaic system design not only the computational skills needed to arrive at
the array and battery size, but also the judgmental skills needed to see when and
where photovoltaics can be a viable solution to power needs, and the integrating
skills required to specify appropriate electrical and mechanical components from a
variety of manufacturers.
For those people preparing to attend the Comprehensive Seminar offered by
Siemens Solar, please answer all the exercises as you proceed through the text.
Take the time to think about the questions, and how the information applies to
problems you may face. Upon completion of the text, send the answers to the
exercises and the seminar application form to Siemens Solar to become eligible to
attend the Comprehensive Seminar.
Siemens Solar Basic PV Technology Course

Copyright 1998 Siemens Solar Industries
1-1
Fundamentals Introduction
How to Proceed Through These

Materials
1) Browse:
Browse through the workbooks to become familiar with the contents and the style of
presentation.
2) Sequence:
Choose to begin with topics of interest, or proceed sequentially through all the
chapters, beginning with the introductory/survey chapters, and continuing with the
technical material.
3) View Video:
Begin work in a new chapter by watching the videotaped presentation with the
workbook open. Follow along and examine the graphics as they are discussed.
4) Read Workbook:
Periodically the video presentation asks you to Stop the tape and refer to your
workbook. Take time to read the pages up to the pause, and review in greater
detail the key ideas discussed in the video. Underline or highlight important phrases
to help reinforce your learning.
5) Do Exercises:
If the section has exercises, first scan all the questions to get a sense of the scope
of detail needed to answer. Then proceed carefully to work on each problem.
(Completion of all the exercises is a pre-requisite to attending the Comprehensive
Seminars). If an exercise is difficult re-read the workbook section or review the
video.

1-2
Caution
Products specified for use in this manual might not conform to the National Electric
Code and may not conform to local requirements if the system is installed on a
building, movable structure or vehicle. Before assembly and installation of any
photovoltaic power system, you should consult with local authorities so that you may
be assured installation will safely conform to all local building code requirements. A
permit may be required.
Also, consult local codes before using products or installation procedures outlined in
this manual. Codes could possibly cover applicable inspections, permits for
electrical wiring, wire size, interconnections, grounding, enclosures, conduits, overcurrent protection, receptacles, load restrictions, disconnects and appliances.
Failure to follow applicable codes constitutes misuse of the products.
Do not attempt installation before reviewing all applicable instructions.
The technical information and suggestions for installation, operation, use and
maintenance made herein are based on Siemens Solar Industries knowledge and
experience and are believed to be reliable, but such information and suggestions do
not constitute a warranty, expressed or implied.
Since the conditions or methods of installation, operation, use and maintenance of
the equipment described in this manual are beyond Siemens Solar Industries
control, Siemens Solar Industries does not assume responsibility and expressly
disclaims liability for loss, damage or expense arising out of or in any way connected
with such installation, operation, use or maintenance.

1-3
(End of Chapter)

1-4
CHAPTER ONE
INTRODUCTION
1-1
Welcome to the World of Photovoltaics
1-1
Purpose of This Manual
1-1
How to Proceed Through These Materials
1-2
Caution
1-3

1-5
Chapter Two
The Basics of Photovoltaics
Photovoltaic refers to the creation of voltage from light, and is often abbreviated as
just PV." A more common term for photovoltaic cells is solar cells," although the
cells work with any kind of light and not just sunlight.
A solar cell is a converter it changes energy of light into electrical energy. A cell
does not store any energy, so when the source of light (typically the sun) is removed,
there is no electrical current from the cell. If electricity is needed during the night,
some form of electrical storage (typically a battery) must be included in the circuit.
In this chapter we will present some of the most fundamental concepts of energy
and power that are the basis for understanding photovoltaic power systems. We will
also discuss some of the common terms used in photovoltaic technology, and
present the prime benefits of using solar electricity for your power requirements.

2-1
Fundamentals - Basics of Photovoltaics
What Are Solar Cells

There are many materials that can be used to make solar cells, but the most
common is the element silicon. (This is not to be confused with silicone," a
synthetic polymer.) Silicon is the second most abundant element in the Earths
crust, next to oxygen, and silicon and oxygen together make quartz or common
sand. It is therefore very abundant, as well as non-toxic and safe. This is the same
silicon that is used to make computer chips, and some of the processing steps
involved in making solar cells are similar to the steps in making computer devices.
However, solar cells are much larger than typical individual computer circuits, and
they must be much less expensive! A typical solar cell used for terrestrial (Earthbased) applications is 3-6 inches in diameter and costs only a few dollars, whereas a
tiny computer circuit device might be only a tenth of an inch in length and width and
cost tens or hundreds of dollars.
The conversion process occurs instantly whenever there is light falling on the
surface of a cell. And the output of a cell is proportional to the input light: the more
light, the greater the electrical output. The cell does not use up any internal fuel to
produce output. The sunlight acts as the fuel for the conversion process. And that
fuel is delivered free everywhere in the world. The solar resource is more uniformly
distributed over the Earths surface than other renewable sources of energy light
wind or hydro. These resources are plentiful in certain specific climates and
geographic locations, but may depend on exact details of land contour and
elevation.

2-2
What Are Solar Cells?

Thin wafers of silicon
Similar to computer chips
But much bigger and much cheaper!
Silicon is abundant (sand)

Non-toxic, safe
Light carries energy into cell

Cells convert sunlight energy
into electric current- they do
not store energy
Sunlight is the fuel


2-3
Key Benefits of Solar Electricity

Energy Independence
There are many other ways to generate electricity, but what are the specific benefits
of solar generated electricity from photovoltaic cells? One of the most attractive
benefits is that you will have energy independence," the ability to create your own
electrical power, independent of fossil fuel supplies or utility connections.
Fuel Is Already Delivered

In a sense you do need sunlight as the fuel, but that is already delivered for free all
over the planets surface. Other conventional generation methods require access to
a site for fuel deliveries. This may limit the choice of suitable sites so as to have
road access, and even then access may be prevented due to poor road conditions,
or vehicle problems. The cost of delivering fuel to remote locations can be
substantial. For example, it has been estimated that it requires one unit of diesel
fuel to deliver one unit of diesel fuel to remote villages along the Amazon River in
Brazil. In other words, the cost of the fuel is doubled!
Minimal Maintenance
Solar electric systems typically require very minimal maintenance because there are
so few moving parts. Contrast this with a diesel-powered system or even other
renewable sources such a wind generators or hydro generators, which often have
costly repairs or regular maintenance of moving parts. Very complex photovoltaic
systems do have more parts and may require some maintenance. But when looking
at small power requirements, such as for home lighting or remote
telecommunications systems, only occasional battery maintenance is required. It
would be a mistake to say that photovoltaic systems require NO maintenance, but
the absolute amount of time and money required for photovoltaic systems is quite
low.

2-4
Benefits of Solar Electricity

Energy independence
Fuel is already delivered free everywhere
Minimal maintenance
Maximum reliability
Generate the energy you need where you need it
Reduce vulnerability to power loss
Systems are easily expanded

Maximum Reliability
This is perhaps the primary advantage of photovoltaics when compared to any other
form of electrical power generation. Because there are typically few or no moving
parts and the complexity of the systems can be kept low, the ultimate reliability of
photovoltaic power systems in the real world is quite high. The photovoltaic
generator typically is not affected by environmental effects such as lightning strikes,
high winds or blowing sand, humidity and heat, or snow and ice. The key to reliability
is quality and simplicity. If high quality components are used with the solid-state
solar generators, and if the component count and complexity of the system design
are kept to a minimum, the chance of any failure occurring is remarkably low.

2-5
Generate Where Needed

You can think differently about designing power systems for your loads, and not
always have to consider a central generator large enough for all your current
demands. You can distribute the generation of power to various sites, such as at
each classroom, or each house, rather than always having to install a large
generator and string power lines to individual users that might be separated by great
distances.
Telecommunications systems designers can look to photovoltaics as a way to
perhaps be more selective with the locations of their repeaters. For example, an
engineer might be considering covering a certain area with repeaters, and think that
he is forced to choose sites that are easily access, so that diesel fuel can be
delivered and maintenance can be performed. But the sites may not be the optimum
for coverage of the area. Instead, by choosing to use photovoltaic power for his
sites, he may now consider more remote, inaccessible sites, and may actually be
able to install fewer total repeaters but end up giving the same coverage that the
more numerous accessible sites would give.
Because photovoltaic generators can be as small as a few watts, you can truly
consider installing just the amount of power that you need at each site. This flexibility
is not available from other forms of generation.
Reduced Vulnerability
Because you can avoid stringing long power lines for many miles or kilometers from
some central generation source, many of the problems with utility power losses can
be avoided. Ice storms or vehicle accidents can cause power lines to go down,
perhaps tens or hundreds of miles from where the power is actually needed. With a
reliable photovoltaic power system at your site you could still have power, while
others around you have none.
And if you have chosen to distribute the generation of power to various load sites at
your location, you can insure even more reliability and less vulnerability to each load
site. For example, if separate homes have their own lighting system, and one user
overdischarges their batteries, or damages their system, the other users will be
unaffected.
This also applies to deliberate vandalism or terrorism. For example, having electrical
generation for lighting and security distributed to each building would make loss of
security for the whole site more difficult or practically impossible.

2-6
Easily Expanded
Photovoltaic power generators are modular by design. More power can be added to
an existing array easily. Old modules can be added to new ones without any penalty
(if the voltages are properly matched). Just enough power can be purchased and
installed today to meet your current needs, and as demand grows more modules
can be added in later years. This also means that financially it is easy to start with a
minimal power system today, and then add to the power as your budget allows later.

2-7
How Solar Cells Work

People often say that solar cells work by magic because there is nothing moving,
the result is instantaneous, and no fuel is apparently needed! The basic process by
which solar cells convert sunlight into electricity can seem magical, but actually is
simple. In a later chapter on Photovoltaic Physics more details will be given. For
now, we can however give a simple explanation.
Internal Field and Electron Flow

Most typical solar cells are made of the element silicon. (For cells made of other
materials, the explanation is still basically accurate). When light shines on a solar
cell the energy of the light actually penetrates into the solar cell, and on a random
basis knocks negatively charged electrons loose from their silicon atoms. To
understand this we can think of light as being made of billions of energy particles
called photons. (These are not the positively charged protons located at the
nucleus of atoms). The incoming photons act much like billiard balls, only they are
made of pure energy! When they collide with an atom the whole atom is energized,
and an electron is ejected or ionized from the atom.
The freed electron now has extra potential energy, and this is what we call voltage
or electrical pressure. The freed electron has energy that could be used to charge
a battery or operate an electric motor for example. But the problem is how to get the
freed electron out of the solar cell. This is accomplished by creating an internal
electro-static field near the front surface of the cell during manufacturing. Other
materials besides the basic silicon are grown into the silicon crystal structure. They
create an electrical imbalance that results in a one-way electrical broom that
sweeps the freed electrons out of the solar cell and pushes them on to the next
cell, or on to the load.
As billions of photons flow into a cell that is exposed to light, billions of electrons are
knocked loose and gain extra energy. They flow though the internal electro-static
field, and out of the cell or module. This flow of electrical charges with extra
potential energy or voltage is what we call electrical current.

2-8
How Solar Cells Change

Sunlight Into Electricity
Light knocks loose electrons
from silicon atoms
Freed electrons have extra
energy, or voltage
Internal electric field pushes
electrons to front of cell
Electric current flows on to
other cells or to the load
Cells never run out of
electrons
photon
h+
internal
field
ee-
P/N junction


2-9
Water Analogy
It is often helpful to give an analogy to water flowing. Imagine a water pump
connected to a circuit of pipes that are already full of water. The pipe circuit also
includes some sort of load like a water wheel, and all the water returns back to the
pump through the pipes, so that no water is ever lost due to evaporation or
splashing. In the analogy the pump is the solar cell, the pipes are the wires
connecting the cell to an electrical load and back to the cell, and the water in the
pipes is like the electrons already in the wires.
When sunlight (or any light for that matter) shines on the cell it delivers the fuel that
is needed by the cell needs to operate. Electrons are freed and set in motion. The
internal electro-static field pushes the freed electrons out of the cell and into the
wire. The analogy would be as if the pump was turned on. The pump begins to
push water into the pipe. But the pipe is already full of water, so water flows almost
simultaneously throughout the whole system.
The water flows on to the load like a water wheel, where its pressure and flow allow
useful work to be done. All of the water is then captured and flows again though
pipes back to the pump. The pump continues to push new water to the load through
the pipes.
In the real case of the solar cell the electrons freed by the incoming sunlight photons
flow out of the cell and on to the electrical load. They give up their extra potential
energy or voltage there and allow useful work to be done. The electrons then
continue to flow to the back of the solar cell, where they become available once
again to be knocked loose and flow on to the load.
The electrical circuit is closed, just like the pipe system was closed, so no electrons
are ever used up in the process. The solar cell never runs out of electrons. It
only needs continuous input of fuel in the form of light energy to keep running.

2-10
Cells Into Modules

Because typical silicon solar cells produce only about 1/2 volt we need to connect
cells together to give more useful voltages. When electrical generators are
connected together in series, or positive to negative, the voltage of each generator
adds up.
Usually 30-36 solar cells are connected together in a circuit to give a final voltage of
about 15-17 volts, which is enough to charge a 12-volt battery. Charging batteries is
the primary use for photovoltaic modules, so most are designed around doing that
job.
Connect Cells To Make

Modules
One silicon solar cell
produces .5 volt
36 cells connected together
have enough voltage to
charge 12 volt batteries and
run pumps and motors
Module is the basic building
block of systems
Can connect modules
together to get even more
power

But manufacturers could produce different module designs to better match other
loads, for example high voltage motors for water pumps or utility connected systems
that often operate at hundreds of volts.
If the voltage or current from one module is not enough to power the load, then
modules can also be connected together, just as the cells were. Manufactures
usually build modules with convenient junction boxes that allow interconnecting in
series or parallel.

2-11
Terms and Definitions

There are some basic terms that we have been using that should be carefully
defined, so we all speak the same language. Often people not familiar with the
technology may use the terms incorrectly. Specifically, we want to be clear on the
difference between modules and panels.
The CELL is the basic building block of a manufacturer of solar modules.
The fundamental physics of the materials used determines the voltage of a cell, and
the size determines the current. Usually manufacturers settle on one or two basic
sizes and designs for their cells and then proceed to make millions of them. They
can be packaged and exported to other module manufacturing facilities, and built
into specialized products such as lanterns, radios and garden lights. Think of the
cell as the smallest unit to work with. But it is fragile, and its voltage is low (typically
1/2 volt). For use in the real world it must be protected and connected to other cells
to give useful voltage.
The MODULE is really the basic building block for real-world remote power
systems.
It is a collection of cells interconnected by usually flat wire, and includes

encapsulation to protect the cells and interconnecting wires from corrosion and
impact. It usually includes a frame to allow easy mounting and a junction box to
allow wiring to other modules or to the battery and loads. The number of cells
connected in series determines the final voltage of the module. Usually this is 30-36
silicon cells to give a voltage suitable for charging 12-volt batteries, but some
modules are made to deliver higher voltages for use in utility power systems. A
photovoltaic power system can be a simple as one module connected to a batter or
a motor. If more current or voltage is needed then modules must be connected
together.
A PANEL is a collection of modules physically and electrically grouped together

on a structure.
This would be the building block for larger power systems. Usually the modules are
wired together on the panel to give the final system voltage (for example 12, 24, 48
volts or higher) and the panels are wired together or individually through field
junction boxes and then on to the system controls and batteries. Individual panels
can be disassembled or maintained while the other panels are operational.

2-12
An ARRAY is the full collection of all solar photovoltaic generators.
Sometimes an array is so large that it is grouped into SUB-ARRAYS, for easier

installation and power management. An array can be as small as one module (for a
simple home lighting system) or as large as 100,000 modules or more for very large
utility connected systems!
Terms Used
CELL
-- basic building block in factory
MODULE -- smallest unit that can do

real-world work; building block
in the field
PANEL -- physically connected modules on a

structure
ARRAY -- all solar generators in one

installation


2-13
Basic Concepts of Energy and

Power
We should take some time to make sure that the meanings of some commonly used
terms are quite clear. All future discussions in this course will depend on
understanding the differences between energy and power, and between voltage and
current. Once again, using a water analogy can help in understanding these terms.
Voltage
We can draw an analogy between the term voltage and water pressure. Imagine
water held behind a large dam where there is tremendous pressure but no forward
movement of the water. An electrical system with high voltage potential means that
the electrons in the components and wires have stored pressure and is capable of
doing work if released to flow, just as the water could do work if released from the
dam.
The unit of measure used for electrical voltage is the volt.
It is important to realize that voltage does not flow. It is a measure of the difference
in force or pressure between two points in an electrical circuit.
Current
The analogy to electrical current is the rate of flow of water. Electrons actually flow
past any point in an electrical circuit, just as water actually flows past when released
from a dam (or any other source of pressure).
The unit of measure for electrical current is the ampere or just amp. The number
of electrons flowing is enormous, amounting to over 1,000,000,000,000,000,000
electrons per second in just one amp!
And dont refer to current flow as amps per second. It is simply amps -- the unit
already includes the notion of number of electrons per second.

2-14
Power
The concepts of power and energy are so close that people often think they are the
same things. But the difference is critical in understanding how photovoltaic
systems are designed and how they work.
Power is the rate of doing work, and is NOT the amount of work done.
If we use the water analogy, the power of flowing water is given by the rate of flow of
the water past a point multiplied by the pressure of that water flow.
Power (water analogy)
Rate of Flow X Pressure of the Flow
In the electrical case power is the current flowing times the voltage of that flow.
Power (electricity) =
Current X Voltage
The units of measure for power are watts. It is incorrect to say watts per hour,
because the unit already has the measure of time included.
From the formula, you can see that increasing either the flow (current) or the
pressure (voltage) increases the power. Also, if there is no current or no voltage,
there is no power.

2-15
Energy
If power is the rate at which work is done, then energy is the AMOUNT of work that
is done during a specific period of time. The rate of doing work (power) must be
multiplied by the time to give the amount of work.
Amount
Rate X Time
Energy
Power X Time
Current X Voltage X Time
The unit of measure of electrical energy is watt-hour or kilowatt-hour (kWh).

Energy is so fundamental that it is hard to define. It is the fundamental stuff of the
universe. Scientists and engineers have learned how to describe the conversion of
energy and how it transforms from one form (heat for example) into another (light for
example). But no one can very clearly define exactly WHAT IT IS. What we do
know is how to use it to do work.
So ENERGY is what you have to work with, while POWER is the rate at which you
convert or use it.

2-16
Battery and Module Work Together
Battery and Module Work

Together
Daytime
Charging
Use
Anytime

The importance of understanding the difference between energy and power can now
be presented in the context of a photovoltaic system. Most photovoltaic systems
use batteries to store the energy converted by the solar modules during a day into
chemical energy for use during the night or on stormy days. The battery acts as a
reservoir of ENERGY and mediates between the POWER that might be available at
any moment from the solar modules and the POWER that the loads might want to
draw at that instant. If the loads need more power than the modules can produce,
then the battery discharges a bit to supply the difference. During the night, for
example, the modules produce no power so the battery must discharge to supply all
the power needed by the loads. During a day, if the loads do not require all the
power available from the modules, then the extra power goes into recharging the
battery.

2-17
Balance the Energy Diet

The amount of time that the loads draw their power determines the total amount of
ENERGY that they draw in a typical 24-hour day. The modules must, on the
average, replace that ENERGY during the few hours of sunlight that are available.
So the loads can operate anytime, day or night, on cloudy or clear days. The
modules replace the used energy only during daylight hours.
Another way to look at this is: The loads can only use up the energy that the
modules can produce.
The power of the loads can exceed the power of the modules, but the energy used
by the loads cannot exceed the energy produced by the modules. If it did, over time
the battery would try to fill in the deficit and would quickly become fully discharged.
For short times the loads can draw more power from the battery than the modules
are producing at that moment. At night the loads can draw power even though the
modules are producing no power at all.
What must be kept in balance is the ENERGY produced and consumed. If the
average daily load energy consumed is balanced with the average daily module
energy produced, then you have a proper solution to your energy problem!
During this course, you will learn how to predict the energy demand of the loads and
determine how many solar modules are needed to meet that energy demand.

2-18
(End of Chapter)

2-19
CHAPTER TWO
THE BASICS OF PHOTOVOLTAICS
2-1
What Are Solar Cells
2-2
Key Benefits of Solar Electricity

Energy Independence
Fuel Is Already Delivered
Minimal Maintenance
Maximum Reliability
Generate Where Needed
Reduced Vulnerability
Easily Expanded
2-4
2-4
2-4
2-4
2-5
2-6
2-6
2-7
How Solar Cells Work

Internal Field and Electron Flow
Water Analogy
Cells Into Modules
Terms and Definitions
2-8
2-8
2-10
2-11
2-12
Basic Concepts of Energy and Power

Voltage
Current
Power
Energy
Battery and Module Work Together
Balance the Energy Diet
2-14
2-14
2-14
2-15
2-16
2-17
2-18

2-20
Chapter Three
Market Overview
In this section, we will examine some of the key players in the worldwide
photovoltaic market, including the major manufacturers and the common
technologies used today. We will look at some trends in application and market
growth that will give you a sense of the excitement we feel for the future. The main
purpose of this section is to give you an orientation to the current and future status
of this power generation technology and how Siemens Solar is positioned to stay the
world leader.

3-1
Fundamentals Market Overview
Major Manufacturers
The number of manufacturers of solar photovoltaic cells and modules peaked during
the late 1970s and early 1980s at almost 30, but has decreased during the 1980s
and 90s to about a dozen major contributors. The cumulative shipments of the top
dozen manufacturers during the last fifteen years are shown below. There is
manufacturing occurring in all regions of the world, including India, Europe, Japan as
well as the US.
Cumulative Shipments
1980-1996
ie
en
140
S
ar
ex
100
Ph
K
P
B
E
S
ur
E
le
So
p
Sh
ar
s
io
el
H
ot
os
40
ol
ow
ar
at
yo
60
20
ce
ra
ol
80
S h ip p e d
M e g a w a tts
120
0
Source: Strategies Unlimited

In 1996 Siemens Solar celebrated a milestone for the world photovoltaic industry by
becoming the first company to ship a cumulative total of 100 Megawatts of solar
cells and modules. The other major manufacturers are Solarex (U.S.), Kyocera
(Japan), and BP Solar (U.K.).

3-2
The overall worldwide market for photovoltaic power is shared by the major
manufacturers. The market share of the key manufacturers is shown below for
1995. Again Siemens Solar has the largest single portion of the worldwide market,
with about 23% of the total.
Market Share
Siemens Solar 21%
89.5 MW
1996
(by shipments)
Solarex 11%
Other 35%
Kyocera 11%
ASE 4%
Sharp 3%
BP Solar 9%
Eurosolare 3%
Astropower 3%


3-3
Market and Application Forecasts

Market Growth Projections
How big is the photovoltaic market and how fast is it growing? These questions are
of interest both to people actively participating in the business and to financial and
governmental institutions looking to photovoltaic technology as one answer to their
application problems.
One prediction of how the market for photovoltaic power will grow in the next few
decades has been developed by Strategies Unlimited, a consulting firm that has
followed the growth of photovoltaics worldwide for many years. The assumptions
behind each of these scenarios are discussed a little later in this section.
The usual growth projection shows 20-22% annual growth through the year 2010,
and anticipates no technical breakthroughs and no increase in government
incentives. Even at this normal pace, the market size is projected to be about 1800
megawatts of annual production worldwide. This would mean an increase of 4000%
from 1990 to 2010!
Long Term Market Forcast

1990-2010
Business As Usual
growth
~22% to 2000
~20% to 2010
3500
3200
3000
2500
Shipments
MWp
2000
Accelerated growth
~29% to 2000
~22% to 2010
1800
1500
1000
500
42.7
0
1990
1995
2000
Accelerated
2005
2010
As Usual
Source:
Strategies Unlimited Report PM-39 March 1993


3-4
Following the accelerated pace of market growth, assuming important technical

breakthroughs in manufacturing and efficiency and aggressive government
involvement in the market with financial incentives, the worldwide annual production
is projected to be greater than 3000 megawatts. This would mean an increase in
production of 8000% in 20 years!
The actual growth of the worldwide market will probably fall somewhere in between
the usual and the accelerated forecasts. But this will still mean an annual
production rate of about 2500 megawatts, an increase of 6000% from 1990!
Turning these production figures into sales revenue estimates, we can anticipate the
worldwide market for photovoltaic to grow into the billions of dollars in the next
decade. One scenario of projected shipments and sales prices is shown below. By
2010 the market could be as large as $6 billion.
Long Term Sales Forcast

1990-2010
7000
6000
6000
5000
4000
Sales Volume
$ Million
3000
2000
1000
271
0
1990
1995
2000
2005
2010
Source:
Strategies Unlimited Report PM-39 March 1993


3-5
Application Growth Forecasts

We turn our attention from general market forecasts to specific application areas to
look at how photovoltaic power is used around the world. The application groups of
photovoltaic power can be divided into five broad groups:
Remote Industrial: This has been the major application area for 30 years, including
telecommunications, cathodic protection, telemetry, navigational systems and other
unmanned installations in harsh remote sites. The load demands are well known
and the requirements for reliable power are the highest.
Remote Habitation / Consumer Power: This segment includes applications that

are typically occupied, such as cabins, homes, villages, clinics, schools, farms, as
well as individually powered lights and small appliances. The load demands in this
segment are not as well defined, and are more flexible.
Grid Connected: These systems are typically multi-kilowatt or megawatt scale

systems that are directly connected to an existing power grid network. Electric
power is generated only during daylight hours, and is either consumed at the site of
generation (as on commercial buildings) or is fed into the general utility grid system
and consumed as a part of the normal power system. Small 4-10 kilowatt rooftop
systems can be located on top of individual homes, while larger 30-100 kilowatt
systems can be associated with commercial or industrial buildings to offset their
daytime lighting or air-conditioning loads. Large 100-500 kilowatt systems can be
installed along utility feeder lines close to their full capacity to improve power quality
and postpone rewiring or installing new larger transformers.
Government Demonstration: This market segment has always been small, and
constitutes projects funded by government or military organizations. These are
typically funded early in a governments learning curve about photovoltaics, to
provide information on installed costs and reliability.
Consumer Indoor: These products use photovoltaic cells to provide the small
amount of power needed for small electronic devices such as watches and
calculators.

3-6
The latest estimate for the way modules are currently used worldwide is shown on
the next page. The remote habitation application area is about 50% of the total
worldwide market, with the traditional remote industrial applications making up the
next largest market segment.
Market Segments
Forcast 1997
Remote
Industrial
27%
50%
Remote
Habitation
10%
5%
8%
Grid
Connected
Consumer
Other


3-7
Projections have been made to the year 2010 estimating the changes that might
occur in the relative size of different application groups. Once again, they present a
business as usual forecast and an accelerated forecast. The two scenarios are
similar in most respects. The remote industrial application group, the traditional
bread and butter of the industry, will continue to be a major segment, but will
continue to yield to remote habitation and grid connected systems. The major
difference in the two scenarios is the timing and the extent of the growth of the grid
connected application group.
Long Term Forcast (business as usual)

Major Application Groups
100%
90%
Remote Industrial
80%
70%
Remote Habitation
Consumer Power
60%
Government
Demonstration
50%
40%
Consumer Indoor
30%
Grid Connected
20%
10%
0%
1990
1995
2000
2005
2010
Source: Strategies Unlimited Report PM-39 March 1993


3-8
Long Term Forcast (accelerated)

Major Application Groups
100%
90%
Remote Industrial
80%
70%
Remote Habitation
Consumer Power
60%
Government
Demonstration
50%
40%
Consumer Indoor
30%
Grid Connected
20%
10%
0%
1990
1995
2000
2005
2010
Source: Strategies Unlimited Report PM-39 March 1993

The future will have photovoltaics used closer to where we live and work, as the cost
of delivered power comes down to compete with generators and utility power.
Remote habitation already constitutes the largest market segment. In developed
countries, this would constitute power for remote homes, cabins, and farms. For
developing regions, these applications would be home lighting, schools, clinics,
farms and village power.
The most important trend to understand is the impending growth of the grid
connected market. As the cost of photovoltaic power continues to decrease, and as
traditionally generated utility power costs continue to increase, the market
acceptance of photovoltaic power will accelerate. In these scenarios, the market
share due to grid connected systems will be near or greater than 50% by 2010. This
will indicate the maturation of the photovoltaic market, as what has been traditionally
considered a remote power solution becomes economic for the vast urban market.

3-9
The assumptions used in the business as usual and accelerated scenarios of the
Strategies Unlimited report need to be presented, so that you can judge the
probability of either case. The differences are presented next.
The major differences arise from how aggressively and how quickly industry,
government and utility organizations become involved with developing photovoltaic
technology.
Factor
Business As Usual
Accelerated
Industry
Investment
Incremental increase in capacity
Several 10MW+ plants
Utility Investment
Modest R&D, limited commercial

purchases
Increased R&D, large distributed power

sites
Incentives
2 /kWh 1995, 4 2000,

6 2010
4 /kWh 1995, 6 2000,

10 2010
PV Infrastructure
Mature at moderate rate
Large players take aggressive position
Financial Sources
Modest rate of participation
National and international agencies

aggressive
National Energy
Policy
Limited to national security
Support renewables for environmental

problems
Government
Support
Flat technical support
Increased technical and legislative

support
Environment
Concern
Incentive policies begin around 2000
Incentive policies begin strongly

around 1995
Petroleum
Changes
Demand at 1%/year, price at 1-1.5%

above inflation
Demand at 1.2%/year, price at 2-3%

above inflation
Natural Gas
Changes
Demand at 5-8%/year, price at 2-3%

above inflation
Demand at 5-10%/year, price at 3-4%

above inflation
Global Economy
Only leaders become strong during

1995-2000
Leaders recover 1995, global

improvement 1995-2000


3-10
Overview of Current Cell

Technologies
The single crystal silicon technology used in Siemens Solar cells is not the only
method for fabricating solar photovoltaic devices. We present here a brief
discussion of the most common competing cell technologies in the field today.
Single Crystal Silicon (Czochralski

or CZ)
The most common type of thick solar cell is made by melting purified chunks of
silicon in a crystal growing furnace, and slowly solidifying the silicon into a large
cylindrical crystal. In this process the atoms of silicon are aligned, and the electrical
properties are optimized. This process is called the Czochralski or CZ method.
Individual round wafers are sawed from the cylindrical crystal. If the sides of the
cylinder are cut first, making a long block, then when the wafers are cut they are
square.
This has been the predominant method for making solar cells for 30 years. Silicon is
abundant, electrically stable, and relatively easy to manufacture. Efficiencies of 12%
on a final module are typical.
Cast Polycrystalline Silicon

Another method of creating a block of silicon is to melt purified silicon rocks in a
rectangular block shaped mold. When the silicon is cooled slowly, it solidifies into
the block shape. But in the solidifying process, the atoms do not align into a large
single crystal as in the CZ method. Small regions of single crystal structure
crystallize next to each other, creating a polycrystalline block of different grains and
orientations.
At the boundaries between the grains, it is possible for incomplete (dangling)
atomic bonds to interfere with the current flow, and typically a slightly lower output is
produced compared to equivalently processed single crystal cells.
As with CZ cells, polycrystalline wafers are produced by slicing the large block using
saws.
Efficiencies of 10-12% are typical.

3-11
Ribbon Silicon
The slicing process in the previous two methods is wasteful, often converting 4050% of the material into dust. This is because the wafers are only approximately
0.015" thick, and the saw blade is about this thickness as well.
One method of producing wafers avoids most of this waste by growing a thin ribbon
from the melted silicon. The ribbon is either pulled sideways off the top of the melt,
or pulled up through a die.
Very fast growth rates are possible, but the speed results in polycrystalline
structures. If the pulling process is done very carefully, near single crystal structure
is possible. The ribbon thickness is approximately 0.010"-0.015", so no further
sawing is necessary. The ribbon is simply scribed and broken to produce
rectangular wafers.
However, the surface of the wafer is not typically flat and often bulges. Waviness in
the surface makes further manufacturing steps and interconnection difficult.
Efficiencies similar to polycrystalline silicon are typical.
Thin Films
All of the previous methods produce a single cell as the basic building block. This
means that many cells must be connected together to produce a module of useful
voltage, because each cell produces approximately 0.5 volts. The interconnecting
and subsequent lamination steps are costly.
In the past 10 years great progress has been made in manufacturing solar modules
by depositing extremely thin films of semiconductors onto glass or metal substrates.
This process has many advantages over the "traditional" methods mentioned above
that produce individual cells.
The semiconductor layers are only a few hundred atoms thick, so expensive material
costs are reduced. The entire module is made as a unit, so interconnecting
machinery and processing are eliminated. And the cell size can be modified easily,
so it becomes easy and cost effective to make modules of different power output for
different applications or products. A unique characteristic of some thin film solar
devices is that the light that does not interact to knock loose electrons can pass
through the device because the layers are so thin. This means semi-transparent
films are possible. Car sunroofs, boat hatches, and building glass, for example, can
be made that produce useful electric power and also serve as windows!
Several approaches are being pursued around the world to develop thin film solar
cells and modules. Some of these are described next.

3-12
Thin Film Silicon:Hydrogen (TFS:H)

Thin films of silicon:hydrogen alloy are the predominant technology mass produced
today, and are found in solar calculators, watches and PV modules.
The atomic structure of the thin film is not totally ordered, like the structure of single
crystal silicon. The probability is very high that an electron freed by light will
recombine with a hole before it can get very far. The light energy would just be
turned into low-grade waste heat. That is why the doped layers are placed on either
side of a large intrinsic, or undoped, layer. Instead of waiting for electrons knocked
loose by photons to wander into a region of electric field, the electric field spans
almost the entire thickness of the semiconductor. Electrons that are knocked loose
by incoming light are pushed immediately by the field, and have a very good chance
of being swept out of the cell before they recombine with a hole. The sandwiching of
an intrinsic layer of silicon:hydrogen alloy between P-type and N-type layers is called
a P/I/N structure.
Currently the efficiency of mass produced large area thin film silicon:hydrogen alloy
modules is about 3-6%, lower than that of the "traditional" technology modules. But
devices of 12% efficiency have been reported.
Copper Indium Diselenide (CIS)

Another combination of semiconductor materials that shows great promise is copper
indium diselenide (CuInSe2 , often shortened to just CIS). When combined with a
thin layer of another semiconductor, usually cadmium and zinc sulfide (CdZn)S, the
difference between the two materials forms a p-type/n-type junction. With both the
single crystal and the thin film silicon cells discussed before, one material (silicon)
was treated with small amounts of impurities (dopants) to create a P/N or P/I/N
junction. This is called a "homojunction" or "single junction" device. In the case of
CIS, the junction is created by placing two different materials in contact with each
other, creating a "heterojunction" device.
The response of CIS to light extends from the middle of the visible range far into the
near-infrared region of the solar spectrum, and allows for more of the available light
to be used by the cells compared to single or polycrystalline or amorphous silicon
devices. Efficiencies of greater than 12% have been achieved on small devices, and
over 10% on one square foot large area modules.

3-13
Tandem TFS/CIS
An exciting prospect for high efficiency and low cost lies in combining thin film
devices of different spectral responses together in one module. One plan for such a
"tandem" module involves placing a module of thin film silicon:hydrogen alloy with a
transparent back conductor in front of a module made from CIS. The front thin film
silicon:hydrogen module absorbs the short wavelength blue, green and yellow light,
and lets the orange, red, and infrared light pass through to the CIS module.
Research devices achieving over 15.5% have been made using this method, and it
is anticipated that 18-20% is achievable. The combination of higher efficiencies and
lower costs than any traditional design makes this a very exciting near-term
technology.
Concentrator Cells
Covering large areas with solar cells and intercepting what light falls on the surface
is called a "flat plate" method. This is the common approach taken by
manufacturers of single and polycrystalline silicon and thin film devices. Another
approach is to focus the light onto a small area and have a specially made cell at the
focus. This seeks to reduce overall cost by reducing the amount of cell material
required and substituting inexpensive lens or mirror materials.
Concentrator cells must be specially made to handle the large currents produced.
From 10-500 times the normal sun intensity can be produced on the cell surface,
resulting in an equally large amount of current compared to what would be produced
by flat plate converters.
To keep the focus on the small cell, the entire module assembly must very
accurately track the sun. Since tracking is essential, concentrator devices are
appropriate only where clear skies are predominant.
Recent advances in concentrator cells have been achieved using a "point contact"
structure. Thousands of p/n junctions are made on the back of a single silicon chip.
Thousands of interconnections must be made to make a working cell, but some of
the losses that occur in single junction devices are avoided and efficiencies of
greater than 30% have been reported. The transfer from laboratory to large-scale
commercial production has yet to be achieved.

3-14
Market Share By Cell Technology

The latest figures for 1995 for shipments worldwide separated by cell technology
show that single crystal silicon cells and modules dominate the photovoltaic market.
Approximately 57% of all shipments of photovoltaic power generators were made up
of single crystal or CZ technology, most of this was produced by Siemens Solar.
The next largest component of the photovoltaic market is polycrystalline technology
with about 25%. And thin film technology, primarily amorphous silicon cells for
consumer products such as solar calculators and watches, constituted a little more
than 13% of the worldwide shipments.
Single Crystal Dominates

World Shipments
Worldwide Shipments By Technology: 1996
S in g le C ry s ta l
Si
54%
T h in F ilm s
13%
11.7 MW
48.3 MW
R ib b o n
89.5 MW
24 MW
P o ly c ry s ta llin e
O th e r
3%
Source: The Solar Letter, February 14, 1997
3%
CdTe,
Concentrators
Si
27%

The total worldwide production of photovoltaic power generators in 1995 was about
81 megawatts. This is large for the photovoltaic manufacturers, but is quite small
when compared to the generating capacity of conventional power plants. One large
power plant can produce 800-2000 megawatts, continuously day and night. This
shows how far the photovoltaic market has to grow to become a major contributor to
the worlds energy needs.

3-15
It is interesting to look at how the primary technologies (single crystal, polycrystalline,

and thin film silicon) have evolved in the marketplace in recent years. Below is
shown the relative market share of each technology in 1991 and 1995. The actual
production level of thin film silicon has decreased by 20%, and the production of
polysilicon has decreased slightly also, while the production level of single crystal
silicon has increased by 137%.
Worldwide Growth Rates Of

Primary Technologies
1991-199684 MWp
W o rld P ro d u ctio n M W p
90
80
70
54 MWp
60
CZ
+145%
50
40
+15%
30
Poly
20
-15%
10
Thin Film
0
1991
1996
T h in F ilm
P o ly
CZ
Source: Solar Letter, 2/17/97


3-16
Cell Technology Characterized by

Efficiency
Perhaps one of the most important features that differentiate these various cell
technologies is their output efficiency. By efficiency, we mean their ability to convert
the solar power falling on the surface of a device into useable electric power. The
general definition of efficiency is the output of a device divided by the input to the
device.
Efficiency Defined
Efficiency
Output
Input
Module Efficiency
Electrical Power out

Solar Power in
Given an amount solar power falling on the

module (or cell or array), how much electrical
power gets produced?

Specifically for solar photovoltaic cells or modules, we define efficiency as follows:
Efficiency
Electrical Power out of device Solar Power into device
Solar devices work only during daylight, when there is solar energy to convert. The
best devices will capture and convert the most of the limited amount of solar
radiation that is made available during daylight hours. There are important
differences between the ability of the three major types of technologies to efficiently
convert sunlight into electricity.

3-17
In general, single crystal silicon devices are the most efficient, with polycrystalline
silicon cells slightly lower in efficiency, and thin film devices of amorphous silicon the
lowest in efficiency. Data that has been gathered over a period of seven years from
actual outdoor testing of these different technologies is shown. This data was
gathered as a part of a US government/utility program called Photovoltaics for Utility
Scale Applications (PVUSA), where multi-kilowatt arrays of various technologies
have been installed and are being monitored for performance in real world
conditions.
Single Crystal Silicon Efficiency

Is Superior
PVUSA testing various systems since 1989
Actual field performance
12
Single Crystal
DC Efficiency %
10
Polycrystalline
8
6
4
Amorphous
Jan-96
Jul
Jan-95
Jul
Jan-94
Jul
Jan-93
Jul
Jan-92
Jul
Jan-91
Jul
Jan-90
Jul
Jan-89
0
Source:
PVUSA Progress
Report March 1996

The PVUSA data shows clearly how the different technologies compare in efficiency.
The average efficiency of the single crystal silicon array is about 10%, while the
efficiency of the polycrystalline modules is about 8%. The thin film amorphous
silicon array averages only about 3% efficient. This comparison shows the normal
seasonal variations in efficiency due to ambient heat in the summer months for the
single and polycrystalline modules. One interesting feature of thin film amorphous
silicon devices is that they are less affected by temperature variations, as shown by
the nearly flat efficiency data throughout the hot and cold seasons.

3-18
Advantages and Weaknesses of

Current Technologies
Every technology available today has some positive features as well as some
negatives. It is useful to put the different technologies side-by-side and make some
general statements about how they compare. In the future, the weaknesses of any
one technology may be overcome, and we may see new manufacturing techniques
that allow for currently weak technologies to become more competitive in the
marketplace. But for now, we seek to identify some very general characteristics of
each technology and give you a highly simplified overview of how they compare.
Cell Technology Comparison Chart

Cell Type
Best Cell
Efficiency
Module Area
Efficiency
Advantages
Weaknesses
Silicon
22.7 %
12-15 %
Well understood;
Receiving
renewed attention
Indirect band gap

limits efficiency;
How thin?
CdTe/CdS
15.8 %
6-8 %
Low cost;
High deposition
rates possible
Cd liability;
Needs more
development
Amorphous
Silicon
13.2 %
4-9 %
Low cost
Looses power
over time;
Low efficiency
CuInSe2
16.9 %
10%
23% potential;
Low cost
Manufacturing yields
are low; Needs more
development
Single Junction
Concentrator
28.7 %
NA
Hybrid PV / thermal;
central power
generation
Lacks production
economy of scale;
Complex BOS
Multijunction
Concentrator
35 %
NA
Hybrid PV/thermal
Space
Lacks production
economy of scale;
Complex BOS


3-19
The bulk silicon (both single crystal and polycrystalline) and thin film amorphous
silicon technologies have been described earlier. The cadmium telluride
(CdTe/CdS) material is another heterojunction thin film technology that shows much
promise and is being pursued by a variety of manufacturers. The multijunction
concentrators are more exotic devices that utilize different cell technologies
deposited on top of one another, to capture different regions of the solar spectrum.
These offer the highest potential efficiencies, but are by far the most complex and
expensive. Their best application would be where space is an expensive premium,
such as in large central power generation plants or in orbiting space facilities.
The primary trade-offs that are evident from this comparison are potentially high
device efficiency vs. reduced efficiency for real modules, and potential low costs vs.
process stability and manufacturing yield issues. There is not one clear choice for
the best technology, as all of them have potential improvements and difficulties to
overcome.
The module area efficiencies mentioned in this comparison are general values for
actual production volume power modules, and are contrasted with the best cell
efficiency values that are often reported in press releases and research papers. It
is very important to draw a distinction between the peak efficiency of a single
research device and the consistently attainable efficiency of a high volume
production module. A single research device may have a high efficiency, and show
the potential that production devices can attain, but actual full production modules
incorporate interconnect space, frame space, gridline coverage and other physical
features that decrease the efficiency of the overall device.
Another concept that is highlighted in this table comparison of technologies is that
concentrator technologies have not achieved mass production status yet. They
have the potential for the highest efficiencies of all the technologies presented, but
involve extensive balance-of-system (BOS) equipment, such as specially designed
sun-position tracking frameworks, concentrating lenses, computer controls, and
special heat sinks. Their ability to compete with flat-plate technologies will depend
on the cost-effectiveness of the entire package needed to create useable power,
and not just on the efficiency of the small individual photovoltaic cells.

3-20
(End of Chapter)

3-21
CHAPTER THREE
MARKET OVERVIEW
3-1
Major Manufacturers
3-2
Market and Application Forecasts

Market Growth Projections
Application Growth Forecasts
3-4
3-4
3-6
Overview of Current Cell Technologies

Single Crystal Silicon (Czochralski or CZ)
Cast Polycrystalline Silicon
Ribbon Silicon
Thin Films
Concentrator Cells
3-11
3-11
3-11
3-12
3-12
3-14
Market Share By Cell Technology
3-15
Cell Technology Characterized by Efficiency
3-17
Advantages and Weaknesses of Current Technologies
3-19

3-22
Chapter Four
General Skills of System
Designers
Many people feel that photovoltaic system design involves using a computer or
calculator and solving some equations that lead to one unique solution. The truth is
that designing solar power systems involves a great deal of judgment before and
after any calculations are made, and any recommendation by a computer program
must be tempered by experience and a keen sense of unpredictable human and
environmental factors. Skilled system designers are critical of their own
assumptions and question their results to see if all contingencies have been
considered.
In this section, we will discuss some of the approaches that should be followed by
system designers to gathering the initial information, dealing with objections and
uncertainties, and thoroughly transferring the final system to the user. These
include:
Considering Tradeoffs
Anticipating and Second-Guessing
Identifying Potential Applications
Handling Customer Objections
Gathering Information
Designing the System
Installation
Transferring the System to the Client
Each of these skills will be discussed next.

Copyright 19978Siemens Solar Industries
4-1
Fundamentals Skills for System Designers
Consider Tradeoffs
Most of the time, system design involves some degree of compromise between
competing and desirable qualities. There are many choices that have to be made:
type and size of equipment, location, amount of backup or redundancy, degree of
protection, level of safety, amount of complexity, initial costs or costs over time, and
so on. Factors that may influence the choices might include the clients budget, the
remoteness of the site, how critical the loads are, the sophistication of the actual
users, and future growth possibilities.
System Design Involves

Tradeoffs
Choices based on budget, remoteness,
how critical is load
Simplicity
and
Reliability
Efficiency
Initial
Cost
Lifetime
Cost
Centralized
Generation
Distributed
Generation

Discussed next is some fundamental tradeoffs that must be made constantly when
designing photovoltaic systems. Both of the extremes of the tradeoffs are desirable,
so the dilemma facing the designer is how far to go towards one side or the other.

4-2
Efficiency vs. Simplicity and Reliability

People designing technology based systems often tend to seek maximum efficiency,
and consider any additional equipment or expense to be justifiable if it results in
measurable efficiency increases. But this tendency must be tempered by real world
concerns for field failures and general unpredictability in weather and load demands.
The primary characteristic of photovoltaic generators is simplicity--no moving parts!
And from this comes an extremely high reliability. Any move away from this
simplicity and associated reliability must be scrutinized carefully. Most users would
prefer a slightly inefficient system that continues to perform year after year over a
sophisticated system that has failed due to a component failure or unanticipated
glitch. It is the users perspective that must be considered first, and not the
designers desire to show analytically under standard conditions a high theoretical
efficiency.
Initial Cost vs. Lifetime Cost

For some clients the most important factor they face is keeping the initial cost of a
system down. It doesnt matter to them that certain components, such as batteries,
are not the highest quality types and will not last more than a few years. They can
afford to buy replacements when needed a few years down the line.
For other clients, the possibility of reducing maintenance costs is extremely
attractive. For example, industrial customers designing critical and very remote
systems understand the high cost of diesel refueling or helicopter time to replace
batteries. These costs over time can be shown to easily justify the increased cost of
choosing the highest quality components from the start.
Centralized vs. Distributed Generation

Photovoltaic technology works well as a distributed source of power, where each
load or home or user can have their own array. However, there may be situations
where a centralized power generation approach is desirable.
A centralized approach may be appropriate when for example a large and complex
hybrid diesel-PV system is needed for a village or commercial site. Trained
personnel can be assigned to operate and maintain the equipment. Responsibility
for proper operation falls on someone other than the final users. But this can also
mean that the users do not appreciate the necessity for conservation in their energy
usage. They may abuse the system, add unnecessary or unauthorized loads, and in
general cause problems for the others in the centralized system.

4-3
On the other hand, if a distributed system is designed, then each site or user must
be responsible for their own energy generation and usage. If they abuse or overuse
their system, their system may fail or shut itself down. But this will not affect the
other users, and the abuser will quickly learn to be more conscious of their energy
usage. Also the geography of the overall site may favor a distributed approach. For
example, Navaho Indians in the central region of the US have their individual homes
located perhaps a kilometer or so from the nearest neighbor. Stringing power lines
to each home would prove uneconomical compared to installing separate and standalone power systems for each home.
Anticipate and Second-Guess

Even when you have seemingly reliable information for your design calculations, you
should think about what could go wrong and try to incorporate equipment that can
help prevent abuse of the system and premature failures. Cultural differences or
climatic conditions that are different from the designers own should be understood.
The final users may not have any appreciation of how to maintain high technology
equipment or they may think that the system can take care of itself, or they may be
afraid of it--all factors that can lead to misuse and failure.
Often photovoltaic power systems are being installed where no electrical power was
available before. Once users become accustomed to using it, they may use much
more that the system designer was told they would. Lights and appliances may be
left on for more hours than planned, or extra lights and loads may be added,
overstressing inverters and wiring as well as draining the batteries.
Simple choices like lights with timers and hard wiring appliances to a central power
distribution center can help to keep the system working within the limits it was
designed for. Feedback to the users from low voltage alarms can help educate
them about their overuse so they stop using non-essential loads before the
batteries are over discharged.
There is not a formula to apply in the system design process that will assure that all
contingencies have been considered. It is up to the designer to stop and think--have
I put myself in the place of the intended user and into the intended climate, and
thought about what could possibly happen? The more thorough the secondguessing, the more robust and reliable the final system design.

4-4
System Design Involves

Second-Guessing
Anticipate what could go
wrong
Incorporate equipment to
prevent improper operation
Question your initial load
and location information


4-5
Identify Potential Applications

The process of system design cannot begin until good cost-effective applications
have been identified. Several key qualities that indicate a good opportunity for
photovoltaic power are given.
Certainly remoteness and any requirement for high reliability clearly point to
considering photovoltaics. And relying on photovoltaic systems instead of utility lines
or generators can reduce vulnerability to failure due to extremely bad weather, such
as ice or windstorms or blowing sand. And having distributed photovoltaic power for
each critical load or site means that if one system is damaged, the others remain
operational.
If there are critical loads at a site, then more than one source of power is desirable.
Photovoltaics can be combined with conventional sources to add reliability and
diversity and insure that critical systems stay on line no matter what may happen.
A unique feature of solar photovoltaic generators is that they can generate small
amounts of power and output at relatively low voltages. Many small loads, for
example a remote telemetry system or an electric fence charger, need only a few
watts of power, while generators cannot be found below a few kilowatts of
generating capacity, and utility line extensions for such small loads would surely be
too costly.
The low voltage of the modules can come in handy when compared to the extremely
high voltages of transmission lines. For example, aircraft warning lighting atop high
voltage transmission towers could tap off the very power lines they support for
power, but the cost and weight of the transformer needed would be prohibitive. But
lightweight solar modules and batteries can be mounted on the tower itself and
provide 12 or 24 volt power to lights without the need for transformers.
Certainly if there is a concern about noise or air pollution or environmental
sensitivity, these are flags that a photovoltaic solution may be appropriate. And
clients may have general social concerns about using renewable sources of power,
or about being independent of utility networks.
The low voltage of typical solar modules makes them safe to handle and install. Of
course once many modules are connected together in series, potentially dangerous
power is present. But most small photovoltaic systems involve only 12 or 24 volt
power, which is safer than the common 110 or higher voltages from utility power.

4-6

Modularity
Safety
High Insolation
Independence
Social Concerns
Environment
Portability
Remoteness
Reliability
Vulnerability
Critical Loads
Low Voltage
Small Power
Noise
Pollution

There may be a desire for a modular approach to power generation, where some
power is needed now and more will be needed in the future. Instead of buying and
installing an oversized generator anticipating long term future power requirements,
you can install only the power that is needed today, and add modules at any time in
the future when budgets or load demands increase. Entire village power systems
can grow this way, starting small and growing as the community grows.
Individual modules are lightweight and can be easily transported if necessary.
Modules have been used for mobile radios, camera battery recharging, lighting for
temporary camps, as well as transported on trucks or by hand from one water pump
to another.
And certainly if there is abundant solar insolation, photovoltaics should be
considered. Why not use a resource that is plentiful, and delivered for free!

4-7
Handle Customer Objections

After potential applications have been identified, designers may face the obstacle of
dealing with objections or questions about the technical aspects of photovoltaic
power systems. Users may not understand how the modules will survive in harsh
environments, or work with other components to give reliable power day or night.
Some typical objections are listed.
Handle Objections
Needs Maintenance
User Education
Breakable
Theft
Safety Concerns
Building Codes
Need the Sun
Initial Cost
Produces DC
Area Limitations

Certainly a common objection is the relatively high initial cost of a photovoltaic power
system compared to a generator only system. This must be handled by presenting
the true costs of operating a generator over time, including regular and unscheduled
maintenance, parts, labor, fuel (including delivery costs) and other invisible costs.
The problem of high initial cost can also be addressed by developing creative
financing arrangements, so that users pay for their energy over time just as they do
for conventional energy.

4-8
Another key objection is that photovoltaic devices make DC electricity, while most
common appliances and other loads are designed to operate from AC power. This
can be dealt with by either proposing the use of DC appliances, or by including a
small inverter that converts DC to AC. The extra cost of the inverter can be
compared to the usually higher prices of DC appliances, and a choice can be made
as to which way to go.

4-9
Gather Load Information

Once any objections have been overcome, it is time to begin gathering critical data
about the situation. There are three areas of information that are important: the
application; the climate; and the client or user. Application information must be as
detailed and quantitative as possible. Otherwise slightly wrong information at the
beginning of the sizing process will result in completely incorrect recommendations.
Gather Information About the

Application
Load requirements
Load profile
Surges
Power quality
DC or AC
Critical loads
Ease of access to site

The individual loads must be well understood, including their current and voltage
requirements, and how many hours each day they will be on. In addition, the profile
of the load demand over a typical week or for each month must be specified. It will
be important to know if the load is greater in the summer or winter. This information
will help determine the proper tilt angle of the modules so that they intercept the
most solar radiation when the loads need it. Also if the load is used only a few days
each week, as in the case of a weekend cabin, can affect the balancing of modules
and battery storage and final system cost.

4-10
Certainly if AC loads are to be included, their requirements for power quality and
surge currents will affect the choice and sizing of a DC-to-AC inverter. The
inefficiency of the inverter must be added into the calculations, and the input voltage
of the inverter chosen will determine the voltage of the solar array and battery bank.
Seasonal Load Profiles

70
Summer peak profile
60
50
40
30
Winter peak profile
20
10
0
Jan Feb Mar Apr May Jun
Jul Aug Sep Oct Nov Dec

If there are any critical or essential loads this will influence the choice of sizing of the
array and battery to handle unexpected bad weather. And the ease or difficulty of
reaching the site may influence the reserve time built into the battery, as well as the
setting of any low voltage alarm signals to prompt maintenance actions.

4-11
Gather Climate Information

Accurate location and weather data is of course critical to proper system sizing.
Data on solar radiation or insolation must be obtained, and not just for a season or
for one year, but accumulated over many years. Specific data for a site taken over a
short period may not reflect the normal weather that can be expected. The usual
format for solar insolation data is in units of kilowatt-hours/square meter (kwh/m2), or
in Langleys (calorie/cm2), or sometimes in megajoules/m2 or btu-hours/ft2.
Measurements of simply hours of sunlight or number of bright days is not
quantitative enough and cannot be used. Common sources of solar insolation data
are airport authorities, agricultural stations, government agencies, and universities.
Siemens Solar has compiled a database of solar insolation for over 2000 sites
around the world. This is used by our computer sizing programs to accurately
predict module output.

Climate
Latitude, longitude
Insolation
Temperature
Variability of weather
Harshness


4-12
Temperature ranges as well must be well know, as modules loose voltage potential
and batteries loose life expectancy with higher temperatures. Temperatures will also
influence the choice of the module design, as 36 cells module circuits are typically
used for the hottest climates, while 33 cells in series are adequate for more
moderate climates.
Local knowledge about the variability or harshness of the weather is important to
gather. Statistical data on solar insolation is usually based on averaged data, so
particular variation patterns from day to day are lost. The severity and duration of
typical seasonal storms would be very helpful in deciding how much reserve capacity
to build into the battery bank, and whether backup generator capacity should be
considered.

4-13
Gather Client Information

A third critical source of information that directly affects the choices you make in
designing a photovoltaic system is the client or user of the system. The information
gathered here may not be as mathematically precise or as easily quantifiable as the
previous application and climate information, but it is quite important.
The clients budget may be a useful starting point for your design process. You may
often have to begin with the budget and work backward to see how much solar
power they can afford at the start. You can always encourage them to add more
solar power as their budget allows.
The reliability of the information that they give to you should be examined. Is the
load information that they have given accurate, or is it just a guess? Are they
repeating what others have said, or are they sources of original data? Sometimes
clients are seeking replacements for fuel-powered generators, and they may make
the simple mistake of thinking that the load is equal to the power rating of the
generator, without looking at detail at the actual loads. If a large generator was
installed for surge power or because it happened to be on hand or for whatever
reason, then any calculations made based on that capability will be far too high for
the actual site.
The technical skill level of the users will have a definite influence on your system
design choices. The sophistication of controls and feedback, and the accessibility to
the system will depend greatly on your assessment of the users abilities and
understanding. If the user is fairly uneducated or unskilled with electricity, then you
will want to make the feedback simple and the access quite limited. This may also
influence your choice of battery type, where you would probably choose a
maintenance-free battery. Whereas, if the user is a sophisticated engineer, you will
probably want to allow greater access to the system controls, and may feel free to
use a battery that requires proper and regular maintenance.

4-14

Client
Budget
Reliability of their data
Level of technical skill
Future growth possibilities
Aesthetics

You should keep in mind the probability of near-term future growth of demand, and
design your system to accommodate that growth. If the load is well known and
fixed, with no possibility of increase, then you can design your system to fit that load
value exactly. But if there is the real chance of load expansion, for example more
channels of radio being utilized, or more hours of lighting needed or more users in a
village, then expansion and increased power flows should be planned. Space
should be left in the module mounting structures, larger wires should be used,
regulators and inverters should be oversized, and so on. This will help to reduce the
need to purchase more components in the near future.
And aesthetics may have an influence on your design choices. Whether the user
wants the array to be visible or not will influence your location options and may affect
your wire lengths and sizes for example. Instead of mounting the array on the
ground at the proper tilt angle, the user may want the array to be mounted on a roof
with a tilt or azimuth angle that is not optimum. This will definitely affect your array
sizing calculations.

4-15
Design the System

It seems that we would never get to the point of finally calculating the array and
battery size! But not we are ready. It takes all this previous data gathering to be in a
position to make the choices needed during system design. Design is not just
plugging in some numbers. It involves balancing calculations with judgments, and
selecting equipment based on the information about the climate, the load and the
client.
Design the System
Array sizing
Battery bank sizing
Wiring
Safety components
User feedback
DC or AC or Hybrid
Mounting
Accomodate future growth

The number of modules in series and parallel for the array can now be calculated,
as well as the capacity and voltage of the battery bank. All of the system wiring
must be chosen, with attention to safety equipment such as circuit breakers, fuses,
disconnects and grounding. Electrical code considerations may dictate the use of
equipment that adds to the overall system complexity and cost, but also adds to
safety.
At this time you should consider the hardware needed to give proper user feedback.
Simple dials or fancy digital meters can be chosen, and even remote transmission of
system parameters can be designed.
4-16
Fundamental decisions need to be made here about whether the system is all DC,
or all AC, or some combination of the two. You have to decide whether the system
will be a pure stand-alone photovoltaic system or a hybrid design with a fuel
generator or wind or hydro assist.
The array mounting design must be made, perhaps including empty space for future
module additions. Components such as inverters charge regulators, and wiring
should be chosen to allow maximum growth, so that costly upgrades or changes can
be avoided.

4-17
Installation
You can perform all the calculations in the world, but if the equipment is not installed
correctly, your system could still fail. Bad communications between the designers
and the installers can lead to wasted time, money and faulty system performance.
As much of the system should be pre-assembled as possible and shipped to the site
ready for installation. This allows you to test the components as they work together
and make repairs or adjustments as needed at the home shop (where tools and
spare equipment is readily available). Charge controls, system disconnects,
inverters and feedback components can often be pre-mounted to a strong
backboard, transported to the site, and installed quickly.
Installation
Pre-assemble as much as possible

Site selection
Safe practices
Codes and inspections


4-18
The installation site should be visited before system design is completed so that
special circumstances can be taken into account. Perhaps there are trees or other
sources of shading at the site. This might require clever mounting solutions or
longer and larger wire than anticipated. Foundation design might depend on the
type of soil, the wind loading, the growth of vegetation and so on.
Safe practices should be followed at all times during the installation phase. All
appropriate local and regional electrical and building codes should be followed.
Each component and the overall system should be fully tested before the installation
is considered complete. Abnormal operations should be corrected, and all system
voltage drops should be carefully measured to insure that all connections are proper.

4-19
Transfer the System to the Client

The system design process is still not complete until the system has been properly
transferred to the client or user. Often this step is left out with the assumption that
the user understands what to do. This can be disastrous. Users might not
understand the importance of no shading on the modules or of regular battery
maintenance. The system can still eventually fail, even if all care has been taken in
designing and installing the components, if the users dont properly maintain their
system.
Transfer System to Client
Simple language or pictorial

Troubleshooting Guide
Service/Maintenance Plan
Get user sign-off


4-20
A Troubleshooting Guide should be created for the user, showing symptoms,

possible causes, corrective actions. A Service and Maintenance Manual should be
on hand, showing the user with simple pictures what to check and when, and how
they should perform the service.
It would be a good idea to get the user to sign-off on these transfer documents. This
adds to the perceived seriousness of the transfer process, and forces them to be
aware of the content of the documents.
Examples of Transfer Documents

By way of example we have included two samples of documents that serve to
transfer understanding and responsibility for proper system operation to
unsophisticated users.
Presented on the next page is a reduction of a large color poster created for a rural
electrification program in South America. It explains in simple words and drawings
the proper operation and maintenance of a home lighting system.
And on the following pages is an owners manual created for small home
electrification systems installed on the Navajo Indian Reservation in the United
States.
Please look at these documents and think about how you would best structure your
transfer documents so that they would be useful to the people involved. Differences
in education and sophistication will require vastly different documentation. Some
documentation for rural systems users may involve only pictures or cartoons, with
very little text, while complex engineered systems may require detailed
specifications. The user should be the starting point for designing transfer
documentation.

4-21

The Homeowners Manual presented on the following pages was prepared for the
owners of simple 12-volt DC home electrification systems installed on the Navajo
Indian Reservation in the United States.

4-22
(End of Chapter)

4-23
CHAPTER FOUR
GENERAL SKILLS OF SYSTEM DESIGNERS
4-1
Consider Tradeoffs
Efficiency vs. Simplicity and Reliability
Initial Cost vs. Lifetime Cost
Centralized vs. Distributed Generation
4-2
4-3
4-3
4-3
Anticipate and Second-Guess
4-4
4-6
Handle Customer Objections
4-8
Gather Load Information
4-10
Gather Climate Information
4-12
Gather Client Information
4-14
Design the System
4-16
Installation
4-18
Transfer the System to the Client

Examples of Transfer Documents
4-20
4-21

4-24
Chapter Five
Applications and Typical
Systems
In this chapter, we look at some of the most popular and economical types of
applications for solar photovoltaic power systems. There are many new applications
that are being developed, but we can discuss here those that have already proven to
be cost effective and reliable.
We can also look briefly at some common system configurations, to become familiar
with some of the equipment that is used in systems. Most all of the equipment
mentioned (such as inverters and charge controls) will be discussed in more detail in
later chapters.

5-1
Fundamentals- Applications & Typical Systems
Flexibility of Application
One of the blessings of photovoltaic technology is that there is a tremendous
flexibility in the kinds of designs and applications that solar power can
accommodate. There is a wide variation in the size of system, the complexity, and
the dependence or independence on natural cycles. There is not one best way to
design a photovoltaic system.
Application Flexibility
1 watt
Megawatts
Near existing
utility power
Remote
Daytime only
Any weather, any

time
Purely PV
Hybrid system
Centralized
generation
Distributed
generation


5-2
Power Flexibility
The inherent modularity of solar photovoltaic systems means that there is
tremendous flexibility in the amount of power you can install. Wind generators begin
at about 500 watts of generating capacity, and fuel generators begin at a few
thousand watts. But you can install as little as a few watts of photovoltaic generating
capacity if that is all that is needed, for example to charge an electric fence energizer
or operate a small telemetry unit.
Hundreds or thousands of watts of power can be installed for homes, farms,
industrial or commercial uses, schools and clinics.
And hundreds of thousands or millions of watts of generating capacity can also be
installed economically as well. Large-scale utility systems serving thousands of
homes or businesses have been installed.
Near Utility Power or Remote

The historical market for photovoltaics has been remote applications where the cost
of extending utility power lines was prohibitively expensive. Even when the distance
for an extension is not very far and the cost is moderate, if the power needed is
small (for example to run a single water pump on a farm) it is still more economical
to install photovoltaic at the site instead of extending the line.
But you should not think of photovoltaics as being appropriate only for remote
applications. The costs involved in utility line hookup even when the distances are
very short can still be higher than the costs of installing a stand-alone photovoltaic
power system. Urban street lighting, bus shelters, advertising lighting and other
commercial uses of power might involve relatively short hookups, but the
transformers and equipment needed, the civil work to trench for power lines or install
power poles, and all the associated costs of utility hookup can make even a short
connection costly.
A perfect example of the cost effectiveness of photovoltaics in an urban electrified
setting is emergency roadside telephones. Thousands of these systems have been
installed along highways, where regular utility power is only a few feet away. But the
costs involved in civil work and equipment make installing photovoltaics the
economic choice.

5-3
Daytime Only or Anytime

Just because the photovoltaic modules use the sun this does not mean that the
loads need to operate during daylight. And it doesnt mean that you cant run your
loads during stormy weather. By using batteries for nighttime operation, and
including backup generators such as diesel, wind or hydro, you can design a system
that will operate in the harshest of environments.
But photovoltaics do offer the unique opportunity to avoid the cost and complexity of
these components if you indeed need only power during daylight! Water pumping is
a perfect example of how solar photovoltaics can match a load requirement with
minimal cost and complexity. A DC pump can be connected directly to a solar array,
and will automatically operate whenever there is sufficient sun. During the summer
when there is greater need for water, the system pumps more water automatically
because there is more sunlight! Attic or ceiling fans are other application areas
where direct coupling to solar modules results in low cost highly reliable operation
during daytime only.
Photovoltaics Only or Hybrid

Generation
Systems can be designed to operate with only photovoltaics as the source of power,
to maximize reliability and minimize maintenance and complexity. Or backup
generators can be included in the system design to provide power day and night and
during seasons of low solar radiation. A diesel, gasoline or propane generator can
charge the battery bank and operate AC loads as well. Wind generators or hydro
generators can also be connected to a system to provide power.
By combining the reliability and simplicity of photovoltaic modules with the 24-hour
availability of standby generators, you can get the best of both worlds.

5-4
Centralized or Decentralized
Generation
Photovoltaics are inherently a decentralized generation technology. The power that
is needed can be installed where it is needed. Each house, schoolroom,
commercial site or transmitter can be outfitted with its own independent system. If
one system fails or needs maintenance, or if the user exceeds the energy available,
the other systems continue to operate without interruption.
But decentralized systems require vigilance by their owners. If the skills or
knowledge of how to operate or maintain the system were not available, then
perhaps a centralized approach would be best.
A centralized photovoltaic power system could have all the modules, batteries,
inverter capacity and controls needed for an entire village. Trained personnel could
maintain the equipment, and would be responsible for proper operation. The users
of the energy are not as directly connected to the source of their energy, but perhaps
this approach is best for their situation.

5-5
Outdoor Lighting Applications

Outdoor lighting applications are often decentralized and relatively small, providing a
perfect match to photovoltaic power advantages. Each lighting unit can have its own
array and battery and controls for maximum reliability and flexibility in site location,
and minimum vulnerability to power loss. Often systems can be designed to operate
from only one or two modules.
Roadside Flashers
Highway Safety Signs

5-6
Bus Shelters
Garden Lights
Portable Lanterns

5-7
Telecommunications Applications
Telecommunications has always been an important market for photovoltaics. The
equipment used in many telecom applications is operated on DC power, making the
match to DC generated solar power simple and economical. Reliability in severe
climates is perhaps the most important attribute of photovoltaics for this market.
Downtime is terribly expensive and may be life threatening, so reliability is critical.
And the modularity and flexibility of photovoltaic system design means that modules
can be used to power even the smallest telemetry station as well as very large
microwave repeaters.
Microwave, TV or
Radio Repeaters

Telemetry Stations
5-8
Radio and Telephones

5-9
Navigation Applications
Similar to the telecom market, aides to navigation require the highest reliability.
Signals or lights must operate under all conditions, in all seasons. They are typically
located in remote sites, and traditional power sources such as utility line extensions
or diesel generators would be quite costly. Each navigational aid can have its own
photovoltaic power supply, making vulnerability to overall power loss extremely low.
And this market is a perfect example of how a relatively small amount of power can
produce a very large benefit.
Railroad Signals
Buoys

5-10
Airport Approach Systems
Offshore Oil Platforms

5-11
Cathodic Protection Applications

Cathodic protection involves the electrical prevention of rusting. Metals in contact
with the earth or moisture will naturally corrode. But this process is an electrochemical reaction, and can be reduced or practically stopped by applying a reverse
current to the natural rusting reaction. The electro-chemical reaction involves DC
current flow, so photovoltaic power is a perfect solution. Small systems can be
placed along remote stretches of buried pipelines, for example, and systematically
and automatically prevent corrosion with minimal maintenance requirements.
Well Heads
Oil and Gas Pipelines

5-12
Water Pumping Applications

Productivity in farming can be drastically improved with greater access to water
supplies. Often photovoltaic power systems are used to replace windmills that have
just broken down too often, or cannot lift the water due to lowering water tables.
Instead of installing a diesel powered pump system in remote reaches of a farm, a
direct-coupled solar system can be installed. Such a water pumping system would
require no batteries, no need for fuel deliveries or engine maintenance, would
automatically produce more water in the summer when it is needed more, and would
greatly improve the productivity of the land. Simple water systems can also be
installed in remote villages or homes. Instead of drinking polluted surface water
villagers can drink fresh, clean ground water, thereby reducing disease related
problems and costs.
Livestock Watering
Irrigation
Village Drinking Supply

5-13
Remote Habitation Applications

One of the most exciting markets for photovoltaics is for home lighting and
appliances. Single homes or whole villages can generate their own electric power
without the need for sophisticated maintenance or regular fuel deliveries. Small
clinics can operate vaccine refrigerators, sterilization equipment, emergency radios
and other critical loads. Typically they are located in very remote locations, and
money is not available for costly generator maintenance or fuel. Vacation homes or
cabins, or even regular residences can have all the comforts of modern life from a
photovoltaic power system, including lights, appliances, radio, TV, VCR, microwave
oven, computers, power tools and refrigeration.
Rural Home Lighting Systems

5-14
Home
Electrification
Schools
Hospitals
and Clinics

5-15
Power for Mobile Applications

Solar modules can be used in mobile applications for battery charging and operation
of appliances and equipment. For example mobile recreational vehicles (RVs) can
have a few modules mounted permanently on the roof, to supply quiet DC power for
lights, radio, TV, appliances, coolers and other small appliances. Boats can use
modules to keep starting batteries fully charged, and also for full operation while
running, such as lights, radio, radar and sonar, and other equipment.
And modules can be used for electric vehicle charging stations. Commuters with
electric vehicles can park under the shade of the solar array, and have their vehicle
battery recharged while shopping or during work hours.
Electric Vehicle Charging Station
Recreational

Boats / Vehicles
5-16
Utility Interconnected Applications

for Demand Side Management (DSM)
As discussed in the previous chapter on Marketing, the application sector for utility
connected photovoltaic power systems is growing, and will become a major segment
of the worldwide market by 2010. Usually these systems are installed without
batteries. The depend on a reliable source of utility power for operation during
stormy weather and during each night. It is interesting that the utilities themselves
see distributed photovoltaic power systems connected directly to their grid as a
useful tool. They realize that such systems can contribute to what is called Demand
Side Management or DSM.
Decreased Need for Costly Peaking

Generation
In their drive to become more economical, utilities cannot just seek to make their
generators more efficient (supply side management), but also must look at how
their generated electricity is being used by their customers (demand side
management). Many utilities seek to help customers consume their electrical
energy more wisely and efficiently. This is especially true during the utilitys peak
demand periods, when costly peaking generators have to be turned on to meet
especially high requirements.
Same Effect as Installing More Efficient

Loads
The essence of DSM involves the user of the utility system being given information
or technology to help better manage and reduce their consumption, without
compromising their comfort or flexibility. One way to achieve DSM is to install more
efficient loads, such as fluorescent lights and computer controlled motors. But by
installing photovoltaics at a commercial or residential site, the customer can actually
generate some of his own energy from his roof, and DISPLACE energy that he
would otherwise consume from the utility. The overall effect is the same from the
utilitys point of view--less energy needs to be generated.
And the wonderful feature of photovoltaics as a DSM technology is that the energy is
generated automatically during the peak demand periods of most utility systems-during summer months and during the middle of the day. This helps especially with
offsetting the added demands of air conditioning, as well as meeting the usual
demands of lighting.
5-17
Commercial Farm DSM

Facility System
Individual Home Rooftop
Commercial Building with Integrated Module Facade

5-18
Reduces Daytime Consumption at the

Customers Site
An example is shown below of the performance of a DSM system using
photovoltaics to reduce the daytime demand of a commercial office building in the
Sacramento, California area. Solar modules generate approximately 15 kW of peak
power during the middle of the day, and act to reduce the overall demand of the
commercial building on the utility system. For large customers, utilities might have
time-of-day pricing, where they would pay a higher price for electricity during peak
demand periods. The added cost of the photovoltaic equipment could be compared
to the reduced demand for exceptionally costly utility electricity.
PV Output Reduces Peak Demand

120
PV Output
Load (kW)
100
80
HVAC (minus PV)

60
Lighting
40
20
Computing
0
0
10
12
14
16
18
20
22
24
Hour of Day
PG&E R&D Building, June 21, 1990


5-19
Utility Applications for Grid

Support and Bulk Power Generation
A further development of utility interconnected photovoltaic systems is in the realm
of grid support. Large amounts of power (greater than 100 kW) are inserted into a
utilitys grid at strategic locations to support the quality of power supplied at the end
of the grid.
Improve Voltage at the End of Long

Lines
Utility lines travel long distances from their original source of generation, and branch
repeatedly to smaller and smaller systems of distribution, until finally relatively small
levels of voltage and power are delivered to individual customers. Along the way
some voltage is lost. If the loss becomes too great this can adversely affect the
customers loads.
A utility company may have to consider restringing new larger conductors for a
particular feeder system if voltage losses are too great. By inserting hundred of
kilowatts or megawatts of photovoltaic power right where it is needed the most, the
utility can improve the voltage delivery of selected feeders in a very economical way
Load Voltage (% of nominal)
Voltage Support Effect

110
105
100
95
With PV
90
Without PV
85
80
0
10
15
20
Distance From Substation (km) .

5-20
6 MW Grid Support Facility
500 KW Grid Support Facility
200 KW Grid Support Facility

5-21
Correct Power Factor

Another problem that utilities face at the end of long feeder lines is that the power
factor of their AC electricity can be poor. The power factor is a measure of how well
the voltage and current waveforms are synchronized with each other. A power
factor of 1.0 would mean that the current and voltage hit their peaks at exactly the
same moment, and maximum real power is transmitted. Over long transmission
lines there is a lot of capacitance, and this tends to push the voltage and current
waveforms out of sync. Also large motors at various commercial sites can also
push the waveforms apart. The more out of sync the current and voltage are, the
less real power is being delivered to the customer, even though the utility must
generate the real power in the first place.
Photovoltaic power systems putting their power through properly adjusted DC-to-AC
inverters could push the waveforms closer together, and result in improved power
factors and better efficiency.
Postpone Costly Transformer

Upgrades
Utility feeders have large multi-megawatt transformers that reduce the voltage of the
electric power from the very high values used for long transmissions down to more
manageable values for local distribution. The transformers are installed with more
power capacity than is needed, but with community growth they may reach near their
full capacity within a few years. At that time the utility might face the need to replace
the old transformer with a larger one. This can cost millions of dollars.
By installing a large utility interconnected photovoltaic power station downline from
the transformer, the required power can be delivered to the customers, and the
costly upgrade can be postponed for years. The cost benefit to the utility is
immediate and substantial, and shows the true cost effectiveness of photovoltaics.

5-22
Delivered Cost Of Utility Power Is Close

To Photovoltaic Cost
Various utilities have recently examined the true DELIVERED cost of their electric
power, as opposed to just looking at the cost of generating the power, and have
found that the dollar value of the electricity at the end of the wire can be substantially
higher than the busbar cost at the generator. This is extremely important because
economic analyses of conventional means of power generation (nuclear, coal, gas)
often look only at the generation costs but do not include all of the costs associated
with transmission and distribution (T & D). This is especially unfair when
comparing the costs of photovoltaic generated power, because photovoltaics is
inherently a distributed form of generation, and power can be generated close to the
customer thereby avoiding many of the T&D losses and costs of conventional
sources. This cost advantage must be looked at to fairly judge the cost
effectiveness of using photovoltaics in utility scale applications.
The results of two economic analyses are shown, one conducted by the Pacific Gas
and Electric (PG&E) utility for a 500 kW grid support system and the other by a utility
in Western Australia. The generation costs of utility electricity are low compared to
the relative costs of photovoltaics, but when all the T&D costs have been added, the
final effective cost of electricity at the customers site is almost doubled!
The cost components used for the analysis are defined below:
Energy:
Capacity:
Min. Load Savings:
QF Savings:
Loss Savings:
Voltage Support:
Substation:
Reliability:
Transmission:
Environment:
The value of fuel savings

The value of deferring new power plants
Avoid power plant operation at inefficient minimum loads
Reduced avoided cost payments to qualified facilities
Locating generation at the load reduces resistance losses
Reducing losses in a high impedance network
Value of avoiding or delaying substation upgrades
Value to customers of avoiding outages
Value of avoiding transmission line upgrades
Value of avoiding emissions
The real cost of large-scale photovoltaic generated power compares favorably to

these utility costs TODAY. There is little need for utilities to wait until the cost of
photovoltaics comes down it is cost effective for them today!

5-23
Annual Levelized Value ($/kW-yr)
Economic Analysis of Grid

Support Systems
800
Environment
700
Transmission
600
Reliability
500
Substation
Voltage Support
400
Loss Savings
300
QF Savings
Min.Load
200
Capacity
100
Energy
0
PG&E
SECWA
Figure 5-4 Two utility analyses show real cost of delivered utility power

5-24
Typical System Configurations

Most photovoltaic power systems are custom designed for their particular situation,
but we can generalize somewhat to show the common approaches to solving power
supply problems. In the following pages we present simple diagrams of the typical
configurations of equipment that are used to serve the applications that we have
discussed previously.
We can group system types into five broad categories:
Simple Single-Module DC Systems
Large DC Systems
AC Power Systems
PV-Generator Hybrid Systems
Utility Connected Systems

5-25
Simple DC System
Configuration
DC Loads

Rural Electric Lighting Systems: A single module is connected to a single low

cost battery through a simple charge regulator. The regulator has terminals for
connection to a fluorescent light. The regulator should have a low voltage alarm
or relay that turns off the light if the battery voltage becomes too low. The user
would have to wait until the module charged the battery back to an acceptable
intermediate voltage before they could turn on the light again. The battery could
be a commonly available automobile battery. It wouldnt be expected to have a
long life, but its low cost would allow the initial system price to be acceptable.
Replacing the battery every year or so could be an acceptable price for clean,
reliable, safe lighting.

5-26
Street Lighting Systems: Single or multiple modules can be connected through

a charge regulator to a deep cycle battery, which powers a streetlight. These can
be used for roadside illumination or for parking lighting or pathway lighting. An
electronic dusk-to-dawn timer is included in the system for automatic operation of
an efficient fluorescent or low-pressure sodium lamp. The light is switched on
automatically at dusk, and either operates for a pre-set length of time or until
dawn. Rugged yet simple charge control and timer circuitry are demanded, for
reliable operation in severe environmental conditions. Sealed batteries are
preferred so that minimal maintenance and no corrosive gassing occur near the
controls. Yet these types of batteries may have shortened life in the outdoors
(hot) conditions expected for this application.
Portable Lanterns: Rugged portable lighting systems are a most affordable

option for low-income rural areas. Lanterns may have the solar module built into
the lantern housing, or more commonly have the module separate, connected by
a removable cord. Sealed maintenance free lead acid batteries are commonly
used, and nickel-cadmium batteries can also be used. Typical arrangements use
a 5-7 watt fluorescent lamp, and a 5-10 watt solar module. Usage time is
typically 4 hours.

5-27
Simple DC Water Delivery Systems
Simple DC System
Configuration

Personal water delivery systems: One or two modules connected directly to a

positive displacement DC water pump can deliver hundreds of gallons from
moderate depths. Typically a waterproof submersible motor and pump is
lowered down a well or borehole into the water. Modules can be directly
connected to the pump, or a maximum power tracking device can be connected
between the modules and the pump to maximize the electrical match and
improve output. During summer months when more water is typically needed for
drinking or irrigation and livestock, greater solar insolation means that more water
is pumped automatically.
Surface mounted pumps: An inexpensive solution to surface mounted water

pumping is to remove an AC motor and replace it with a DC motor that can be
directly connected to solar modules. An inexpensive jet pump conversion can be
made to surface mounted pumps to allow pumping from greater depths, although
the total output will be low due to the internal return flow of water to operate the
jet.
Floating pump: A floating pump is appropriate for canals, lakes, ponds and
even open wells.

5-28
Large DC Systems
Large DC System Configuration
DC
Loads

Multiple module arrays: Larger systems can be designed by adding more

modules and batteries. A larger charge controller would be needed to handle the
increased current flow from the array. Again, some DC loads could be
connected directly to the regulator.
Load distribution box: If the load circuits were too numerous, a DC circuit
breaker distribution box could be used to allow connection of multiple load
circuits.
Multiple charge controllers: If the array is larger than the current capacity of
one single charge controller, multiple charge controllers can be connected in
parallel. Slightly different charge voltages can be set to allow the battery to
gradually reach full charge.

5-29
AC Power Systems
AC and DC System Configuration
DC
Loads
Inverter
Inverter
AC
Loads

Inverter operates AC loads: AC appliances can be powered by adding an ACto-DC inverter. This complex component switches electronically the DC voltage
and current from the battery to produce alternating current (AC) and voltage,
commonly used by household and office equipment. Such a power system could
then operate any load that might be needed, including computers, fax machines,
radio, TV, VCR and CD players, refrigerators and freezers, power tools and
kitchen and bathroom appliances. DC loads could still be included in the system
as well.

5-30
Photovoltaic-Generator Hybrid
Systems
Hybrid System Configuration
AC
Loads

Back-up generator: A common choice is a fuel powered motor generator, either

gasoline, propane, or diesel. By combining the reliability and quiet operation of
photovoltaic modules with the assured availability of generator power during any
season, you can assure power availability upon demand.
Transfer switch: Generator AC output power can be passed directly on to AC

loads. A transfer switch is needed to prevent generator power from feeding
backwards into the inverter. The transfer switch could be a fast acting electronic
design, or a simple manual switch that the user operates when needed.
Rectifier charges battery: The generator can also pass power through a
rectifier that changes the AC back to DC current and voltage. This DC power is
passed to the battery bank, to recharge the batteries after a long period of below
average weather. A smaller battery bank can be installed compared to standalone systems because system autonomy is now provided by the generator.

5-31
Hybrid System Configuration
AC
Loads

!"
DC charging only: A hybrid system can be designed to have the generator act
only as a battery charger. No AC output from the generator is used to run loads.
Instead all the AC power for the loads is output only from the inverter. The
generator is turned on, either automatically by the charge control system or
manually by the user, whenever the battery voltage gets too low during bad
weather.
Avoid transfer switch: Operating all loads from the inverter means there is no
transfer switch glitches that could harm electronic equipment.
Operates generator at full power: The generator is sized so that it can operate
at its full rated output to charge the battery. When the batteries are sufficiently
recharged, the generator is turned off, and the finishing charge is supplied by the
solar modules. Operating at full output means maximum fuel efficiency and long
life.
Battery life extended: The generator is turned whenever needed to keep the
battery from staying discharged for more than a few days. Battery life is
maximized due to minimal sulfation.

5-32
Comparison Between Photovoltaic Only and

Diesel Only Power Systems
We can better understand the benefits of hybrid systems by looking at some
characteristics of photovoltaic-only and diesel-only systems. By combining the two
technologies into a hybrid system, we can usually achieve the best that both have to
offer.
Characteristic
PV only
Dependence on
natural cycles
Cost effective
size
highly
dependent
below 20-30
kwh/day
Reliability
excellent
Fuel
none
Maintenance
annually or
semi-annually
frequent
moderate
good
slow
fast
Initial cost
high
low
Operating cost
low
high
Handling load
peaks
Battery charge
rate

5-33
Diesel only
independent
above 150
kwh/day
depends on
maintenance
frequent
deliveries
Utility Interactive Systems
Utility Connected Configuration

!"# $!%&
Instead of having a system independent of the utility grid a system can be designed
to work with the grid. A specially designed utility-interactive inverter is needed, and
many models are available worldwide. The basic arrangement of a utilityinterconnected system is quite simple.
The solar array is connected to the inverter, as are the utility lines. The output is
connected to the normal distribution box for the house or business.
During the day power generated by the array is fed into the inverter and changed
into pure sinusoidal AC power that is synchronized with the grid frequency.
If that power is needed in the home it is passed on. If the load demand is less
than what the array is producing, the excess is fed into the utility grid system, and
energy is credited to the home.
If more power is needed than the array can produce at a particular moment then
power is drawn from the utility to add to the array power.
Typically there are no batteries in utility-interconnected systems, so at night all
the power needed flows from the utility.

5-34
Recently, new designs of bi-directional inverters have been created that allow utilityinterconnected systems to have battery backup as well. AC power from the utility
passes through the normal meter and into a standard distribution center.
Synchronized sinewave power from the inverter is connected to this distribution
center as well. Inverter power is also available from a second AC output connection
and can be sent to a dedicated distribution box for critical loads.
Bi-Directional Utility Configuration
Critical Loads
(Operate from
battery)
Regulator
Battery
Inverter
Meter
Non-Critical Loads .
(Operate only
from utility
power)

!"%$&' $!%&
The inverter continuously shares between AC utility power and DC battery power.
AC power from the utility can be rectified into DC power and used to recharge the
battery bank during bad weather.
Battery power is drawn upon instantly if there is a utility power failure or
brownout, much like an uninteruptable power supply (UPS) system.
The non-critical load distribution center looses power when there is utility failure,
but the critical center continues to draw power from the inverter and battery.
Bi-directional utility interconnected system offer both independent power during

utility failures and utility backup during normal utility operation.

5-35
(End of Chapter)

5-36
CHAPTER FIVE
APPLICATIONS AND TYPICAL SYSTEMS
5-1
Flexibility of Application
Power Flexibility
Near Utility Power or Remote
Daytime Only or Anytime
Photovoltaics Only or Hybrid Generation
Centralized or Decentralized Generation
5-2
5-3
5-3
5-4
5-4
5-5
Outdoor Lighting Applications

Roadside Flashers
Highway Safety Signs
Bus Shelters
Garden Lights
Portable Lanterns
5-6
5-6
5-6
5-7
5-7
5-7
Telecommunications Applications
Microwave, TV or Radio Repeaters
Telemetry Stations
Radio and Telephones
5-8
5-8
5-8
5-9
Navigation Applications
Railroad Signals
Buoys
Airport Approach Systems
Offshore Oil Platforms
5-10
5-10
5-10
5-11
5-11
Cathodic Protection Applications

Oil and Gas Pipelines
5-12
5-12
Water Pumping Applications

Irrigation
Livestock Watering
Village Drinking Supply
5-13
5-13
5-13
5-13

5-37
Remote Habitation Applications

Rural Home Lighting Systems
Home
Electrification
Schools
Hospitals and Clinics
5-14
5-14
5-15
5-15
5-15
5-15
Power for Mobile Applications

Electric Vehicle Charging Station
Boats / Vehicles
Recreational
5-16
5-16
5-16
5-16
Utility Interconnected Applications for

Demand Side Management (DSM)
Decreased Need for Costly Peaking Generation
Same Effect as Installing More Efficient Loads
Reduces Daytime Consumption at the Customers Site
5-17
5-17
5-17
5-19
Utility Applications for Grid Support & Bulk Power Generation

Improve Voltage at the End of Long Lines
Correct Power Factor
Postpone Costly Transformer Upgrades
Delivered Cost Of Utility Power Is Close To Photovoltaic Cost
5-20
5-20
5-22
5-22
5-23
Typical System Configurations

Simple DC Water Delivery Systems
Large DC Systems
AC Power Systems
Photovoltaic-Generator Hybrid Systems
Utility Interactive Systems
5-25
5-26
5-28
5-29
5-30
5-31
5-34

5-38
Chapter Six
The Physics of
Solar Cells
The process of direct conversion of light into electricity seems almost magic. In this
chapter, you will learn about the basics of how this process occurs, without the use
of intimidating formulas or confusing jargon. The purpose of this explanation is to
make you more comfortable with the fundamentals of how silicon solar cells work.
Other semiconductor materials can be used to make PV devices, but describing a
silicon-based device will illuminate the general principles common to all.

6-1
Fundamentals Photovoltaic Physics
Holes and Conduction Electrons

All matter is made of atoms which consist of a small dense nucleus containing
positive and neutral particles (protons and neutrons) and a surrounding cloud of
fast moving negatively charged particles (electrons). The outer most electrons seem
to be arranged in symmetrical elongated shells or orbitals like stretched out
clouds. What is shown in the figure below is a single silicon atom, which has four
outer electrons.
Silicon Atom
nucleus
outer electron "cloud"
ee-
ee-


6-2
When single silicon atoms are combined, they connect together to form a solid.
Neighboring atoms share outer electrons, forming bonds. These bonds where
electrons are shared between atoms is what holds all matter together. Crystalline
silicon consists of orderly bonding of each silicon atom with 4 neighboring silicon
atoms. Such a highly ordered structure of atoms is also called a crystal lattice.
A model of crystalline silicon is shown below. Each atom is represented by a large
ball, and the interconnecting rods represent the shared electron bonds. Each
silicon atom is connected or bonded to four other silicon atoms, forming a beautiful
repeating lattice structure. The lattice is the same arrangement as found in a
diamond, except that silicon atoms are present instead of the carbon found in a
diamond.


6-3
At the atomic level, light acts as a flux of discrete particles called photons. Photons
carry momentum and energy but are electrically neutral. When semiconductor
material is illuminated by light photons of light actually penetrate into the material,
traversing deep into the solid.
Photons with enough energy that collide with electrons can dislodge them from their
bond. The photon disappears and its energy is transferred to the electron, which
becomes free to wander throughout the semiconductor material as a conduction
electron, carrying a negative charge and usable energy.
It is at the moment of releasing the electron that sunlight energy has been converted
into electrical energy.
Holes Created by Photon

photon
e-
e-
e-
e-
eeee-
eeee-
e-
ee-
e-
eee-
e-
eee-
e-
e-
eeee-
eeee-
e-
eee-
e-
eee-
e-
eee-
e-
e-
eeee-
eeee-
eeee-
eeee-
eeee-
e-
e-
e-
e-
e-
e-
e-
e-
e-


6-4
Whenever an electron is knocked loose it leaves behind a bond that is missing an

electron. Such an incomplete bond is called a hole.
A nearby electron can jump from its bond into the hole and fill it, but this leaves a
hole where the electron came from. In this way the hole moves in the material. But
wherever there is a hole, an electron is missing, resulting in a localized net positive
electrical imbalance there. The hole appears to be a positive charge moving in the
solid, although it is really an absence of an electron moving about. Overall the net
electrical charge of the material is neutral.

6-5
P-Type Dopants
In the absence of any external electric field electrons freed and energized by
photons will wander for a short time and then recombine with a wandering hole.
The energy originally transferred to the electron from the photon is simply lost to the
semiconductor lattice as heat. The key to producing usable output current is to
sweep the freed electrons out of the material before they recombine with a hole.
When there is just silicon atoms in the lattice, the material is called intrinsic or pure.
One way to alter the electrical properties of silicon is to introduce elements into the
intrinsic semiconductor that contribute excess holes and electrons. Materials that
significantly alter the properties of semiconductors are called dopants and the
process of placing them into the semiconductor is called doping.
Boron P-Type Doping

Defecit of
one electron
leaves "hole"
Nearby electrons
can move in and
"fill" hole.
ee-
e-
e-
e-
e-
ee-
e-
e-
e-
e-
e-
e-
e-
ee-
ee-
e-
e-
e-
e-
e-
e-
e-
e-
e-
e-
e-
e- B eee-
e-
e-
e-
ee-
e-
e-
e-
e-
e-
e-
e-
e-


6-6
One dopant used with silicon is boron, which has only three outer electrons, one less
bonding electron than silicon. Each boron atom can only bond with three
neighboring silicon atoms, leaving one bond half-complete. A nearby electron from
a silicon atom can vibrate into the hole site next to the boron, filling the hole. But it
leaves behind an absence of an electron, a hole, where it originally came from.
Another nearby electron can vibrate into that hole, but leaves behind a hole where it
came from. So there exists in the semiconductor structure a wandering absence of
an electron.
Wherever this absence of electron (hole) is there is one too few electrons to balance
the positive charges of the nucleus of that particular atom. This results in a net
positive charge at that atomic site. As the hole site moves around, so does the net
positive charge. Since the moving charges in this material have a positive charge,
the semiconductor material is called positive-type or P-type.
The concentration of boron is quite low, usually around one boron atom to every
10,000,000 silicon atoms.
It is important to understand that the overall net charge in the semiconductor is
neutral, but if you look at small regions, there will be net negative charges at the
boron atom sites and net positive charges moving around with the holes. The boron
atoms will have a negative charge because a neighboring electron has fallen into the
incomplete bond and brings with it a net negative charge. The local negative charge
at the boron atom is balanced by the local positive charge near the hole, this being
now a silicon atom that has temporarily lost one of its outer electrons.

6-7
N-Type Dopants
Another dopant used with silicon is phosphorous. Each atom of phosphorous has
one more outer bonding electron than silicon. The fifth electron breaks away from
the phosphorous atom easily because there is not bond to hold it. It becomes a
freely moving negatively charged particle. Since the moving charges in this type of
semiconductor carry a negative charge, this type of doped semiconductor is called
negative-type or N-type.
The concentration of phosphorous atoms is again quite low, but typically greater
than the boron concentration, usually around one phosphorous atom for every 1000
silicon atoms.
Phosphorous N-Type Doping

Extra electron
can wander
in crystal
ee-
e-
e-
e-
e-
e-
e-
e-
e- ee- e- e- P eeee-
e-
e-
e-
e-
e-
e-
e-
e-
e-
e-
e-
e-
e-
e-
e-
ee-
e-
e-
e-
e-
ee-
e-
e-
e-
e-
e-
e-
e-
e-

Once again, the net charge of the entire semiconductor is neutral. But in small local
regions you can find fixed net positive charges where phosphorous atoms are
permanently missing their fifth outer electron. And there will be moving negative
charges that are those fifth outer electrons, freely wandering throughout the lattice.

6-8
Creating An Internal Electric Field

Regions of p-type boron-doped silicon and n-type phosphorous-doped silicon are
created adjacent to one another. Some free electrons in the n-type region cross
over and fall into the holes in the p-type region where they remain permanently.
As this cross over process continues, every boron site that contributed a hole
becomes permanently negatively charged, and every phosphorous site that gave up
an electron becomes permanently positively charged. Two equivalent but oppositely
charged regions grow on either side of the p-type/n-type interface, creating an
electric field.
The electric field is oriented to push electrons in one direction, toward the n-type
region, and any holes are swept toward the p-type region. Any freely moving
charges that enter the zone of influence of the field are immediately swept out of that
zone, so that zone is called the depletion region. A name given to this type of
electric interface is a p/n junction.
Creating Electric Field

P-type silicon region
ee-
e-
e-
e-
e-
e-
ee-
e-
e-
N-type silicon region
e-
e-
e-
e-
e-
e-
e-
e-
e-
e-
ee- B
e-
e-
e-
e-
e-
e-
e-
e-
e-
e-
e-
e-
e-
e-
e-
ee-
e-
e-
e-
ee-
e-
e-
e-
e-
e-
e- ee- P ee-
e-
ee-
ee-
e-
e-
e-
e-
e-
e-
e-
e-
e-
e-
e-
e-
e-
e-
e-
e-
Permanent internal electric field and lines of force


6-9
The strength of the internal electric field is quite strong. The distance across the
region of the field is only about one micron (1/10,000 of a centimeter, or 1/25,000
inch), but the field potential is about one volt. This means that if the field were to
extend across one inch it would be about 25,000 volts! This strong field is
equivalent to an electronic broom that can sweep freed electrons out of the cell
and create the one-way flow of electrons that can be called electric current.

6-10
Internal Electric Field Sweeps

Electrons Out of Cell
It is this internal electric field that sweeps electrons out of a cell. When light
penetrates into the semiconductor material, knocking free electrons and giving them
potential energy, the freed electrons wander until they are pushed by the electric
field across the P/N junction. They are forced out of the cell, and are available for
useful work.
Internal Field Sweeps Electrons

Out of Solar Cell
Newly created hole
Lines of force of
permanent internal
electric field
h+
photon
eElectrons
return from
eexternal
Freed
electron
circuit to
fill holes.
Electrons
flow on to
next cell
or out of
module
and into
external
circuit.
P/N Junction region

In a module, a number of cells are connected together in series. The electrons flow
from one cell into conductors that carry them to the next cell. In the next cell they
are once again struck by photons, given more potential energy, and swept out of the
cell. Finally the electrons leave the last cell in the module and flow to the load.

6-11
For every electron that leaves a cell, there is another that is returning from the load
to replace it. The wire that is used to make the circuit from the module to the battery
or load and back to the module contains electrons, so as soon as an electron leaves
the last cell in a module and enters the wire, an electron at the other end of the wire
moves into the first cell in the module.
So the PV device cannot run down like a battery, nor can it run out of electrons.
It produces output (electrical energy) in response to input fuel (light energy). A PV
cell cannot store electrical energy, it can only convert light energy into electrical
energy.

6-12
Current Is Regenerated In Each

Cell
The current that is generated in a cell does not simply flow on through all the other
cells in series with it and eventually on to a load. The electrons that are liberated in
one cell flow through the internal p/n junction and on to the second cell, but they fall
into holes there and stop. They must be energized by photons again in that second
cell in order to move out of that cell and on to yet another, and so on. This is why it
is so important for all cells in a solar module to be equally illuminated by light, with
no shadows or dark regions. The current flow from one cell must be regenerated in
each and every cell in series with it in order to have the electric current flow properly.
We will discuss the severe negative effect of shading on cell and module output later
in the program.

6-13
Voltage Increases As Current

Passes Through Series Cells
In each cell electrons gain about one volt of potential energy (or voltage) when they
are energized and ionized by the photon. In passing through the p/n junction, they
lose about one half volt through collisions and accelerations, so they are left with
about one half volt of potential as they leave the front of a cell and move on to a
second cell. They enter into the back of the second cell, and fall into holes in the
silicon structure. This flood of electrons with about one half a volt of potential energy
raises the entire potential of the second cell. The voltage potential of electrons in
the second cell is about one half volt higher than it was in the first cell.
When electrons in this second cell are again struck by photons, they add the
approximate one volt of energy from the photon to their new higher base energy of
one half volt, so now they have about 1 volts of potential. But again in passing
through the p/n junction of this second cell they lose about one half volt again, so
they are left with a net voltage gain of about one volt when leaving this second cell.
They carry this one-volt on to the third cell, where they again fall into holes, and
await yet more photons to energize them again, and the cycle repeats for all the
cells in series.
Manufacturers connect enough cells in series to produce a final voltage that is
useful. Typically modules are used to charge 12-volt batteries, which actually need
about 14-15 volts to be fully charged. So 30-36 cells are usually connected together
in series to make a single module that can produce about 15-18 volts.
If higher voltages are needed, for example to charge higher voltage battery banks or
operate high voltage motors or interface with the high voltages of utility power, then
modules are connected together in series. Some photovoltaic systems that are
connected to inverters that produce AC power to feed into the standard utility power
grid operate at voltages of 450 volts! This would mean that over 1000 cells are
connected together in series! Remember though that each cell contributes only
about one half volt, whether it is the first cell in the series or the last.

6-14
Exercises

!


6-15

"
#

$
%


6-16
Spectral Response
Different sources of light may appear equal in brightness to the human eye but will
contain different amounts and intensities of colors. For example, fluorescent light is
typically stronger in the blue than incandescent lights. Also throughout a day there
is a difference in the spectral content of morning, noon, and evening sunlight, as can
be observed by looking at the sky.
Light is just a narrow range of all electromagnetic radiation that is emitted by the
sun. Radiation is a moving electric-magnetic field, and the field vibrates regularly at
a very rapid pace. The distance between the peaks of the vibrations of radiation is
called the wavelength of that radiation. Light is radiation between approximately
400 nanometers or nm (violet color) and 800 nm (red color).
Standardized Solar
Spectral Distribution
UV
Power
Density
Visible
Light
IR
1.2
( mw/cm2/nm)
0.8
0.6
0.4
0.2
200 400 600 800 1000 1200 1400
Wavelength ( nanometers, nm )


6-17
The spectral response is a measurement of the response (measured by generated

current) when a PV device is exposed to a spectrum or range of light. A 100%
response would mean for 100 photons of a certain wavelength that are absorbed,
100 electrons would be freed and swept out for use.
Spectral Response
Q
u
a
n
t
u
m
E
f
f
i
c
i
e
n
c
y
1.00
CZ
0.80
CIS
TFS:H
0.60
0.40
0.20
0.00
0
0.2
0.4
0.6
0.8
1.2
Wavelength (m)

Light of the same color or wavelength will produce different amounts of current in
different semiconductor devices. The spectral response of a typical CZ silicon cell
begins about 350 nm, peaks around 800 nm and falls off rapidly beyond 1100 nm.
The response range spans the entire visible spectrum and reaches into the near
infrared. The spectral response of a typical cell made from thin film
silicon:hydrogen, on the other hand, is narrower, beginning around 300 nm, peaking
around 500 nm, and falling off beyond 700 nm. And a third type of PV device
material, copper indium diselenide or CIS, has yet another spectral response range,
beginning just short of the other two but extending further into the invisible infrared
region.
Because a cells response to light depends on the wavelength of that light, just
knowing the total energy of the light is not enough information to predict cell output.
Two sources of light can have the same total energy density but one source could
have its energy in the form of a few high energy blue photons, while another could

6-18
contain many low energy red or infrared photons. Also, two light sources can
appear to be similar in brightness to the human eye, but one may emit a great deal
of extra radiation beyond the visible range where our eye will not notice but to which
the PV device will respond. A standard typical outdoor spectrum has been defined
as the spectrum from the sun that filters through 1.5 thickness of our atmosphere,
and is referred to as Air Mass 1.5. This serves as a common reference spectrum
for comparing device output.
Exercises

"
&'((#

)* !+,-$./

!

0
+,-
$.


6-19
(End of Chapter)

6-20
CHAPTER SIX
THE PHYSICS OF SOLAR CELLS
6-1
Holes and Conduction Electrons
6-2
P-Type Dopants
6-6
N-Type Dopants
6-8
Creating An Internal Electric Field
6-9
Internal Electric Field Sweeps Electrons Out of Cell
6-11
Current Is Regenerated In Each Cell
6-13
Voltage Increases As Current Passes Through Series Cells
6-14
Spectral Response
6-17

6-21
Chapter 6 Answers
The Physics of Solar Cells

The photovoltaic effect describes the property of certain materials to produce electrons
when light falls on (and is absorbed by) the material's surface.

Electron, positive (+)

b. An electron around a silicon atom

c. Get the electrons out before they recombine with a hole

N, negative, electron

P, positive, "hole" (absence of an electron)

Negative, positive

N, phosphorous

b. Convert energy
Siemens Solar Basic Photovoltaic Technology
6-1
Photovoltaic Physics

c. Light particles (photons)

d. Receive electrons from the circuit as other electrons leave the cell and pass into the
circuit.

b. CZ Cell (refer to spectral response graph)
6-2
Photovoltaic Physics
Chapter Seven
Module Manufacturing
and Testing
The manufacturing process used by Siemens Solar involves many proprietary elements
and is quite involved. But it is important to discuss the manufacturing process so that
you can appreciate the technical features and benefits of the modules, and can have
confidence in the durability and longevity of the products.
The manufacturing process can be presented as three segments or phases: crystal and
wafer fabrication; cell manufacturing; and module assembly. Each of these processes
is discussed next.
Manufacturing Flow
Crystal and
Wafer Fabrication
Cell Manufacturing
Change quartz
into pure silicon
Diffuse with
phosphorus
Solder cells
together
Melt and grow

into single crystal
Coat with antireflective coating
Laminate
Crop, slab and

saw into wafers
Print contacts
Attach frame,
j-box, diodes
Etch surface
Electrical test
Final electrical
test


7-1
Components Module Manufacturing & Testing
Wafer Manufacturing
Raw Materials
The process of making a solar module ultimately begins by mining the raw material.
This usually consists of high quality silica or quartz (SiO2) from mines around the world.
Silicon (Si) is an abundant element on the earth, and can be found in common sand,
but it is more economical today to begin the process with silica that has been somewhat
naturally purified. In large furnaces, the silica is reduced to silicon and then purified
until it becomes 99.9999% pure! Examples of these purified rocks of silicon are shown
below.


7-2
Crystal Growing
Chunks of purified silicon are carefully loaded into a crystal growing furnace, along with
a small amount of the element boron. The furnace is sealed and the chunks are heated
to greater than 1400 deg.C. (over 2500 deg.F.) until they melt (Figure 7-4). A small
seed crystal of pure silicon about the size of a pencil is lowered inside the furnace until
it touches the liquid. The cooler seed acts as a template and atoms of silicon and
boron freeze onto it, and the seed begins to grow (Figure 7-5). By carefully controlling
temperatures and the speed of growth, the diameter of the growing crystal is increased
to about 5" and then held steady by computerized controls. The crystal is then grown in
length until almost all of the original silicon material is exhausted.
This process of growing a single crystal structure from a small seed crystal is called the
Czochralski method, usually abbreviated as just CZ.


7-3
Cropping
After removal from the crystal growing furnace, the ingot top and bottom (crown and
tail) are cropped off and recycled (Figure 7-6). The ingot is cut into short sections for
easier handling. The sides of the section are cropped along its length to make a
squared-off block. The M line of cells is cut almost square for the greatest packing
density in a module. The Pro line of cells is cut less, leaving more of a rounded shape
(Figure 7-7).


7-4
M and Pro Series Cells
M series cell pack

together m ore densly
Pro series cell requires

less overall processing


7-5
Slicing Wafers
The section is mounted into a wire saw to cut individual wafers. The wire saw involves
a single long thin wire wrapped many hundreds of times around four rotating drums.
The ingot section is placed in the center of the four drums, and is slowly pushed up
through the web of the wires. Hundreds of thin silicon wafers are cut all at the same
time this way.
Surface Etching
The wafers are then etched to remove some surface damage caused by the wire saw
abrasion, and to create a surface that helps absorb light (Figure 7-9). The final wafer
surface is made of millions of tiny 4-sided pyramids, following the pure crystal structure
in the original seed. These pyramids reflect light amongst themselves and allow more
light to be absorbed (Figure 7-8).


7-6
Cell Manufacturing
Diffusion
Wafers are next loaded into diffusion tubes which are long tubes of silica glass,
surrounded by large resistive heaters, and that will contain a hot gas of phosphorous. A
computer controls the entire process, slowly moving wafers into the chamber, allowing
gases to flow for precise amounts of time, and then removing the wafers (Figure 7-10).
This process is conducted in clean room conditions similar to the semiconductor
industry. Clean room conditions are important in optimizing the performance of the
cells.

Inside the tube atoms of phosphorous penetrate or diffuse into the wafers a short
distance, only 12 millionths of an inch or 0.3 microns, and embed themselves in the
crystal structure. This embedding of select impurities into a semiconductor like silicon
is called doping.
The differences between phosphorous-doped silicon and the base material of borondoped silicon set up a permanent electric field in each wafer. This field sweeps out
electrons that are knocked loose by light, and makes the wafer into an active solar cell
when light shines on it.

7-7
Anti-Reflective Layer Deposition

The diffused wafers are coated with thin films of metal oxide to effectively reduce the
amount of light reflected from the surface, and therefore increase the amount of light
absorbed and converted into electricity. This AR coating turns the cell surface dark
blue. This is because the cells are most effective with green, yellow, and red light, and
do not use blue light very well. So the design of the AR coating is optimized for the
most effective light, and the blue light is allowed to be partially reflected, so the cells
look blue.
Screen Printing of Gridlines

The cells are loaded onto a series of automated high-speed screen-printers that deposit
a pattern of metal paste onto the front and back (Figure 7-11). The pattern of thin lines
and contact pads is designed to efficiently collect electrons that reach the cell surface,
without blocking out too much sunlight.


7-8
The metallic gridlines are the final structure that makes up a complete cell.
Structure of CZ Cell
Metallic
Gridlines
Anti-reflective
Coating
150 nm
300 nm
N-type layer
Electric Field
at the p/n
junction
P-type layer
250000 nm
Metallic
Gridlines

Cell Operation
e-
Thin N-type Layer

P/N Junction
region of
electric field
Thick P-type Layer

7-9
Every cell is tested under simulated noonday sunlight to determine its electrical output,
and cells of similar output are sorted and grouped together (Figure 7-14).


7-10
Interconnecting Cells
Cells are connected together in series to add voltage. This is done by an automated
soldering machine that connects the back of one cell to the front of the next with two
ribbons of tin-coated copper ribbon (Figure 7-15). The cells are soldered to the ribbons
at multiple points to allow for expansion and contraction of the copper against the
silicon. Two ribbons are used to add reliability to this critical interconnection.
Lamination
Strings of 10, 11, or 12 cells are soldered together to create final groups of 30, 33, or 36
cells, and then are laminated into a sandwich of support materials designed to insure
long operating life. The sandwich consists of tempered glass, then one layer of EVA
(ethylene vinyl acetate), the cell circuit, another layer of EVA, and finally a back cover of
multiple polymer sheets that prevent moisture penetration. The interconnected cell
circuit is therefore floating in a sea of plastic. The sandwich is first sealed in vacuum
laminators that remove moisture and air and soften the plastics, and is then thermally
set in large ovens.
Final Module Assembly Steps

An edge gasket is placed along the perimeter of the laminate to create a soft cushion
for the final metal frames. A strong frame is attached to the edges of the laminate for
strength.
In modules which have three strings of cells (such as the SM55, SM50-H and SM46),
separate positive and negative junction covers are attached at each end. Each box is
secured to the laminate with strong adhesive, and contains a bypass diode, which
decreases the negative effects of any shading of cells that might occur.
In module designs with four strings of cells (such as the SP75), a single junction box is
attached at one end of the module. Dual bypass diodes can be installed in the single
junction box, with each diode protecting one half of the module circuit.

7-11
Every completed module is tested again under simulated noonday sunlight to insure it
meets minimum power output specifications. The output is recorded on computer for
quality control and customer assurance. The modules also are tested for safe high
voltage operation by putting 3000 volts between the circuit and the metal frame. All
modules are also visually inspected for blemishes or flaws before being packaged 4 to
a box for shipment.
Cells Connected In Circuit
e-
CZ cell
Interconnecting
copper ribbons


7-12
Exercises

! "

#

$ "

%

&'
(

))))))))))))))))))))))))))))))))))))))))))))

7-13
Qualification Testing of
Module Designs
PV power systems must be reliable even in the harshest environments and operate for
many years to be economically justified. Siemens Solar currently guarantees module
output for 10 years, and life cycle cost analyses are done based on module life of 20
years or greater. To be confident to make such long-range projections, module
manufacturers subject their modules to intense testing to qualify their electrical and
physical design. These tests are not normally done on every production module
because they are destructive tests and they can cause some discoloration or
degradation to the tested modules. These tests described below are qualification
tests which means that they are used to qualify an existing or new module design as
a truly robust, long-life design, suitable for installation in any climate for many years.
These tests are typically done on any new module design or when changes are made
to existing designs. Module manufacturers may have their Quality Assurance
Department perform some of these tests on an ongoing basis on samples of regular
production modules, to continually check that manufacturing processes are producing
quality products.

!"#

7-14
CEC Tests (JPL Block V Tests)

The first work done on creating standards for testing terrestrial solar modules was done
by the Jet Propulsion Laboratory (JPL) of the California Institute of Technology for the
U.S. Department of Energy back in the late 1970s. JPL was asked to develop a set of
tests to demonstrate module durability and reliability. Over a period of a few years, the
tests evolved with modules being obtained from U.S. manufactures in block buys, so
the tests became know at the Block I Test, then the Block II Test, and so on. The
last version of the JPL test design was Block V, and these are detailed in JPL
Document No. 5101-162, Block V Solar Cell Module Design and Test Specification for
Residential Applications - 1981. The basic test design developed by JPL has been
modified by various organizations around the world.
The Commission of European Communities (CEC) and the International European
Community (IEC) have developed updated qualification tests based on the initial work
done by the Jet Propulsion Laboratory. The latest version of the CEC test is CEC 503,
which supersedes CEC 502 and 501. To pass modules may not exhibit greater than
5% loss of electrical power in any test.

7-15
CEC Tests Summary:

UV Exposure:
Modules are exposed to 15 hours continuous of 1000 watts/m2 of UV light, to

accelerate any degradation of the encapsulant plastic.
Damp Heat:
Modules subjected to 1000 hours continuous of 85% relative humidity at 85oC. This
simulates severely humid tropical environments and will reveal any weaknesses to
moisture penetration.
Thermal Cycling:
Modules are heated to 85oC then cooled to -40oC for 200 cycles. This is a greater
temperature swing than any real module on the earth would experience. This tests for
effects of expansion and contraction of contacts and lamination.
Humidity-Freezing:
Modules are thermal cycled from 85oC to -40oC in 85% relative humidity for 10 cycles.
This tests for ingress of moisture that could result in corrosion and deterioration of the
laminating plastics.
Mechanical (Wind) Loading:

Modules are mounted in a frame and subjected to 50-pound/square foot (psf) or 244
kg/m2, equivalent to 125 mph (200 km/h) winds, on alternate front and back for 10,000
cycles. This is to simulate severe wind conditions and to check for loosening of
contacts and possible cell breakage.
Twisting:
Modules are fixed at three corners and the forth is lifted approximately 1" (3 cm), to
again simulate wind conditions and torquing, to check for cell breakage and loss of
electrical contact.
Hail Stone Impact:

Ice balls are shot directly at the module glass front surface from a cannon using
compressed air. Various impact points are chosen, including above cell edges, the
center of the glass, and above interconnects. Siemens Solar modules withstand 1 1/2"
hailstones traveling at 63 mph in this test (4cm diameter at 100 km/h). In a real
hailstorm, many stones would impact the glass at angles, and would glance off.
High Electrical Potential:

3000 volts DC is applied between the cell circuit and the metal frame to test for any
current leakage. A maximum of 50 microamps is allowed to flow at room temperature.

7-16
UL Approval of Siemens Solar Modules

Siemens Solar modules have qualified for the Underwriters Laboratory (UL) listing
(Figure 7-16). This listing assures consumers as well as inspectors, installers, lenders
and other professionals that the module design is safe. A wide variety of tests were
performed to insure that the modules are safe for consumers. Some are presented
below:
Construction Integrity:
Insulating materials, current carrying parts, internal wiring wireways, assembly and
installation factors, connecting means, bonding, material compatibility, spacing, wiring
compartments, corrosion resistance, sharp edges, accessibility, fire resistance and
encapsulation.
Performance Tests:
Temperature, voltage and current measurement, leakage current, strain relief, dielectric
voltage withstand, inverse current overload, installation/maintenance, impact, fire,
exposure to water spray, accelerated aging of gaskets and seals, temperature cycling,
humidity cycling, metallic coating thickness, hot spot endurance, arcing, mechanical
loading, junction cover crush resistance.
Underwriter's Lab Tests
Consumer and Inspector

confidence in UL label
Construction Integrity
materials used, safe design, fire
resistance
Performance Tests
impact, inverse current, water,
crushing, gaskets
Siemens Solar Industries

Camarillo, CA 93011
MODEL M55
PHOTOVOLTAIC MODULE
AT 1000 W/ M 2 SOLAR IRRADIANCE
AND 25oC CELL TEMPERATURE
MAX. POWER
53 WATTS
30B9 LISTED
SHORT CKT.
3.35 A
RATED
3.05 A
MAX. SYST. OPEN CKT. V.

600 VOLTS
OPEN CKT.
21.7 V
RATED
17.4 V
FIRE RATING
CLASS C
SERIES FUSE
5A
FIELD WIRING
BYPASS DIODE
COPPER ONLY, 14 AWG MIN. INSTALLATION GUIDE
INSULATED FOR 75 C MIN.
233-701500-20
MADE IN U.S.A.

7-17
U.S. Coast Guard Tests

The US Coast Guard has developed its own rigorous qualification test program detailed
in Specification 401 (III), Solar Photovoltaic Arrays for 12 volt DC Marine Aids to
Navigation. Some are variations of the JPL type of tests, while some, like the PIT
test, are unique to the USCG. Special modules (MAR 10, 20, and 35 modules) are
manufactured by Siemens Solar to meet these extremely tough requirements:
Pressure Immersion Temperature Cycling (PIT):

Modules are immersed in salt water and alternately pressurized from 0-5 psig (0-3500
kg/m2). The water temperature is cycled from 40-50oC. to 3-9oC. for 500 cycles. This
is a severe test of a modules ability to handle operating in harsh marine environments,
where navigational buoys might be tossed by large waves.
Shock:
Modules are lifted at one edge to a height of 4" (10 cm) and dropped.
Twisting:
This twist test is similar to the CEC test where one corner is lifted approximately 1
(3cm).
Steel Ball Impact:

This is a more severe test than the ice ball impact test, where a 2.36 oz. 1" diameter
steel ball (67 gm 2.5 cm diameter) is dropped onto the module glass front from a height
of approximately 3 feet (1 meter).
Termination Robustness:
The module is suspended by its electrical output cable.
Temperature Shock:
Modules are cycled 3 times between 71oC. and -57oC. with a transition time of less than
5 minutes. This is more severe than the gradual temperature transitions of the thermal
cycling described earlier, and would simulate modules mounted on offshore buoys
being tipped into freezing arctic waters.
Vibration:
Modules are subjected to varying frequency of vibrations from 5 to 200 Hz for 84
minutes along each of the three axes.

7-18
Specific Features and Benefits of

Siemens Solar Module Technology
The entire goal of the manufacturing engineering done at Siemens Solar is to produce a
module that can last in the real world. The preceding discussion of module
manufacturing and testing cannot be of value unless it is connected to benefits to the
final user of the modules. Some of the ways that the technical features of the modules
can result in direct user benefit are listed below.
Single Crystal (Czochralski Method) Silicon Cells

The extremely quick output response to even dim light means high efficiency during
dawn and dusk and in overcast environments, and superior performance during the
most critical time of the year. This superior performance leads to lower cost per usable
amp-hour delivered to loads.
Textured Cell Surface

Tiny pyramids etched on the cell surface scatter all wavelengths of light, and increase
absorption and reduce reflection inexpensively across the full solar spectrum. This
feature is unique to single crystal lattice materials, as polycrystalline surfaces do not
have uniform lattice orientation and therefore cannot achieve uniform pyramidal
reflection enhancement.
Anti-reflective Coating on Cells

Multiple layers of thin metal oxides cover the front surface of each cell, to more
gradually change the index of refraction from glass to silicon, thereby decreasing
reflection and increasing light absorption and efficiency. The Siemens Solar TOPStm
process is our proprietary method for optimizing and matching both a texturized surface
and an antireflective coating, which produces a superior absorbing surface.
Emissive Back Surface

A final back layer of highly emissive white Tedlar minimizes module and cell
temperature, and therefore maximizes cell output.

7-19
Choice of Single or Separate Junction Boxes

The SM-series of modules have separated positive and negative junction boxes, which
makes wiring easy and reduces the possibilities of wiring errors in the field.
The SP-line modules (and SM55-J module) have a large single junction box that
accepts either conduit or single conductor cable pushed through a waterproof
elastomer seal. The large J-box contains dual bypass diodes that are attached with
screws for easy replacement, as well as allowance for a screw-in blocking diode. All
stainless steel screw terminals and two isolated floating terminals insure easy and
reliable connections.
6 or 12 volt Configuration
The new large single J-box allows for field adjustable operation at either 6- or 12-volts.
A wiring diagram is molded into the terminal cover for permanent reference.
White Background Around Cells

The space between cells is filled with an extremely bright white background. Light that
does not hit a cell is scattered by this specular surface and is internally reflected within
the glass until it reaches a cell edge. Thus light that might not have been used is
converted into current. This is more efficient than other module designs with a blue or
less reflective surface.

$"#% "

7-20
Cell Crystal Planes Oriented at 45o Angle to Cell Edges

The single crystal structure is oriented so that the orthogonal crystal planes are not
parallel to the cell edges, but are angles at 45o. Any small cracks will probably
propagate under the dual interconnect ribbons and not cut off an edge section from
contributing power. This reduces the chances of local hot spots and increases the
reliability of output over a long module life. Polycrystalline cells cannot be oriented in
this way.
External Grounding Screw

Positive grounding of module frame is easy and increases chances of survival in
lightning strikes.
Dual Cell Interconnect Ribbons

Two solder coated copper ribbons fully span both the top and bottom of each cell. This
insures cell interconnection even when a cell is cracked or when glass is shattered.
Cells Soldered at Multiple Points

Interconnect ribbons are not solidly soldered to the silicon but have multiple contact
points. This reduces tension from the difference in thermal expansion of silicon and
copper, and increases the life expectancy of the module.
Tempered Water-White Low Iron Glass

Maximum protection and maximum transmission of light are possible. The glass does
not appear green when viewed on edge, due to extremely low amounts of iron that
would scatter and absorb light.
Glass Roughness (Stipple) Faces Inside

The rough surface from the tempering is placed on the inside, so that a smooth surface
is presented to the environment. This allows for the natural cycles of wind and rain to
efficiently clean the module, virtually eliminating the necessity of cleaning.
Junction Box Lid Attached to Junction Box

A strong connecting strand keeps the lids with the modules, preventing loss during
installation or maintenance.

7-21
Dual Bypass Diode Protection

Some manufactures allow for only one bypass diode to be installed in their junction box.
All Siemens Solar power modules have two bypass diodes installed at the factory. The
number of series cells removed from the circuit in the case of local shading is not the
full 36 cells in one module, but is limited to 24 cells in the SM-series and only 18 cells in
the SP-series of modules. This reduces the heating of any shaded cells, and preserves
array output power even better during periods of local cell shading.
High Dielectric Polymer

Current leakage from the cells to the metal frame is limited well below 50 microamps
even at 3000 volts, for added safety in high voltage arrays.
Back Contact Pattern is a Grid

An open grid of silver lines gives equivalent conductivity to a solid back contact, but
reduces cell cost and allows for unused long wavelength light (infrared) to pass through,
reducing module temperature and maximizing voltage.
UL Approval
The Underwriter's Laboratory approval gives extra assurance to developers, installers,
lending institutions, and customers that the technology is proven and safe.

7-22
Modules Are Extremely Reliable

Perhaps the most essential benefit of using photovoltaic power is high reliability. The
technical features described earlier combined with high overall manufacturing quality
result in an extremely reliable product, as our module return experience illustrates:
Module Return History
Solar modules are extremely reliable!

During 1983-1992
Total installed time
Total returns
+ Failure rates:
181,611 SM50-H sold

637,558 years
385 modules
.069 failures/ million hours

.0006 failures/ year

An analysis was done in 1994 for our popular SM50-H module. The total returns due to
module failure were accumulated for a ten year period, and compared to the total hours
of operation worldwide for all the SM50-H modules shipped and installed during that
time. With only 385 modules returned for failure during the ten years, and a total
estimated operating time of over 637,000 hours for all the 181,000 modules shipped,
the failure rate calculates to be only .069 failures in one million hours of operation! Or
about one failure in 14 million hours!
It must be emphasized that a complete power solution usually involves more than just
modules. Batteries, charge regulators and control systems, fuses and circuit breakers,
diodes, and of course load devices are involved. All of these other components have
failure rates higher than solar modules, and care must be taken when designing and
installing photovoltaic power systems to minimize their effect on overall reliability.
7-23
Exercises

*
!

! +

!

%

,

-.

))))
/

))))

))))
(
012( ))))
3

))))

7-24

))))
))))
))))
))))
))))
World Class Module Specifications

The previous discussion illustrates the extent that Siemens Solar seeks to achieve the
highest quality products for the world market. We seek to set the standard for solar
module performance and quality. Many of you will be involved in the process of writing
or reviewing proposals and tenders for solar equipment. We would hope that you
would apply the highest standards to module manufacturing, testing and performance
measurements.
Below are listed the various standards of performance and testing that Siemens Solar
modules have met. They should be the standard for consideration to insure a world
class project or system!
Include In Tender Specifications...

Performance tested in accordance with IEC
904 and ASTM E1036
Pass environmental tests required by JPL
5101-161 (Block V)
Meet UL 1703 requirements and listed under
UL file No. E79555
Qualification certificate issued by CEC Joint
Research Center, Ispra Establishment
registration No. PV-MT-503-83/95 according to
requirements of CEC specification No. 503


7-25
(END OF CHAPTER)

7-26
CHAPTER SEVEN
MODULE MANUFACTURING AND TESTING
7-1
Wafer Manufacturing
7-2
Cell Manufacturing
7-7
7-11
Qualification Testing of Module Designs

CEC Tests (JPL Block V Tests)
UL Approval of Siemens Solar Modules
U.S. Coast Guard Tests
7-14
7-15
7-17
7-18
Specific Features & Benefits of Siemens Solar

Module Technology
7-19
Modules Are Extremely Reliable
7-23
World Class Module Specifications
7-25

7-27
Chapter 7 Answers
Module Manufacturing & Testing

The incorrect steps are:
f. 20% of all modules made are tested for output power (100% of the modules are
tested for output power)
j. Boron is added to each wafer before printing (boron is added when the silicon is
melted to make crystals)

The proper order for the steps is:
l - d - b - m - g - i - a - e - c - k - h

Condition
A.
B.
C.
D.
E.
Test
Modules in a moist tropical climate

Modules bobbing on buoy in ocean
Modules on alpine mountain top
Modules in large 450 VDC array
Modules in thunderstorm area
h
e
g
a
b
Another test perhaps

g, a
c, f, g, h
d, g
d, g
c, d
Note: other combinations are possible.
7-1
Module Manufacturing and Testing
Chapter Eight
Output Curves
The single most important technical aspect of photovoltaic cells and modules is the
current-voltage output curve. Understanding the curve allows a system designer to
anticipate how a module will be influenced by the environment, and how a particular
load will interact with the module. The output curve also shows how the module
output is "bounded" and therefore safer than battery or generator in the case of short
circuits. As you read this chapter sketch the curve often, label points on the curve,
and sketch how the curve might be influenced by the environment. This will help in
understanding sizing and wiring concepts presented later in this book.

8-1
Components Output Curves
Basic Current-Voltage Curve

(The I-V Curve)
Standard Output Curve
4.00
60.00
Isc
50.00
3.00
Imp
2.50
40.00
2.00
30.00
Power
1.50
20.00
1.00
Voc
Vmp
0.50
0.00
0.00
5.00
10.00
15.00
20.00
Power (watts)
Current (amps)
3.50
Current
10.00
0.00
25.00
Voltage (volts)

The standard representation of the output of a PV device (cell, module, or array) is

called the current-voltage curve. The letter "I" is usually used to represent electrical
current, so this is also called the "I-V curve". The output current is fairly constant
over most of the operating voltage range, and is in fact considered as a type of
constant current source in this range. Eventually though at high enough voltage, the
relatively level current rapidly falls off past a "knee" region on the curve.
The curve represents a "snap-shot" of all the potential combinations of current and
voltage possible from a module under specified environmental conditions. The
particular current and voltage at which the module operates are dictated by the load.
This is discussed later in this chapter.

8-2
Output Curve Terminology

Short Circuit Current (Isc):
The maximum current produced by a device under given conditions of light and
temperature, corresponding to no output voltage, so the power at this point is zero.
Open Circuit Voltage (Voc):

The maximum voltage from a device under given conditions of light and
temperature, corresponding to maximum voltage potential but zero current flow, so
the power at this point is zero.
Current at Maximum Power (Imp):

The current that results in maximum power under given conditions of light and
temperature used as the "rated" current of a device. This value usually occurs at the
knee of the I-V Curve.
Voltage at Maximum Power (Vmp):

The voltage that results in maximum power output under given conditions of
temperature and light, used as the rated voltage of a device and to determine how
many cells or modules are needed to match a load voltage requirement. This value
usually occurs at the knee of the I-V Curve.
Maximum Power (Pmax):

The maximum output power from a device under given conditions of light and
temperature. The product of the current at maximum power times the voltage at
maximum power.
Pmax = Imp X Vmp

8-3
Module Output Curves

Most photovoltaic module manufacturers produce a variety of modules with different
voltage and current characteristics to suit different markets and customers. As an
example, the standard module I-V Curves for modules made by Siemens Solar are
shown below.
SM Modules
1000 w/m 2 and 25 deg.C.
5.00
4.50
Current (amps)
4.00
3.50
3.00
SM55
2.50
SM50-H
2.00
SM1
SM6
SM46
SM20
1.50
1.00
0.50
0.00
0.00
5.00
10.00
15.00
20.00
25.00
Voltage (volts)

It is the number of solar cells connected in SERIES that determines the maximum
voltage of a module. Each silicon solar cell can produce a maximum of about 0.6
volts at full sunlight. So 36 cells in series will result in a maximum voltage of about
21 volts. Some modules are made with 33 cells for a maximum of about 19 volts.
And modules made with only 30 series cells will produce a maximum voltage of only
18 volts.
And it is the SIZE (area) of a single cell that determines the maximum current
output. The Siemens Solar modules are made from two basic sizes of cell. The
square cell used in the SM-series of modules is about 104 mm in across, and
produces a maximum current of about 3.4 amps. The larger cell used in the SP
series of modules is about 125 mm across and produces about 4.8 amps maximum.

8-4
To create a module that produces less current than the standard cell, cells are cut
into half or quarter, to give 1/2 or 1/4 the standard current. This is the case for the
Siemens Solar SM20, SM10, and SM5 (with 1/8 cell size). And the Siemens Solar
SP36 and SP18 are made of 1/2 and 1/4 cells cut from the larger SP75 cells.
SP Modules
1000 w/m 2 and 25 deg.C.
5.00
4.50
SP75
Current (amps)
4.00
3.50
3.00
SP36
2.50
2.00
1.50
1.00
SP18
0.50
0.00
0.00
5.00
10.00
15.00
20.00
25.00
Voltage (volts)


8-5
Exercises

Manufacturer
and model

Isc
Voc
Imp
Vmp
Pmax
# Series
Cells
Cell Size

!
"
#$

% %

!
"
#$

&% %

'
(
)

& *

+

&

&


8-6
Effect of Environment on
Module Output Potential
The IV curve is really just a "snapshot" view of the potential output of a photovoltaic
device under static environmental conditions. If the environmental conditions are
changed, the output potential of the device changes. The three main environmental
conditions that we will examine are:
Light Intensity or Irradiance

Cell Temperature
Light Spectral Content
Effect of Light Intensity (Irradiance)

When the intensity, or irradiance level, of light changes the number of photons
entering the PV device changes, and the number of electrons released changes. So
the direct result of a change in light intensity is a change in output current at all
voltages.
Voc drops
slowly with
lower
irradiance
Lower irradiance
reduces current


8-7
The Isc is DIRECTLY PROPORTIONAL to the light intensity, and the Voc varies
more slowly in a logarithmic relationship. In other words, the ratio of the Isc to the
light intensity or irradiance will be the same. If the light intensity is halved, the Isc
will drop to half.
Isc1
Light intensity1
Example:
Isc2
Light intensity2
A module has a rated Isc of 3.4 amps at 1000 w/m2. We can calculate
the Isc at another light level by taking the proportion of the new light
level to the standard level of 1000. If the actual light intensity on a
module were 850 w/m2, then the actual Isc would be given by the
proportion:
Isc1
Light intensity1
Isc2
Light intensity2
3.4 amps
1000 w/m2
Isc2
850 w/m2
Isc2
3.4 X ( 850 / 1000 )
2.9 amps
Standard Condition for Irradiance

A standard condition of light intensity has been established so that PV device output
can be compared. The internationally accepted standard is 1000 watts/m2 (or 100
mw/cm2). This is called "one sun" or "peak irradiance".

8-8
1000 watts/m2
Circumstances For Exceeding the

Standard
Most locations will measure less than this at noon, but many will measure more than
this. Sometimes white clouds will reflect more light onto the module surface
producing irradiance values of 1400 w/m2 or higher.
Measurements at high altitudes may be greater than the standard 1000 w/m2
because less air is between the sun and the surface, so more light gets through.
Sometimes ground reflectance from white sand or rocks or reflective building or
water surfaces will also raise the total intensity of light on a surface to greater than
the standard 1000 w/m2.
Cloudy Day
Clear Day
1000 w/m
1000 w/m


8-9
Effect of Temperature
As the cell temperature rises, the main effect is to reduce the voltage available at
most currents. There is a slight rise in current at very low voltages. The change in
voltage is DIRECTLY PROPORTIONAL to the rise in temperature. It is important to
note that these factors refer to cell temperature, not just ambient or air temperature.
The relationship between air temperature and cell temperature depends on light
intensity, and is discussed later in this chapter.
Isc rises slightly as temperature goes up
Higher temperature
reduces voltage
!
Manufacturers of modules anticipate the lost of voltage in real world hot conditions,
and compensate by building modules with enough cells in series so that even when
very hot, the module has enough voltage to charge batteries or operate the load
device.
The general formula for determining the change in voltage with temperature is given
on the next page. The effect of temperature on Voc for one cell is multiplied by the
amount of temperature change and the number of cells in series. This will give the
voltage change for the whole module. Subtracting this voltage change from the
voltage at one temperature gives the voltage at another temperature.
8-10
Example Cell Temperature Coefficients:

Siemens Solar
125 mm cell (140 cm2)
Siemens Solar
103 mm cell (104 cm2)
Parameter
Change
per oC
% Change
per oC
Change
per oC
% Change
per oC
Voc
-2.15 mV
-0.36 %
-2.15 mV
-0.36 %
Vmp
-2.18 mV
-0.44%
-2.19 mV
-0.45 %
Isc
2.06 mA
-0.04%
1.20 mA
-0.04 %
Imp
-4.37 mA
-0.10%
-3.23 mA
-0.10 %
Pmax
-9.53 mW
-0.45%
-7.08 mW
-0.47 %
"#
Example Module Temperature Coefficients:

Parameter
SM55
SP75
SM50-H30
(33 cells)
Voc
-0.077 volt / oC
-0.077 volt / oC
-0.071 volt / oC
Vmp
-0.079 volt / oC
-0.078 volt / oC
-0.072 volt / oC
Isc
1.20 mA / oC
2.06 mA / oC
1.20 mA / oC
Imp
-3.22 mA / oC
-4.42 mA / oC
-3.25 mA / oC
Pmax
-0.255 watt / oC
-0.345 watt / oC
-0.234 watt / oC
"#

8-11
Calculating Voltage Changes
Voltage Change
Voltage Loss Factor X Temperature Change

and
Voltage2 = Voltage1 - Voltage Change
Example:
A module has 36 cells in series for a Vmp of 17.3 volts at 25 oC. In real
world conditions, the cells will easily heat up to 50oC. Using the factors
presented, and multiplying by 36 cells and 50 - 25 = 25 oC.
temperature difference, we get
Vmp Change
= Factor X Temperature Change

= -0.079 v/oC X 25 oC
= -1.98 volts
Vmp2
= 17.3 volts Vmp (at 25 oC) - 1.98 volts

= 15.3 volts Vmp (at 50 oC)
which is still enough to fully charge a typical "12 volt" battery that
actually needs up to 15 volts to reach full charge.

8-12
Calculating Power Changes

The overall reduction in maximum power (Pmax) is a general factor to use to
estimate the effect of temperature on output power, not just current or voltage. The
power reduction factor for solar modules as indicated in the table presented
previously is generally about -0.45% to -0.5%, and this includes changes in current,
voltage, and the curve shape. In other words, a cell or module will loose a little less
than 1/2% of its peak power with every degree in temperature rise.
Typical Maximum Power Reduction Factor
- 0.45 % / oC
The formula for determining the effect of temperature on overall device power is
given below. Use the overall reduction in Pmax factor given above and multiply by
the amount of temperature change. This will give the percentage of power change.
% Power Change = Pmax Reduction Factor X Temperature Change
Example:
A module is rated at 55 watts Pmax at 25oC. If it is operating outdoors

and heats up to 50 oC., then the Pmax will be reduced.
% Power Change
= -0.45% / oC
= -.45% / oC
X
X
(50 oC - 25 oC)
25 oC
= -11.25% change
Pmax2
= Pmax1
(1 - % Power Change)
100
= 55 watts Pmax (at 25 oC)
= 48.8 watts Pmax (at 50 oC)

8-13
( 1 - 11.25 )
100
Standard Condition for Cell Temperature

Since PV device output is affected by cell temperature, a standard cell temperature
o
of 25 C has been accepted internationally. Stating the output of modules and cells
with reference to this common temperature allows for proper comparisons to be
made.
Standard Condition for Temperature =
25 oC
Temperature Effect Depends on the

Number of Series Cells
The factors given previously help to calculate the change in Voc (and Vmp), but they
dont really help to understand the effect temperature has on module output into a
battery. Battery charging is the most common use for photovoltaic modules, and it is
the battery voltage that determines where on the IV curve the module operates. An
average battery voltage of 13.5 volts can be used to illustrate the point.
The number of series cells determines the final module voltage and the voltage of
the knee. The difference between a 33-cell module and a 36-cell module becomes
critical when we look at the effect of temperature. The IV curve is reduced in voltage
at higher temperatures. This reduction causes the knee of the curve to move
inward, and since the battery is holding the module operating point to around 13.5
volts, the operating point actually moves downward along the knee. The amount of
decrease in current depends on how many cells in series.

8-14
SM55
36 cells
Current (amps)
~13.5 volts average battery voltage

4 .0 0
3 .0 0
2 5 d e g .C .
2 .0 0
47 deg. C.
1 .0 0
65 deg C.
0 .0 0
SM50-H
33 cells
Current (amps)
0 .0 0
5 .0 0
1 0.00
2 0.00
2 5.00
Volta ge
4.00
3.00
2.00
1.00
0.00
0.00
1 5.00
25 deg. C.
47 deg. C.
65 deg. C.
5.00
10.00 15.00
Voltage
20.00
25.00
$
We can see in the figures that the 36-cell module (Siemens SM55 in this example)
are less affected by high operating temperatures than the 33-cell module (Siemens
SM50-H). The 36-cell module drops from 3.3 amps at 25 oC to only 3.0 amps at 65
o
C, while the 33-cell module drops from 3.25 amps at 25 oC to less than 2.5 amps at
65 oC. However the 33-cell module drops to only 3.0 amps at a more moderate 47
o
C. Thus we see that a 36-cell module is designed to operate well in the hottest of
climates, while a 33-cell module performs well in more moderate temperatures.

8-15
Exercises
2

%
% ,$
#$$$- .
2
%!$- //////////

0

*1
#

, % *
2
2- 1
,
*
2
3$-
% ,#

%

2
#$$$-

#

,
2
*%#$$$-

% 445&6

#"" ,&.
&7 - &-
8
///////

0 445&6

////////
*#$&


8-16
Spectral Content of Light

The spectral distribution of light refers to how much of the light's energy is delivered
at each different wavelength or color. The spectral distribution of the sun's light
above the atmosphere is nearly that of a "black body" or perfect radiator at 6000 oC.
The greatest output from the sun is in the visible range of wavelengths (roughly from
violet at 380 nm to red at 750 nm), but there is substantial energy output in short
wavelength ultra-violet (UV) and in long wavelength infrared (IR) and radio waves.
As the light passes through our atmosphere certain wavelengths are scattered and
absorbed more than others by air, moisture and aerosols. For example, the large
dips in the spectral distribution graph shown are due primarily to absorption by water
vapor. The proportion of reds, greens, blues, etc., will change depending on the
thickness of atmosphere that sunlight has to penetrate. That thickness varies hourly
as the sun rises in the sky. The thickness of atmosphere light must penetrate is
called the Air Mass (AM). If the sun were directly overhead, light would pass through
1 thickness of atmosphere, or 1 AM to reach the earth's surface. In the afternoon or
morning, the sun is at a lower angle, and the distance light must pass through might
be 2 or more times the noon time thickness (i.e.. 2 AM or more).
Effect of Light Spectrum

Sun at noon
Sun at
mid-morning
One Air Mass .

(AM 1 )
Earth
Atmosphere
1.5 Air Masses

( AM 1.5 )

8-17
A standard set of atmospheric conditions has been established for the photovoltaics
industry by NREL, the US National Renewable Energy Laboratory (formerly SERI,
Solar Energy Research Institute) in Golden, Colorado, for the U.S. Department of
Energy. This standard is listed as ASTM #E892, and among other factors is a
spectral distribution that is equivalent to light passing through 1.5 thickness of
atmosphere, or Air Mass 1.5 (AM 1.5).
Solar Spectral Distribution
ASTM #E892 or Air Mass 1.5
By standardizing the amount of atmosphere under which measurements of cell and

module output are made, we also fix the proportion of red, green, blue, etc. light, or
the spectral distribution. This standard is necessary because photovoltaic cells
respond differently to different wavelengths of radiation. (Recall the discussion of
spectral response in the chapter on Photovoltaic Physics). If a CZ cell were
exposed to blue light and red light of the same total energy, the current produced
would be higher for the red light. The standard spectral distribution of AM 1.5 gives
a common reference condition so modules from different manufactures and of
different construction can be fairly compared.

8-18
Summary of Standard Conditions

Because the electrical output of a photovoltaic device is so affected by the
environment, we must standardize the conditions under which they are measured
and compared. Once those conditions are set, the output potential of the device is
absolutely determined.
The three standard conditions that are used to specify the environmental conditions
for photovoltaic devices are summarized below.
Light Intensity or Irradiance
1000 watts/m2
Cell Temperature
25 oC
Solar Spectral Distribution
ASTM #E892 or
Air Mass 1.5

8-19
Exercises

14*

.

*
* * (9

&
)

: ;
9 :;

//////
<
//////
&
//////
=

//////
>
//////
5

//////
?

//////
@

//////
,&
3&
#$$$-

5$28-&
#9 1
#9 1
#$$$
#$$$-
A(9

B
,#
B :
,4 ( ;
C
D:
,4)
;
):
$;


8-20
Other Rating Conditions

The relationship of module output as rated under Standard Operating Conditions
(SOC) and the actual output that a user would be able to measure under real
outdoor conditions is very poor. Typically the irradiance level will be lower and the
cell temperature will be higher than SOC. Other test conditions have been defined
over the years to try to account for this discrepancy. Some of these sets of
conditions are described next.

Nominal Operating Cell Temperature (NOCT)
Irradiance
Ambient Temperature
Wind Speed
Module at Open Circuit
800 w/m2
20 oC
1.0 m/s maximum
Standard Operating Conditions (SOC)

Irradiance
Cell Temperature
1000 w/m2
NOCT
Nominal Operating Conditions (NOC)

Irradiance
Cell Temperature
800 w/m2
NOCT
%
Nominal Operating Cell Temperature

The NOCT is not a condition but rather an empirically determined value for a specific
module design and construction. The NOCT is intended to represent the typical real
cell operating temperature for a module. The actual temperature will depend on the
thermal characteristics of the cell and the module packaging. For example, glasson-glass modules will typically have a higher NOCT than white background modules
due to the greater thermal mass of the two layers of glass. Blue background
modules may have a slightly higher NOCT than white background modules due to
absorption. Generally a module manufacturer will seek to have as low an NOCT as
possible.

8-21
Effect of Operating Conditions

on Output Curve
Current (amps)
4
3.5
STC
SOC
2.5
NOC
2
1.5
1
0.5
0
0
10
15
20
25
Voltage (volts)

Standard Operating Conditions (SOC)

Using the NOCT as the first order correction to STC conditions, the SOC conditions
have irradiance still at the peak 1000 w/m2, but have the cell operating at the NOCT.
This represents a clear day with the cell heated to some realistic temperature.
Nominal Operating Conditions (NOC)

Moving still further toward real conditions, the specifications of NOC have the cell
heated to NOCT, and also the irradiance reduced to 800 w/m2. This approximates a
non-optimal day in most respects, with the cell heated and the light diminished,
perhaps due to high clouds, pollution, or atmospheric haze. Module peak power is
typically reduced by about 25% under NOC conditions compared to STC conditions.

8-22
Defining Module Output Efficiency

One common measure of the quality of a solar cell or module is the efficiency. But
there is not just one way to define efficiency for solar devices. This should be
made clear, because statements are often made comparing manufacturers
efficiencies, and it can be deceiving if you are not clear on exactly what efficiency
is being quoted.
In general, efficiency is defined as the ratio of output from a device compared to the
input to the device.
Efficiency
Output
Input
Total Area Efficiency

This definition involves the ratio of maximum electrical power output compared to the
total light power incident on the ENTIRE device. The device area includes module
frame, interconnects and gridlines on the surface of the cells. This is the "real world"
efficiency of a module. This value indicates what you really have to work with.
Example:
Pmax
Total Device Area X Input Light Power
Calculate the total area efficiency of the following module:
Pmax =
55 watts

Length
Width
=
=
1.293 m
0.330 m
55 watts
1.293m X 0.33 m X 1000 w/m2
0.129 or 12.9%
8-23
From the example, we see that the overall efficiency of the example module is
12.9% under full sun conditions. In other words, of the total solar radiation that falls
on the total module are (including all the inactive areas like frames, inter-cell spaces,
and gridlines) 12.9% of that energy is output as electrical power. This may seem
like a low figure, with about 87% of the incident power not being converted, but
remember that the fuel for the electrical power is free sunlight. This efficiency ranks
amongst the highest module efficiencies in the world, and is the direct result of the
single crystal technology used in the Siemens Solar cells.
Aperture Area Efficiency

Another way to calculate efficiency is called aperture area efficiency. This refers to
the module efficiency calculated without any frame dimensions. This would include
all area inside of the frame, including any gridlines or cell interconnects. This
modified way to calculate efficiency removes the effect of a large frame, which really
doesnt reflect the quality of the cell efficiency anyway.
Active Area Efficiency

This definition involves calculating efficiency based only on the area of the device
that is exposed to light or active semiconductor. Light incident on shaded areas
like interconnect wires, cell gridlines, and frame area is not included. It is always
higher than the total area efficiency will be for a completed device with
interconnected cells. Active area efficiency is often announced by research groups
or by manufacturers and confused with total module area efficiency. It should be
understood to refer to fundamental physics of a single cell only, and not to final
manufactured modules where cells will be covered with gridlines and interconnect
wires.

Pmax
Active Area Only X Input Light Power
8-24
Exercises

&

*

:
;

:

;

:

;

7

.
*

E
Module Manufacturer and Model

Total Area Efficiency (%)
8-25
Combining Cells and Curves

Solar cells can be combined in series and parallel to increase voltage and current.
When cells are connected in series, the current flow is the same through each cell
and the resulting voltage is the sum of the voltages of each cell. When cells are
connected in parallel, the voltage across each cell is the same and the currents add
to produce a final current.
The exact output curve for a combination of cells can be created by adding the
curves for all the single cells. For series connections, the current flowing through
each of the cells is the same. So at any current level, the voltage of one cell is
added to the voltage of the next. The output curves are added horizontally or
voltage-wise, as shown. For parallel connections, the voltage across each cell is the
same and the currents flow together. So at every voltage value, the current of each
cell is added. The curves are added vertically or current-wise, as shown.
Combining Cells To Make A Module

Parallel connected cells
increase current potential
Series connected cells

increase voltage potential


8-26
A solar module is a collection of cells connected in series and sometimes in parallel

to produce a basic building block with enough voltage to do useful work. The most
common load is a "12 volt" battery, which I place in quotes because it actually needs
approximately 14-15 volts to be fully charged. So most modules are made of
enough cells in series to produce at least 14.5 volts at maximum power, to be able to
effectively charge the batteries. To achieve this voltage 30-36 single crystal silicon
are needed in series. A composite curve for a 36-cell CZ module is shown. The IV
curve for a single cell is added along the voltage axis 36 times. The final IV curve
for the entire module reaches out to a Voc of more than 21 volts.
Cells Combine To Make Output

Curve For Module
3.4 amps
...36 cells in series...
0.6 volts
each cell
21 volts

If the main application for the modules were not 12-volt battery charging, then some
other number of cells might be more appropriate. For example, if the ultimate
application was to be direct connection to utility power at voltages of 120 or 240
volts, then there would be no reason to keep the voltage of the modules around 12
volts. Modules with a Vmp of 40 or 60 volts could be designed. The ultimate
application voltage determines the number of cells needed in series. The most
common market for large power photovoltaic modules is still 12 volt battery charging,
so that is why most manufacturers produce modules with 30-36 CZ cells in series.

8-27
If more voltage or current than one module can produce is needed, modules can
also be connected in series and parallel to achieve practically any final voltage and
current. Systems producing 600 volts DC and hundreds of amps have been
installed, and successfully operated for years.
Combining Modules To Make

An Array
6.8 amps
84 volts

For example, in the drawing above there are 4 modules connected in series to make
a nominal 48-volt string, and two strings are connected in parallel to increase current
output. The composite curve for the entire 8-module array can be constructed by
either adding voltage-wise first and then adding current-wise, or visa versa.

8-28
If we add voltage first, the 4 modules in series are added voltage-wise to reach out
to beyond 80 volts at Voc, and then the 2 parallel strings are added current-wise to
show that the current adds.
If we use the values calculated previously for a 36-cell module, the resulting curve
for the entire array has values given below:
New parameters for array of four series X two parallel modules
21.7 volts Voc (module)
X 4 series modules
= 86.8 volts Voc (array)
17.3 volts Vmp (module)
X 4 series modules
= 69.2 volts Vmp (array)
3.4 amps Isc (module)
X 2 parallel modules = 6.8 amps Isc (array)
3.05 amps Imp (module)
X 2 parallel modules = 6.1 amps Imp (array)

8-29
Exercise

9?

* !

*2
9 .

*:)

;
1

1 E///////////////////////////////////////////
%//////
//////
% //////

&

* 4
*,
10
16
10
14
16
14
12
//////
12
10
10
2
1
0
0
20
40
60
0
0
80
a.
20
40
60
0
0
b.

20
40
60
80
c.
20
40
60
d.
0%

1

9 *

16
14
(1 )
12
10
(3 )
8
6
4
s in g le c e ll
(4 )
(2 )
0
0
10
20
30
40
50

/////

///// , /////

///// %

/////

8-30
/////
Determining Output Over Time

The IV curve, adjusted for environmental conditions, is only one half of the
information required to determine what a module will output. The other critical
information is the LOAD operating characteristics. It is the load that determines
where on the IV curve the module or array will operate, not the other way around.
Different types of loads will interact with the PV device in different ways, and the
output of a module under the same environmental conditions will vary depending on
the load.
Three factors determine output potential:
Cell Temperature
Irradiance
Spectral Distribution
Forth factor determines actual output:
Load Interaction
An IV curve gives only a snapshot of potential output of a PV cell or module or array

under static conditions of light, spectrum and temperature at a given instant. A
module in the real world will experience changing conditions throughout a day. A
more relevant measure of module output comes from measuring or predicting the
varying output over a typical day. And the output is determined by the load.
The three basic types of loads that photovoltaic devices interact with are:
Maximum Power Tracking Device

Battery
DC Motor
How these different types of loads interact with a solar module and determine the
actual output are discussed next.

8-31
Cell Temperature Is Affected By Irradiance

The cell temperature, and therefore the voltage available from the PV device, is
related to light level or irradiance. The greater the irradiance, the more energy is
brought to the module and the greater the internal heating. The exact relationship
between irradiance and cell temperature depends on the module construction. The
thickness of glass, the emissivity of the back layer, the darkness of the cells -- all this
has an impact on the final cell temperature. The relationship between light level and
cell temperature for Siemens Solar modules is shown as an example. The rise in
cell temperature above the ambient air temperature is linear with irradiance.
Cell Temperature Affected by

Irradiance
Temperature
(Deg C)
40
30
20
10
0
0
200
400
600
800
1000
1200
Irradiance (Watts/m2)


8-32
Knowing the air temperature and the irradiance level, the cell temperature can be
determined, as given below.
Cell Temperature = Ambient Temperature + Temperature Rise
Example:
A 36-cell module is rated at 25oC. (the standard temperature discussed

earlier). If in a real situation, the insolation level is 900 watts/m2, the
temperature rise is given by the graph as approximately 25oC. above
ambient. If the air is at 35oC., the cell temperature will be
Cell Temperature
= 35oC (ambient air) + 25oC (temperature rise)

= 60oC.
Once the actual cell temperature has been calculated, it can be used to predict how
the IV curve voltage potential is reduced. All computer programs used for array
sizing use some sort of adjustment for cell temperature to better predict actual
module performance in the real world.
It is important to see that the irradiance and ambient temperature work together to
predict the actual cell temperature in a module. For example, we just calculated that
if the ambient air temperature were 35 oC. and the irradiance was 900 w/m2, then
the cell temperature would be 60 oC. But what if the same module was located in a
cold alpine climate with an ambient temperature of only 5 oC. The same irradiance
level of 900 w/m2 would result in the same temperature rise value of 25 oC., but the
final cell temperature would be less.
Example:
The same solar module in a cold alpine climate with only 5oC ambient
in 900 w/m2
Cell temperature
= 5oC (ambient air) + 25oC (temperature rise)

= 30oC
The cell temperature is much lower, and therefore the module voltage would be
much higher in the alpine climate than in the hot climate. So just knowing the
irradiance on the module is not enough to predict module performance. The
ambient air temperature must be known as well.

8-33
Exercise

% :;
4, &
2

!$$- .

//////////&

9 *

9 .

3! &0

.

#
&' *
////////-

445

) &
E
F#
F
F
#"!
#"
9F2$ &
% F////////-
% F $$- 9F4 &

F////////

8-34
Maximum Output Potential During a Day

By predicting how the air temperature and irradiance vary during a day, the
instantaneous output potential of a cell or module, as represented by the IV curve
can be accurately predicted for any time of the day. By continuously varying the
environmental conditions on the module, a series of IV curves can be calculated.
Since power is a rate of delivery of energy, adding the maximum power possible
from a module (Pmax) over the day gives the maximum total energy that can be
drawn from a module during a day.
This might actually occur for modules connected to a utility interconnected inverter
for example. Usually utility interactive inverters have Maximum Power Tracking
(MPT) circuitry that continuously adjusts the operating voltage of an array so that the
maximum power is output at any instant. Such MPT devices are also popular for
direct coupled water pumping systems. The solar array is connected to the MPT
device, and its maximum power is extracted and converted into power at the
optimum voltage for the pump at any instant.
The variation of Pmax for a 36-cell 55-watt module is shown over a typical summer
day. The area under the curve represents the total maximum power that one
module could actually deliver to a load. In this case, that is 253 watt-hours (Wh), or
approximately 1/4 kilowatt-hour (1/4 kWh).
The previous discussion has talked about IV curves and how they represent the
potential output at any instant under fixed environmental conditions. But the output
added up over time is the energy that could go into a load during a day. The total
energy available to a load is the value we have been seeking. If we know the total
energy needed by a load in a typical day, we can now determine how many modules
would be needed to supply all that energy. For example, using the above figure of
1/4 kWh from one module, if the daily load was 5 kWh/day for a remote application,
then approximately 20 modules would be needed to produce all that energy. This is
the principle behind the array sizing that will be presented in the chapter on System
Sizing.

8-35
Output Into Maximum Power

Tracking Device
Current (amps)
Maximum Power (watts)
12:00 noon
(B)
2.8
3:00 p.m.
(C)
8:00 a.m.
0
10
(A)
12
14
16
18
40
35
30
25
20
15
10
5
20 0 4
Voltage (volts)
(B)
(C)
Total Energy
253 Wh
(A)
6
10
12
14
16
18
20
Time of Day


8-36
Exercise

0

:G; 5435

** :
*;

:B
;? *
:%

!5#9
5
*
:
*

; *
*
*:

*;
H

*//////////////
H

B
//////////////////
Hour
of Day
4
5
6
7
8
9
10
11
12
1
2
3
4
5
6
7
8

Watts
Feb
Watts
June
0.48
5.60
10.79
14.09
15.18
14.00
10.67
5.51
0.47
0.81
2.04
3.91
9.28
16.27
22.27
26.31
28.65
29.36
28.51
26.06
21.96
15.98
9.08
3.8
2.0
.8
8-37
Graph the watts against time to show the output over a typical day.
Watts
35
30
25
20
15
10
0
4
10
11
12
Time of Day

8-38
Interaction with a Battery Load

The output value given previously was from a module operating at its maximum
power point all day long. This can be achieved by using some sort of electronic
maximum power tracking circuitry, that will be described later in chapter on Energy
Enhancement Through Tracking. However, the most common load for a PV module
or array is a battery. A battery has a voltage, depending on its state of charge, age,
temperature, rate of charge or discharge and other factors. This voltage determines
the voltage of everything connected to the battery, including the module or array.
Even if the system has motors, inverters, and other devices connected to the
battery, or even other loads connected directly to the array, the battery will set the
operating voltage for every component in the system.
It is sometimes thought that a module with a maximum voltage potential of 20 volts
will push a 12-volt battery to 20 volts and harm the battery. But it is the battery that
fixes the operating voltage of the module, not the other way around.
A detailed discussion of battery characteristics and performance is presented later in
the chapter on Battery Technology. For now, we only need to realize that as a
battery is charged its voltage rises, and as it is discharged the voltage falls. The
voltage of a battery varies within a fairly narrow range, usually between 11-15 volts
for a "12 volt" battery, with average operation around 13-14 volts.
This is why modules are designed with up to 36 solar cells in series, giving a Vmp at
25oC of more than 17 volts. At first this may seem too high for a 12-volt battery. But
a module in a hot climate will loose approximately 2 volts due to heat, bringing the
Vmp down to approximately 15 volts. A "12 volt" battery actually needs about 14.5
volts to reach full charge.
So it is the battery voltage that determines the operating point on the IV curve, and
therefore the charging current into the battery. As shown in the figure, the module or
array is affected by temperature and light level throughout a typical day. Since the
battery holds the module operating point within a narrow range of voltage around 1314 volts, the resulting output will be seen as varying current into the battery. The
operating voltage will rise slowly over a day as the battery is charged. But compared
to the wide changes in the module curve, the voltage of the battery changes only
slightly.
As shown in the figure, the total energy into a battery is slightly less than the
maximum possible energy that could have been delivered. This is because the
battery held the operating point of the module at a lower voltage than the Vmp of the
module. Module manufacturers try to create modules that are good matches in
voltage to the ultimate load, in this case a battery. However, because of
environmental uncertainties and variations, the match of module Vmp and load
voltage cannot always be perfect. This is a case where the match is close, but not
perfect.

8-39
Output into Battery
Current (amps)
2.8
Power (watts)
40
Changes Over a Day (M55 in Phoenix)
Maximum power possible
35
12:00 noon
30
(B)
25
2
3:00 p.m.
20
(C)
Power into battery
15
8:00 a.m.
247 wh
10
1
(A)
5
0
0
0
10
12
14
16
18
20
Voltage (volts)
10
12
14
16
18
Time of Day

!

8-40
20
Exercise

0

445
2 5
*

:9 ; :B
*;
? *
:%

!5#39
5
*
:
*

; *
*
*:

*;
@ *

H

9 /////////////////
H

B
*//////////////////
Hour
of Day
6
7
8
9
10
11
12
1
2
3
4
5
6

Watts
April
Watts
July
7.4
20.3
32.2
39.0
41.7
41.6
40.1
37.2
31.0
19.9
7.1
1.3
8.1
18.9
28.0
32.1
33.1
32.7
31.4
29.3
25.7
17.7
7.6
1.0
8-41
Graph the watts against time to show the output over a typical day
Watts
45
40
35
30
25
20
15
10
0
4
10
11
12
Time of Day

8-42
Interaction with a DC Motor

A DC motor also has an "I-V curve", usually called a load curve. A typical curve
ramps up in current with increasing applied voltage. When overlaid on a module IV
curve, this defines a precise intersection point with the I-V curve. This intersection
point sets the operating current and voltage for the module or array. So the motor
determines the voltage and current output of the module or array.
A direct coupled motor system has the module connected directly to a DC motor and
pump without any controls or battery. It has the advantage of the fewest
components and thus high reliability. But the system suffers in overall efficiency due
to mismatch each morning and evening. A direct coupled motor/PV array system is
usually designed so that has the motor operates the modules near their point of
maximum power during the middle of the day (see point B in the figure). But when
the module current output is lower in the morning and evening, there will be a very
poor match, as shown in the figure. The motor operates the module at far lower
voltage than the module's Vmp (points A and C). So the motor may operate, but it is
getting less than the maximum possible power the module or array could deliver.
A maximum power tracking device can be placed between the array and the motor
to force the array to operate at its maximum power point all day long. This would be
one way to increase the efficiency of the overall system. Maximum power trackers
are discussed more in the chapter on Energy Enhancement Through Tracking.

8-43
Output into DC Motor

Changes Over a Day (M55 in Phoenix)
Current (amps)
12:00 noon
2.8
(B)
3:00 p.m.
(C)
8:00 a.m.
1
(A)
0
0
10
12
14
16
18
20
Voltage (volts)

$

8-44
Interaction with a Resistive Load

A pure resistive load is yet another type of load for a PV device, but it is not very
common. Connecting an array directly to a water heater coil is one possible real
application. This is sometimes done by a charge regulator with power diversion as
an option, whereby array power is diverted when the batteries are fully charged.
Current
Module
or
Resistance
(ohms)
array
Voltage
of system .

Using a resistor in the field to test the curve shape of a PV array is another useful
application of a pure resistive load. A resistor also has a load curve or an "I-V
curve". It is a straight line. This defines a single intersection point with the module
or array I-V Curve.

8-45
A resistive load obeys Ohm's law, that is V = I X R, or R = V/ I. For a fixed value of

resistance (R), as the voltage is increased, the current increases by the same
proportion. A high value of R operates a module or array nearer to the Voc point. A
low value of R operates it nearer to the Isc point.
Current (amps)
3
load line
for low R
load line for

high R
1
0
8
12
16
Voltage (volts)
20

%
A resistance can be calculated that will operate a module or array at or very near its
maximum power point. This can be a useful tool in the field. Using standard digital
volt meters in the field, it is easy to measure Isc and Voc, but you are not really
looking at the output of a module or array anywhere near its maximum output point.
By using a resistor to operate the module or array somewhere near its knee, you
can get a complete picture of the curve shape.

8-46
In real outdoor conditions, the light level would probably be lower than the standard
level of 1000 w/m2, and the cell temperature would probably be higher than the
standard temperature of 25 oC. A 36 cell 36-watt module at outdoor conditions of for
example 800 watts/m2 and 50 oC cell temperature has Vmp = 17 volts and Imp = 2.1
amps (approximate). Dividing these values will give a resistance that will operate
the module near its maximum power point in typical field conditions:
Resistance needed to operate at Pmax
(35 Pmax, typical outdoors)
Vmp
Imp
17 volts
2.1 amps
8.1ohms
So connecting an 8.1-ohm resistor to a 36-watt module in the field will operate the
module very near its maximum power point, to allow checking of output other than
just Voc and Isc.
Operating at Maximum Power

with Resistive Load
Current (amps)
(Vmp, Imp)
3
R = Vmp/Imp
2
1
0
12
16
20
Voltage (volts)


8-47
To calculate the size of resistor needed for an array of modules, multiply the above
resistance value by the number of modules in series, and divide by the number of
modules in parallel. The final resistance value will operate the array near its
maximum power point.
Resistance array =
Example:
Resistance module
# Modules in Series
# Modules in Parallel
You wish to do field diagnostics on a small array of four 36-watt

modules in series by two in parallel. What value of resistance would
you need to operate the array near its maximum power point in actual
field conditions?
Resistance array
10 ohms
20 ohms
4 in series
2 in parallel
The example above shows that to operate an array of four 36-watt modules in series
by two in parallel, you would need a resistor of approximately 20 ohms. This is
shown graphically below.
20-ohm resistor operates array near Pmax


8-48
Exercise:

9
* * 5

*2 0
* '

* *
D
+
*
*

* G7

%

/////////////

8-49
(End of Chapter)

8-50
CHAPTER EIGHT
OUTPUT CURVES
8-1
Basic Current-Voltage Curve (The I-V Curve)
8-2
Output Curve Terminology
8-3
Module Output Curves
8-4
Effect of Environment on Module Output Potential

Circumstances For Exceeding the Standard
Standard Condition for Cell Temperature
Temperature Effect Depends on the Number of Series Cells
Spectral Content of Light
8-7
8-7
8-8
8-9
8-10
8-14
8-14
8-17
Summary of Standard Conditions

8-19
8-21
Defining Module Output Efficiency

Aperture Area Efficiency
8-23
8-23
8-24
8-24
Combining Cells and Curves
8-26
Determining Output Over Time

Cell Temperature Is Affected By Irradiance
Maximum Output Potential During a Day
8-31
8-32
8-35
Interaction with a Battery Load
8-39
Interaction with a DC Motor
8-43
Interaction with a Resistive Load
8-45

8-51
Chapter 8 Answers
Output Curves

SM55
SM50-H
SM46
SM20
SM10
SM6
SP75 (12V)
SP36 (12V)
SP18 (12V)
Isc
Voc
Imp
Vmp
Pmax
(Amps)
3.45
3.35
3.35
1.6
0.71
0.42
4.8
2.4
1.2
(Volts)
21.7
19.8
18.0
18.0
19.9
19.5
21.7
21.7
21.7
(Amps)
3.15
3.15
3.15
1.38
0.61
0.39
4.4
2.1
1.1
(Volts)
17.4
15.9
14.6
14.5
16.3
15.0
17.0
17.0
17.0
(Watts)
55
50
46
20
10
6
75
36
18

c. 0.8

d. 0.9

b. No

b., c.
8-1
Output Curves

Isc1
Light intensity1
Isc2
Light intensity2
2.0 amps
1000 w/m2
Isc2
850 w/m2
Isc2
2.0 X ( 850 / 1000 )
1.7 amps

If the two days are the same temperature, then we can calculate the Isc at standard
irradiance for each module.
Module 1:
2.5 amps
745 w/m2
Isc2
1000 w/m2
Isc2
2.5 X ( 1000 / 745 )
3.36 amps for Module 1 at 1000 w/m2
2.1 amps
650 w/m2
Isc2
1000 w/m2
Isc2
2.1 X ( 1000 / 650 )
3.23 amps Module 2 at 1000 w/m2
Module 2:
Under standard irradiance, Module 1 has an Isc of 3.36 and Module 2 has an Isc of
3.23. Therefore the correct answer is:
a. Module 1
Note: If the temperatures were significantly different on the two days that the
measurements were taken, we could not answer the question without knowing the
actual temperatures.
8-2
Output Curves

Using the cell temperature coefficients for either the SP cell or the SM cell, we see that
the change in Voc per oC is
-2.15 mV/ oC
or
-0.36%/ oC
We calculate the voltage loss factors for 33 cells as follows:

Voltage loss factor = -2.15 mV/ oC X 33 cells
= -0.00215 V/ oC X 33 cells
= -0.071 V/ oC
Voc Change

= -0.071 V/ oC X [55 oC - 25 oC]
= -0.071 X 30
= -2.13 Volts
Voc2
= 19.9 Volts Voc (at 25 oC) - 2.13 Volts

= 17.77 Volts Voc (at 55 oC)
Using the percentage loss figures:

Voltage loss factor = -0.36 %/oC
Voc Change

= -0.36 %/oC X [55 oC - 25 oC]
= -0.36 X 30
= -10.8 %
Voc2
= 19.9 Voc (at 25 oC) X (1 - % Voc Change)

100
= 19.9 Volts X ( 1 -
10.8 )
100
= 19.9 X (1-0.108)
= 19.9 X (0.892)
= 17.75 Voc (at 55 oC)
8-3
Output Curves

The voltage loss factor for the 33-cell module was calculated in the previous problem to
be:
Voltage loss factor = -0.071 V/oC
So for every 10 oC, the voltage change is:
Voc Change

= -0.071 V/oC X 10 oC
= -0.71 Volts Voc
As the temperature increases by 10 C, the module voltage decreases by 0.71 Volts

Condition
A. Air (ambient temperature
B. Insolation
C. Irradiance
D. Cell temperature
E. Tilt angle
F. North-south orientation
G. Spectral distribution
H. Module active area
Standard Value
j. N/A
j. N/A
c.
a.
j. N/A
j. N/A
f.
j. N/A

a. Jawaloa (latitude 23.5 deg. North)

b. Linoleum footage in house (where the base area of walls is excluded)
8-4
Output Curves

The work below shows the calculations and results for the SM10, SM50, SM55, SP75,
and SR100 modules. The method for other modules will be the same.
The datasheet for the SM10 module gives the following information:
Pmax =
10 watts
Length
Width
=
=
0.360 m
0.330 m
10 watts
0.360 m X 0.330 m X 1000 w/m2
0.084 or 8.4%
For the SM50 module:

Pmax =
50 watts
Length
Width
=
=
1.293 m
0.329 m
50 watts
1.293 m X 0.329 m X 1000 w/m2
0.118 or 11.8%
For the SM55 module,

Pmax =
55 watts
Length
Width
=
=
1.293 m
0.329 m
55 watts
1.293 m X 0.329 m X 1000 w/m2
0.129 or 12.9%
For the SP75 module:

Pmax =
75 watts
Length
Width
=
=
1.200 m
0.527 m
75 watts
1.200 m X 0.527 m X 1000 w/m2
0.119 or 11.9%
8-5
Output Curves
For the SR100 module:

Pmax =
100 watts
Length
Width
=
=
1.498 m
0.594 m
100 watts
1.498 m X 0.594 m X 1000 w/m2
0.112 or 11.2%

The data sheet for the SM55 module gives the following information:
Voc: 21.7 Volts
Vmp: 17.4 Volts
Isc:
Imp:
3.45 Amps
3.15 Amps
With 8 modules in series, the voltages will be multiplied by a factor of 8:

Voc = 21.7 X 8
= 173.6 Volts
Vmp = 17.4 X 8 = 139.2 Volts
Four strings of modules in parallel will increase the currents by a factor of 4:
Isc = 3.45 X 4
= 13.8 Amps
Imp = 3.15 X 4
= 12.6 Amps

The correct representation is sketch c.

a. Most cells in series _(4)_
c. Next fewest cells in series _(2)_
e. Most cells in parallel _(1)_
b. Fewest cells in series

_(1)_
d. 2 strings of cells in parallel
_(3)_
f. Same Imp as another module (2) or (4)

Reading the graph of cell temperature rise vs. irradiance, we estimate that at 800 watts/
m2 the cell temperature rise will be approximately 22 C. Then the cell temperature will
be:
Cell Temperature = 32 C (ambient air) + 22 C (rise)
= 52 C
8-6
Output Curves

The thermocouple reports the cell temperature to be 68 C. The rise in cell temperature
can then be calculated as:
68 C (Cell )
= 40 C (ambient air) + Rise in Temperature
Rise in Temperature = 68 C - 40 C = 28 C
Looking at the graph of cell temperature rise vs. irradiance, we estimate that a rise in
temperature of 28 C results when the irradiance is equal to 950 watts/m2.

a. Voc = 17.5 volts
Ambient temp. = 40 oC
From the module literature, we know the following data for the SM-50H module.
Voc: 19.8 Volts
The Voc change is therefore:
Voc Change
= 19.8 - 17.5 = 2.3 Volts.
The temperature coefficient for an SM50-H module is given in the chapter as -0.071
Volts / C. We can calculate the temperature change of the cell as:
2.3 Volts

= -0.071 V/oC X Temperature change
Change in T
= 2.3 Volts / (-0.071 V/ oC)

= 32 oC above the standard temperature
So the cell temperature must be 32 oC + 25 oC (standard temperature) = 57 C. This is

17 oC above the ambient temperature of 40 oC. So the cell temperature rises above
ambient is 17 oC. We refer to the graph of Temperature Rise vs. Irradiance and
determine that the irradiance would be about 650 watts/m2.
b. Irradiance = 700 watts/m
Ambient temp. = 35 oC
From the module literature, we know the following data for the SM-50H module.
Vmp: 15.9 Volts
8-7
Output Curves
Looking at the graph of cell Temperature Rise vs. Irradiance, we determine that an
irradiance of 700 watts/m2 would result in a temperature rise of about 20 oC. If the
ambient temperature is 35 oC, then the cell temperature is 35 oC + 20 oC = 55 oC. The
temperature coefficient stated in the chapter for Vmp is -0.072 V / oC. We calculate the
change in Vmp as:
Vmp change

= -0.072 V/oC X [55 oC - 25 oC]
= - 2.16 Volts
Vmp (at 55 oC)
= 15.9 - 2.16 Volts = 13.74 Volts

By summing the each hour's energy, we get the output for each month:
Output in February =
Output in June
=
76.8 Wh
247.1 Wh
Watts
35
30
25
Feb
Jun
20
15
10
5
0
4
10
12
Time of Day
8-8
Output Curves

By summing the each hour's energy, we get the output for each month:
Output in April
Output in July
=
=
317.5 Wh
266.9 Wh
Watts
45
40
35
30
25
Apr
Jul
20
15
10
5
0
6
9 10 11 12 1
Time of Day
8-9
Output Curves

The module literature gives the following values for the SP75 module:
Vmp = 17.0 volts
Imp = 4.4 amps
The resistance that will operate a single SP75 module at the maximum power point
under standard conditions is then:
Resistance
Vmp =
Imp
17.0
4.4
3.86 Ohms
Therefore, connecting a 3.86-ohm resistor to the 75 watt module in the field will operate
the module very near its maximum power point. To operate the whole array at the
maximum power point, we calculate the resistance as:
Resistance (array) =
3.86 ohms
5.79 ohms
6 in series
4 in parallel
Note that under real field conditions, the values of Vmp and Imp may be slightly
different. This is because the irradiance may be less than 1000 watts/m2 and the cell
temperature will likely be hotter than 25 C.
8-10
Output Curves
Chapter Nine
Load Estimation
The loads cannot be taken as given while all other calculations are carefully
scrutinized. The loads influence every aspect of system design, and must be as
efficient and reliable as possible. As you read, think about what influences load
efficiency, and how you might have to argue with a client to replace their existing or
proposed load with a more efficient one.
The entire system design is based on the size of the load. If the information is
inaccurate the initial costs will be too high or the array and battery could be too small
and the system will eventually fail. It is, therefore, essential that time be taken to
look carefully at the load requirements and the expected usage pattern.
Using literature values for load consumption is common, but it is more accurate to
have the load demand measured to be sure. Often nominal numbers are presented
in literature, and a particular piece of equipment may require more or less power
than stated.
If an existing application is being retrofitted with a PV power system it is very
important to not just look at the old generator capacity and try to recreate it with PV.
Very often an oversized diesel generator was installed, perhaps because that size
was used elsewhere or to allow for future growth. A PV power system can be
designed to accurately match the current load requirement without limiting the ability
to expand in the future to meet greater demand.
The load profile throughout the year must be accurately determined. Any seasonal
variation might influence the choice of tilt angle or battery size for autonomy. The
"duty cycle" or hours of operation for intermittent loads must be estimated carefully.
In the case of telecommunications equipment, not only the hours of transmitting but
also the hours of standby or quiescent operation need to be included in the load
calculations.
Improving load efficiency is the quickest way to reduce PV power system cost. A
more efficient load device may even be slightly more expensive than an existing or
conventional load device. But if the power consumption of the efficient load is
significantly lower than the other, the cost savings in the array and battery may offset
the higher load cost in many cases, and result in a total system cost (modules +
battery + loads) that is lower than if inefficient loads are used.

9-1
Components Load Estimation
Using oversized diesel generators may have allowed load device efficiency to be
neglected in the past. But when a small load is operated by a large capacity
generator, fuel efficiency is low so fuel is wasted. Simply retrofitting a new PV power
system to an existing inefficient load is a mistake if a more efficient load device can
be installed.
Load Sizing Forms

We have developed a form to use for calculating the average daily load. Both AC
and DC loads can be added. The Daily Load Sizing Form has you add up all the
loads to estimate the largest daily load demand on your batteries. The WeekAveraged Load Sizing Form has you average the load demands over a 7-day period.
This allows for slightly smaller array sizing if your loads vary during the week. The
battery should be sized to meet the largest daily load demand, while the array can
be sized to meet a more averaged load value. The sizing forms are presented next.

9-2
Daily Load Sizing Form

System Description :
DC Loads
Qty.
Amps
Hours/day
Daily Demand (Ah)
DC Loads (Ah) =
AC Loads
Qty.
Watts
Hours/day
AC Sub-Total (Wh)
Continuous Watts
Surge Est.
Daily Demand (Wh)
Inverter Choice:
[
][
][
]
+
AC Sub-Total Efficiency Input Voltage

9-3
[
] =
DC Loads
____________
Daily Load (Ah)
Week-Averaged Load Sizing Form

System Description:
DC Loads
Qty.
:
Amps
Hours
X
X
X
X
X
X
______
______
______
______
______
______
Days/wk
Weekly Demand (Ah)
:
:
:
:
:
X ______
X ______
X ______
X ______
X ______
X ______
X
X
X
X
X
X
=
=
=
=
=
=
X ______ X ______ X
X ______ X ______ X
Weekly DC Loads (Ah)
AC Loads
Qty.
:
:
:
:
:
:
:
:
Watts
X ______
X ______
X ______
X ______
X ______
X ______
X ______
X ______
Hours
X
X
X
X
X
X
X
X
______
______
______
______
______
______
______
______
Days/wk
= ______
Surge Est.
= ______
=
=
=
=
=
=
=
=
=
Inverter Choice:
[
][
][
]
+
AC Sub-Total Efficiency Input Voltage
[
] =
DC Loads
[
] 7 days
Total Weekly
Demand (Ah)

Weekly Demand (Wh)
X
X
X
X
X
X
X
X
Weekly AC Sub-Total (Wh)

Continuous Watts
9-4
Total Weekly
Demand (Ah)
=
Week-Averaged
Daily Load (Ah)
Daily Load Demand

You calculate the total daily load demand by multiplying each load demand times the
time that the load operates in a typical 24-hour period. DC loads are estimated by
using amp-hours, while AC loads are estimated by using watt-hours. This is
because the DC loads can draw their current directly from the battery, while the AC
loads must draw their power from a DC-to-AC inverter that will change the voltage
from the (usually 12, 24 or 48 volts) DC level to the (usually 110 or 220 volts) AC
level. The input and output voltage of the inverter must be included in the
calculations separately.
DC Load Demand
DC Load Current (amps) X Hours of Operation
AC Load Demand
AC Load Power (watts)
X Hours of Operation
If DC loads are given in watts instead of amps, then you can easily convert by
dividing by the nominal operating voltage. This is usually 12 volts, and sometimes
24 volts.
DC Load Current (amps) =
Example:
A typical 12 volt 40-watt fluorescent light has a current given by

DC Load Current
Example:
DC Load Power (watts)

Nominal DC Voltage
40 watts
12 volts
3.3 amps
It is very common for microwave repeater equipment to operate at 24

volts nominal. If the operating power of a repeater is 200 watts (at 24
volts) the load current is given by
DC Load Current

200 watts
24 volts
8.3 amps
9-5
Loads May Have Different Levels

Not all loads have just one value. Some loads may have a full-power value and a
standby value. For example, telecommunications systems often have a transmit
power and a standby or quiescent value. The time that the load operates at its
different levels must be determined, and the energy demand for each mode of
operation added up.
Variable Load Profile

Load may have a stand-by or quiescent
current
Must include this component in total dem and
operating
current
quiescent
current

Another example of a load with different consumption is some instant on

televisions. These will continue to draw a small phantom load even when they are
turned off. But the primary candidate for load variable demand is
telecommunications systems. It is very important to include the quiescent or
standby load demands in your total load demand calculations.
Example:
A telecom repeater operates at 24 volts and draws 10 amps when it is

transmitting. It has a standby mode that consumes .5 amps. The transmit
time is estimated to be 5 hours a day.
Load Demand (transmit)
=
=
10 amps
50 Ah
5 hours
Load Demand (standby)
=
=
.5 amps
9.5 Ah
(24 - 5 hours)
The standby daily load is almost 1/5 of the transmit daily load, even though
the instantaneous current draw is only 1/20th.

9-6
Loads May Vary Seasonally

Most loads will not be operated the same amount of time every day and every
month. There will be a seasonal variation to the load demand, and this will influence
the choice of tilt angle for the array as well as the array and battery size. An
example of two seasonal profiles is shown below, one peaking in the summer
months, and the other peaking in the winter.
Seasonal Load Profiles

70
Summer peak profile
60
50
40
30
Winter peak profile
20
10
0
Jan Feb Mar Apr May Jun
Jul Aug Sep Oct Nov Dec

An example of the variation in insolation throughout a year is shown on the next

page. The insolation on a flat surface varies greatly with the least during the winter
and the greatest during the summer. The best system design has the array tilted to
an angle so that the profile of insolation throughout the year matches the load
profile. In this way, the fewest modules are needed to meet the load demand.
If the load demand is small in winter and large in summer, as with air conditioning
loads or water pumping for irrigation, then tilting the array near a flat angle will give
an insolation profile that best matches the load requirements.

9-7
New Delhi Insolation Profiles

Latitude 29 deg. N
Insolation (Langleys)
600
500
400
300
200
100
0 deg.
15 deg.
30 deg.
45 deg.
Dec
Nov
Oct
Sep
Aug
Jul
Jun
May
Apr
Mar
Feb
Jan
60 deg.

If the load is relatively constant every month of the year, as might be the case for a
navigational aid or a constantly transmitting repeater, then a different angle is better.
Tilting the array up increases the insolation intercepted during the winter months and
sacrifices some during the summer months, with a resulting profile that is more
constant throughout the year to better match the requirements of a constant load.
If a tilt is chosen that does not match the insolation profile to the load profile, then
the array will have to be quite large to produce enough output when insolation is low
and may produce excess power that is just wasted when insolation is high. By
choosing a tilt angle for the array that gives the best match, array size is minimized,
and a reliable fixed angle mounting structure can be designed for the system.
Computer models or repetitive hand calculations can predict the best angle to give a
match between load demand and insolation.

9-8
Adding Up The Load Demand

Use the Daily Load Sizing Form to add up all the loads in a system to give the total
daily demand. For an example of DC load sizing, lets select some loads for a small
remote home or cabin, including lights and small appliances.
Example:
A small remote cabin owner wants to install some lights and

appliances. He will have two 40-watt fluorescent lights for bright
kitchen lighting, three compact fluorescent (PL lights) at 11 watts each
for other rooms, a small 40-watt TV and a 24-watt ceiling fan. All these
loads are to be DC so no inverter is required. All loads will operate at
12 volts nominal.
DC Loads
Lights
PL lights
TV
Fan
Qty.
:
:
:
:
2
3
1
1
Amps
X
X
X
X
3.3
0.92
3.3
2
Hours
X
X
X
X
4
5
4
8
DC Loads (Ah)

9-9
Weekly
Demand (Wh)
=
26.4
13.8
13.2
16
=
=
=
=
69.4
For practice at adding AC loads, lets estimate the demand for a remote school with
a variety of AC appliances, lights and other loads.
Example:
A small school in a remote area wants to remove their generator and

use only photovoltaics to power all their loads. The school has eight
40-watt fluorescent lights, small lights in the bathrooms, four
computers, an overhead projector, and a small microwave oven for
heating lunches, and a refrigerator.
AC Loads
Qty.
Watts
Lights
PL lights
Computer
Projector
:
:
:
:
8
2
2
1
Microwave
Refrigerator
:
:
1
1
X
X
X
Hours
40
11
200
300
800
200
X
X
X
8
2
4
3
2
12
AC Sub-Total (Wh)

9-10
Weekly
Demand (Wh)
2,560
44
1,600
900
=
=
=
1,600
2,400
=
=
9,104
Exercise

!"#

$ %
& "#

%
' $"#
(
" %

)*#

+ ,

,

&

- % * ./ %*0
,
! % * . %*0 &
( , " %
.1 %*0

&
, +
(

2 3
& 4
- (
, 4
"/ (
& &
4
%
2 3

& 4
, 4
& &
4
/ (
"5 (
- %
( ( ( ,

(

+

,

& 3
& &
4
- %
& 3
& &
4
/ %
6
( (

+ ,
6
+

6

,

+
,


9-11
Choosing An Inverter
Before you can complete the sizing forms for AC loads, you must choose an
inverter. This choice will determine the average inverter efficiency and the nominal
DC voltage of the array and battery, both of which are needed to finish the
calculations indicated at the bottom of the sizing forms.
Meeting Continuous Power Demand

There are two parameters that will help choose an adequate inverter to handle the
AC load demand. First the inverter must be able to operate all the AC loads that
might be on at one time continuously. This value is calculated by adding up the
watts of the AC loads and filling in the Continuous Watts line in the sizing form. Be
sure to multiply by the quantity to get the correct total load requirement.
Estimate of Continuos Watts: Add all AC loads that might be on at same time
Meeting Surge Power Demand

The second parameter that the inverter must meet is the surge requirement.
Inductive loads, such as motors (example: washing machines, water pumps),
compressors (example: refrigerators), and even certain types of fluorescent light
ballasts. In an inductive load, there are coils of wire that must be loaded with
energy. During this short period of loading, the current that can flow can be 4-6
times the continuous running current! So any inverter chosen to operate such loads
must be able to supply this surge of current, for fractions of a second or sometimes
for seconds.
To estimate the total surge power that the inverter might have to deliver, you might
add up all the surge powers of all the inductive loads and add this to the continuous
power demand of the non-inductive loads. But this approach is much too severe.
The probability of two or more inductive loads turning on at exactly the same time is
quite low. And the result of such an approach would perhaps be the choice of an
unnecessarily large and expensive inverter.

9-12
A more practical approach to estimating total surge power is to identify the load with
the largest surge power and add this value to the continuos power demand of all the
other AC loads that might be on at the moment that the large load is turned on.
Thus inverter would have to operate all the loads (except the one with the largest
surge) and be able to turn on the load with the largest surge as well.
Estimate of Surge Watts:
Add the surge of the largest load

to the power of all the other AC loads
that might be on at the same time
If you are given no information on the surge demand of a particular inductive load, it
is safe to assume that six times the continuous current will be drawn as a surge to
start the load.
An example of a residential refrigerator load profile is shown. As time passes
(moving from right to left) the compressor turns on and then off, trying to keep the
refrigerator box cool. A refrigerators efficiency is measured by not just the
continuous power level, but also by how many hours in a typical day the compressor
actually operates. Each time it turns on, there is a spike or surge in the current that
is drawn. The continuous running current of this particular refrigerator is about 2
amps, but the surge current needed is about 12 amps, a factor of 6 times greater!
Refrigerator Load Profile -- Continuous and Surge Levels


9-13
Estimating Continuous and Surge Power

for Complex Systems
Sometimes it is not practical to add up the continuous and surge requirements of
each load. This is particularly true when dealing with residential systems, where
there might be a large number of AC appliances and loads, but their time usage is
unknown. It would be improbable that you or the user could accurately predict
exactly how much time each and every load would be used, and exactly what loads
would be operating at the same time. For large AC systems it is better to estimate
the continuous and surge power demands by using statistical averages.
To estimate the continuous demand of a complex system, an Energy Consumption
Chart such as the one shown below is quite useful. These are often prepared by
local utilities or energy departments. The annual energy consumption of various
common loads is indicated. This number is based on a wide variety of users, and
represents only a statistical average. You would add up the annual energy
consumption of the loads, and then divide by 365 days/year to get the estimated
daily load.
Example:
A home is being built in a remote area, and is to be solar powered.

The owner has made a list of the appliances that will be used in the house.
Annual energy values have been taken from the chart.
Refrigerator/Freezer
Microwave oven
Toaster
Dishwasher
Coffee Maker
Washing machine
Room air conditioner
Bathroom lights
Bedroom lights
Dining room lights
Hall lights
Kitchen light
Stereo
Television
Various personal care items
1135 kWh/year
190
39
363
106
103
1350
60
200
144
108
100
108
320
35
Total Annual Energy Requirement
4361 kWh/year
Estimated Daily Load
4361 kWh/year
365 days/year
12 kWh/day
The system designer and the client could look at this list and see what items contribute most
to the energy burden, and perhaps discuss alternative or more efficient loads to reduce the
overall energy and therefore the solar array and battery size as well.
9-14
Source: Photovoltaic User Guide, California Energy Commission, July 1984


9-15
Using another statistical tool presented next, you can estimate the continuous and
surge power that might be required from an inverter. A chart has been prepared
showing how much continuous power and surge power might be needed to serve a
range of daily loads.
Estimating Surge
Estimated Power
Requirement (kW)
24
22
20
18
16
14
12
10
8
6
4
2
0
surge
range
continuous
range
12
16
20
24
Average Daily Load (kWh)

Read up from the estimated daily load requirement to find the approximate
continuous and surge demand. For example, if the total daily load was about 20
kWh/day, you would read up from 20 to find that the continuous power output of the
inverter may be between 4-7 kW, and that the surge demand may be between 10-20
kW. There is quite a range allowed, and this method is not precise. But it is a quick
way to estimate values that otherwise could be tedious to calculate and which might
be ultimately inaccurate anyway.
Example:
Using the daily load demand of 12 kWh/day already estimated for the
remote home, the inverter chosen should be able to supply
Continuous Watts: 2-4 kW
Surge Watts:

6-12 kW
9-16
Reading Inverter Literature

With continuous and surge load power values in hand, you can now select a
particular inverter for your system and determine the average efficiency and the
input DC voltage that it requires. These pieces of information will then be used to
complete the load sizing calculations.
Use actual manufacturers literature to determine the pertinent factors and values.
First locate the continuous power output capability of the various models, and then
try to determine the maximum surge capability as well. If these meet or exceed your
calculated values, then this particular model is suitable for your load demands.
(Other technical features such as output waveform, or voltage and frequency
regulation, will also guide you to choose one inverter over another. These aspects of
inverter technology are discussed in the chapter on Inverter Technology).
Once you have chosen an inverter, you can read from the literature the average
efficiency and the input DC voltage. Sometimes an inverter may be available in
more than one input DC voltage. The higher the DC voltage, the smaller the DC
current that is needed for the same amount of power, and therefore the smaller the
wires, fuses, circuit breakers and the lower the resistive losses in these components
as well.
If you are only operating AC loads in the entire system, then you are free to choose
any input DC voltage that you wish. To minimize costs and resistive losses, you
may want to choose the higher voltage, such as 24 or 48 volts. (Some large inverters
even come with 110 volts input).
However if you are planning on having both DC and AC loads operating from the
same photovoltaic power system, then you are more constrained in your inverter DC
input voltage. Most DC loads operate at 12 volts. Some fluorescent light ballasts,
refrigerators, and pumps operate at 24 volts, as do many telecommunications
devices. If you want to operate 12 or 24 volt loads you will have to choose an
inverter model with that same input DC voltage. (Sometimes a DC-DC converter
can be used to allow multiple DC voltages in the same system. For example, a 48volt inverter and battery may be installed to operate AC loads, and a DC-DC
converter can be installed as well to convert from 48 to 12 volts for some small DC
appliances).
Exercise


9-17
Calculate Ampere-Hours For AC Loads

With average efficiency and voltage values from manufacturers literature, you can
complete the sizing forms for AC loads. The sizing forms show that to calculate the
Daily Load (Ah), you must take the AC Subtotal (Wh) and divide by the inverter
efficiency and input DC voltage.
AC Load (Ah)
AC Subtotal (Wh)
Efficiency X Voltage
Why do we divide by the efficiency? We want to determine the amount of DC amphours that the solar array must produce. What we have been calculating so far in
the Load Sizing Forms is the amount of AC watt-hours that the loads need. Since
the inverter is not 100% efficient, more energy must be put into the inverter than is
output to the AC loads. How much more is determined by the efficiency.
Inverter Efficiency
DC power in
AC power out
Inverter
power out = Efficiency

power in
power in = power out

Efficiency


9-18
Efficiency is defined as the ratio of output power or energy divided by the input
power or energy. Therefore the energy or power in is given by the energy or power
out divided by the efficiency.
So dividing the AC Subtotal (Wh) by the efficiency CONVERTS the watt-hours out to
the AC loads into the watt-hours needed to be put into the inverter on the DC side.
DC Watt-Hours Into Inverter
AC Subtotal (Wh)
Efficiency
Then to complete the calculation, divide the DC watt-hours into the inverter by the
DC voltage to get the DC amp-hours into the inverter. This is the Ah load demand
that we need to deliver from the batteries and array.
DC Amp-Hours Into Inverter
DC Watt-Hours Into Inverter

DC Input Voltage of Inverter
If the DC voltage of the inverter and any DC loads are the same, then both can be
operated from the same battery bank and the same array. The Ah values for the DC
loads and AC loads can be added to give the grand total for the whole system. If the
voltages are different, then either a DC-DC converter must be used, or separate
systems can be designed for each.

9-19
Exercise:

!
"
! # $

%&
'

(

"

)

! *
+)&
,
-

#


9-20
Daily Load vs. Occasional Load

If a load is the same every day, then the solar array should be sized to meet that
load demand. But if a load is used heavily some days of a week and less on other
days, then the solar array can be sized smaller than the heaviest demand and larger
than the smaller demand, and on the average during the week meet the load
demand. The Week-Averaged Load Sizing Form is intended to more economically
size a solar array to operate a load that is not the same every day.
A good example of how this approach is applied is a weekend cabin. The loads are
used heavily on the Saturday and Sunday (2-day/week) and are not even operated
at all Monday-Friday (five days/week). The battery should be sized to meet the
heavy weekend demands, but the array has all five weekdays to catch up on what
it could not produce during the weekend.
If the load varies during the week, then the array can be sized based on the average
demand spread through the seven days of the week, using the Week-Averaged
Load Sizing Form. At the end of the week, the small array has still fully recharged
the battery bank. The battery is always sized using the Daily Load Sizing Form
If Load Varies During Week

Array Output Spread Through Week
100
90
80
70
60
50 Battery Net
40
30
20
10
0
1
Array
Load
Battery Net
At End Of
Day


9-21
If the load is the same every day, then the array and battery are sized to meet that
constant daily demand, using the Daily Load Sizing Form.
If Load Is Same Every Day

Array Matches Load Each Day
100
90
80
70
60
50 Battery Net
40
30
20
10
0
1
Array
Load
Battery
Net At End
Of Day

To see the impact of approaching load estimation from an averaging method, on the
next pages we redo the examples for the small remote cabin and school, this time
estimating what the averaged load would be. This might lead to a substantially
smaller load value and therefore a smaller array.

9-22
Example:
The small remote cabin that was previously used (page 11) will only be
occupied on the weekends. Recalculate the Ah demand using the WeekAveraged Load Sizing Form. Compare the result to the daily usage value
originally calculated.
DC Loads
Lights
PL lights
TV
Fan
Qty.
:
:
:
:
2
3
1
1
Amps
X
X
X
X
Hours
X
X
X
X
3.3
0.92
3.3
2
4
5
4
8
Days/wk
X
X
X
X
=
=
=
=
2
2
2
2

[ 138.8 ] 7 days
Total Weekly
Demand (Ah)
Weekly
Demand
(Wh)
52.8
27.6
26.4
32
138.8
19.8
Week-Averaged
Daily Load (Ah)
On the next page an example should be worked out for the remote school discussed
earlier. Recalculating the loads using the Week-Average Load Sizing method, as if
they were spread out over all seven days of the week, can be done to see if the load
is substantially smaller taking this approach.
Array and battery sizing calculations will be discussed in the chapter System Sizing.
But you can see that by averaging the load demand over the entire week that the
energy demand from the array is reduced. The difference between the values for
daily loading and week averaging is greater for the remote cabin than for the school.
This is because the cabin changes from a regular daily loading to only a two day per
week loading. The change for the school design from daily loading to having the
weekends off is not so great.
If the load distribution during a typical week is fairly even, then use the Daily Load
Sizing Form value for both array and battery sizing. Use the Week-Averaged Load
Sizing Form value for array sizing only if the distribution during a 7-day week is really
skewed to a few days. Then you might see some economics in sizing the array for
the average. In all cases, used the Daily Load Sizing Form value for battery sizing.
The battery has to be sized to handle the actual heavy daily loads.

9-23
Example:
The school also previously used (page 12) is to be occupied only during
the week (Monday-Friday). Recalculate the Ah load demand if it is
spread out over the full seven days of the week using the WeekAverage Load Sizing Form. Note that the refrigerator is not turned off
and must be kept operating during all seven days. How does this value
compare to the previously calculated load demand?
AC Loads
Qty.
Watts
Hours
Days/wk
Weekly
Demand
(Wh)
8
2
2
1
X
X
X
X
40
11
200
300
X
X
X
X
8
2
4
3
X
X
X
X
5
5
5
5
=
=
=
=
12,800
222
8,000
4,500
Microwave : 1
Refrigera : 1
tor
X
X
800
200
X
X
2
12
X
X
5
7
=
=
8,000
16,800
Lights
PL lights
Computer
Projector
:
:
:
:
Weekly AC Sub-Total (Wh)

[ 50,320 ] 7 days
Total Weekly
Demand (Wh)

9-24
50,320
7,188
Week-Averaged
Daily Load (Wh)
Exercise

.%/ 0*'
*
. -.)
*
# 12

12)
.

3 %.

'456#78#
"
.
4
8
9:
:;
91
:=
;>

8
#
?;;
@=
*
2;

9-25
12<

><
%/ )*'
2<
%< '

><.
%/ )*'

><.
%/ )* '
Efficient Lighting Loads

As was mentioned at the beginning of this chapter, it is very important that system
designers seek the most efficient loads for photovoltaic systems. Investing in
efficient loads will most often pay for itself in cost savings through reduced array and
battery size. In residential systems, two types of loads in particular are strong
candidates for efficiency improvements--lighting and refrigeration. Customers are
often unaware of the efficient counterparts to conventional equipment for lighting
and refrigeration, and a bit of education can go a long way toward reducing total
system costs.
Energy for lighting can be the largest component of the total load of a remote site
using PV. There are many considerations in choosing lights including color,
efficiency, availability, and power requirements.
Types of Lamps
Incandescent
This is the common "light bulb". The incandescent and quartz-halogen lights
develop their light by passing current through a thin high resistance tungsten wire
which gets hot and radiates light. Along with the visible light comes a great deal of
heat, so the lights are the lowest in efficiency. Only 5% of the input energy is
converted to visible light. The tungsten filament gradually evaporates and weakens,
and the metal that evaporates gradually coats the inside of the bulb, reducing
brightness. Incandescents are available in a wide range of output powers and are
available for 12-volt DC operation.
Quartz-Halogen
The quartz-halogen light differs from the standard incandescent by using a tungsten
filament surrounded by a halogen gas (bromine). The wire glows hotter than the
standard incandescent, so high temperature resistant quartz is used for the bulb.
The halogen gas aids in re-depositing evaporated tungsten back onto the filament,
so the bulb life is longer, and little coating occurs on the inside of the bulb,
increasing end-of-life brightness. The lamps last almost twice as long as
conventional incandescents, and the overall efficiency is almost double.
Quartz-halogen lamps are available in a wide range of power, and often have an
integral reflector attached to the bulb to enhance output, and can operate from 12
volt DC power. Many automobile headlights are quartz halogen, and can be used
for home lighting applications. Bulbs used for slide projectors can easily be used for
household lighting, for example in track lighting.

9-26
Fluorescent
In fluorescent lamps mercury vapor is excited inside a tube by an alternating current,
and light is emitted by the gas. The light strikes the inner coating of the tube, which
is some type of fluorescing material, producing a soft glow. The light produced falls
well within the visible range, so the efficiency is good.
Traditionally fluorescent bulbs have been long cylinders ranging from a few watts to
40 watts. New "parallel length" or PL lamps are now available that are much more
compact and can often times replace incandescent bulbs in standard light fixtures.
The PL lamps fold back the long cylinder to make a compact "H" shape. Double "H"
or "quad" PL lamps are also available. PL type lamps are available in 5, 7, 9, 13, 18,
24 and 36-watt models. To operate from DC power a ballast is needed to produce
high frequency AC current. These are discussed further in this section.
New types of traditionally shaped fluorescent lamps are available that produce more
lumens/watt, better color, and longer life (up to 24,000 hours). These lamps are a
smaller tube diameter (1 5/16") and are called "T-10" tubes.
High Intensity Discharge (HID) Lamps:

Mercury Vapor
Metal Halide
Low Pressure Sodium
High Pressure Sodium
These lamps are ideal for outdoor applications like security lighting where lamps will
be operated for extended periods of time.
The mercury vapor lamp produces a harsh bluish-white light and is the least efficient
of the HID lamps. They have a low initial cost, but higher operating costs than other
HID choices.
The low-pressure sodium lamps produce a harsh yellow-gold color and are suitable
only for general outdoor lighting. They are the most efficient of the light sources, but
the harshness of the light makes their applications limited.
High-pressure sodium lamps are the most efficient of the white light sources. They
produce a soft amber color and are used widely for street lighting and other exterior
lighting applications.
Metal halide lamps have good color rendering capabilities and are widely used for
sporting event lighting.

9-27
Electrical Efficiency and Lamp Efficacy

The efficiency of a lamp is the ratio of the total luminous power radiated from the
lamp, AT ALL WAVELENGTHS, compared to the electrical power input to the lamp.
Lamp Efficiency
Total Power Radiated (watts)

Electrical Power (watts)
However, if a lamp produces light beyond the visible range, that radiation does not
contribute to apparent brightness of the lamp. A better measure of the ability of a
lamp to convert electrical power into visible light is called efficacy, which compares
the APPARENT BRIGHTNESS OF A LAMP, measured in lumens, against the
electrical power input, measured in watts.
Lamp Efficacy
Apparent Brightness (lumens)

Electrical Power (watts)
A lamp with twice the efficacy of another will appear twice as bright with the same
input power, or can use half as much power to produce the same apparent
brightness.
Fluorescents are an excellent choice compared to conventional incandescents, and
typically give 4-5 times the brightness compared to incandescents for the same input
power. Bulbs that use less power can be used to achieve the same lighting, so a PV
power system can be smaller and less costly.
Low-pressure sodium lamps are the winners regarding efficacy, but the amber color
is too harsh for indoor lighting. They are chosen for general outdoor area lighting,
where color may not be as important as area coverage. High-pressure sodium lamps
would still have very high efficacy and would deliver a much more tolerable light.

9-28
Light Efficacy
Efficacy (lumens/watt)
100
50
0
Incandescent
High Pressre
Low Pressure
Sodium
Sodium
Fluorescent
Mercury Vapor
Tungsten Halogen


9-29
The life of the lamp must also be a consideration in system design. It is encouraging
that the more efficient or efficace lamps are also the more long life products as well.
Fluorescents can last 8,000-12,000 hours (some even up to 24,000 hours)
compared to typical incandescents that last only about 500-1500 hours. The entire
family of high intensity discharge lamps (HID) typically lasts about 15,000-25,000
hours.
Lamp Life
25000
Life (hours)
20000
Hours of Life
15000
10000
5000
0
Incandescent
Tungsten
Halogen
Mercury Vapor
Fluorescent
High Pressure
Sodium
Low Pressure
Sodium

It is critical that customers and users of photovoltaic generated power understand

the value of more efficient lights. The least efficient incandescent light is perceived
as being cheaper because the lamp bulb cost is so low, whereas a compact
fluorescent may cost 10 times as much. But with the fluorescent lasting 6-8 times as
long and consuming 4-6 times less, the cost over time of the incandescent can be
shown to be much higher than the cost of the more expensive fluorescent choice.

9-30
Fluorescent Ballasts
Incandescents need only to have adequate voltage and current to operate. They will
glow immediately. All the other forms of lamps need some form of ballast to give an
initial pulse to start the lamp, or to convert the power to an alternating current to
keep the gas glowing.
Traditional ballasts have been made using a "core and coil" technology. These
operate using the available 60 Hz utility power, and are not particularly efficient.
There is a slightly noticeable flicker to the light, and sometimes a delay or pulsing
when starting lamps. Recently available improved high frequency electronic ballasts
(20,000 Hz output) help to reduce these problems. They are almost twice as
efficient, turn on the lamps instantly, operate the lamps with no flicker or electronic
interference, and some even offer a degree of dimming to fluorescent lamps.
The overall efficiency of the lamp must have the efficiency of the ballast included to
determine final power requirements. The small electronic ballasts for PL lamps
consume approximately 2 watts, while the traditional ballasts use almost 4 watts.
The cost of the ballast must also enter into the cost effectiveness equation. A small
PL lamp may cost $5 and the ballast may add $15-25 to that. When compared to a
$1-3 incandescent lamp, this looks bad. But the PL is expected to last 10 times as
long as the incandescent and consumes 4-5 times less power to give the same
brightness. Over the life of the PL lamp and ballast, the total cost of the
incandescent choice becomes much greater.
The ballasts typically do not operate more than one PL lamp however. Each PL
lamp for example must have its own ballast, so usage in track lighting is difficult. An
option is to use small 12-volt Quartz-Halogen bulbs, although they are not as
efficient.
Ballast and fluorescent lamp design may have to be modified to allow starting in cold
climates. The ballast may not be able to pulse strongly or long enough to start a
cold lamp. Special pre-heaters can be installed in the bulb to allow for starting cold
lamps.

9-31
Color Quality
One measurement that relates to color quality is the color temperature of the light.
At high temperatures, matter emits a spectrum of light, and the spectrum changes
with higher temperature. The surface of the sun is approximately 5900 deg. C. or
6200 deg.Kelvin, another temperature scale similar to the Celsius (C) scale but
beginning at "absolute zero" degrees. Our sunlight spectrum is very close to the
spectrum emitted by a body at that temperature. The hotter the temperature of a
radiating source of light, the higher the percentage of high-energy blue light is
emitted. The colder the temperature, the higher the percentage of low energy red
light is present in the spectrum. So the spectrum emitted from lamps can be
compared to the spectrum that would be emitted by a perfect radiator. A color
temperature is given to a lamp to give a measure of how "red" or "blue" the lamp
spectrum appears. The small energy efficient PL lamps are usually available with
color temperature of 2700 deg.K, but are also available at 3000, 3500, and 4100
deg.K.
Another measure of the quality of light is the color rendition index (CRI). This is a
figure of merit that compares how well lamp light matches the balance of colors in
daylight. Two lamps may be compared only if they share similar color temperatures.
Incandescents can give fairly good color rendition. Fluorescents have traditionally
been more blue than incandescents, but now "warm" and "full spectrum"
fluorescents are available. The mercury and sodium lamps do not produce a broad
gradually changing spectrum of wavelengths and so their color is quite harsh.
Using Outdoor Light

One consideration to make when determining lighting needs is how much natural
sunlight can be brought into the space. The cost of the light is free, and the color
rendition is excellent. Glass is not as good an insulator as wall materials, and
adjustments may have to be made to the heating plans for a living space. But
passive solar heating and cooling can be anticipated and often controlled. There are
rooftop mechanisms that track the sun and reflect light down into a living or
workspace efficiently throughout a day. These could be operated using PV
modules, adding no burden to the power requirements of the space.

9-32
Exercise

!

8

A
@=
(PL)
:=

B;=;
B:=;;

::;;
+@=

:;;;
>;;;
7
B;:;<.-
B;:;<.-

>;;; "


9-33
Efficient Refrigeration Loads

A second major residential load that warrants scrutiny is refrigeration. People are
often unaware of how much the compressor is running in their refrigerator or of how
much power it is consuming. Efficiency improvements here can result in
tremendous savings in array and battery costs.
Types of Refrigerators
Refrigeration units operate from two basic sources of power, electricity or liquid fuel.
Electric units can be DC or AC or switchable between the two. Fuel types can use
kerosene or propane.
There are units that automatically switch from DC to AC to gas power, most
commonly used on recreational vehicles (RV's). These units are typically very
inefficient and use a great deal of power (up to 400 watts) on DC power. They are
therefore not a good choice for a remote PV powered site.
It is not uncommon for remote sites using PV for lighting; communications and tools
to use a gas powered refrigeration unit. Gas may already be on site for cooking, and
avoiding module and battery costs for refrigeration keeps the overall system more
cost effective.
However there are many manufacturers of DC powered refrigeration units. These
vary from small portable units of only 0.4 cubic feet (10 liters) capacity, to full sized
17 cubic feet (480 liters) capacity units for a home or village.
Designs for Efficiency

It is especially important for PV powered refrigeration units to be designed for
efficient operation. The walls should be thick and well insulated. And the storage
chamber should be no larger than required for the expected usage. For example, a
small unit for vaccine storage does not need to be as large as a unit intended for
storing food or milk.
The seals around the doors should seal tightly. A double seal helps to insure no
leaking.
Top loading designs are best at keeping the cold in whenever the lid is opened. If a
front door style is used, efforts should be made to keep the cold from escaping. This
can be done by hanging plastic baffles with slits, to allow access to foods but
prevent gross movement of air. Or sections can be enclosed with their own doors,
so that access to one area does not expose another area to loss of cold.
9-34
The condenser coils should be located to allow maximum escape of heat without
heating the walls of the unit. Many common AC units are inefficient because they
locate the condenser coils underneath the refrigerator, forcing all the escaping heat
to pass up and around the refrigerator walls! Coils can be located on the top of the
unit, or ideally they can be located away from the unit, perhaps outside where the
ambient temperature may be lower than inside the home.
There are available on the market super efficient refrigerator designs. A particular
model of PV powered refrigerator, manufactured by SunFrost, needs only five 50watt solar modules to operate a 17 cubic foot capacity unit, even at 90o F ambient air
temperature. The unit alone costs more than double what a conventional AC
powered design costs, but the total cost of the modules, batteries, and inverter
required for the inefficient conventional AC unit far surpass the total cost of the
efficient DC unit plus its 5 modules and small battery bank. The SunFrost product
also comes ready for AC power and is only slightly less efficient. It can be
incorporated into a conventional all-AC residential system design if so desired.

9-35
Exercise

C

" $

% '
*
"
%
'
12) "

* ! 8

D% '
B1=;;
A
11;
&;
6 #
:1<
:?<
7 6 /
:&;-<
:&;-<
/
B?1;
B?1;

9-36
Question Your Load Information

All of your accurate and precise calculations for load sizing will not overcome
incorrect initial data! You should be vigilant and always on the lookout for incorrect
load data. Incorrect assumptions, inaccurate estimations, unrealistic forecasts,
unexpected increases--all of these and more can be lurking behind your nice looking
calculations. You should at the very least ask yourself some of the following
questions when dealing with load estimation.
Is Future Growth Anticipated?

Telecom channels can be added, and hours of operation can increase. Populations
can increase, bringing more demand. Enjoyment of the quiet and reliable nature of
photovoltaic power can lead to expansion.
Think ahead when doing your system design. This also applies to component
selection as well. Choose charge regulators, wires, inverters and other components
that are adequately oversized to accommodate some future growth without major
changing out of hardware.
Is The Load Profile Well Known or

Is It A Guess?
Some clients may not know an existing sites real energy or power consumption, and
may guess for purposes of your calculations. Remember that most users of
electricity are aware of their POWER requirements but not of their ENERGY
requirements. They may confuse the two, and assume that if an existing generator
or source of power is present, that that must be the actual power used by the load.
For example, if a 10 kW diesel generator is on site for whatever historical reason,
this does not necessarily mean that the load demand is 10kW.
Has seasonal variation been taken into account? Lighting demand, for example,
may increase in the winter months, while water demand may increase in the
summer.

9-37
Have All Loads Been Included?

Watch out for quiescent or phantom loads! They may seem insignificant when
compared to the continuous power demand of the load, but if the off time of the
load is great, the small quiescent load demand can add up to be significant.
This is especially true for telecommunications systems, where there may be a small
but significant standby load that is often neglected by specifying engineers that are
used to lots of utility or diesel power being available.
Is This A New User of Electricity?

Initially, you may expect that a new user of electricity will be cautious and sparing
with their usage. But relatively quickly over time, they become accustomed to the
convenience and demand more. Lights may be left on longer. Radio/telephone
conversations may increase in length or in frequency. Television viewing may
expand.
It is difficult to estimate or anticipate exactly how much demand will increase, but
you should be aware that it probably will, and should try to account for this in your
system design.

9-38
(End of Chapter)

9-39
CHAPTER NINE
LOAD ESTIMATION
9-1
Load Sizing Forms
9-2
Daily Load Demand

Loads May Have Different Levels
Loads May Vary Seasonally
Adding Up The Load Demand
9-5
9-6
9-7
9-9
Choosing An Inverter
Meeting Continuous Power Demand
Meeting Surge Power Demand
Estimating Continuous and Surge Power for Complex Systems
Reading Inverter Literature
Calculate Ampere-Hours For AC Loads
9-12
9-12
9-12
9-14
9-17
9-18
Daily Load vs. Occasional Load
9-21
Efficient Lighting Loads

Types of Lamps
Electrical Efficiency and Lamp Efficacy
Fluorescent Ballasts
Color Quality
Using Outdoor Light
9-26
9-26
9-28
9-31
9-32
9-32
Efficient Refrigeration Loads

Types of Refrigerators
Designs for Efficiency
9-34
9-34
9-34
Question Your Load Information

Is Future Growth Anticipated?
Is The Load Profile Well Known or Is It A Guess?
Have All Loads Been Included?
Is This A New User of Electricity?
9-37
9-37
9-37
9-38
9-38

9-40
Chapter 9 Answers
Load Estimation

We first calculate the current for each load type:
Fluorescent lights
40 Watts
12 Volts
3.3 amps
Ceiling Fans
20 Watts
12 Volts
1.7 amps
Vaccine refrigerator
60 Watts
12 Volts
5.0 amps
This information is used to complete the Daily Load Sizing Form:

System Description: Remote Clinic - Daily Loads
DC Loads
Fluorescent Lights
Ceiling Fans
Vaccine Refrigerator
:
:
:
Qty.
6
3
1
X
X
X
Amps
3.3
1.7
5
X
X
X
Hours
/day
6
12
10
DC Loads (Ah)
9-1
=
=
=
Daily
Demand (Ah)
118.8
61.2
50.0
230.0
Load Estimation

System Description:
DC Loads
Remote Clinic - Weekly Loads
Qty.
Amps
Hours
Weekly
Demand (Ah)
Days/W
k
Weekday LoadsFluorescent Lights :

Ceiling Fans
:
6
3
X
X
3.3
1.7
X
X
8
8
X
X
5
5
=
=
792.0
204.0
Weekend LoadsFluorescent Lights :

Ceiling Fans
:
6
3
X
X
3.3
1.7
X
X
4
4
X
X
2
2
=
=
158.4
40.8
Constant LoadsVaccine Refrig
10
350.0
1545.2

0
AC Subtotal
Effic.
9-2
+ 1545.2
Input
DC
Voltage
Loads
1545.2
Total Weekly
Demand (Ah)
1545.2
Weekly
Ah/day
220.7
Avg Load
(Ah/day)
7
Days
Load Estimation

To clarify the power requirements, look at the amount of time that the loads are in
transmit and standby. The combined time on and off must equal 24 hrs (1 day).
Radio transmitter #1
Transmit
Standby
12 hrs / day
12 hrs / day
Transmit
Standby
8 hrs / day
16 hrs /day
System Description: Telecom site with two transmitters
DC Loads
Trans. #1 Transmit
Trans. #1 Standby
Trans. #2 Transmit
Trans. #2 Standby
Qty.
1
1
1
1
:
:
:
:
X
X
X
X
Amps
8
0.5
5
0.3
X
X
X
X
Hours
/day
12
12
8
16
=
=
=
=
Daily
Demand (Ah)
96.0
6.0
40.0
4.8
146.8
DC Loads (Ah)

During the winter months, the stated transmit times are reduced:
Transmit
Standby
8 hrs / day
16 hrs / day
Transmit
Standby
5 hrs / day
19 hrs /day
System Description: Telecom site with two transmitters
DC Loads
Transm. #1 Transmit
Transm. #1 Standby
Transm. #2 Transmit
Transm. #2 Standby
:
:
:
:
Qty.
1
1
1
1
X
X
X
X
Amps
8
0.5
5
0.3
=
=
=
=
Daily
Demand (Ah)
64.0
8.0
25.0
5.7
102.7
DC Loads (Ah)
X
X
X
X
Hours
/day
8
16
5
19
9-3
Load Estimation

Refer to the manufacturer's literature.

The refrigerator will be the largest surge load. We will first take all the other AC load
watts, then add six times the refrigerator load. As a worst case, we assume that all of
the loads can be on at the same time.
Continuous AC Loads
Load
Lights
PL lights
Computer
Projector
Microwave
Qty
8
2
2
1
1
Watts
40
11
200
300
800
Total continuous load
Total
320
22
400
300
800
1842
The refrigerator is 200 Watts continuous. At 6 times the rating, this gives a surge
power of 1200 Watts for the refrigerator.
The total surge requirement is 1842 + 1200 = 3042 Watts.
Select an inverter with a surge capacity of at least 3042 Watts.

Results will vary depending on student input.
9-4
Load Estimation

We first calculate the load size during the week:
System Description: Telecom site - load during the week
DC Loads
Trans #1 - Transmit
Trans #2 - Transmit
Trans #2 - Standby
:
:
:
Qty.
1
1
1
X
X
X
Amps
10
15
0.8
X
X
X
Hours
/day
24
8
16
=
=
=
372.8
DC Loads (Ah)
AC Loads
Computers - On
Computers - Off
Lights
:
:
:
Qty.
2
2
4
X
X
X
Watts
300
75
40
X
X
X
Hours
/day
8
16
8
=
=
=
8,480
AC Sub-total
90%
Effic
24
Input
Daily
Demand (Wh)
4,800
2,400
1,280
8,480
AC Loads (Wh)
Continuous Watts =
Surge Est. =
Daily
Demand (Ah)
240.0
120.0
12.8
760
1560
372.8
DC
Loads
765.4
Daily
(Ah/Day)
Note: We calculated the Continuous Watt load to be two computers (on) plus all four
lights. So, 2 X 300 + 4 X 40 = 760. The Surge Estimate was calculated by taking
6 times the power for the lights and adding it to the computer loads: 2 X 300 + 6 X
4 40 = 1560.
9-5
Load Estimation
We next calculate the average weekly load:

System Description:
DC Loads
Repeater #1
Repeater #2:
Transmit (M-F)
Standby (M-F)
Transmit
(Weekend)
Standby
(Weekend)
Telecomm Site - Weekly Average Loads
Qty.
Amps
Hours
10
24
Days/W
k
7
X
:
:
:
1
1
1
X
X
X
15
0.8
15
X
X
X
8
16
4
X
X
X
0.8
20
Weekly
Demand (Ah)
=
1680.0
5
5
2
=
=
=
600.0
64.0
120.0
32.0
2496.0
AC Loads
Computer Use:
On - (M-F)
Off - (M-F)
Off - Weekend
Fluorescent Lights
Qty.
:
:
:
:
2
2
2
4
Watts
X
X
X
X
300
75
75
40
Hours
X
X
X
X
8
16
24
8
Weekly
Demand (Wh)
Days/W
k
X
X
X
X
5
5
2
5
24,000
12,000
7,200
6,400
=
=
=
=
49,600
Weekly AC Loads (Wh)

Continuous Watts =
Surge Est. =
49,600
AC
Load
0.85
Effic
24
Input
760
1560
+ 2496.0
DC
Load
4927.4
Weekly
Ah
7
Days
4927.4
Weekly Ah
703.9
Avg Load
(Ah/day)
The Continous Watt load and Surge Estimates are the same as above. Note that the
load averaged out over the week is slightly less than the load from Monday to Friday,
704 Ah/day vs. 765 Ah/day. This is because the weekend loads are lower, which
reduces the average.
9-6
Load Estimation
When sizing a battery for this system, we should use the higher load (765 Ah/day) for
Monday through Friday. This is because the higher load during the week represents a
significant portion of the amount of energy storage. In other words, a 5-day battery
needs to be sized for the worst 5 days of load. We might size the array based on the
weekly load, however, since over a 7-day period, we only need to replace the average
load each day (704 Ah/day).

We first consider the costs of the incandescent bulbs. Over the total time period of
8000 hours, the total amount of electricity used by the bulbs is:
8000 hours X 75 Watts = 600,000 Watt-hrs = 600 kilowatt-hours
The cost of this energy is:
600 kilowatt-hours X $0.10 / kWh = $60.
The lifetime of a single incandescent bulb is 1000 hours, so we will need to purchase 8
total bulbs for the 8000 hour period. The cost of the bulbs and the energy is:
8 X $0.50 + $60 = $64.
Now we calculate the same costs for the compact fluorescent. The electricity used is:
8000 hours X 15 Watts = 120,000 Watt-hrs = 120 kilowatt-hours
120 kilowatt-hours X $0.10 / kWh = $12.
One compact fluorescent should last for 8000 hours, so we only need a single bulb.
The cost of the bulbs and energy is then:
$15 + $12 = $27.
We see that over the lifetime of the compact fluorescent, the total cost is less than half
of the using incandescent bulbs.
9-7
Load Estimation

We first calculate the costs of using the conventional refrigerator. The energy used in
one day by the conventional refrigerator is:
220 Watts X 12 hrs / day = 2640 Watt-hrs
The number of modules required to produce this energy is:
2640 Watt-hrs / 160 Wh per day = 16.5, rounded up to 17 modules.
17 modules X $320 = $5440
The energy required for the super-efficient refrigerator is found the same way:
60 Watts X 13 hrs / day = 780 Watt-hrs
And we need to buy a number of modules equal to:
780 Watt-hrs / 160 Wh per day = 4.9, rounded up to 5 modules.
5 modules X $320 = $1600
We can now compare the total costs:
Conventional
Refrigerator
Refrigerator $0 (already purchased)
Modules
$5440
Total $5440
Super-efficient
Refrigerator
$2500
$1600
$4100
So even with the purchase of a new refrigerator, the cost savings in the energy result in
a lower overall cost.
9-8
Load Estimation
Chapter Ten
Battery Technology
In stand-alone photovoltaic power systems, the most important and most poorly
understood component next to the PV modules is the battery. There is no one
perfect type of battery for all remote photovoltaic power systems. There are many
factors that influence the choice and performance of a battery in a photovoltaic
system. And battery technology is changing, with new constructions and
performance levels available to choose from. The following discussion is intended to
give an overview of battery technology and to equip a photovoltaic system designer
with the questions needed to determine which battery is best suited for a particular
application.

10-1
Components Battery Technology
Purpose of Batteries in Photovoltaic

Systems
A storage battery is an electrochemical cell that stores energy in chemical bonds.
When a battery is connected in a circuit to dc electrical load, such as an
incandescent light or resistor, the chemical energy within a battery is converted to
electrical energy, and there is a flow of current through the circuit. An understanding
of storage battery design, terminology and performance characteristics is essential
in the design of stand-alone PV systems.
In stand-alone photovoltaic systems the electrical energy produced by the PV array
cannot always be used when it is produced. Because the demand for energy does
not always coincide with its production electrical storage batteries are commonly
used in PV systems. The three primary functions of a storage battery in a PV
system are:
Energy Storage Capacity and Autonomy to store electrical energy when it is

produced by the PV array and to supply energy to electrical loads as needed or
on demand.
Voltage and Current Stabilization to supply power to electrical loads at stable

voltages and currents, by suppressing or 'smoothing out' transients that may
occur in PV systems.
Supply Surge Currents to supply surge or high peak operating currents to

electrical loads or appliances.

10-2
Energy Storage Capacity and Autonomy

Because the production of energy from a photovoltaic array may not always coincide
with the energy demand for electrical loads, batteries are required in most standalone PV systems. This is perhaps the most important function of batteries in standalone PV power systems to allow the loads to operate when the PV array by itself
cannot supply enough power. Energy storage is required if electrical loads are
required to operate at nighttime or during extended periods of cloudy or overcast
weather.
In the system sizing process the PV array is generally sized to satisfy the average
daily load demand during the period with the lowest insolation to electrical load ratio
(usually during winter months), to ensure that sufficient energy is available at all
times of the year. The battery storage capacity is generally sized to meet the
average daily electrical load for a specified number of days without input from the PV
array, or for a specified autonomy period.
Autonomy or the days of storage are often referred to when speaking about the
battery storage capacity of a stand-alone PV system. A stand-alone PV system is
described as having "autonomy" if sufficient battery storage capacity is available to
operate the electrical loads directly from the battery, without any energy input from
the PV array. The greater the design autonomy period the larger the battery
capacity required for a given load demand. For common, less critical PV
applications, autonomy periods are typically designed for between two and six days.
For critical applications involving essential loads or public safety or where weather
patterns dictate, autonomy periods may be greater than ten days. Longer autonomy
periods result in a lower average daily depth of discharge, and lower the probability
that the allowable depth of discharge, or minimum load voltage is reached.

10-3
Voltage and Current Stabilization

Another purpose for batteries in stand-alone PV power systems is to stabilize or
level out the potential wide variations in voltage and current that may occur in a PV
electrical system. As discussed previously in this manual a PV array can
theoretically operate at an infinite number of operating points between the shortcircuit current and open-circuit voltage. When electrical loads are directly connected
to a PV array, the load impedance dictates the operational voltage of the PV array,
which may not be optimal to operate the load at its prescribed conditions, or to
permit full utilization of the maximum power available from the PV array. Batteries
are used to allow the loads to operate within a prescribed voltage and current range,
as well as to ensure that the PV array is operated near its maximum power voltage.
By acting as a buffer between the PV array and loads, a battery can also stabilize
the voltage and current supply to electrical loads in which the load power
requirement oscillates or varies with respect to time.

10-4
Supply Surge Currents

Another important function of batteries in stand-alone PV power systems is to supply
surge or peak currents required by electrical loads that are higher than normal
steady-state operating currents. Since PV devices are inherently current-limited in
their output by short-circuit current and irradiance, a PV array by itself may not be
able to supply enough current to meet the surge requirements of some electrical
loads. While the PV array may be large enough to supply the total energy needed
by a load over a day, it may not be large enough to meet a momentary power
demand by a load at any particular time. A battery is capable of delivering high
currents, and can supply large currents to the loads for short periods, while being
charged by the array at lower currents over the course of a day.
Electric motors are common loads that often require large surge currents for starting.
For example in refrigerators, compressors, power tools and other motor loads, the
surge current may be 5 to 10 times the normal running or operating current level. A
battery can supply hundreds of amperes for short periods, thereby meeting these
momentary surge requirements.

10-5
Battery Design and Construction

Battery manufacturing is an intensive, heavy industrial process involving the use of
hazardous and toxic materials. Batteries are generally mass produced, combining
several sequential and parallel processes to construct a complete battery unit. After
production, initial charge and discharge cycles are conducted on batteries before
they are shipped to distributors and consumers.
Manufacturers have variations in the details of their battery construction, but some
common construction features can be described for most all batteries. Some
important components of battery construction are described below.
Cell
The cell is the basic electrochemical unit in a battery consisting of a set of positive
and negative plates divided by separators, immersed in an electrolyte solution and
enclosed in a case. In a typical lead-acid battery, each cell has a nominal voltage of
about 2.1 volts, so there are 6 series cells in a nominal 12-volt battery. Figure 10-1
shows a diagram of a basic battery cell.
Active Material
The active materials in a battery are the raw composition materials that form the
positive and negative plates, and are reactants in the electrochemical cell. The
amount of active material in a battery is proportional to the capacity a battery can
deliver. In lead-acid batteries, the active materials are lead dioxide (PbO2) in the
positive plates and metallic sponge lead (Pb) in the negative plates, which react with
a sulfuric acid (H2SO4) solution during battery operation.

10-6
Battery Cell Composition

Electrical load
Negative plate
Positive plate
Grid
Grid
Separator
Active material
Active material
Case
Electrolyte

Electrolyte
The electrolyte is a conducting medium that allows the flow of current through ionic
transfer or the transfer of electrons between the plates in a battery. In a lead-acid
battery the electrolyte is a diluted sulfuric acid solution, either in liquid (flooded) form,
gelled or absorbed in glass mats. In flooded nickel-cadmium cells, the electrolyte is
an alkaline solution of potassium hydroxide and water. In most flooded battery
types, periodic water additions are required to replenish the electrolyte lost through
gassing. When adding water to batteries it is very important to use distilled or demineralized water, as even the impurities in normal tap water can poison the battery
and result in premature failure.

10-7
Grid
In a lead-acid battery the grid is typically a lead alloy framework that supports the
active material on a battery plate, and which also conducts current. Alloying
elements such as antimony and calcium are often used to strengthen the lead grids,
and have characteristic effects on battery performance such as cycle performance
and gassing. Some grids are made by expanding a thin lead alloy sheet into a flat
plate web. Others are made of long spines of lead with the active material plated
around them forming tubes, or what are referred to as tubular plates.
Plate
A plate is a basic battery component, consisting of a grid and active material,
sometimes called an electrode. There are generally a number of positive and
negative plates in each battery cell, typically connected in parallel at a bus bar or
inter-cell connector at the top of the plates. A pasted plate is manufactured by
applying a mixture of lead oxide, sulfuric acid, fibers and water on to the grid.
The thickness of the grid and plate affect the deep cycle performance of a battery.
In automotive starting or SLI type batteries many thin plates are used per cell. This
results in maximum surface area for delivering high currents, but not much thickness
and mechanical durability for deep and prolonged discharges. Thick plates are used
for deep cycling applications such as for forklifts, golf carts and other electric
vehicles. The thick plates permit deep discharges over long periods, while
maintaining good adhesion of the active material to the grid, resulting in longer life.
Separator
A separator is a porous, insulating divider between the positive and negative plates
in a battery, used to keep the plates from coming into electrical contact and shortcircuiting, and which also allows the flow of electrolyte and ions between the positive
and negative plates. Separators are made from microporous rubber, plastic or
glass-wool mats. In some cases, the separators may be like an envelope, enclosing
the entire plate and preventing shed materials from creating short circuits at the
bottom of the plates.

10-8
Element
An element is defined as a stack of positive and negative plate groups and
separators, assembled together with plate straps interconnecting the positive and
negative plates.
Terminal Posts
Terminal posts are the external positive and negative electrical connections to a
battery. A battery is connected in a PV system and to electrical loads at the
terminal posts. In a lead-acid battery the posts are generally lead or a lead alloy, or
possibly stainless steel or copper-plated steel for greater corrosion resistance.
Battery terminals may require periodic cleaning, particularly for flooded designs. It is
also recommended that the clamps or connections to battery terminals be secured
occasionally as they may loosen over time.
Cell Vents
During battery charging, gasses are produced within a battery that may be vented to
the atmosphere. In flooded designs the loss of electrolyte through gas escape from
the cell vents it a normal occurrence, and requires the periodic addition of water to
maintain proper electrolyte levels. In sealed or valve-regulated batteries the vents
are designed with a pressure relief mechanism, remaining closed under normal
conditions, but opening during higher than normal battery pressures, often the result
of overcharging or high temperature operation. Each cell of a complete battery unit
has some type of cell vent.
Flame arrestor vent caps are commonly supplied component on larger, industrial
battery systems. The venting occurs through a charcoal filter, designed to contain a
cell explosion to one cell, minimizing the potential for a catastrophic explosion of the
entire battery bank.
Case
Commonly made from a hard rubber or plastic, the case contains the plates,
separators and electrolyte in a battery. The case is typically enclosed, with the
exception of inter-cell connectors which attach the plate assembly from one cell to
the next, terminal posts, and vents or caps which allow gassing products to escape
and to permit water additions if required. Clear battery cases or containers allow for
easy monitoring of electrolyte levels and battery plate condition. For very large or
tall batteries, plastic cases are often supported with an external metal or rigid plastic
casing.

10-9
Battery Types and Classifications

Many types and classifications of batteries are manufactured today, each with
specific design and performance characteristics suited for particular applications.
Each battery type or design has its individual strengths and weaknesses. In PV
systems, lead-acid batteries are most common due to their wide availability in many
sizes, low cost and well understood performance characteristics. In a few critical,
low temperature applications nickel-cadmium cells are used, but their high initial cost
limits their use in most PV systems. There is no perfect battery and it is the task of
the PV system designer to decide which battery type is most appropriate for each
application.
In general, electrical storage batteries can be divided into to major categories,
primary and secondary batteries.
Primary Batteries
Primary batteries can store and deliver electrical energy, but cannot be recharged.
Typical carbon-zinc and lithium batteries commonly used in consumer electronic
devices are primary batteries. Primary batteries are not used in PV systems
because they can not be recharged.
Secondary Batteries
A secondary battery can store and deliver electrical energy, and can also be
recharged by passing a current through it in an opposite direction to the discharge
current. Common lead-acid batteries used in automobiles and PV systems are
secondary batteries. The table lists common secondary battery types and their
characteristics that are important to PV system designers. A detailed discussion of
each battery type follows.

10-10
Secondary Battery Types and Characteristics

Battery Type
Cost
Deep Cycle
Performance
Maintenance
Flooded Lead-Acid
Lead-Antimony
low
good
high
Lead-Calcium Open Vent
low
poor
medium
Lead-Calcium Sealed Vent
low
poor
low
medium
good
medium
medium
fair
low
medium
fair
low
Nickel-Cadmium
Sintered-Plate
high
good
none
Pocket-Plate
high
good
medium
Lead Antimony/Calcium Hybrid

Captive Electrolyte Lead-Acid
Gelled
Absorbed Glass Mat


10-11
Lead-Acid Battery Classifications

5.3.3.1
Lead-Acid Batteries
Many types of lead-acid batteries are used in PV systems, each having specific
design and performance characteristics. While there are many variations in the
design and performance of lead-acid cells they are often classified in terms of one of
the following three categories.
SLI Batteries
Starting, lighting and ignition (SLI) batteries are types of lead-acid batteries designed
primarily for shallow cycle service, most often used to power automobile starters.
These batteries have a number of thin positive and negative plates per cell designed
to increase the total plate active surface area. The large number of plates per cell
allows the battery to deliver high discharge currents for short periods. While they
are not designed for long life under deep cycle service, SLI batteries are sometimes
used for PV systems in developing countries where they are the only types of battery
locally manufactured. Although not recommended for most PV applications, SLI
batteries may provide up to two years of useful service in small stand-alone PV
systems where the average daily depth of discharge is limited to 10-20%, and the
maximum allowable depth of discharge is limited to 40-60%.
Motive Power or Traction Batteries

Motive power or traction batteries are a type of lead acid battery designed for deep
discharge cycle service, typically used in electrically operated vehicles and
equipment such as golf carts, fork lifts and floor sweepers. These batteries have a
fewer number of plates per cell than SLI batteries, however the plates are much
thicker and constructed more durably. High content lead-antimony grids are
primarily used in motive power batteries to enhance deep cycle performance.
Traction or motive power batteries are very popular for use in PV systems due to
their deep cycle capability, long life and durability of design.
Stationary Batteries
10-12
Stationary batteries are commonly used in non-interruptible power supplies (UPS) to

provide backup power to computers, telephone equipment and other critical loads or
devices. Stationary batteries may have characteristics similar to both SLI and
motive power batteries, but are generally designed for occasional deep discharge,
limited cycle service. Low water loss lead-calcium battery designs are used for most
stationary battery applications, as they are commonly float-charged continuously.
Types of Lead-Acid Batteries

There are several types of lead-acid batteries manufactured. The following sections
describe the types of lead-acid batteries commonly used in PV systems.
Lead-Antimony Batteries
Lead-antimony batteries are a type of lead-acid battery which use antimony (Sb) as
the primary alloying element with lead in the plate grids. The use of lead-antimony
alloys in the grids has both advantages and disadvantages. Advantages include
providing greater mechanical strength than pure lead grids, and excellent deep
discharge and high discharge rate performance. Lead-antimony grids also limit the
shedding of active material and have better lifetime than lead-calcium batteries when
operated at higher temperatures. Disadvantages of lead-antimony batteries are a
high self-discharge rate, and as the result of necessary overcharge, require frequent
water additions depending on the temperature and amount of overcharge.
Most lead-antimony batteries are flooded, open vent types with removable caps to
permit water additions. They are well suited to application in PV systems due to
their deep cycle capability and ability to take abuse, however they do require
periodic water additions. The frequency of water additions can be minimized by the
use of catalytic recombination caps or battery designs with excess electrolyte
reservoirs. The health of flooded, open vent lead-antimony batteries can easily be
checked by measuring the specific gravity of the electrolyte with a hydrometer.
Lead-antimony batteries with thick plates and robust design are generally classified
as motive power or traction type batteries, are widely available and are typically used
in electrically operated vehicles where deep cycle long-life performance is required.

10-13
Lead-Calcium Batteries
Lead-calcium batteries are a type of lead-acid battery which use calcium (Ca) as the
primary alloying element with lead in the plate grids. Like lead-antimony, the use of
lead-calcium alloys in the grids has both advantages and disadvantages.
Advantages include providing greater mechanical strength than pure lead grids, a
low self-discharge rate, and reduced gassing resulting in lower water loss and lower
maintenance requirements than for lead-antimony batteries. Disadvantages of leadcalcium batteries include poor charge acceptance after deep discharges and
shortened battery life at higher operating temperatures and if discharged to greater
than 25% depth of discharge repeatedly.
Flooded Lead-Calcium, Open Vent

Often classified as stationary batteries, these batteries are typically supplied as
individual 2-volt cells in capacity ranges up to and over 1000 ampere-hours.
Flooded lead-calcium batteries have the advantages of low self-discharge and low
water loss, and may last as long as 20 years in stand-by or float service. In PV
applications, these batteries usually experience short lifetimes due to sulfation and
stratification of the electrolyte unless they are charged properly.
Flooded Lead-Calcium, Sealed Vent

Primarily developed as 'maintenance free' automotive starting batteries, the capacity
for these batteries is typically in the range of 50 to 120 ampere-hours, in a nominal
12 volt unit. Like all lead-calcium designs, they are intolerant of overcharging, high
operating temperatures and deep discharge cycles. They are maintenance free in
the sense that you do not add water, but they are also limited by the fact that you
cannot add water that generally limits their useful life. This battery design
incorporates sufficient reserve electrolyte to operate over its typical service life
without water additions. These batteries are often employed in small stand-alone
PV systems such as in rural homes and lighting systems, but must be carefully
charged to achieve maximum performance and life. While they are low cost, they
are really designed for shallow cycling, and will generally have a short life in most PV
applications
An example of this type of battery that is widely produced throughout the world is the
Delco 2000. It is relatively low cost and suitable for unsophisticated users that might
not properly maintain their battery water level. However, it is really a modified SLI
battery, with many thin plates and will only provide a couple years of useful service
in most PV systems.
Lead-Antimony/Lead-Calcium Hybrid
10-14
These are typically flooded batteries, with capacity ratings of over 200 amperehours. A common design for this battery type uses lead-calcium tubular positive
electrodes and pasted lead-antimony negative plates. This design combines the
advantages of both lead-calcium and lead-antimony design, including good deep
cycle performance, low water loss and long life. Stratification and sulfation can also
be a problem with these batteries, and must be treated accordingly. These batteries
are sometimes used in PV systems with larger capacity and deep cycle
requirements. A common hybrid battery using tubular plates is the Exide Solar
battery line manufactured in the United States.

10-15
Captive Electrolyte Lead-Acid Batteries

Captive electrolyte batteries are another type of lead-acid battery. As the name
implies, the electrolyte is immobilized in some manner and the battery is sealed
under normal operating conditions. Under excessive overcharge, the normally
sealed vents open under gas pressure. Often captive electrolyte batteries are
referred to as valve regulated lead acid (VRLA) batteries, noting the pressure
regulating mechanisms on the cell vents. Electrolyte cannot be replenished in these
battery designs, therefore they are intolerant of excessive overcharge.
Captive electrolyte lead-acid batteries are popular for PV applications because they
are spill proof and easily transported, and they require no water additions making
them ideal for remote applications were maintenance is infrequent or unavailable.
However, a common failure mode for these batteries in PV systems is excessive
overcharge and loss of electrolyte, which is accelerated in warm climates. For this
reason, it is essential that the battery charge controller regulation set points are
adjusted properly to prevent overcharging.
This battery technology is very sensitive to charging methods, regulation voltage and
temperature extremes. Optimal charge regulation voltages for captive electrolyte
batteries vary between designs, so it is necessary to follow manufacturers
recommendations when available. When no information is available, the charge
regulation voltage should be limited to no more than 14.2 volts at 25 oC for nominal
12-volt batteries. The recommended charging algorithm is constant-voltage, with
temperature compensation of the regulation voltage required to prevent overcharge.
A benefit of captive or immobilized electrolyte designs is that they are less
susceptible to freezing compared to flooded batteries. Typically, lead-calcium grids
are used in captive electrolyte batteries to minimize gassing, however some designs
use lead-antimony/calcium hybrid grids to gain some of the favorable advantages of
lead-antimony batteries.
In the United States, about one half of the small remote PV systems being installed
use captive electrolyte, or sealed batteries. The two most common captive
electrolyte batteries are the gelled electrolyte and absorbed glass mat designs.

10-16
Gelled Batteries
Initially designed for electronic instruments and consumer devices, gelled lead-acid
batteries typically use lead-calcium grids. The electrolyte is 'gelled' by the addition of
silicon dioxide to the electrolyte, which is then added to the battery in a warm liquid
form and gels as it cools. Gelled batteries use an internal recombinant process to
limit gas escape from the battery, reducing water loss. Cracks and voids develop
within the gelled electrolyte during the first few cycles, providing paths for gas
transport between the positive and negative plates, facilitating the recombinant
process.
Some gelled batteries have a small amount of phosphoric acid added to the
electrolyte to improve the deep discharge cycle performance of the battery. The
phosphoric acid is similar to the common commercial corrosion inhibitors and metal
preservers, and minimizes grid oxidation at low states of charge. Gelled batteries
represent over 90% of the captive electrolyte batteries used in small PV systems in
the United States.
Absorbed Glass Mat (AGM) Batteries

Another sealed, or valve regulated lead-acid battery, the electrolyte in an AGM
battery is absorbed in glass mats that are sandwiched in layers between the plates.
However, the electrolyte is not gelled. Similar in other respects to gelled batteries,
AGM batteries are also intolerant to overcharge and high operating temperatures.
Recommended charge regulation methods stated above for gelled batteries also
apply to AGMs.
A key feature of AGM batteries is the phenomenon of internal gas recombination.
As a charging lead-acid battery nears full state of charge, hydrogen and oxygen
gasses are produced by the reactions at the negative and positive plates,
respectively. In a flooded battery these gasses escape from the battery through the
vents, thus requiring periodic water additions. In an AGM battery the excellent ion
transport properties of the liquid electrolyte held suspended in the glass mats, the
oxygen molecules can migrate from the positive plate and recombine with the slowly
evolving hydrogen at the negative plate and form water again. Under conditions of
controlled charging the pressure relief vents in AGM batteries are designed to
remain closed, preventing the release of any gasses and water loss.
While these batteries are successfully used in PV systems, unfavorable
performance of AGM batteries in early PV applications has limited their use to less
than 10% of the captive electrolyte batteries currently used in small PV systems in
the United States.

10-17
Lead-Acid Battery Chemistry

Now that the basic components of a battery have been described, the overall
electrochemical operation of a battery can be discussed. Referring to Figure 10-1,
the basic lead-acid battery cell consists of sets positive and negative plates, divided
by separators, and immersed in a case with an electrolyte solution. In a fully
charged lead-acid cell the positive plates are lead dioxide (PbO2), the negative
plates are sponge lead (Pb), and the electrolyte is a diluted sulfuric acid solution.
When a battery is connected to an electrical load, current flows from the battery as
the active materials are converted to lead sulfate (PbSO4).
Lead-Acid Cell Reaction

The following equations show the electrochemical reactions for the lead-acid cell.
During battery discharge the directions of the reactions listed goes from left to right.
During battery charging, the direction of the reactions is reversed, and the reactions
go from right to left. Note that the elements as well as charge are balanced on both
sides of each equation.
Lead-Acid Cell Electrochemical Reactions

At the positive plate or electrode:
PbO2 + 4 H + + 2e Pb 2 + + 2 H 2 O
Pb 2 + + SO42 PbSO4
At the negative plate or electrode:
Pb Pb 2 + + 2e
Pb 2 + + SO42 PbSO4
Overall lead-acid cell reaction:
PbO2 + Pb + 2 H 2 SO4 2 PbSO4 + 2 H 2 O

10-18
Some consequences of these reactions are interesting and important. As the

battery is discharged, the active materials PbO2 and Pb in the positive and negative
plates, respectively, combine with the sulfuric acid solution to form PbSO4 and
water. Note that in a fully discharged battery the active materials in both the positive
and negative plates are converted to PbSO4, while the sulfuric acid solution is
converted to water. This dilution of the electrolyte has important consequences in
terms of the electrolyte specific gravity and freezing point that will be discussed later.
Formation
Forming is the process of initial battery charging during manufacture. Formation of a
lead-acid battery changes the lead oxide (PbO) on the positive plate grids to lead
dioxide (PbO2), and to metallic sponge lead (Pb) on the negative plates. The extent
to which a battery has been formed during manufacture dictates the need for
additional cycles in the field to achieve rated capacity.
Stratification
Stratification is a condition that can occur in flooded lead-acid batteries in which the
concentration or specific gravity of the electrolyte increases from the bottom to top of
a cell. Stratification is generally the result of undercharging, or not providing enough
overcharge to gas and agitate the electrolyte during finish charging. Prolonged
stratification can result in the bottom of the plates being consumed, while the upper
portions remaining in relatively good shape, reducing battery life and capacity. Tall
stationary cells, typically of large capacity, are particularly prone to stratification
when charged at low rates. Periodic equalization charges thoroughly mix the
electrolyte and can prevent stratification problems.

10-19
Specific Gravity
Specific gravity is defined as the ratio of the density of a solution to the density of
water, typically measured with a hydrometer. By definition, water has a specific
gravity of one. In a lead-acid battery, the electrolyte is a diluted solution of sulfuric
acid and water. In a fully charged battery, the electrolyte is approximately 36%
sulfuric acid by weight, or 25% by volume, with the remainder water. The specific
gravity of the electrolyte is related to the battery state of charge, depending on the
design electrolyte concentration and temperature.
In a fully charged flooded lead-acid battery, the specific gravity of the electrolyte is
typically in the range of 1.250 to 1.280 at a temperature of 27o C, meaning that the
density of the electrolyte is between 1.25 and 1.28 times that of pure water. When
the battery is discharged, the hydrogen (H+) and sulfate (SO42-) ions from the sulfuric
acid solution combine with the active materials in the positive and negative plates to
form lead sulfate (PbSO4), decreasing the specific gravity of the electrolyte. As the
battery is discharged to greater depths, the sulfuric acid solution becomes diluted
until there are no ions left in solution. At this point the battery is fully discharged,
and the electrolyte is essentially water with a specific gravity of one.
Concentrated sulfuric acid has a very low freezing point (less than -50o C) while
water has a much higher freezing point of 0o C. This has important implications in
that the freezing point of the electrolyte in a lead-acid battery varies with the
concentration or specific gravity of the electrolyte. As the battery becomes
discharged, the specific gravity decreases resulting in a higher freezing point for the
electrolyte.
Lead-acid batteries used in PV systems may be susceptible to freezing in some
applications, particularly during cold winters when the batteries may not be fully
charged during below average insolation periods. The PV system designer must
carefully consider the temperature extremes of the application along with the
anticipated battery state of charge during the winter months to ensure that lead-acid
batteries are not subjected to freezing. Table 10-2 shows the properties and
freezing points for sulfuric acid solutions.

10-20
Properties of Sulfuric Acid Solutions

Specific Gravity
1.000
1.050
1.100
1.150
1.200
1.250
1.300

H2SO4 (Wt%)
H2SO4 (Vol%)
Freezing Point
(oC)
0.0
7.3
14.3
20.9
27.2
33.4
39.1
0.0
4.2
8.5
13.0
17.1
22.6
27.6
0
-3.3
-7.8
-15
-27
-52
-71
Adjustments to Specific Gravity

In very cold or tropical climates the specific gravity of the sulfuric acid solution in
lead-acid batteries is often adjusted from the typical range of 1.250 to 1.280. In
tropical climates where freezing temperatures do not occur, the electrolyte specific
gravity may be reduced to between 1.210 and 1.230 in some battery designs. This
lower concentration electrolyte will lessen the degradation of the separators and
grids and prolong the batterys useful service life. However, the lower specific
gravity decreases the storage capacity and high discharge rate performance of the
battery. Generally, these factors are offset by the fact that the battery is generally
operating at higher than normal temperatures in tropical climates.
In very cold climates, the specific gravity of the electrolyte may be increased above
the typical range of 1.250 to 1.280 to values between 1.290 and 1.300. By
increasing the electrolyte concentration, the electrochemical activity in the battery is
accelerated, improving the low temperature capacity and lowers the potential for
battery freezing. However, these higher specific gravities generally reduce the
useful service life of a battery.
While the specific gravity can also be used to estimate the state of charge of a leadacid battery, low or inconsistent specific gravity reading between series connected
cells in a battery may indicate sulfation, stratification, or lack of equalization between
cells. In some cases a cell with low specific gravity may indicate a cell failure or
internal short-circuit within the battery. Measurement of specific gravity can be a
valuable aid in the routine maintenance and diagnostics of battery problems in
stand-alone PV systems.

10-21
Sulfation
Sulfation is a normal process that occurs in lead-acid batteries resulting from
prolonged operation at partial states of charge. Even batteries that are frequently
fully charged suffer from the effects of sulfation as the battery ages. The sulfation
process involves the growth of lead sulfate crystals on the positive plate, decreasing
the active area and capacity of the cell. During normal battery discharge, the active
materials of the plates are converted to lead sulfate. The deeper the discharge, the
greater the amount of active material that is converted to lead sulfate. During
recharge the lead sulfate is converted back into lead dioxide and sponge lead on the
positive and negative plates, respectively. If the battery is recharged soon after
being discharged the lead sulfate converts easily back into the active materials.
However, if a lead-acid battery is left at less than full state of charge for prolonged
periods (days or weeks), the lead sulfate crystallizes on the plate and inhibits the
conversion back to the active materials during recharge. The crystals essentially
lock away active material and prevent it from reforming into lead and lead dioxide,
effectively reducing the capacity of the battery. If the lead sulfate crystals grow too
large they can cause physical damage to the plates. Sulfation also leads to higher
internal resistance within the battery, making it more difficult to recharge.
Sulfation is a common problem experienced with lead-acid batteries in many PV
applications. As the PV array is sized to meet the load under average conditions,
the battery must sometimes be used to supply reserve energy during periods of
excessive load usage or below average insolation. As a consequence, batteries in
most PV systems normally operate for some length of time over the course of a year
at partial states of charge, resulting in some degree of sulfation. The longer the
period and greater the depth of discharge, the greater the extent of sulfation.
To minimize sulfation of lead acid batteries in photovoltaic systems the PV array is
generally designed to recharge the battery on the average daily conditions during the
worst insolation month of the year. By sizing for the worst months weather, the PV
array has the best chance of minimizing the seasonal battery depth of discharge. In
hybrid systems using a backup source such as a generator or wind turbine, the
backup source can be effectively used to keep the batteries fully charged even if the
PV array can not. In general, proper battery and array sizing, as well as periodic
equalization charges can minimize the onset of sulfation.

10-22
Nickel-Cadmium Batteries
Nickel-cadmium (Ni-Cad) batteries are secondary, or rechargeable batteries and
have several advantages over lead-acid batteries that make them attractive for use
in stand-alone PV systems. These advantages include long life, low maintenance,
survivability from excessive discharges, excellent low temperature capacity
retention, and non-critical voltage regulation requirements. The main disadvantages
of nickel-cadmium batteries are their high cost and limited availability compared to
lead-acid designs.
A typical nickel-cadmium cell consists of positive electrodes made from nickelhydroxide (NiO(OH))and negative electrodes made from cadmium (Cd) and
immersed in an alkaline potassium hydroxide (KOH) electrolyte solution. When a
nickel-cadmium cell is discharged, the nickel hydroxide changes form (Ni(OH)2) and
the cadmium becomes cadmium hydroxide (Cd(OH)2). The concentration of the
electrolyte does not change during the reaction so the freezing point stays very low.
Sintered Plate Ni-Cads

Sintered plate nickel cadmium batteries are commonly used in electrical test
equipment and consumer electronic devices. The batteries are designed by heat
processing the active materials and rolling them into metallic case. The electrolyte
in sintered plate nickel-cadmium batteries is immobilized, preventing leakage,
allowing any orientation for installation. The main disadvantage of sintered plate
designs is the so called 'memory effect', in which a battery that is repeatedly
discharged to only a percentage of its rated capacity will eventually 'memorize' this
cycle pattern, and will limit further discharge resulting in loss of capacity. In some
cases the 'memory effect' can be erased by conducting special charge and
discharge cycles, regaining some of its initial rated capacity.
Pocket Plate Ni-Cads

Large nickel cadmium batteries used in remote telecommunications systems and
other commercial applications are typically of a flooded design, called flooded
pocket plate. Similar to flooded lead-acid designs, these batteries require periodic
water additions, however, the electrolyte is an alkaline solution of potassium
hydroxide, instead of a sulfuric acid solution. These batteries can withstand deep
discharges and temperature extremes much better than lead-acid batteries, and they
do not experience the 'memory effect' associated with sintered plate Ni-Cads. The
main disadvantage of pocket plate nickel cadmium batteries is their high initial cost,
however their long lifetimes can result in the lowest life cycle cost battery for some
PV applications.

10-23
Nickel-Cadmium Battery Chemistry

Following are the electrochemical reactions for the flooded nickel-cadmium cell:
Nickel-Cadmium Cell Reactions

2 NiO(OH ) + 2 H 2 O + 2e 2 Ni(OH ) 2 + 2OH
Cd + 2OH Cd (OH ) 2 + 2e
Overall nickel cadmium cell reaction:
Cd + 2 NiO(OH ) + 2 H 2 O Cd (OH ) 2 + 2 Ni(OH ) 2
Notice these reactions are reversible and that the elements and charge are balanced
on both sides of the equations. The discharge reactions occur from left to right,
while the charge reactions are reversed.
The nominal voltage for a nickel-cadmium cell is 1.2 volts, compared to about 2.1
volts for a lead-acid cell, requiring 10 nickel-cadmium cells to be configured in series
for a nominal 12-volt battery. The voltage of a nickel-cadmium cell remains relatively
stable until the cell is almost completely discharged, where the voltage drops off
dramatically. Nickel-cadmium batteries can accept charge rates as high as C/1, and
are tolerant of continuous overcharge up to a C/15 rate. Nickel-cadmium batteries
are commonly subdivided in to two primary types; sintered plate and pocket plate.

10-24
Exercises

!

!

" #

#

# "
!
! ! !

$

% &' ! $

(
))))))))))
(
))))))))))

* !

!

&'
(
)))))))))) !

"
'+,
! !

-+,


10-25

.

/

0

%

1
2$
#
$
$

.


10-26
Battery Strengths and Weaknesses

Each battery type has design and performance features suited for particular
applications. Again, no one type of battery is ideal for PV system applications. The
designer must consider the advantages and disadvantages of different batteries with
respect to the requirements of a particular application. Some of the considerations
include lifetime, deep cycle performance, tolerance to high temperatures and
overcharge, maintenance and many others. The table below summarizes some of
the key advantages and disadvantages of the different battery types discussed in the
preceding section.
Battery Type
Flooded Lead-Acid
Lead-Antimony
Advantages
Disadvantages
low cost, wide availability,

good deep cycle and high
temperature performance,
can replenish electrolyte
low water loss, can
replenish electrolyte
high water loss and

maintenance
Lead-Calcium Sealed Vent

low water loss
Lead Antimony/
Calcium Hybrid
medium cost, low water

loss
poor deep cycle performance,

intolerant to high
temperatures and
overcharge, can not replenish
electrolyte
limited availability, potential
for stratification
medium cost, little or no

maintenance, less
susceptible to freezing,
install in any orientation
medium cost, little or no
maintenance, less
susceptible to freezing,
install in any orientation
fair deep cycle performance,

intolerant to overcharge and
high temperatures, limited
availability
fair deep cycle performance,
intolerant to overcharge and
high temperatures, limited
availability
wide availability, excellent

low and high temperature
performance,
maintenance free
excellent deep cycle and
low and high temperature
performance, tolerance to
overcharge
only available in low

capacities, high cost, suffer
from memory effect
Lead-Calcium Open Vent
Captive Electrolyte Lead-Acid

Gelled
Absorbed Glass Mat
Nickel-Cadmium
Sealed Sintered-Plate
Flooded Pocket-Plate
poor deep cycle performance,

intolerant to high
temperatures and overcharge
limited availability, high cost,

water additions required


10-27
Battery Cycling
One of the main parameters that distinguishes battery types is their ability to cycle.
A battery cycle refers to the process of charging and discharging a battery. Battery
discharge is the process that occurs when a battery delivers current, quantified by
the discharge current or rate. Charging is the process when a battery receives or
accepts current, quantified by the charge current or rate. A discharge followed by a
recharge is considered one cycle.
The discharge can be very small or shallow, or it can be very severe or deep. A 100
percent depth of discharge cycle provides a measure of the total battery capacity at
a given discharge rate. All batteries can be cycled, but the question is how deeply
and how many times before a permanent loss of capacity occurs. Batteries used in
photovoltaic applications will definitely be subjected to cycling on a daily basis, and
perhaps deeply cycled occasionally.
We can think of a battery as being full of charge, even though it is actually full of
chemicals that hold potential energy. We use the capacity of a battery bank in a
photovoltaic power system to operate the loads during each night, and during
periods of heavy load use or below average insolation. If a series of below average
weather days occurs in a row, then the battery is not fully recharged at the end of
each daily cycle, and the capacity and state of charge of the battery reduces daily.
B attery C ycling
B attery capacity
O ne discharge and
recharge is one cycle
B attery life depends on
how deep and how
m any tim es the battery
is cycled
B atteries designed for
shallow cycling w ill
w ork, but for short tim e
Solar system s need
deep cycling batteries
+
available
for use
on e d ays use
leave unused


10-28
Battery Discharging
When a battery is discharged the chemical reaction on the plates proceeds inward
toward the grid. The deeper the discharge, the deeper the chemical reaction occurs.
In lead-acid batteries the lead-sulfate molecules that are formed are larger than the
lead or lead-oxide molecules, and the bonding of the active material to the plates is
gradually weakened due to grid growth. Figure 10-3 illustrates the discharge
process.
Discharge Process
Discharge reaction
proceeds inward toward
grid
100% discharge
weakens adhesion
Increased resistance
produces heat
Degradation accelerates
grid
active material
coated onto grid


10-29
Charge/Discharge Rates Expressed as

Time
Battery manufacturers often refer to rates of charge or discharge not in amperes but
by the time it would take to completely charge or discharge a battery at a specific
current. The rate of charge or discharge of a battery is expressed as a ratio of the
nominal battery capacity to the charge or discharge time period in hours. For
example, a 100 Ah battery is discharging at the rate of 2 amps. The time to
completely discharge a fully charged battery at this rate would be the capacity
divided by the current, or 100 Ah / 2 amps = 50 hours. So we would say that the
battery is discharging at the "50 hour rate" or at "C/50".
Rate in Amperes =
Capacity in Amp - Hours C

=
Time in Hours
T
This notation is helpful because it allows us to talk about relative rates of battery
charge and discharge, without referring to the exact size of a battery. For example,
most manufacturers recommend charging their batteries no faster than the C/5 rate
to limit gassing and overcharge. This means 20 amps for a 100 Ah battery, and 100
amps for a 500 Ah battery. Moderate charge rates are around C/20 or C/30, while
trickle charging at C/100 will hardly produce any gassing at all in most batteries.
This notation is used for discussing discharge rates as well. For example, batteries
used for UPS systems generally have their capacity measured at the C/5 or C/2 hour
rate, because in the event of a utility power failure, the UPS system is expected to
operate for only 2 or 5 hours. Batteries used in electric forklift operations will have
their capacity rated at the C/8 hour rate, because it is anticipated that they will be
discharged during a typical 8-hour shift.
Batteries used in typical PV systems experience very low rates of charge and
discharge compared to these common industrial applications. For example, the
maximum charge rates from the PV array to battery are commonly about C/40, and
typical discharge rates supplied to the loads may be as low as C/100 to C/200.

10-30
Depth of Discharge (DOD)

The depth of discharge (DOD) of a battery is defined as the percentage of capacity
that has been withdrawn from a battery compared to the total fully charged capacity.
By definition, the depth of discharge and state of charge of a battery add to 100
percent. The two common qualifiers for depth of discharge in PV systems are the
allowable or maximum DOD and the average daily DOD and are described as
follows:
Allowable Depth of Discharge

The maximum percentage of full-rated capacity that can be withdrawn from a battery
is known as its allowable depth of discharge. The allowable DOD is the maximum
discharge limit for a battery, generally dictated by the cut off voltage and discharge
rate. In stand-alone PV systems the low voltage load disconnect (LVD) set point of
the battery charge controller dictates the allowable DOD limit at a given discharge
rate. Furthermore, the allowable DOD is generally a seasonal deficit, resulting from
low insolation, low temperatures and/or excessive load usage. Depending on the
type of battery used in a PV system, the design allowable depth of discharge may be
as high as 80% for deep cycle, motive power batteries, to as low as 15-25% if SLI
batteries are used. The allowable DOD is related to the autonomy, in terms of the
capacity required to operate the system loads for a given number of days without
energy from the PV array. A system design with a lower allowable DOD will result in
a shorter autonomy period.
Average Daily Depth of Discharge

The average daily depth of discharge is the percentage of the full-rated capacity that
is withdrawn from a battery with the average daily load profile. If the load varies
seasonally, for example in a PV lighting system, the average daily DOD will be
greater in the winter months due to the longer nightly load operation period. For PV
systems with a constant daily load, the average daily DOD is generally greater in the
winter due to lower battery temperature and lower rated capacity. Depending on the
rated capacity and the average daily load energy, the average daily DOD may vary
between only a few percent in systems designed with a lot of autonomy, or as high
as 50 percent for marginally sized battery systems. The average daily DOD is
inversely related to autonomy; meaning that systems designed for longer autonomy
periods (more capacity) have a lower average daily DOD.

10-31
Autonomy Determines Capacity and Depth

of Discharge
The number of days of reserve or autonomy is the main factor that determines the
size of the battery and therefore the magnitude of the daily battery depth of
discharge. The greater the number of days of autonomy sized into the battery, the
larger the total capacity and therefore the smaller the percentage used each day for
a typical daily cycle.
The relationship between days of autonomy and depth of discharge is shown below,
for both deep cycling batteries (maximum allowable DOD = 80%) and for shallow
cycling batteries (maximum allowable DOD = 50%).
Autonomy (Reserve)
Determines Daily Discharge
Depth of
Discharge (%)
80
70
60
80% Maximum D.O.D.
50
40
30
50% Maximum D.O.D.
20
10
0
11
13
15
17
19
Number of Days of Autonomy


10-32
Example:
A photovoltaic power system is designed with deep cycling batteries to

give 5 days of autonomous operation to a final discharge depth of
80%. Reading up from 5 days on the graph to the 80% Maximum
DOD line, the typical daily depth of discharge would be only 15%.
If the battery were a shallow cycling type and could only be 50%
discharged at the end of 5 days, the average daily depth of discharge
would be only 10%.
Since most remote PV power systems are designed with at least 4 days of
autonomy, batteries in remote photovoltaic systems will be shallow cycled on an
average daily basis, whether deep or shallow cycling type batteries are used.
However, we recommend using deep cycling batteries in remote photovoltaic power
systems, even though they will typically be shallow cycled, perhaps 15-10% daily or
even less. By using deep cycling type batteries, the system can withstand the
expected seasonal drops in capacity due to below average insolation. Shallow
cycling batteries can be used as well, but shorter life should be expected.

10-33
Exercises

3

3

4

3 !

3

!4


10-34
State of Charge (SOC)

The state of charge (SOC) is defined as the amount of energy in a battery,
expressed as a percentage of the energy stored in a fully charged battery.
Discharging a battery results in a decrease in state of charge, while charging results
in an increase in state of charge. A battery that has had three quarters of its
capacity removed, or been discharged 75%, is said to be at 25% state of charge.
The figure below illustrates the seasonal variation in battery state of charge and
depth of discharge for a typical PV system.


10-35
Temperature Limits Discharge Depth

As discussed earlier, if the internal temperature of a battery reaches the freezing
point of the electrolyte, the electrolyte can freeze and expand, causing irreversible
damage to the battery. In a fully charged lead-acid battery, the electrolyte is
approximately 35% by weight and the freezing point is quite low (approximately -60
o
C). As a lead-acid battery is discharged, the becomes diluted, so the concentration
of acid decreases and the concentration of water increases as the freezing point
approaches the freezing point of water, 0 oC. The figure below shows the
relationship between the maximum allowable depth of discharge and corresponding
minimum temperature to avoid freezing in a typical flooded lead-acid battery. Notice
that this is the similar to the information presented in Table 10-2 on the discussion of
specific gravity.
Temperature Limits Maximum

Discharge Depth
Maximum D.O.D. (%)
80
60
40
20
0
-60
-40
-20
Lowest Battery Temperature (deg.C)


10-36
The maximum depth of discharge must not be allowed to exceed the point that
would allow freezing. The photovoltaic power system must be designed with enough
battery capacity so that after the maximum expected days of autonomous operation,
the depth of discharge does not exceed the danger point. If the coldest expected
battery temperatures are not known, the coldest 24-hour average temperature for
the location may be used.
You can see that if the coldest temperature does not go below -8 oC, then there is no
effect on the allowable maximum depth of discharge to prevent freezing. This factor
is only applicable in very cold climates.
Example:
At a remote telecommunications site on a mountain top, the coldest

24-hour temperature is approximately -20 oC. Read up from -20 oC in
Figure 10-6 to a value of approximately 50%.
This means that the maximum allowable depth of discharge would be
approximately 50%, even if "deep cycling" type batteries were used
and the manufacturer had indicated that they can be designed for
maximum discharges of 80%.
The example system should be designed so that at maximum discharge, the

batteries would be only 50% discharged (at the end of the autonomy period). If
allowed to discharge more during this coldest time of the year, they might be in
danger of freezing.

10-37
Exercises

" !

$

&5
6 *

(
)))))))))) ,

"

'+ 6

$

$

7(
)))))))))) ,
6 #
#
!

(

10-38
Shallow and Deep Cycle Batteries

Batteries are sometimes classified as deep or shallow cycling, but there are only
general classifications and are not always truly indicative of actual battery
performance and cell design. Shallow cycle batteries generally a type of SLI or
automotive starting battery. These batteries are made with lead-calcium grids and
should not be cycled greater than 15% on a daily (or nightly) basis. They should not
be allowed to discharge more than 50% under any circumstances, because they
become very difficult to recharge from beyond that depth of discharge. This is
especially true for low cost automobile batteries that might be used for simple rural
electrification systems. These types of batteries are designed to provide large
currents for short periods, and are not designed to sustain deep discharges. They
will give about 500-1000 cycles to 15% depth of discharge before loosing too much
capacity and requiring replacement.
Shallow Cycle Batteries

500 - 1000 cycles to 15% maximum daily discharge
50% maximum allowable depth of discharge
Deep cycle batteries are often traction of motive power types and can handle greater
discharges for longer times than typical shallow cycle batteries. Even so they should
not be cycled to 100% discharge. Most manufacturers recommend that deep cycling
batteries be discharged to no more than 80% of the rated capacity, to prevent
reactions from occurring close to the grid. The expansion from lead and lead dioxide
molecules into large lead sulfate molecules can severely weaken the bonding of the
active materials to the grid, causing increased internal resistance and heating.
Typically deep cycle batteries can deliver 1500-1800 cycles to 80% depth of
discharge before needing replacement, and will deliver 3000-4000 cycles if
discharged more moderately to 25% depth of discharge or less.
Deep Cycle Batteries

3000 - 4000 cycles to 25% depth of discharge
500 - 1800 cycles to 80% maximum allowable depth of discharge

10-39
The difference in cycle life between deep and shallow cycle batteries is shown in the
figure below. The general trend for both types of batteries is that a greater depth of
discharge yields a fewer number of cycles. All batteries deliver fewer cycles at
greater depths of discharge.
Depth of Discharge Affects Cycle Life

Number of Cycles to 20% Loss of Capacity
5000
4000
Deep Cycling Batteries
3000
2000
1000
0
Shallow Cycling Batteries

10
30
50
70
Depth of Discharge (%)

What determines when a battery should be replaced? Most all battery

manufacturers recommend that when a battery has lost 20% of its capacity it has
reached the end of its useful life, and should be replaced. The cycle graph
presented above is based on the number of cycles until 20% capacity has been lost,
or in other words until only 80% of the original capacity is left. However, this chart
should only be used as a rough guideline for battery cycle life. The specific battery
type and design, temperature, rates, charging methods, maintenance and other
factors all have an effect on battery cycle life.
Example:
If a shallow cycling battery is discharged to approximately 15% each cycle,

you could expect about 500 cycles before it would need replacing. This
would correspond to about 1-2 years of operation if the cycling were once a
day. If it were discharged to about 50% each cycle, you could expect a few
hundred cycles, corresponding to less than a year.
If a deep cycling battery were discharged to 80% each cycle, you could
expect about 1500-1800 cycles, corresponding to about 4-5 years of
operation if there were one cycle each day. Cycling to only 20% each cycle
could give about 4000 cycles, or about 10 years of operation.

10-40
Battery Capacity
Capacity is a measure of a battery's ability to store or deliver electrical energy,
commonly expressed in units of ampere-hours.
Ampere-Hour Definition
The ampere-hour (Ah) is the common unit of measure for a battery's electrical
storage capacity, obtained by integrating the discharge or charge current in amperes
over a specific time period. An ampere-hour is equal to the transfer of one ampere
over one hour, equal to 3600 coulombs of charge. For example, a battery that
delivers 5 amps for 20 hours is said to have delivered 100 ampere-hours.
Factors Affecting Battery Capacity

Capacity is generally specified at a specific discharge rate, or over a certain time
period. The capacity of a battery depends on several design factors including; the
quantity of active material, the number, design and physical dimensions of the
plates, and the electrolyte specific gravity. Operational factors affecting capacity
include; the discharge rate, depth of discharge, cut-off voltage, temperature, age
and cycle history of the battery. Sometimes a battery's energy storage capacity is
expressed in kilowatt-hours (kWh), which can be approximated by multiplying the
rated capacity in ampere-hours by the nominal battery voltage and dividing the
product by 1000. For example, a nominal 12 volt, 100 ampere-hour battery has an
energy storage capacity of (12 x 100)/1000 = 1.2 kilowatt-hours.

10-41
Effects of Temperature on Capacity

Cold temperatures decrease the total capacity available from a battery. During
discharge, the electrolyte does not penetrate as deeply into the active material on
the plates, and the cut-off voltage is reached sooner at cold temperatures.
PV system designers must be acutely aware of the effects of temperature on battery
capacity. Battery manufacturers generally rate capacity at a temperature of 25 oC.
If the required battery size in a PV system is calculated based on the expected
capacity at 25 oC, the battery may be too small to provide the storage necessary to
achieve the design autonomy period during cold temperatures. As a result the
battery could be severely discharged and the system loads may not be satisfied.
Additional battery capacity must be installed in stand-alone PV power systems to
compensate for the expected reduction in capacity at low temperatures.
Standard Temperature for Rated Battery Capacity
25 oC
Battery Capacity Must Be Derated For Low Temperature Operation

Conversely, a battery operated at temperatures greater than 25 oC will deliver more
than the rated capacity. However, under no circumstances should a battery be
heated or operated at elevated temperatures to increase the available capacity.
Higher than rated operating temperatures significantly reduce battery life. Most
battery manufacturers recommend their batteries be operated in temperature ranges
of between 20 and 30 oC.
How is this information used in the sizing and design of batteries in PV systems?
We apply this data during our battery sizing calculations to insure that even when
the batteries are at their coldest, the system has the capacity it requires. In other
words, when we calculate the amount of capacity we need to give the days of
autonomy desired, the calculated value must be increased by an appropriate factor if
the battery is operated below 25 oC.

10-42
Example:
Assume that 1000 ampere-hours of battery capacity is required to

operate a specific load for the required days of autonomy. Also
assume that the average discharge rate is C/120 or about 8 amps.
If the battery temperature could get as low as -20 oC during operation,
1000 Ah of capacity is still required at these lower temperatures. Since
most battery manufacturers rate capacity at 25 oC, the rated capacity
must be derated to the available capacity at -20 oC. How much
capacity is required at 25 oC to ensure that 1000 Ah are available at 20 oC at the C/50 rate?
To determine the increase in capacity required, we refer to the figure
below which shows the percent of standard rated capacity (at 25 oC) as
a function of the discharge rate and temperature. At -20 oC at the
C/120 rate, the capacity of a typical lead-acid battery is about 82% of
the rated capacity at 25 oC. To calculate the increased capacity
required, simply divide the 1000 Ah required at 25 oC by 0.82 to give
1000 / 0.82 = 1220 Ah. Therefore, by installing about 1220 Ah of
battery capacity (rated at 25 oC), we still have 1000 Ah available when
the battery is at -20 oC.
Temperature and Discharge Rate

Effects on Lead-Acid Battery Capacity
Percent of Rated Capacity
120
110
100
90
80
70
60
C/500
C/50
C/0.5
50
40
C/120
C/5
30
-30
-20
-10
10
20
30
40
Battery Operating Temperature - C

10-43
Discharge Rate Affects Capacity

When a battery is discharged, the lead sulfate and water reaction products get in the
way of further reaction. At high discharge rates, the reactions are confined to the
layers of the active material that are in immediate contact with the free electrolyte,
limiting the cell capacity. This is because there is insufficient time for the electrolyte
to diffuse into the pores of the plates, and the sulfate molecules forming at the
surface clogs the pores, preventing full use of all the active material. This effect is
amplified as the discharge rate increases. The result is that final cut-off voltage is
reached sooner and less total capacity is usable at faster rates of discharge. So
battery capacity is not a fixed value, but depends on how fast the battery is
discharged. Discharging a battery slowly delivers more capacity from the battery,
while discharging it quickly delivers less total capacity.
This effect is not permanent. For example, if a battery is discharged at a fast rate, it
only delivers a fraction of its rated capacity. If it is then fully recharged and it is
discharged at a slower rate, then more capacity will be available. However, a few
cycles may need to be performed on the battery to achieve stable capacities at the
new rates.
The rate of discharge must therefore be included in any statement of battery
capacity. Rates are not usually stated in amperes, but in hours for full discharge to a
specified cut-off voltage. Most manufacturers use the ten-hour rate for their nominal
ratings (some use eight hours or twenty hours). For example, if a battery capacity is
presented by the manufacturer as "180 Ah at the 10 hour rate", this means that if the
battery were discharged for ten hours at a constant C/10 rate (180/10 = 18 amps),
then 180 Ah would have been discharged from it to the cut off voltage. Batteries
used for industrial or motive power applications often have capacity measured at the
6 or 8 hour rate, because the time of a normal working shift is that long.
The terminology of the rate of discharge in hours, for example the 10 hour rate ,
tells us how fast the battery is being discharged. It does not necessarily mean that a
battery will in fact actually be discharged for ten hours and drop in voltage to the final
cut-off voltage. It means only that it is being discharged at a rate that would fully
discharge it in that time. In practice, it may only be discharged for a few hours.
Capacity ratings for longer periods are appropriate for photovoltaic power systems
because the battery is usually designed to discharge steadily over many days of bad
weather. For example, discharging a battery bank for a remote home designed with
five days autonomy would take 120 hours (5 days x 24 hours/day). And a battery
designed for a critical remote telecom repeater with 14 days of autonomy would take
over 300 hours (14 days X 24 hours/day) to completely discharge.

10-44
Typical Discharge Times

Industrial, motive applications
10 hours
Photovoltaic applications
100-300 hours
The time we need to specify is the time to TOTAL discharge (100% DOD). But we
do not allow batteries to fully discharge every cycle. In fact, most manufacturers
recommend that shallow cycling type batteries only be discharged to a maximum of
50% of their full capacity, and that deep cycling batteries be discharged to a
maximum of 80% of their capacity. This is to prevent weakening of the bond
between the active materials and the grids.
Maximum Recommended Depth of Discharge for Lead Acid Batteries

Shallow cycling types
50%
Deep cycling types
80%
Thus we need to take into account the maximum allowable depth of discharge in
calculating the time it would take to fully discharge a battery. The time (hours) to
discharge to the maximum DOD limit is simply the number of days of autonomy
reserve times the number of hours each day that the loads operate.
Time to discharge to max. DOD = Days of Reserve X Load Operating Time
For continuous loads, such a microwave repeaters or navigational systems, the
operating time might be 24 hours/day. For simple home lighting systems, the load (a
single fluorescent light for example) might be specified as 4 hours/day. For systems
with a variety of intermittent loads, a weighted average for the load operating time
can be calculated by summing each load (amps DC, or watts AC) multiplied times its
operating time and dividing this by the sum of the load values alone. This weights
the average load operating time for a collection of different loads based on the size
of the load and its operating time.

10-45
Load Operating Time:
Continuous Loads:
Single Load Systems:
Multiple Load Systems:
use 24 hours
use load operating time
use weighted average load operating time
Weighted Average Load Operating Time =
load time
loads
The formula for the time to total discharge (not just to max. DOD as described
above) is time to discharge to the maximum DOD divided by the maximum allowed
discharge.
Time to fully discharge
Example:
Days of Reserve X Load Operating Time

Maximum Depth of Discharge (%)
The small remote cabin used in the Load Estimation Chapter has a
variety of loads. The weighted average load operating time would be
given by dividing the sum of the load X time by the sum of the loads
only:
Weighted Average =
Load Time
=
69.4 Ah
(2 x 3.3) + (3 x .92) + (1 x 3.3) + (1 x 2)
4.7 hours
The system will be designed with 4 days of reserve autonomy in the

battery bank. If low cost shallow cycling batteries are to be used, the
limit to maximum DOD is 50%. The battery discharge rate, given in
time to total discharge would be given by:
Time to full discharge

4 days X 4.7 hours/day

0.5
37.6 hours
10-46
Example:
A small remote telemetry system is to be designed with 5 days of

reserve battery capacity. Deep cycling industrial batteries are to be
used that can discharge to 80%. The load is a continuous (24-hour)
load. The average discharge rate, expressed in hours, is given by
Time to full discharge
5 days X 24 hours/day
0.80
150 hours
In the example a battery is designed to give five days of reserve or autonomy. We

choose a deep cycling type battery, so the maximum depth of discharge (DOD) can
be up to 80%, or in other words down to 20% state of charge. As the battery is
discharged over time the state of charge decreases. By the end of 5 days or 120
hours, we would have discharged the battery to 80% depth of discharge (20% state
of charge) if there were no input from the PV array.
But the figure shows that the time to fully discharge the battery is actually 150 hours,
or about 6.25 days. So we are really discharging this particular battery at the 150
hour rate or the C/150 rate, and not the C/120 rate. We would not actually
discharge this battery over 150 hours to 100% depth of discharge or to 0% state of
charge, but this is the time that it would take if we did. Although it is this number that
is used to determine the available capacity from the battery with this autonomy
period.

10-47
Hour Rate Indicates Time to

Total Discharge
5 days (120 hours) reserve to 80% DOD gives 150 hours discharge time
1 2 0
Time to total discharge:

1 0 0
State
of
Charge
(%)
#Days X 24 hrs
Max % Discharge
8 0
# Days X 24 hours
6 0
4 0
Limit of discharge
recommended by manufacturer
2 0
2 0
4 0
6 0
8 0
1 0 0
1 2 0
1 4 0
1 6 0
Time Duration of Discharge (hours)

Literature Example Showing Capacity at Different Rates of

Discharge
A portion of a typical battery capacity specification sheet is shown on the next page
to demonstrate the dependence of the rate of discharge on battery capacity.
As an example, battery type 3-35A27 listed at the bottom of the sheet will deliver
468 Ah if discharged in 8 hours, but will deliver 610 Ah if discharged more slowly
taking 48 hours, and will deliver 624 Ah if discharged over 100 hours. Note that the
cut-off voltage of 1.75 volts per cell (VPC) is included in the specification. This
would mean that when a 12-volt battery, made of two of the 3-35A27 battery
modules connected in series, has reached 10.5 volts (6 x 1.75) the manufacturer
considers the battery to be fully discharged.
This manufacturer has listed a "Nom AH Cap" or nominal ampere-hour capacity for
each battery. The numbers shown are slightly higher than those found in the 8 Hr
column, so they reflect capacity obtained by discharging a bit more slowly than the 8
hour rate. Most likely they are using a common battery industry standard 10-hour

10-48
rate. So the manufacturers nominal capacity for the 3-35A27 is 470 Ah at the 10
hour rate, but the capacity obtained at the slower 100 hour rate would be 624 Ah. If
a range of rates is given, use the formula presented previously for calculating the
time for full discharge to determine the proper rate to use. If the exact discharge
time is not given by the manufacturer just use the time that is closest or less than the
actual time to conservatively size the battery.
Example:
In our example of the remote telemetry system presented above, the

rate of discharge, stated in hours, was given as 150 hours. The
manufacturers literature presented here goes only to the 100-hour
rate. So use that column to select the capacity available from the
battery. The capacity at the 150 hour rate would be slightly greater,
but not that much different.
If 600 Ah at 12 volts were required for the telecom project, you could
select two 6-35A15 batteries and connect them in parallel, to get 2 x
336 Ah = 672 Ah at the 100 hour rate.
Or you could select the 3-35A27 that has 624 Ah at the 100 hour rate,
but it is only 6 volts. So you would need to use two of them and
connect them in series to get 12 volts.

10-49
Type
Volts
Nominal
AH
Capacity
Horizontal Stacking Dimensions
AH Capacities of Single Cell

to 1.75 vpc average
at 25 deg.C.
Width
Height
Depth
Cell Type
8h
24
h
48
h
100
h
35A05
72
86
94
96
in
mm
in
mm
in
mm
6-35A05
12
75
17.23
438
8.60
218
14.2
361
6-35A07
12
110
21.67
550
35A07
108
129
141
144
6-35A09
12
145
26.17
665
35A09
144
172
188
192
6-35A11
12
180
30.67
779
35A11
180
215
235
240
6-35A13
12
215
35.17
893
35A13
216
258
282
288
6-35A15
12
250
39.67
1008
35A15
252
301
329
336
3-35A17
290
24.52
623
35A17
288
344
376
384
3-35A19
325
26.77
680
35A19
324
387
423
432
3-35A21
360
29.02
737
35A21
360
430
470
480
3-35A23
395
31.27
794
35A23
396
473
517
528
3-35A25
430
33.52
851
35A25
432
516
563
576
3-35A27
470
35.77
909
35A27
468
559
610
624

!"#$%&!''$$(
)'#*

10-50
Cut-Off Voltage Affects Capacity

The cut off voltage is the lowest voltage that a battery system is allowed to reach in
operation, and defines the battery capacity at a specific discharge rate.
Manufacturers often rate capacity to a specific cut off, or end of discharge voltage at
a defined discharge rate. For lead-acid batteries the cut off voltage used to rate
capacity is generally 1.75 volts per cell, or 10.5 volts for a nominal 12-volt battery.
The cut off voltage for nickel-cadmium cells is typically 1.0 volt. If the same cut off
voltage is specified for different discharge rates, the capacity will generally be higher
at the lower rate. Note that the cut off voltage defined by a battery manufacturer
most often represents a fully discharged battery. Typically, batteries used in PV
systems are never allowed to reach this low of a cut off voltage, and in practice are
generally limited to no more than an 80% depth of discharge as determined by the
low voltage load disconnect point of the battery charge controller.
Voltage (volts)
Battery Voltage Over Time

2.2
2.1
2
1.9
1.8
1.7
1.6
1.5
0
20
40
60
80
100
Time (hours)
C/100

10-51
C/10
Effects of Self Discharge Rate on Capacity

In open-circuit mode without any load or charging a battery undergoes a natural
reduction in state of charge, due to internal mechanisms and losses within a battery.
Different battery types have different self-discharge rates, the most significant factor
being the active materials and grid alloying elements used in the design. Higher
temperatures result in higher discharge rates, particularly for lead-antimony designs.
The figure below shows the effects of battery cell temperature on the self-discharge
rate for lead-acid batteries with lead-calcium and lead-antimony grids.
Lead-Acid Battery Self Discharge Rate
Self Discharge Rate

(% of rated capacity per week)
20
Lead-Antimony Grid (end of life)

Lead-Antimony Grid (new)
Lead-Calcium Grid (typical)
15
10
5
0
-50
-25
25
50
75
Battery Operating Temperature ( C)


10-52
Exercises

.!

/

0

!

3

*

&5 6

4

+ 6 68-&+(
))))))))))
+ 6 685+(
))))))))))
-+ 6 68-&+(
))))))))))
&5 6 685++(
))))))))))
-+ 6 68-&+(
))))))))))
"

!

#
-5++ " &' !
! -' !

685++ 3

-5 6

*

$

(
)))))))))) " )))))))))) 6

"

&++ "

3 !

685+ 3
5 6

#
(
)))))))))) "

10-53
Voltage Changes During Charging

and Discharging
We will now begin to examine some of the many factors that affect battery
performance. We begin with the most basic of characteristics--battery voltage.
Nominal values for battery voltage were given previously, namely 2.0 volts for a lead
acid battery cell and 1.2 volts for a nickel-cadmium cell. But these values are only
nominal, and the actual voltage of a cell or a battery will depend on several factors.
The fact that batteries operate over a range of voltages has direct impact on at least
three other components of a typical photovoltaic power system. The DC loads and
the inverter must be designed to accept the wide voltage range typical in
photovoltaic systems. If there is a problem with the extent of the range, a DC power
supply or voltage regulator must be included in the system, to keep the voltage
range with the limits of the inverter or DC loads. And the current or charge
regulators commonly used in photovoltaic systems must be designed to anticipate
the wide battery voltage range.

10-54
Battery Voltage and State of Charge are

Related
A fully charged battery is said to be at 100% state of charge (SOC). As the battery
is discharged, the chemical balances change, and it moves away from 100% SOC to
some partial SOC. The change in SOC is reflected in both the composition and
density of the electrolyte and in the voltage of the battery.
If the electrolyte could be accessed the state of charge could be determined by
measuring the specific gravity as discussed earlier. In a fully charged lead-acid cell,
the specific gravity is typically about 1.265, while a completely discharged cell would
have a specific gravity close to water, or 1.000.
While specific gravity can be used to approximate the battery state of charge in
remote PV systems, it is seldom recorded except during periodic maintenance
checks. However, the open-circuit battery voltage (without a charge or discharge
load, battery at rest for some time) may be used to estimate the state of charge.
The figure below illustrates that there is a fairly linear relationship between the SOC
of a lead-acid battery and the open-circuit voltage.
For example, if a battery were 50% discharged, or at a SOC of 50%, we could
measure a specific gravity of approximately 1.17, or a battery open-circuit voltage of
about 12.0 volts. So in this case, reading 12.0 volts on a "12 volt" battery means
that the battery is about 50% discharged! Precise digital meters are needed to
accurately determine the voltage and SOC of a battery.

10-55
It is essential to know the immediate history of the battery prior to measurement of

the battery voltage, since a charge or discharge immediately preceding the
measurement may significantly alter the accuracy of the measurement.
Lead-Acid Battery Voltage and

Specific Gravity as a Function of
State of Charge
1.300
13
Open Circuit Voltage
Voltage
(volts)
Specific
Gravity
12
1.200
11
1.100
Specific
Gravity
1.000
10
100
80
60
40
20
State of Charge (%)

Note that there is little variation in specific gravity with state of charge for nickelcadmium cells and voltage trends are different than lead-acid batteries.

10-56
Battery Voltage Depends on Rate of

Charge or Discharge
Discussing battery open-circuit voltage refers to a battery that is at rest, neither
being charged nor discharged. When a battery is being charged, the voltage is
temporarily elevated above its open-circuit voltage. And conversely when under
discharge, the voltage is temporarily depressed below the open-circuit voltage. This
is primarily due to the internal resistance of the battery. Other reasons are due to
the variation in electrolyte concentration at the immediate vicinity of the pores of the
plates, which is a result of the diffusion of electrolyte from an area of higher
concentration (at the bulk of electrolyte) to an area of lower concentration (near the
plates) while the battery is under discharge. During the charging process the
electrolyte concentration is higher at the plates when compared to the bulk of the
electrolyte. So not only is voltage affected by the state of charge of the battery, but
an added factor is the rate of discharge or charge.
The voltage measured at the battery terminals when the battery is being charged or
discharged is called the terminal voltage. As the rate of charge or discharge is
increased, the deviation from the pure open circuit voltage is increased. For
example, when charging, the voltage measured at the 20 hour rate of charge (C/20)
is always higher than for the 50 hour or 100 hour rate at a given battery SOC.
It is difficult therefore to determine battery state of charge simply by measuring
voltage, because in remote PV power systems the battery may be charging or
discharging. The voltage measured (terminal voltage) may be elevated or
depressed with respect to the accurate measure of SOC, the open circuit voltage.
One practical application of this occurs with charge regulator design in photovoltaic
systems. A typical regulator will allow full array current to flow into the battery bank
until the voltage begins to rise steeply, usually around 14.5 volts for a 12 volt battery.
In Figure 10-13, this corresponds to about 80% SOC. At this steep rise area, the
voltage of the battery could continue to go up, but no progress toward 100% SOC
would occur. This is the region of "gassing", where there is enough voltage to break
up water molecules into hydrogen and oxygen gas. A typical regulator design will
reduce the array current at this point. For example, the regulator might reduce the
current down from the C/20 rate to the C/50 rate. At the lower rate, the voltage
drops below 14.5, and begins to rise more slowly as the battery approaches 100%
SOC. The regulator will continue to decrease the array current as the voltage rises
again, and in this way very gently approach 100% SOC. Referring again to Figure
10-13, the battery depicted will be very near 100% SOC at the C/100 rate and still be
below 14 volts.

10-57
In addition to estimating the state of charge from the battery terminal voltages at
different rates of charge and discharge, these curves can also be used to determine
the voltage set point for the charge controller low voltage load disconnect set point
or alarm settings. For example, if it is desired to limit the maximum battery depth of
discharge to 80% at the 40 hour rate, the corresponding load disconnect voltage
should be about 11.5 volts.
Battery Charging and

Discharging
Voltage (volts)
17
20 hour rate
16
50 hour rate
15
Open Circuit Voltage

100 hour rate
14
40 hour rate
13
5 hour rate
12
11
cut-off voltage
10
0
20
40
60
80
State of Charge (%)
20
40
60
80
100
Depth of Discharge (%)


10-58
Battery Charging
Methods and procedures for battery charging vary considerably. In a stand-alone
PV system the ways in which a battery is charged are generally much different from
the charging methods battery manufacturers use to rate battery performance. The
various methods and considerations for battery charging in PV systems are
discussed in the next chapter on battery charge controllers.
Battery manufacturers often refer to three modes of battery charging: normal or bulk
charge, finishing or float charge and equalizing charge.
Bulk or Normal Charge: Bulk or normal charging is the initial portion of a

charging cycle, performed at any charge rate which does not cause the
cell voltage to exceed the gassing voltage. Bulk charging generally occurs
up to between 80 and 90% state of charge.
Float or Finishing Charge: Once a battery is nearly fully charged, most

of the active material in the battery has been converted to its original form,
and voltage and or current regulation are generally required to limit the
amount over overcharge supplied to the battery. Finish charging is usually
conducted at low to medium charge rates.
Equalizing Charge: An equalizing or refreshing charge is used

periodically to maintain consistency among individual cells. An equalizing
charge generally consists of a current-limited charge to higher voltage
limits than set for the finishing or float charge. For batteries deep
discharged on a daily basis, an equalizing charge is recommended every
one or two weeks. For batteries less severely discharged, equalizing may
only be required every one or two months. An equalizing charge is
typically maintained until the cell voltages and specific gravities remain
consistent for a few hours.
Battery Gassing and Overcharge

Gassing occurs in a battery during charging when the battery is nearly fully charged.
At this point, essentially all of the active materials have been converted to their fully
charged composition and the cell voltage rises sharply. The gas products are either
recombined internal to the cell as in sealed or valve regulated batteries, or released
through the cell vents in flooded batteries.

10-59
In general, the overcharge or gassing reaction in batteries is irreversible, resulting in

water loss. However in sealed lead-acid cells, an internal recombinant process
permits the reforming of water from the hydrogen and oxygen gasses generated
under normal charging conditions, allowing the battery to be sealed and requiring no
electrolyte maintenance. All gassing reactions consume a portion of the charge
current that cannot be delivered on the subsequent discharge, thereby reducing the
battery charging efficiency.
In flooded lead-acid and nickel-cadmium batteries, gassing results in the formation
of hydrogen at the negative plate and oxygen at the positive plate, requiring periodic
water additions to replenish the electrolyte. The following electrochemical reactions
show the overcharge process in typical lead-acid and nickel-cadmium cells.
Lead-Acid Cell Overcharge Reaction

2 H + + 2e H 2
H 2 O 2e 12 O2 + 2 H +
Overall lead-acid cell overcharge reaction:
H 2 O H 2 + 21 O2
Nickel-Cadmium Cell Overcharge Reaction

4H 2 O + 4e 2H 2 + 4OH
4OH 2H 2 O + O2 + 4e
Overall nickel-cadmium cell overcharge reaction:
2H 2 O 2H 2 + O2
Flooded Batteries Require Some Gassing

10-60
Some degree of gassing is required to agitate and prevent stratification of the

electrolyte in flooded batteries. When a flooded lead-acid battery is charged, heavy
sulfuric acid forms on the plates, and falls to the bottom of the battery. Over time the
electrolyte stratifies, developing greater acid concentrations at the bottom of the
battery than at the top. If left unmixed the reaction process would be different from
the bottom to the top of the plates, greater corrosion would occur, and battery
performance would be poor. By gently gassing flooded batteries, the electrolyte is
mixed preventing electrolyte stratification. However, excessive gassing and
overcharge dislodges active materials from the grids, reducing the battery life.
Excessive gassing may also lead to higher temperatures, which accelerates
corrosion of the grids and shortens battery life.
Captive Electrolyte Batteries Should Avoid Gassing

Gassing control is especially important for captive electrolyte or sealed batteries.
These are not flooded, and electrolyte cannot be replaced if allowed to escape due
to overcharging. These batteries do not need for their electrolyte to be mixed, as in
flooded batteries. For these types of batteries, the charging process should be
controlled more carefully to avoid gassing.

10-61
Charge Regulation Voltage Affects

Gassing
The charge regulation voltage, or the maximum voltage that a charge controller
allows a battery to reach in operation plays an important part in battery gassing.
Charge controllers or regulators are used in photovoltaic power systems to allow
high rates of charging up to the gassing point, and then limit or disconnect the PV
current to prevent overcharge. The highest voltage that batteries are allowed to
reach determines in part how much gassing occurs. To limit gassing and electrolyte
loss to acceptable levels proper selection of the charge controller voltage regulation
set point is critical in PV systems. If too low of a regulation voltage is used, the
battery will be undercharged. If too high of a regulation voltage is used, the battery
will be overcharged. Both under and overcharging will result in premature battery
failure and loss of load in stand-alone PV systems. In general, sealed maintenance
free valve-regulated batteries (using lead-calcium grids) should have lower charge
regulation voltage set points than flooded deep cycling batteries (using leadantimony grids).
Other Factors Affecting Battery Gassing

The onset of gassing in a lead-acid cell is not only determined by the cell voltage,
but by the temperature as well. As temperatures increase, the corresponding
gassing voltage decreases for a particular battery. Regardless of the charge rate,
the gassing voltage is the same, however gassing begins at a lower battery state of
charge at higher charge rates. The grid design, whether lead-antimony or leadcalcium also affects gassing. Battery manufacturers should be consulted to
determine the gassing voltages for specific designs. The figure on the following
page shows the relationships between cell voltage, state of charge, charge rate and
temperature for a typical lead-acid cell with lead-antimony grids.

10-62
Lead-Acid Battery Charging Voltage as a

Function of State of Charge
3.0
2.9
Lead-Antimony Grids
Charge Rate
Cell Voltage (volts)
2.8
C/2.5
2.7
2.6
Gassing Voltage at 0
C/5
2.5
C/20
Gassing Voltage at 27 oC
2.4
2.3
2.2
2.1
2.0
0
20
40
60
80
100
Battery State of Charge (%)

By examining the figure, one can see that at 27o C and at a charge rate of C/20, the
gassing voltage of about 2.35 volts per cell is reached at about 90% state of charge.
At a charge rate of C/5 at 27o C, the gassing voltage is reached at about 75% state
of charge. At a battery temperature of 0o C the gassing voltage increases to about
2.5 volt per cell, or 15 volts for a nominal 12-volt battery. The effects of temperature
on the gassing voltage is the reason the charge regulation voltage is sometimes
temperature compensated - to fully charge batteries in cold weather and to limit
overcharge during warm weather. This type of information is needed to properly
select battery charge controller voltage regulation set points in order to limit the
amount of gassing for a specific battery design and operational conditions.
Some recommended ranges for charge regulation voltages (at 25o C) for different
battery types used in PV systems are presented in Table 10-4 below. These values
are typical of voltage regulation set points for battery charge controllers used in
small PV systems. These recommendations are meant to be only general in nature,
and specific battery manufacturers should be consulted for their suggested values.

10-63
Battery Type
Charge
Regulation
Voltage at 25 oC
Flooded
LeadAntimony
Flooded
LeadCalcium
Sealed, Valve
Regulated
Lead-Acid
Flooded
Pocket Plate
NickelCadmium
Per nominal 12
volt battery
14.4 - 14.8
14.0 - 14.4
14.0 - 14.4
14.5 - 15.0
Per Cell
2.40 - 2.47
2.33 - 2.40
2.33 - 2.40
1.45 - 1.50

The charge regulation voltage ranges presented in Table 10-4 are much higher than
the typical charge regulation values often presented in manufacturers literature.
This is because battery manufacturers often speak of regulation voltage in terms of
the float voltage, or the voltage limit suggested for when batteries are float charged
for extended periods (for example, in non-interruptible power supply (UPS) systems).
In these and many other commercial battery applications, batteries can be trickle or
float charged for extended period, requiring a voltage low enough to limit gassing.
Typical float voltages are between 13.5 and 13.8 volts for a nominal 12-volt battery,
or between 2.25 and 2.30 volts for a single cell.
In a PV system however, the battery must be recharged within a limited time (usually
during sunlight hours), requiring that the regulation voltage be much higher than the
manufacturers float voltage to ensure that the battery is fully recharged. If charge
regulation voltages in a typical PV system were set at the manufacturers
recommended float voltage, the batteries would never be fully charged.

10-64
Exercises

.!

/

0

1

! 2 ! 3

9:

;
$

!

!

-++

!

/

!

! 3 !

! -<5 !

3
=

*

!

10-65
Balancing Life Cycle and Initial Costs

First costs often determine which battery is selected by a designer; however,
recurring costs, including battery maintenance and replacement are burdened by the
system owner/operator. The sum of the initial cost and amortized recurring costs
are what determine a battery's life cycle cost. In many cases, a battery with high
initial costs may have the lowest life cycle costs, all factors being considered. This is
sometimes the case in the use of higher cost nickel-cadmium cells used for critical
applications.
When deciding between the lower cost of shallow cycling batteries, or the higher
initial cost of deeper cycling batteries, photovoltaic system designers are faced with
the common decision of pay now or pay later. There is not one perfect battery
choice for all users and applications, and paying now may be best in one situation
while paying later may be more appropriate in another.
Battery Trade-Off
Pay Now or Pay Later
Pay less initially for starting type batteries, but get short life (1-2 years)
Pay more initially for deep discharge batteries, and get longer life (4-8
years)
Starting types may be more available, but may not be the best value in
the long run
For low-income rural villagers, the high cost of a better longer life deep cycling
battery may be prohibitive. But they may be able to afford a lower cost, more readily
available and replaceable shallow cycling, starting type battery. In a year or two,
they can afford to buy a replacement. For them, the more affordable choice may be
to not pay now for the best (longest life) but to pay later (for replacements).
On the other hand, a remote telecom site in a difficult or dangerous location would
be best designed with the longest life battery possible. The cost of poor
performance would be much greater than the small savings of a cheaper battery.
The budget would probably be present for purchasing the better longer life battery
initially. The cost savings in reduced maintenance and replacement costs over time
would be of primary consideration in this case.

10-66
Factors Affecting Battery Life

Lower than expected battery life is a common problem in small stand-alone PV
applications. The battery bank is typically the second most expensive component in
most photovoltaic power systems, next to the PV array. Over a twenty to thirty year
life of a PV system, the replacement and maintenance costs for batteries may be the
highest life cycle component in a PV system. For these reasons, PV system
designers and users should have a good understanding of the issues affecting
battery life.
Battery lifetime is dependent upon a number of cell design and operational factors,
including the components and materials of battery construction, temperature,
frequency and depth of discharges, average state of charge and charging methods.
As long as a battery is not overcharged, overdischarged or operated at excessive
temperatures, the lifetime of a battery is proportional to its average state of charge.
A typical flooded lead-acid battery that is maintained above 90% state of charge will
provide two to three times more full charge/discharge cycles than a battery allowed
to reach 50% state of charge before recharging. This suggests limiting the
maximum allowable and average daily DOD to prolong battery life. The actual point
that determines the end of life for a battery is arbitrary, but most often it is assumed
to be when the battery will no longer deliver 80% of its full rated capacity. It is
important that the PV system designer understand the consequences of capacity
loss as a battery ages, as this may affect the reliability of system operation.
Lifetime can be expressed in terms of cycles or years, depending on the particular
type of battery and its intended application. Exact quantification of battery life is
difficult due to the number of variables involved, and generally requires battery test
results under similar operating conditions. Often, battery manufacturers do not rate
battery performance under the conditions of charge and discharge experienced in
PV systems. This makes the accurate estimation of battery life in PV systems very
difficult, if not impossible. However, there are well known operating practices and
procedures that tend to maximize battery life. A discussion of these considerations
follows.
What Maximizes Battery Life?

There are several general guidelines that can lead to maximizing the useful life of
batteries in PV systems. Each of these is reviewed next.
Cool operating temperatures

Shallow depth of discharge
Prevention of overcharging and discharging
Proper and frequent maintenance and water additions
Full recharges after discharging as soon as possible

10-67
Temperature Affects Battery Life

Temperature has important effects on battery life that are important to PV system
designers. The main enemy to battery life is operation at high temperatures. In
general, as the temperature increases by 10 oC, the rate of an electrochemical
reaction doubles, resulting in the recommendations from battery manufacturers that
battery life decreases by a factor of two for every 10 oC increase in average
operating temperature. Higher operating temperatures accelerate corrosion of the
positive plate grids, result in greater gassing and electrolyte loss, and shortened
battery life. Lower operating temperatures generally increase battery life, however
the capacity is reduced significantly, particularly for lead-acid batteries. Where
severe temperature variations from room temperatures exist, battery are sometime
located in an insulated or other temperature regulating enclosure to minimize battery
temperature swings.
Effects of Temperature on Battery Life
Battery Life
(% life at 25 oC)
1000
Lead-Antimony Grids
Lead-Calcium Grids
Nickel-Cadmium
100
10
5
10
15
20
25
30
35
40
45
Battery Operating Temperature (oC)


10-68
As sown in the figure, the grid alloying elements play an important part in battery life.
The lifetime of batteries with lead-calcium grids is more affected by temperature than
lead-antimony grids or nickel-cadmium batteries.
The importance of temperature can be seen by looking at an example of a
manufacturers warranty statement. A portion of the warranty for GNB Absolyte IIP
batteries is presented on the next page. The portion of interest to us now is
highlighted below.
The wording of the warranty statement has been condensed to focus attention on
the temperature factor.
GNB also warrants ...that GNB ABSOLYTE IIP batteries used in photovoltaic
service will furnish... 80% of the specified capacity...for m years...provided that the
following conditions are satisfied.
.....The average daily ambient temperature in the area of use in any year is
established:
not to exceed 25 degrees C
m equal to 10 years
m equal to 7 years
m equal to 5 years
Notice that the warranty period decreases substantially as the average operating
temperature of the battery increases. Notice that with a temperature change from
25 to 35 oC the warranty period decreases by 50%, as per the general rule stated
previously.

10-69

+ ,!"#$%-,!$'$(,!

10-70
Effects of Discharge Depth on Battery Life

The battery depth of discharge also affects battery life. If the discharge depth is kept
shallow (less than 25%), then the adhesion of the active materials to the grids
remains strong, prolonging the life and performance of the battery. In PV systems,
the average daily depth of discharge is generally low because the battery capacity is
sized to provide a large number of days of reserve or autonomy. Typically, the
battery bank in PV system is designed to provide 5 to 10 days of reserve operation
during the lowest insolation to load ratio period, so the average daily depth of
discharge is generally only 10 to 20% of the total installed capacity. The more days
of reserve or autonomy that are designed in to the battery, the shallower the average
daily depth of discharge is, and the longer the battery life. Of course, the greater
capacity installed to give more reserve days results in higher initial costs. Once
again there is a tradeoff of initial cost against long life.
Effects of Depth of Discharge on

Lead-Acid Battery Cycle Life
Battery Life
(Cycles)
10000
1000
100
Motive Power Battery (deep cycle)
Automotive (SLI) Battery (shallow cycle)
10
0
20
40
60
80
100
Battery Depth of Discharge (%)


10-71
Overcharging Shortens Battery Life

Overcharging leads to excessive gassing and loss of electrolyte. The gas bubbles
generated can actually scrub the outside of the plates as they rise, and can
accelerate the erosion of active material from the plates, reducing battery life.
Excessive heating during overcharging accelerates the natural corrosion process.
Corrosion is the electrochemical activity resulting from the immersion of two
dissimilar metals in an electrolyte, or the direct contact of two dissimilar metals,
causing one material to undergo oxidation, or lose electrons, and the causing the
other material to undergo reduction, or gain electrons. Corrosion of the grids
supporting the active material in a battery is an ongoing process, and may ultimately
dictate the battery's useful lifetime.
Overcharging is prevented in PV systems by using a charge controller or regulator
between the PV array and batteries that limits battery gassing, the maximum voltage
and state of charge.
Maintenance Affects Battery Life

Any component of a system will last longer if it is frequently maintained, and this
certainly applies to batteries. Keeping the electrolyte level topped up means the
acid concentration is kept constant and at the correct value, and that the plates are
always fully immersed in liquid. If the electrolyte level is allowed to drop below the
top of the plates, corrosion and loss of capacity and life are accelerated greatly.
Keep the electrical terminals clean and free of corrosion, and make sure the battery
cables are firmly connected to the terminals. Loose or corroded terminals
connections can lead to voltage drops and hot spot heating, and batteries can
experience unequal charging. With unequal voltage potentials across batteries,
ones with good connections will be overcharged while ones with bad connections will
be undercharged. Both of these conditions will lead to loss of battery life over time.
Fully Recharging Affects Battery Life

If lead-acid batteries are left at partial state of charge for weeks, sulfation will set in
and capacity can be permanently lost. Sulfate crystals can grow and warp or short
out the plates. To the degree that systems can be designed to fully recharge the
batteries, the battery life will be extended.

10-72
Battery Efficiency
Battery efficiency is dependent on the type of battery, charging methods, rates of
charge and discharge, depth of discharge and temperature. In general, the
efficiency of a battery is much greater at lower states of charge than when the
battery is nearly fully charged. The total or round-trip battery energy efficiency is
composed of two types of efficiencies. There is a voltage or voltaic efficiency, and a
charge or coulombic efficiency. They measure different characteristics of batteries,
and are often confused.
Battery Voltage (Voltaic) Efficiency

The voltaic efficiency of a battery is determined by the charge and discharge rates
and the battery temperature. The voltaic efficiency is expressed as the ratio of the
battery voltage under discharge to the voltage under charge. High rates and low
temperatures act to decrease battery voltaic efficiency.
Battery Voltaic Efficiency =
Voltage During Discharge

Voltage During Charge
Recall how battery voltage increases during charging and decreases during
discharging as compared to the open-circuit voltage. Battery voltage in a PV system
may vary considerably, depending on state of charge and rate of charge or
discharge. An overall average voltaic efficiency can be calculated by assuming that
on the average, a battery is charged at about 14 volts and discharged at about 12
volts at the given charge and discharge rates in the system. This approximation
yields an overall voltaic efficiency of about 85%.
Approximate Voltaic Efficiency =

10-73
12 volts
= 0.85 = 85%
14 volts
The charge and discharge curves presented below illustrate battery voltaic
efficiency. During charging the voltage is elevated above the open circuit voltage,
and during discharging the voltage is depressed below it. Since the rate of charge
or discharge determines the battery voltage at a certain state of charge, the voltaic
efficiency is higher at lower rates of charge and discharge.
Battery Voltaic Efficiency

16.0
Charge Rates
C/5
Battery Voltage
15.0
C/20
14.0
13.0
C/20
Open-Circuit Voltage
12.0
11.0
Discharge Rates
C/5
10.0
0
20
40
60
80
100


10-74
Battery Charge (Coulombic) Efficiency

The coulombic or charge efficiency of a battery is defined as the ratio of the amperehours withdrawn from a battery during discharge to the ampere-hours provided to
recharge the battery. Due to electron losses from gassing and other internal
mechanisms, a battery can not deliver as many amp-hours as it takes to charge it at
the same rate. At low states of charge, a battery accepts current readily, there is
little gassing and the coulombic efficiency is high. As a battery nears full state of
charge, gassing and internal heating tend to reduce the coulombic efficiency.
Battery Coulombic Efficiency =
Discharge Amp - Hours Output

Charge Amp - Hours Input
Typical industrial applications using motive power or traction batteries must fully
recharge deeply discharged batteries in a limited time, usually less than 8 to 10
hours. To accomplish this, high charge rates (C/5 to C/10) are required resulting in
low coulombic efficiencies. In PV systems, the coulombic efficiency of batteries is
generally very good due to the relatively low charge rates used. Typical charge rates
in PV systems are often C/20 or lower, because the amount of battery storage
required for autonomy is relatively large with respect to the PV array charging
currents. While an accurate estimation of coulombic efficiency is difficult to
determine, 90% is a typical value for most batteries used in small stand-alone PV
systems.
Using Coulombic Efficiency in Sizing Calculations

The coulombic efficiency is used to calculate the PV array size required in a
photovoltaic system. In the load estimation chapter, we showed how to calculate an
average daily load demand in units of ampere-hours. The PV array must not only be
sized to meet the load amp-hour demand, but must also overcome the battery
coulombic inefficiency. To determine the total amp-hours that the PV array must
produce, we divide the total load demand by the estimated battery coulombic
efficiency.
Amp - Hours PV Array Must Produce =

10-75
Daily Load Demand in Amp - Hours

Battery Coulombic Efficiency
Example:
The load demand for a remote home is estimated to be 200 Ah/day at

12 volts. Using an average battery coulombic efficiency of 90%, the
total Ah that the array must produce is given by:
Ah From Array
200 Ah Daily Load Demand

0.90
222 Ah
Battery Round-trip or Energy

Efficiency
The energy or round-trip efficiency of a battery is defined as the product of the
coulombic and voltaic efficiencies. This efficiency defines the ratio of the energy
withdrawn from a battery during discharge to the amount of energy supplied to bring
a battery back to full state of charge.
Battery Energy
Efficiency
Voltaic Efficiency X Coulombic Efficiency
.85
.80 or 80%
.95
The round-trip battery energy efficiency is often used when discussing batteries.
However it is the coulombic efficiency that is most commonly used in photovoltaic
system sizing methods using the amp-hour approach. In cases where energy or
watt-hour approach is used in system sizing, the energy or round-trip battery
efficiency is sometimes used to account for the voltage as well as the charge
efficiency of the battery, especially for quantifying losses due to gassing. For the
sizing calculations presented in the chapter on System Sizing, the only battery
efficiency that needs to be considered will be the charge or coulombic efficiency,
which is usually about 90% for most batteries.

10-76
Exercise

6

-&!

4
3

&+
-++
>
8
-&
8
" $ ! !

?+,
)))))))))) "


10-77
Battery Selection Criteria

The selection of a battery for use in PV systems involves many decisions and trade
offs, and depends on many factors. While no specific battery is appropriate for all
PV applications, common sense and a careful review of the battery literature with
respect to the particular application needs will help the designer narrow the choice.
Some decisions on battery selection may be easy to arrive at, such as physical
properties, while other decisions will be much more difficult and may involve making
tradeoffs between desirable and undesirable battery features. With the proper
application of this knowledge, designers should be able to differentiate among
battery types and gain some application experience with batteries they are familiar
with. The list below summarizes some of the key considerations in battery selection.
Nominal system voltage

Charge regulation requirements
Required capacity or autonomy
Ampere-hour capacity at discharge rate
Daily and maximum depth of discharge
Self-discharge rate
Gassing characteristics
Efficiency
Temperature effects
Size, weight and structural needs
Susceptibility to freezing
Susceptibility to sulfation, stratification
Electrolyte type, concentration
Auxiliary equipment
Maintenance requirements
Terminal configurations
Battery life (cycles/years)
Availability and shipping requirements
Cost and warranty

10-78
Desirable Battery Construction Features

Selection of a durable battery design is as equally important as the proper sizing and
treatment in improving the longevity of batteries in PV systems. A few of the
important battery design characteristics promoting deep-cycle, long-life performance
are listed as follows:
Low-antimony, thick plates for reducing water loss and providing adequate
mechanical strength for long cycle life.
Separator envelopes around the positive plates, and appropriate plate edge
protection to minimize internal short-circuit potential.
Large electrolyte volume below plates to allow for shed materials to accumulate
without causing short circuits.
Sufficient electrolyte volume above plates to minimize frequency of water

additions.
Transparent battery containers to allow for visual inspections.
Size and Weight

Due to their large size and weight, the physical characteristics of batteries often
place restrictions on battery selection. A typical 12 volt, 100 ampere-hour flooded
lead-acid battery weights anywhere between 20 and 40 kilograms, depending on the
weight of active material, grid and interconnect design and amount of electrolyte.
Special lifting devices may be required for large capacity batteries.
For remote PV applications where the transportation of heavy equipment is
cumbersome or otherwise difficult, the designer may choose to use several smaller
batteries or cells rather than a few larger batteries. The physical properties of
batteries also require consideration of the strength and dimensions of the enclosure
or area in which they are to be installed.
Availability
Due to the weight and hazardous nature of batteries, shipping costs are high, and
results in most batteries produced at regional plants being used to supply to local
markets. In developing regions, the optimal battery for a given application may not
be locally available. In these cases, the best battery available should be used, with
special attention paid to properly sizing and treating the battery for maximum life.

10-79
Battery Auxiliary Equipment

Battery auxiliary equipment includes any systems or other hardware necessary to
safely and effectively operate a battery system. Some of the more important battery
auxiliary systems and equipment are discussed below.
Ventilation
Batteries often produce toxic and explosive mixtures of gasses, namely hydrogen,
and adequate ventilation of the battery enclosure is required. In most cases,
passive ventilation techniques such as vents or ducts may be sufficient. In some
cases, fans may be required to provide mechanical ventilation. Required air change
rates are based on maintaining minimum levels of hazardous gasses in the
enclosure. Under no circumstances should batteries be kept in an unventilated area
or located in an area frequented by personnel.
Catalytic Recombination Caps

A substitute for standard vented caps on lead-antimony batteries, catalytic
recombination caps (CRCs) primary function is to reduce the electrolyte loss from
the battery. CRCs contain particles of an element such as platinum or palladium,
which surfaces adsorb the hydrogen generated from the battery during finishing and
overcharge. The hydrogen is then recombined with oxygen in the CRC to form
water, which drains back into the battery. During this recombination process, heat is
released from the CRCs as the combination of hydrogen and oxygen to form water
is an exothermic process. This means that temperature increases in CRCs can be
used to detect the onset of gassing in the battery. If CRCs are found to be at
significantly different temperatures during recharge (meaning some cells are gassing
and others are not), an equalization charge may be required. The use of CRCs on
open-vent, flooded lead-antimony batteries has proven to reduce electrolyte loss by
as much as 50% in subtropical climates.
Battery Monitoring Systems

Monitoring and instrumentation for battery systems can range from simple analog
meters to more sophistication data acquisition systems. Lower level monitoring of
battery systems might include voltage and current meters or battery state of charge
indicators, while higher level monitoring may include automated recording of voltage,
current, temperature, specific gravity and water levels. For small stand-alone PV
systems, monitoring of the battery condition is generally done only occasionally
during routine maintenance checks, or by simple meters or indicators on the battery
charge controller.

10-80
Enclosures
Batteries are generally required by local electrical codes and safety standards to be
installed in an enclosure separated from controls or other PV system components.
The enclosure may also be insulated, or may have active or passive cooling/heating
mechanisms to protect the batteries from excessive temperatures. Battery
enclosures must be of sufficient size and strength hold the batteries, and can be
located below ground if needed to prevent freezing. If the enclosure is located
above ground, care should be taken to limit the direct exposure to sunlight, or some
type of shading or reflective coating should be provided.
Passive Cooling Enclosures

We have shown that temperature is a critical factor affecting battery performance
and life expectancy. Any actions taken by the system designer to reduce
temperature swings will be rewarded with better battery performance, longer life, and
lower maintenance.
One approach to moderating the influence of ambient temperature swings on battery
temperature is the use of passive cooling enclosures, without the need for active
components such as motors, fans or air conditioners. The use of active temperature
regulation means generally requires additional electrical power, and adds
unnecessarily to the complexity, size and cost of the PV system. By using a
thermodynamically passive approach, maximum benefits are gained with minimal
complexity and maximum reliability -- key features of any PV system installation.
An example of a passive cooling enclosure is the Zomeworks Cool Cell TM that is
designed for small to medium sized battery banks (illustrated on the next page).
During the heat of the day, cool water in the enclosure absorbs heat from the
ambient air and batteries, effectively moderating battery temperature. During the
night, the warmed water naturally rises to a specially designed lid where the heat is
radiated from the lid to the cool night sky. As the water in the lid cools, the natural
thermosiphon action allows the heavier cooled water to circulate back into the
enclosure as it is replaced by warmer, less dense water from the enclosure. This
circulation loop relies on the density differences between warm and cool water,
eliminating the need for a circulation pump.

10-81

",!'$$ '$
The performance of the Cool CellTM depends on the ambient temperature and the
extent and type of cloud cover. In areas with exceedingly high ambient
temperatures, this enclosure can provide a significant reduction in battery
temperatures compared to a standard box or insulated enclosure. Cloud cover plays
an important effect on the performance of the Cool CellTM as this affects the
radiation heat transfer from the lid. Clear skies generally have an effective
temperature (for purposes of radiation calculations) of below -40 oC, however cloudy
skies have a much higher effective temperature. Since the radiation heat transfer is
proportional to the effective temperature raised to the fourth power, overcast skies
can reduce the cooling effectiveness of this type of battery enclosure.

10-82
The relative performance of the Cool Cell is illustrated below. Four conditions are
compared: (a) the widely varying case of an uninsulated steel box; (b) the ambient
air; (c) a concrete vault with an insulated lid; and (d) the moderated temperature
swings of the passively cooled enclosure.


10-83
Battery Safety Considerations

Due to the hazardous materials and chemicals involved, and the amount of electrical
energy which they store, batteries are potentially dangerous and must be handled
and used with caution. Typical batteries used in stand-alone PV systems can deliver
up to several thousand amps under short-circuit conditions, requiring special
precautions. Depending on the size and location of a battery installation certain
safety precautions are required.
IMPORTANT:
Exercise extreme caution and follow recommended practices when
working with batteries!
Handling Electrolyte
The caustic sulfuric acid solution contained in lead-acid batteries can destroy
clothing and burn the skin. For these reasons protective clothing such as aprons
and face shields should be worn by personnel working with batteries. To neutralize
sulfuric acid spills or splashes on clothing, the spill should be rinsed immediately
with a solution of baking soda or household ammonia and water. For nickelcadmium batteries, the potassium hydroxide electrolyte can be neutralized with a
vinegar and water solution. If electrolyte is accidentally splashed in the eyes, the
eyes should be forced open and flooded with cool clean water for fifteen minutes. If
acid electrolyte is taken internally, drink large quantities of water or milk, followed by
milk of magnesia, beaten eggs or vegetable oil. Call a physician immediately.
If it is required that the electrolyte solution be prepared from concentrated acid and
water, the acid should be poured slowly into the water while mixing. The water
should never be poured into the acid. Appropriate non-metallic funnels and
containers should be used when mixing and transferring electrolyte solutions.

10-84
Personnel Protection
When performing battery maintenance, personnel should wear protective clothing
such as aprons, ventilation masks, goggles or face shields and gloves to protect
from acid spills or splashes and fumes. If sulfuric acid comes into contact with skin
or clothing, immediately flush the area with a solution of baking soda or ammonia
and water. Safety showers and eye washes may be required where batteries are
located in close access to personnel. As a good practice, some type of fire
extinguisher should be located in close proximity to the battery area if possible. In
some critical applications, automated fire sprinkler systems may be required to
protect facilities and expensive load equipment. Jewelry on the hands and wrists
should be removed, and properly insulated tools should be used to protect against
inadvertent battery short-circuits.
Dangers of Explosion
During operation, batteries may produce explosive mixtures of hydrogen and oxygen
gasses. Keep spark, flames, burning cigarettes, or other ignition sources away from
batteries at all times. Explosive gasses may be present for several hours after a
battery has been charged. Active or passive ventilation techniques are suggested
and often required, depending on the number of batteries located in an enclosure
and their gassing characteristics. The use of battery vent caps with a flame arrestor
feature lowers the possibility of a catastrophic battery explosion. Improper charging
and excessive overcharging may increase the possibility of battery explosions.
When making and breaking connections to a battery from a charging source or
electrical load, ensure that the charger or load is switched off as to not create sparks
or arcing during the connection.
Battery Disposal and Recycling

Batteries are considered hazardous items as they contain toxic materials such as
lead, acids and plastics that can harm humans and the environment. For this
reason laws have been established which dictate the requirements for battery
disposal and recycling. In most areas batteries may be taken to the local landfill,
where they are in turn taken to approved recycling centers. In some cases battery
manufacturers provide guidelines for battery disposal through local distributors and
may in fact recycle batteries themselves. Under no circumstances should batteries
be disposed of in landfills, the electrolyte allowed to seep into the ground, or the
battery burned.

10-85
Selected References
Institute of Electrical and Electronics Engineers, "IEEE Recommended Practice for Installation and
Operation of Lead-Acid Batteries for Photovoltaic (PV) Systems", ANSI/IEEE Std. 937-1987, New
York, NY, 1987.
Institute of Electrical and Electronics Engineers, "IEEE Recommended Practice for Installation and
Operation of Nickel-Cadmium Batteries for Photovoltaic (PV) Systems", ANSI/IEEE Std. 11451990, New York, NY, 1990.
Stand-Alone Photovoltaic Systems - A Handbook of Recommended Design Practices, Sandia National
Laboratories, SAND87-7023, revised November 1991.
Naval Facilities Engineering Command, Maintenance and Operation of Photovoltaic Power Systems,
NAVFAC MO-405.1, December 1989.
Exide Management and Technology Company, Handbook of Secondary Storage Batteries and Charge
Regulators in Photovoltaic Systems - Final Report, for Sandia National Laboratories,
SAND81-7135, August 1981.
Bechtel National, Inc., Handbook for Battery Energy Storage in Photovoltaic Power Systems, Final Report,
SAND80-7022, February 1980.
S. Harrington and J. Dunlop, "Battery Charge Controller Characteristics in Photovoltaic Systems",
Proceedings of the 7th Annual Battery Conference on Advances and Applications, Long Beach,
California, January 21, 1992.
H.A. Kiehne, Battery Technology Handbook, Marcel Dekker, Inc., 1989.
G.W. Vinal, Storage Batteries, John Wiley & Sons, Inc., Fourth Edition, 1954.
D. Linden, Handbook of Batteries and Fuel Cells, McGraw Hill, Inc., 1984.

10-86
(End of Chapter)

10-87
CHAPTER TEN
BATTERY TECHNOLOGY
10-1
Purpose of Batteries in Photovoltaic
10-2
Systems
Energy Storage Capacity and Autonomy
Voltage and Current Stabilization
Supply Surge Currents
10-2
10-3
10-4
10-5
Battery Design and Construction

Cell
Active Material
Electrolyte
Grid
Plate
Separator
Element
Terminal Posts
Cell Vents
Case
10-6
10-6
10-6
10-7
10-8
10-8
10-8
10-9
10-9
10-9
10-9
Battery Types and Classifications

Primary Batteries
Secondary Batteries
10-10
10-10
10-10
Lead-Acid Battery Classifications 5.3.3.1
10-12
Lead-Acid Batteries
SLI Batteries
Motive Power or Traction Batteries
Stationary Batteries
10-12
10-12
10-12
10-12
Types of Lead-Acid Batteries

Lead-Antimony Batteries
Lead-Calcium Batteries
Lead-Antimony/Lead-Calcium Hybrid
Captive Electrolyte Lead-Acid Batteries
10-13
10-13
10-14
10-14
10-16

10-88
Lead-Acid Battery Chemistry

Lead-Acid Cell Reaction
Formation
Stratification
Specific Gravity
Sulfation
10-18
10-18
10-19
10-19
10-20
10-22
Nickel-Cadmium Batteries
Sintered Plate Ni-Cads
Pocket Plate Ni-Cads
Nickel-Cadmium Battery Chemistry
10-23
10-23
10-23
10-24
Battery Strengths and Weaknesses
10-27
Battery Cycling
Battery Discharging
Charge/Discharge Rates Expressed as Time
Depth of Discharge (DOD)
Autonomy Determines Capacity and Depth of Discharge
State of Charge (SOC)
Temperature Limits Discharge Depth
Shallow and Deep Cycle Batteries
10-28
10-29
10-30
10-31
10-32
10-35
10-36
10-39
Battery Capacity
Ampere-Hour Definition
Factors Affecting Battery Capacity
Effects of Temperature on Capacity
Discharge Rate Affects Capacity
Cut-Off Voltage Affects Capacity
Effects of Self Discharge Rate on Capacity
10-41
10-41
10-41
10-42
10-44
10-51
10-52
Voltage Changes During Charging and Discharging

Battery Voltage and State of Charge are Related
Battery Voltage Depends on Rate of Charge or Discharge
10-54
10-55
10-57
Battery Charging
Battery Gassing and Overcharge
Charge Regulation Voltage Affects Gassing
10-59
10-59
10-62
Balancing Life Cycle and Initial Costs
10-66

10-89
Factors Affecting Battery Life

What Maximizes Battery Life?
Temperature Affects Battery Life
Effects of Discharge Depth on Battery Life
Overcharging Shortens Battery Life
Maintenance Affects Battery Life
Fully Recharging Affects Battery Life
10-67
10-67
10-68
10-71
10-72
10-72
10-72
Battery Efficiency
Battery Voltage (Voltaic) Efficiency
Battery Charge (Coulombic) Efficiency
Battery Round-Trip or Energy Efficiency
10-73
10-73
10-75
10-76
10-78
Battery Auxiliary Equipment

Ventilation
Catalytic Recombination Caps
Battery Monitoring Systems
Enclosures
10-80
10-80
10-80
10-80
10-81
Battery Safety Considerations

Handling Electrolyte
Personnel Protection
Dangers of Explosion
Battery Disposal and Recycling
10-84
10-84
10-85
10-85
10-85
Selected References
10-86

10-90
Chapter 10 Answers
Battery Technology

a. Positive Plate
PbO2 + 4H+ + 2e- Pb2+ + 2H2O
Element / Ions
Lead (Pb)
Oxygen (O)
Hydrogen (H)
Ion Charges
Qty on Left
1
2
4
+4 -2 = +2
Qty on Right
1
2
4
+2
Qty on Left
1
4
1
+2 -2 = 0
Qty on Right
1
4
1
0
Qty on Left
1
0
Qty on Right
1
+2 -2 = 0
Qty on Left
1
4
1
+2 -2 = 0
Qty on Right
1
4
1
0
Pb2+ + SO42- PbSO4

Element / Ions
Lead (Pb)
Oxygen (O)
Sulfur (S)
Ion Charges
b. Negative Plate
Pb Pb2+ + 2eElement / Ions
Lead (Pb)
Ion Charges
Pb2+ + SO42- PbSO4

Element / Ions
Lead (Pb)
Oxygen (O)
Sulfur (S)
Ion Charges
c. Overall lead - acid cell reaction

PbO2 + Pb + 2H2SO4 2PbSO4 + 2H2O
Element / Ions
Lead (Pb)
Oxygen (O)
Sulfur (S)
Hydrogen (H)
Qty on Left
2
10
2
4
10-1
Qty on Right
2
10
2
4
Battery Technology

For a 24-volt (nominal) battery, we need
24 volts / 2.1 volts per cell = 11.4 = 12 lead acid cells
24 volts / 1.2 volts per cell = 20 nickel cadmium cells
Note: in most cases a lead-acid cell is assumed to have a nominal voltage of 2.0 volts.

We assume the nominal voltage of a lead-acid cell is 2 volts. Then 24 cells will have a
total nominal voltage of 24 X 2 Volts = 48 Volts.

b. Lower than

b. Left at partial state of charge for a long period


b. Less

a. A smaller

Referring to the chart of Temperature Limits for Maximum Depth of Discharge, we see
that at a temperature of -25 C, the maximum recommended depth of discharge is
approximately 45%.
10-2
Battery Technology

Referring to the chart of Temperature Limits for Maximum Depth of Discharge, we see
that at a temperature of -40 C, the maximum recommended depth of discharge is
approximately 25%.
At first, we might be tempted to answer that we need a battery that is four times as
large. However, this would only be true if we assumed that we could use 100% of the
original battery. Since the maximum recommended depth of discharge is 80% at
normal temperature, the battery for the mountain application needs to be larger by a
factor of
80% = 3.2 times larger
25%
An example helps make this clearer. Assume we need 800 Ahr of storage. A 1000 Ahr
battery designed for use at normal temperatures can provide 800 Ahr storage to 80%.
Now, if we can only discharge the battery to 25% DOD, the battery must provide the
same 800 Ahr capacity. Using the ratio of 3.2 that we calculated above, we choose a
3,200 Ahr battery. Notice that 25% discharge = 25% X 3200 = 800 Ahrs as is required.
Since low temperatures can have a significant impact on the required battery size, we
would be very interested in any methods that might keep the battery from getting cold.
Some possible techniques to do this might include:
Keeping the batteries in a heated area (e.g. an equipment shelter)
Locating the batteries in a buried vault
Providing insulation to minimize the temperature changes


a. 0 oC and C/120?
92%
b. 0 oC and C/50?
85%
c. -10 oC and C/120?
90%
d. 25 oC and C/500?
100%
e. 10 oC and C/120?
95%
These values are taken from the chart showing Temperature and Discharge effects on
Lead Acid Capacity.
10-3
Battery Technology

Referring to the chart, we see that at -15 C and C/500, the battery will have about 90%
of its standard rated capacity (at 25 C). The increased capacity required is
1500 Ahr
0.90
1667 Ahr at 25 C

At 5 C and C/50, the battery will have about 87% of its standard rated capacity (at 25
C). The increased capacity required is
200 Ahr
0.87
230 Ahr at 25 C


At 12 volts, the current required by the loads is:
Lights
20 watts / 12 volts = 1.7 amps
Refrigerator 100 watts / 12 volts = 8.3 amps
System Description:
12 Volt Rural Electrification System
DC Loads
Lights
Refrigerator
:
:
Qty.
2
1
X
X
Amps
1.7
8.3
X
X
Hours
/day
6
12
DC Loads (Ah)
Ah From Array
120.4 Ah Daily Load Demand

0.90 coulombic efficiency
133.8 Ah
10-4
=
=
Daily
Demand (Ah)
20.4
100.0
120.4
Battery Technology
Chapter Eleven
Inverter Technology
The photovoltaic array and battery produce DC current and voltage. If the loads require
AC power an inverter can be used to convert from DC to AC. Commonly available
inverters can output in 1- or 3-phase, 50 or 60 hertz, and 117 or 220 volts and can
range in continuous output power from a few hundred watts to thousands of kilowatts.
Large utility scale inverters are made to output at 480 volts AC or higher and have
capacities exceeding 1000 kilowatts.
In this section we will examine some of the technical features and characteristics of
inverter technology commonly used in photovoltaic systems. The three most important
characteristics of inverters that we will discuss will be:
Output power and surge power
Output efficiency
Output waveform

11-1
Components Inverter Technology
Inverter Power Rating

Power Is A Function Of Time
An inverter's capability to output power depends on its ability to dissipate the heat
generated within the inverter. An inverter's rated output power is generally the power
level it can maintain for a long period. An inverter can put out greater power than its
nominal rating before it overheats and shuts off, but only for a short period of time.
Some inverters can output a great deal more power than their nominal rating for a few
minutes, allowing very large loads to operate, for example large power tools and or
pump motors.
However, there is no standard period of time accepted by all manufacturers for inverter
output rating. There is a trend for manufacturers to rate their inverters on a
continuous basis indicating the power the inverter could output for many hours. This
is a safe, conservative rating method, although most inverters would not be operating at
their full rated output level for 24 hours a day. In most real situations the output is
required for a few hours or minutes.
Some manufacturers rate their inverter output for 30 minutes and some even rate them
based on 15-minute output. The problem with this situation is that you as a system
designer have the burden of discovering the exact method of output rating so that you
can make a fair comparison.
Instead of reading a single power number it is best to find a chart or graph that
describes the output against different times, as shown below.
An example of an inverter with a fairly generous power vs. time curve is shown in the
figure. The literature mentions that the inverter is rated at 2000 watts output, but
notice that this is for 30 minutes. For periods longer than 30 minutes, the power level
will be less. Continuous output would be more like 1700 watts.

11-2
Surge Power
Notice that the output power of the inverter in the figure above can exceed the nominal
2000 watts for a few minutes. Often this is referred to as the surge capability of the
inverter, although strictly surge power may be defined as the power that the inverter can
output for less than a second to start large inductive (motor) loads. Recall from our
discussions in the chapter on Load Estimation that many inductive loads (motors,
compressors, refrigerators, washing machines, and large water pumps) can draw 5-6
times their nominal running power when they start!
For example, a remote home might have non-inductive loads of 1000 watts and a
3/4-hp water pump motor. Normally the water pump might draw about 700 watts, so
the total continuous power needed from the inverter would be about 1700 watts. But
when the pump turns on it draws about 3500 watts. If it turned on when the other 1000
watts of load were operating, the inverter would have to output about 4500 watts (1000
non-inductive + 3500 inductive surge) for a few moments to start the motor. The
inverter example used above in the figure could output about 6000 watts for a few
seconds, so this would be adequate. An inverter with a lower surge capacity might not
be able to start the pump.

11-3
Power vs Time
Output Power (watts)
Power of 2000 watt inverter
6000
5000
4000
3000
2000
1000
0
20
40
60
Time (minutes)

The surge power of an inverter is again not a single fixed number, but varies with the
amount of time. A good inverter can typically output 200-300% of its nominal power for
a few seconds. Some low cost inverters can only output 110% of their output for a
short time. These would have limited surge capacity and would not be useful for
systems where inductive loads would operate.

11-4
Temperature Affects Power Rating

Another source of confusion in inverter rating is the temperature of the rating. The
output can be maintained so long as the inverter design adequately dissipates the
internal heat generated. But this ability will depend on the ambient temperature. If the
air temperature is hotter, the inverter design will not be able to passively dissipate as
much heat, and the power rating will be lower.
Most US manufacturers rate their output at 25 o C. Some others (often international
manufacturers) will rate at 20o C. And some manufacturers take the extreme case of
rating their inverter output at 40o C. This must be looked at to fairly compare product
capability.
The use of a high temperature such as 40o C is conservative and safe. An inverter
rated at 40o C will output more power for a longer period of time at lower temperatures.
But it may not be difficult for the electronics of an inverter to see 35-40o C conditions
inside the inverter case, so this rating is not unreasonable.

11-5
Exercise

!

"
#

" $#%

11-6
Inverter Efficiency
An inverter consumes some power itself. This adds to the total load that the
photovoltaic array must operate and increases array size and initial cost. The inverter
should be as efficient as possible, certainly above 90% over most of its normal
operating range. Many moderately priced inverters can achieve 94% efficiency. The
efficiency may depend on the nature of the load. Pure resistive loads may operate the
inverter at a higher efficiency than inductive loads that absorb the power differently.
Efficiency Depends On Load

Inverter efficiency is not a single fixed value, but varies depending on the amount of
power being generated. A curve of efficiency vs. output power is the best way to
examine how the inverter will perform in a variety of situations. It is difficult or
impossible to predetermine exactly what will be the power demand on the inverter at
any moment, so guesses must be made. You must estimate at what power levels your
system will operate and determine average efficiency from a curve such as the one
presented below.
Older technology using silicon controlled rectifiers (SCR) were not as efficient as more
current designs using new power transistor technology. Efficiencies can now reach
above 90% over much of the range of operation. Notice how the efficiency curve rises
quickly even for small loads, and then gradually drops as power continues to increase.
This shows the effect of the internal power losses increasing with higher currents.
For example, reading from the new transistor efficiency curve, you can see that the
efficiency at full rated load is about 85% while the efficiency at only 10% of full load is
about 93%. This is good news compared to older technology. Because it means that
you can choose a large inverter to handle the heavy loads, you might have in a complex
residential system and yet not sacrifice on efficiency when operating only a few small
appliances or lights in the evening for example. With older technology you would pay a
steep efficiency penalty when running loads much smaller than the full rated capacity of
the inverter.

11-7
Inverter Efficiency
Efficiency
100
new Transistor type
90
80
old SCR type
70
60
50
40
30
20
10
0
0
20
40
60
80
100
Percent of Full Load (%)


11-8
Inductive vs. Resistive Loads

Resistive loads are the simplest for inverters to operate. Another measurement would
use inductive loads, such as motors, and the efficiency curve would probably be slightly
different. Inductive loads tend to push some power back into the inverter, and some
manufacturers designs may handle the inductive loads more efficiency that others.
Measuring the AC output to inductive loads becomes more complex because the
current and voltage will not be in phase as they are with pure resistive loads. You must
use a power meter on the output, or measure the current and voltage and also the
phase difference between them. This is the power factor and it can reduce the real
power delivered to an inductive load 10-30% or more. This is why most manufacturers
measure their inverter output power and efficiency into pure resistive loads. This
method however does not give the total picture of what efficiency you would actually get
in a complex residential or industrial system with many large inductive loads.
Test Set-up For Measuring Efficiency

It is fairly easy to create your own inverter efficiency curve by plotting power into a
variable load against the power consumed by the inverter. A circuit of incandescent
lights or heaters can serve as a variable pure resistive load. Measure the current and
voltage output to the loads, and compare to the current and voltage drawn from the
battery bank by the inverter. Operate loads from about 10% of the full rated output to
well beyond 100% of the rating, to create an efficiency curve similar to that shown.
(+)
12 v DC
(+)
120 v AC
Inverter
battery
variable
load
(ex: lights) .
(-)
(-)
shunt #2
shunt #1


11-9
Exercise

& & &
&'
' &
!
&
&( ' '%

)
&

* ++,-
./0/ ,
%
1 &2
3333333333/
45&2
3333333333/
&
& &
6 7
&./

&


11-10
Inverter Output Waveform

The waveform of the inverter output can be an important factor in matching inverter to
load. The waveform describes the way the current and voltage vary over time. There
are three general classes of waveform: squarewave; modified squarewave or modified
sinewave; and sinewave.
Squarewave
Generally the most inexpensive inverters are "squarewave" types. The input DC power
is chopped and boosted in voltage, with little filtering or modulation of the output. The
resulting output contains many unwanted harmonics, or waves of various frequencies
that fight against each other. The measure of this unwanted waveform content, the
total harmonic distortion, can be 40% or greater. Typically squarewave inverters cannot
surge significantly, perhaps only 10-20 % above the maximum continuous power. The
efficiency can be as low as 50-60%. And they have little output voltage regulation.
However these inverters may be useful for small inductive loads or resistive loads.
Squarewave Inverters


11-11
Modified Squarewave or
Modified Sinewave
Another type of inverter output is available that produces a "modified squarewave"
output. Sometimes these are referred to as modified sinewave in manufacturers
literature, although the form more closely resembles a squarewave with a delay through
zero. The peak voltage of the waveform is only about 150 volts, while utility sinewave
power peaks at about 175 volts. The total harmonic distortion is significantly less than a
squarewave, perhaps less than 20%. Also they can surge 300-400% above the
continuous power, and have good voltage regulation. Efficiencies of greater than 90%
are common. These types of inverters are quite popular in remote homes, and can
operate a wide variety of common appliances and electronic loads including home
computers and stereos. However there are reported problems with operating some
specific loads. Some models of laser printers malfunction or are damaged; microwave
ovens may not heat as much and may have to be operated longer; and small electronic
clocks may run twice as fast. You should check with the manufacturer to see if any
incompatibility with your planned loads have been reported.
Modified Squarewave Inverters


11-12
Sinewave
Finally there are types of inverters that output a nearly pure "sinewave. These types of
inverters may involve extensive and carefully tuned filtering or digital synthesis. In the
past these types of inverters tended to have efficiencies and surge capabilities slightly
below the modified sinewave types, but recently models have been introduced with
efficiencies above 90% and costs below some modified squarewave models. In
general, it is always best to choose a sinewave inverter and supply a sinewave output
to your AC loads. This is what they were designed for, and little or no problems will
occur with respect to harmonic distortions or inadequate peak voltages. Some classes
of electronic loads, such as telecommunications or delicate instrumentation may require
a sinewave power waveform.
Sinewave Inverters


11-13
Exercise

8$

&

5& &

9 &

:

;


11-14
Inverter Circuit Topologies

The topology of inverter design refers to the electrical approach to the conversion
process and the electrical principals applied to designing the control and power circuits.
One fundamental topology difference is where and how the inverter circuit changes
from DC to AC and from the input voltage to the output voltage. Basically there are two
choices
First convert to the final output frequency of 60/50Hz AC and then transform voltage
Upconvert the voltage first and then convert to 60/50 Hz AC for final output.
Final Frequency First, Then Transform

Voltage
In the first topology a 60/50 Hz switching circuit is used to chop the input DC voltage
at the desired final output frequency. The input voltage (typically 12 or 24 volts)
remains the voltage of the output of this first stage of the inverter circuitry. Then this
low voltage AC waveform is put on one side of a large transformer. The ratio of the
number of windings on either side of the transformer determines the change in voltage
through the transformer. For example, to transform 12 volts AC to 120 volts AC, the
output side (120 volts) of the transformer would need to have ten times the number of
windings of the input side (12 volts). The output of this stage of the inverter would then
be a desired frequency and would be at the desired voltage. The transformer is
between the switching transistors and the AC output side of the inverter. This may be
seen as an advantage as the transformer can absorb high energy spikes coming IN to
the inverter (from lightning, load spikes, etc.) and act to protect the more delicate
transistor switching components.

11-15
First Transform Voltage, Then Final

Frequency
In the second topology, very high frequency switching of the order of 25,000 Hz is
performed on the input DC voltage. The output at this stage is still 12 or 24 volts but
the frequency is very high. This waveform is then applied to one side of a transformer
with the winding ratio needed to output at 165 volts. This voltage is the peak-to-peak
voltage of nominal 120-volt power. This high frequency high voltage waveform is then
rectified, or returned to DC power. It is still at 165 volts. The power is now switched at
the desired final output frequency (60/50 Hz), to give the final output waveform at the
desired voltage and frequency. The advantage of the high frequency switching is that
the size of the transformer, and the associated weight and cost, is less than that
needed for 50/60 Hz switching. However the transformer is now no longer between the
AC output and the final switching transistors, and they are more vulnerable to high
energy spikes and noise coming into the inverter.
Pulse Width Modulation

An approach to creating a sinewave output is to apply Pulse Width Modulation (PWM)
switching to the power. Pulses are first made very short, then gradually longer until in
the middle of the wave the pulse is the longest. Then the pulses gradually get shorter.
In this way the time-average of the energy of the pulses forms a sinusoidal voltage
curve. This PWM switching can be done either before or after the main transformer, as
described above.
Multiple Transformers
Another newer topology (used in the sinewave utility interactive Trace Engineering
inverters for example) involves not one but three transformers. One transformer ratio is
3:1, another is 9:1 and another is 27:1. These are switched on and off in an additive or
subtractive manner to create a variable ratio configuration producing many small steps
(around 50 for the Trace model, depending on battery voltage). This allows for a
stepped approximation to a sinewave output waveform.

11-16
Utility Interactive Inverters

There is a class of inverters used in photovoltaic systems that interact with the utility
grid. They draw power from the grid when the solar array cannot supply all the loads,
and feed power into the grid when excess power is available from the array. In order to
synchronize with the utility power waveform, their output must be sinewave also. Until
recently they have been completely dependent on utility power to operate and have not
allowed battery interaction. But recent developments have produced models that are
bi-directional and that can feed utility power into batteries or can instantaneously draw
from batteries if there is a utility power failure.
These types of inverters are primarily used in situations where the grid is reliable, as
different from traditional markets for remote photovoltaic power systems where grid
power is non-existent or unreliable. Power from an array can be fed directly into a
home or business and displace power that would be needed to run daytime loads, such
as air-conditioning or lighting. This approach is especially cost effective where utilities
have established time-of-day rates. Typically they charge higher rates during the peak
of the day, right when photovoltaic systems would be displacing the most power and
saving the most. Then at night when power would be draw from the utility, the rates are
lower and the costs to the user are lower.
Systems that are utility interactive are simpler than typical stand-alone power systems.
They do not involve batteries or charge regulators and are therefore lower in total cost
per watt.
Some models of utility interactive inverters operate at nominal 45 volts so they would
need strings of only three modules in series, while others operate at up to +/- 240 volts
and require strings of 32 modules in series.

11-17
Bi-Directional Utility Interactive Inverters

Until recently utility interactive inverters have been completely dependent on utility
power to operate and have not allowed battery interaction. But new developments have
produced models that are bi-directional and that can feed utility power into batteries or
can instantaneously draw from batteries if there is a utility power failure. For example,
Trace Engineering in the US has introduced a 4000-watt bi-directional inverter that
gives new possibilities to the utility interactive market. Their inverter operates from 24
or 48 volts DC input and outputs a synthesized sinewave output waveform suitable for
the utility grid. Their design draws power instantaneously from either the grid or the
battery and solar array to power AC loads, and can pass power back to the utility or the
battery as needed. If there is a utility power failure, power is seamlessly drawn from the
battery and there is no transfer switch glitch.
A critical requirement of utility interactive inverters is to stop supplying power into the
utility grid when the grid is shut off. This is an important safety requirement, because
when utility linemen shut off a section of the grid for maintenance or repairs, they do not
want any sources of power to be feeding into those lines! Any photovoltaic power
system connected to those lines must also be shut off immediately. Otherwise a small
island of dangerous power would exist on the line. In the past concerns over
islanding would lead utilities to require an external switch that the linemen could
manually switch off outside of any utility connected residence. Now, however, utility
interactive inverters use a variety of methods to detect that grid power is absent and
stop inverting automatically. These methods include (a) voltage over or under detect;
(b) frequency over or under detect; and (c) waveform steps detection.
The market for utility interactive photovoltaic systems is still in its infancy with only a few
thousand systems worldwide. But new developments in inverter technology are making
that market more cost effective and exciting.

11-18
Exercise

8$ &

5
'

' &'
&

11-19
Summary of Inverter Features

Below is presented a summary table listing various features of inverters, and discussing
some of the most important issues related to photovoltaic system design.
Feature
Explanation
Considerations in PV System
Design
Output
Voltage and
Current
Inverters manufactured in the

US are 120 or 240 vac 60 Hz
output to match utility output.
Almost all US inverter makers

also make "export" models with
220 or 240 vac 50 Hz output.
Some inverters are "cascadable"
such that their power outputs
can be combined.
Waveform
Type
See previous discussion in this

manual. There are three types;
sine wave, modified sine wave,
and square wave
Sine waves are best for

applications requiring low noise;
e.g. audio/video, computers, and
communications. Modified sine
waves are best for situations
requiring high efficiency. Square
waves are cheap and best suited
for resistive loads.
Inverters are designed for a

PV Input
Voltage Limits specific DC input voltage (i.e.
12, 24, etc.). For the specified
input voltage, there will be upper
and lower limits. For a typical
12-volt inverter, it might range
from 11 to 16 volts.
If one is using alkaline batteries

(Nicads), this figure may be of
important as alkaline batteries
may reach 16 VDC while under
recharge.


11-20
No Load
Consumption
This is the amount of power that

the inverter consumes when
operating, but not powering any
loads.
Total
Harmonic
Distortion
THD is a measure of how

The lower the THD, the less
perfectly an inverter's waveform amount of noise and radio
matches the ideal sine wave.
frequency interference.
Sine wave inverters have a THD
of 1 - 5%. Modified sine wave
have THD of 10-30% and square
wave of a THD > 30%. As a
reference, in the US, commercial
utility power has a THD of
around 3 - 8 %
Output Power
This refers to the "continuous"

output power for the inverter
A few inverter manufacturers

may still publish output power
that is not continuous, but
derated over time.
Surge Power
The surge capability of an

inverter allows it to have a
starting wattage of 2 to 4 times
its rated output wattage. The
type of motor that you are
running can require a surge of 3
times (Brush Type) to 7 times
(split phase). The bigger the
inverter surge wattage, the larger
the electric motor you can run.
The new 25,000 Hz

"transformerless" inverters have
lower surge capabilities
compared to the traditional 60
Hz inverters. This means that
you may need more inverter
capacity in order to run motors if
you are using 25,000 Hz
inverters.

11-21
Many inverters have sleep

circuits, which puts the inverter
to "sleep" in a mode where no
power is consumed.
RMS Voltage
and Peak
Voltage
To be close to emulating grid

power, inverter RMS voltage
should to be within 10% of the
standard utility RMS voltage and
peak voltage should be within
15% of the standard utility peak
voltage.
Low RMS and peak voltage will

cause poor operation of motors
and appliances that have
transformers (TVs, stereos, etc.).
RMS voltage and peak voltages
that are too high will damage
appliances
Power Factor
Ideally an inverter should be

able to deal with all power
factors, from -1 to 1, where +1
represents a pure resistive load
and -1 representing a pure
inductive load.
The inverter industry has been

successful in producing products
to accommodate a wide range of
power factors. Most reactive
loads (appliances) in homes
have power factors greater than
0.5.
Protective
Circuitry
Inverters are now designed with

a broad range of protection (e.g.
fault, overload, overtemperature, and high/low
voltage).
Good design will have this

protection.
Battery
Charger
Some manufactures build in the

circuitry for battery charging into
the inverter housing. Since the
transformer already exists, the
inverter drives the transformer
backwards as a battery charger.
Three-stage charging is
preferred; Stage One: Bulk
Charge with maximum charge
amps to maximum voltage.
Stage Two: Constant
Voltage/Absorption, where
current tapers off at rate required
to hold voltage constant. Stage
Three: Float Voltage, which
holds full charge without
gassing. In addition an
equalization stage should be
included which allows charging
to a voltage high enough to gas
batteries in order to remove
sulfation.

11-22
Utility
Transfer
Relay
Many inverters with battery

chargers are capable of
transferring their load to the
utility or an engine/generator.
If the transfer is performed with

electro-mechanical relays, then
the time of transfer may
unsuitable for continuous
operation of computers and
other electronic loads.
UL Listing
This tells the user that the

inverter is not a fire hazard.
Important feature to have in

meeting NEC code for
residential systems.
Grounding
NEC requires that the green

equipment grounding conductor
and the white neutral conductor
be grounded on the AC side.
Current should not flow in the
grounding wire. The equipment
grounding conductor should be
sized for the DC input.
Proper grounding of the inverter

is important to meet code.
Many non-UL listed inverters do
not have a provision for the
equipment ground connection.
Some inverters ground their
chassis to the negative
conductor that will not pass UL
standards.
Efficiency
To date efficiency has been

measured using resistive loads,
which has little meaning
because most loads being run
are primarily reactive like motors
and appliances.
This is considered by many a

meaningless figure until inverters
are tested for efficiency with
reactive loads.

11-23
Further Discussion of Selected

Inverter Characteristics
When choosing an inverter for a particular system, the factors of surge capacity,
efficiency and waveform are perhaps the most fundamental. The inverter must be able
to surge to start all the expected inductive loads. It must have enough continuous
output power to handle all the expected loads that might be on at one time. And the
waveform must be appropriate for all the electronic equipment. There are some other
design parameters that should be considered as well. These are described below.
DC Input Voltage
The input DC voltage tends to be a function of the size of the inverter. As the power
through the inverter increases, more current flows and there is greater internal heating.
Manufacturers tend to increase the DC voltage for larger inverters to keep the currents
at a manageable level, around 100 amps or so. So small inverters from 100-2500 watts
continuous tend to be available in 12 volts, while inverters in the 2000-3500 watt range
are made to operate at 24 volts, and models from about 2400-5000 watts are designed
with 48 volt input. Inverters with output greater than 5 kW can have input DC voltages
of 120 to 240 volts and even higher.
If all the loads in a system will be AC then the choice of input voltage is fairly free to
make. The higher the DC voltage the lower the currents into the inverter. This means
that smaller wire can be used, as well as smaller and less expensive disconnect
switches or circuit breakers and fuses. The DC voltage of the inverter will set the
voltage of the battery bank and the array as well.
Do not make the common mistake of thinking that a 24-volt inverter will need twice as
many modules as a 12-volt inverter. The 24-volt choice will use just as many total
modules, but they will be configured differently. Instead of all the modules being
connected in parallel for 12 volts, half as many will be connected in parallel but each
module will be connected to another in series. The total will be the same. And
comparing a 48-volt choice to a 24-volt choice would be similar. Strings of four
modules in series would be used to develop the 48 volts instead of strings of two in
series for 24 volts, but there would be half as many strings in parallel compared to a 24volt case. If DC and AC loads are to be operated in the same system, then the choice
of input voltage is pretty well restricted to 12, or perhaps 24 volts in some cases. Most
all DC loads are designed to operate at 12 volts. Some ballasts for fluorescent lights
and some refrigerator models come in a 24 volt version.
11-24
Voltage Regulation
At high power levels, the inverter draws larger currents from the battery. This causes
the battery voltage to fall. The inverter should be able to compensate for this voltage
drop and maintain output AC voltage fairly well. This information may not be presented
with the usual efficiency curves and is quite important. If output voltage drops too
much, then loads may not operate correctly.
Serviceability
The inverter design should allow easy servicing in the field, or allow for cards to be
swapped and exchanged to minimize down time.
Adjustable Threshold
Most inverters have some threshold of load power requirement before they actually
"turn on" and commutate to produce AC power. This may be a fixed power level, or
may be adjustable by the user. If the load threshold is higher than some small loads in
a house, for example a small electronic clock, the inverter may not sense the load and it
will not operate alone. Some other loads must be on so that the total power is greater
than the threshold value. Some inverter models offer an adjustable threshold level.
However, the lower the threshold level the slightly lower the efficiency of the inverter.
Paralleling For Greater Power

Some inverters can be connected together in parallel and synchronized to increase the
maximum continuous power output. This allows for a small inverter to be installed
initially, and then more power installed later if loads increase or if budget allows.

11-25
(End Of Chapter)

11-26
CHAPTER ELEVEN
INVERTER TECHNOLOGY
11-1
Inverter Power Rating

Power Is A Function Of Time
Surge Power
Temperature Affects Power Rating
11-2
11-2
11-3
11-5
Inverter Efficiency
Efficiency Depends On Load
Inductive vs. Resistive Loads
Test Set-up For Measuring Efficiency
11-7
11-7
11-9
11-9
Inverter Output Waveform

Squarewave
Modified Squarewave or Modified Sinewave
Sinewave
11-11
11-11
11-12
11-13
Inverter Circuit Topologies

Final Frequency First, Then Transform Voltage
First Transform Voltage, Then Final Frequency
Pulse Width Modulation
Multiple Transformers
11-15
11-15
11-16
11-16
11-16
Utility Interactive Inverters

Bi-Directional Utility Interactive Inverters
11-17
11-18
Summary of Inverter Features
11-20
Further Discussion of Selected Inverter Characteristics

DC Input Voltage
Voltage Regulation
Serviceability
Adjustable Threshold
Paralleling For Greater Power
11-24
11-24
11-25
11-25
11-25
11-25

11-27
Chapter 11 Answers
Inverter Technology



The average efficiency of the transistor based inverter is about 92%. The average
efficiency of the SCR based inverter is about 75%.


Compare waveforms to those shown in the text.

11-1
Inverter Technology
Chapter Twelve
Charge Regulators and
System Controls
Photovoltaic modules are highly reliable and virtually maintenance-free; but,
modules alone do not solve a customers power problem. Other components are
generally required to properly control, distribute and store the energy produced by
the array to power an electrical load.
In stand-alone PV systems the array is usually connected to batteries that store the
array energy and supply power to the electrical loads on demand. In most cases,
when batteries are included in a PV system, a charge regulator or controller is
required to protect the batteries from being overcharged by the array and
overdischarged by the system loads. In some designs the charge regulator or
system controller may also provide status information to the user/operator on system
performance and battery state of charge.
In this chapter we will discuss the purpose of battery charge regulators and controls
in PV systems, their important features, and how different types operate to control
the flow of energy in a remote PV power system.

12-1
Components Charge Regulators/Controls
Purpose of Charge Regulators and

System Controls
Battery charge regulation and control of the energy produced by the PV array is a
critical function in PV systems. System control equipment can consist of simply a
battery charge regulator or can include a complex collection of equipment designed
to serve many purposes. Seven of the more important functions of battery charge
regulators and system controls are listed below.
1. Prevent Battery Overcharge
to limit the energy supplied to the

battery by the PV array when the
battery becomes fully charged.
2. Prevent Battery Overdischarge
to disconnect the battery from
electrical loads when the battery
reaches low state of charge.
3. Provide Load Control Functions
to automatically connect and
disconnect an electrical load at a
specified time, for example operating
a lighting load from sunset to sunrise.
4. Provide Status Information to
to display or indicate system
System Users/Operators
operational information such as
battery voltage and current, charging
status, load operation, and other
data.
5. Interface and Control Backup
to integrate alternative sources, such
Energy Sources
as a wind turbine and backup
generator with the PV electrical
system.
6. Divert PV Energy to an Auxiliary
to supply energy to non-critical or
Load
secondary loads when the main
battery bank reaches full state of
charge and the array energy would
otherwise be wasted.
7. Serve as a Wiring Center
to provide a termination or connection
point for other components in the
system, including the PV array,
battery and electrical load.


12-2
Prevent Battery Overcharge

A remote stand-alone photovoltaic system with battery storage is designed so that it
will meet the system electrical load requirements under reasonably determined
worst-case conditions, usually for the month of the year with the lowest insolation to
load ratio. When the array is operating under good-to-excellent weather conditions
(typically during summer) energy generated by the array often exceeds the electrical
load demand. To prevent battery damage resulting from overcharge a charge
regulator or charge controller is used to protect the battery. A charge regulator
should prevent overcharge of a battery regardless of the system sizing/design and
seasonal changes in the load profile, operating temperatures and solar insolation.
Charge regulation is the primary function of a battery charge controller and perhaps
the single most important issue related to battery performance and life. The purpose
of a charge controller is to supply power to the battery in a manner that fully
recharges the battery without overcharging. Without charge control the current from
the array will flow into a battery proportional to the irradiance, whether the battery
needs charging or not. If the battery is fully charged unregulated charging will cause
the battery voltage to reach exceedingly high levels, causing severe gassing,
electrolyte loss, internal heating and accelerated grid corrosion. In most cases if a
battery is not protected from overcharge in PV system, premature failure of the
battery and loss of load are likely to occur.
Charge regulators prevent excessive battery overcharge by interrupting or limiting
the current flow from the array to the battery when the battery becomes fully
charged. Charge regulation is most often accomplished by limiting the battery
voltage to a maximum value, often referred to as the voltage regulation (VR) set
point. Sometimes other methods such as integrating the ampere-hours into and out
of the battery are used. Depending on the regulation method the current may be
limited while maintaining the regulation voltage, or remain disconnected until the
battery voltage drops to the array reconnect voltage (ARV) set point. A further
discussion of charge regulation strategy set points is contained later in this chapter.

12-3
Prevent Battery Overdischarge

During periods of below average insolation and/or during periods of excessive
electrical load usage the energy produced by the PV array may not be sufficient
enough to keep the battery fully recharged. When a battery is deeply discharged the
reaction in the battery occurs close to the grids, and weakens the bond between the
active materials and the grids. When a battery is excessively discharged repeatedly
loss of capacity and life will eventually occur. To protect batteries from
overdischarge most charge regulators include an optional feature to disconnect the
system loads once the battery reaches a low voltage or low state of charge
condition.
In some cases the electrical loads in a PV system must have sufficiently high
enough voltage to operate. If batteries are too deeply discharged the voltage falls
below the operating range of the loads, and the loads may operate improperly or not
at all. This is another important reason to limit battery overdischarge in PV systems.
Overdischarge protection in charge regulators is usually accomplished by opencircuiting the connection between the battery and electrical load when the battery
reaches a pre-set or adjustable low voltage load disconnect (LVD) set point. Most
charge regulators also have an indicator light or audible alarm to alert the system
user/operator to the load disconnect condition. Once the battery is recharged to a
certain level, the loads are again reconnected to a battery.
Non-critical system loads are generally always protected from overdischarging the
battery by connection to the low voltage load disconnect circuitry of the charge
controller. If the battery voltage falls to a low but safe level, a relay can open and
disconnect the load, preventing further battery discharge. Critical loads can be
connected directly to the battery, so that they are not automatically disconnected by
the charge regulator. However, the danger exists that these critical loads might
overdischarge the battery. An alarm or other method of user feedback should be
included to give information on the battery status if critical loads are connected
directly to the battery.

12-4
Provide Load Control Functions

In some cases regulators may have optional features that allow regulation or control
of the PV system electrical load. Load control in PV lighting system regulators is a
popular feature. This control can take place at sunset or sunrise as sensed by a
photosensor or the array current or voltage output. In other cases the controller may
have a timing function to cycle the load operation for a specified period or at a
certain time of day. Experience has shown that these load control functions may
require adjustment and proper specification for the array type and site conditions
such as temperature and background lighting. While these features add to the cost
and complexity of the controller they can also greatly simply the use and operation of
the PV system.
Where load voltage fluctuations may be detrimental to certain loads, load voltage
regulation may be desirable to limit the voltages at which the loads operate. A few
advanced PV system controllers are designed to perform this function.
Provide Status Information To Users

Many charge regulators used in PV systems can provide status information on the
operation of the system and condition of the battery. These optional features of
regulators can allow users to intelligently manage their use of energy and gain a
better understanding of how the system operates to fully utilize its potential.
Battery voltage and/or state of charge is an essential piece of information that can
be indicated by charge regulators. This can be incorporated as a gas gauge with a
simple dial voltmeter showing green, yellow or red regions corresponding to different
battery voltage ranges. Or a digital readout of exact battery voltage can be provided
for more sophisticated users. While random battery voltage readings are interesting,
the user must know how to interpret this information. For example, they must know
at what voltage levels they should begin to curtail their energy usage, to prevent
battery overdischarge.
Many regulators also indicate with a lamp or ammeter whether the array is charging
the battery, or if the array or load is disconnected. Load currents or even the net
battery amp-hours can also be displayed on some regulator/controller designs.
Knowledge of the array and load currents gives users a sense of how much energy
they are producing with the array and consuming with electrical loads. This may
allow them to plan their load usage to better correspond with the energy availability
of their system during low insolation periods.

12-5
Interface and Control Backup Energy

Sources
In the case of a hybrid PV power systems using one or more backup energy sources
in addition to the array more advanced system control centers may be designed to
interface these alternate sources with the PV electrical system. One example would
be a controller that activates a backup generator at low battery state of charge. The
control center would start the generator at a pre-set low battery voltage, and turn it
off when the battery is recharged or reaches a higher voltage limit.
The control of backup energy sources can also be performed by other components
in the PV system. For example, some stand-alone inverters used in PV systems will
start a backup generator or divert the loads to utility power when the battery reaches
low state of charge. Once the battery has been recharged to a pre-set level, the
loads are again connected to and powered by the battery bank.
Divert PV Energy To Auxiliary Load

Batteries in photovoltaic systems are often fully recharged by the middle of the day
during the summer. Normally, the charge regulator disconnects the array to prevent
battery overcharging, wasting valuable array energy. To utilize this excess energy,
some controllers allow the diversion of array energy to power an auxiliary load once
the primary battery bank is fully charged. In most cases the regulator will be
bypassed, directly-coupling the PV array to the auxiliary load. The auxiliary load is
typically a backup battery, a DC water pump, a resistive element in a water heater,
or a fan or some other simple motor load that can be operated from the array without
voltage regulation. In this way all the energy that the array can produce is being
utilized for some purpose.
When the battery voltage falls, and array power is once again needed to run the
regular loads, the auxiliary load is then disconnected. So the auxiliary load must be
an optional, non-critical load that can be operated whenever excess array energy is
available.

12-6
Serves As Wiring Center

In most cases the charge regulator or system controller serves as the termination
and connection point between the conductors leading to the various components in
a PV system. For example, the charge regulator in a small residential lighting
system commonly has the PV array, battery and load all connected to the regulator
terminals. A fuse or circuit breaker for array and battery protection can also be
included in the regulator design.
Larger PV systems generally have overcurrent protection and disconnect devices
included as part of the system control center. With these devices included in the
system controller, the conductors leading to the PV array, battery and loads are
connected at a centralized point in the system. The control center may also be the
principle grounding point in the system and include surge suppression devices. With
this centralized configuration for the system connection and controls, the installation,
operation and maintenance of the system is greatly simplified.

12-7
Terminology
While the specific regulation method or algorithm may vary among different charge
regulators, all have basic functions and characteristics. Charge regulator
manufacturer's data generally provides information about these functions and their
specifications.
Nominal System Voltage

The nominal system voltage is the voltage at which the battery and charge regulator
operate in a PV system. Most charge regulators are designed for operation at a
specific nominal system voltage, while some may allow operation at multiple
voltages, for example with 12 or 24-volt systems.
As mentioned other places in this manual, the selection of the nominal system
voltage has important ramifications on many aspects of design and equipment
selection. For systems with higher load power demands a higher nominal system
voltage is generally used to lower the peak operating currents, reducing the size and
ratings of conductors, overcurrent and disconnect devices and the charge regulator.
However, most equipment for small stand-alone PV systems is widely availability in
12 or 24-volt dc models, and this may dictate the designers selection of the nominal
system voltage.
Nominal Load and PV Array Current

Charge regulators are rated for their ability to handle certain maximum and nominal
currents for the PV array and load. Often, surge conditions may exist from the array
and loads, and the regulator must be tolerant of these conditions. A further
discussion of selecting and sizing charge regulators is presented later in this
chapter.

12-8
Charge Regulator Set Points

The battery voltage levels at which a charge regulator performs control or switching
functions are called the regulator set points. Four basic control set points are
defined for most charge regulators that have battery overcharge and overdischarge
protection features. The voltage regulation (VR) and the array reconnect voltage
(ARV) refer to the voltage set points at which the array is connected and
disconnected from the battery. The low voltage load disconnect (LVD) and load
reconnect voltage (LRV) refer to the voltage set points at which the load is
disconnected from the battery to prevent overdischarge. Figure 12-1 shows the
basic regulator set points on a simplified diagram plotting battery voltage versus time
for a charge and discharge cycle. A detailed discussion of each charge controller
set point follows.
Battery Voltage
Voltage Regulation (VR)

Voltage Regulation Hysteresis (VRH)
Array Reconnect Voltage (ARV)
Load Reconnect Voltage (LRV)

Low Voltage Disconnect Hysteresis (LVDH)
Low Voltage Load Disconnect (LVD)
Charging
Discharging
Time


12-9
Voltage Regulation (VR) Set Point

The voltage regulation (VR) set point is one of the key specifications for charge
regulators. The voltage regulation set point is defined as the maximum voltage that
the charge regulator allows the battery to reach, limiting the overcharge of the
battery. Once the controller senses that the battery reaches the voltage regulation
set point, the controller will either discontinue battery charging or begin to regulate
(limit) the amount of current delivered to the battery. In some regulator designs, dual
regulation set points may be used. For example, a higher regulation voltage may be
used for the first charge cycle of the day to provide a little battery overcharge,
gassing and equalization, while a lower regulation voltage is used on subsequent
cycles through the remainder of the day to effectively float charge the battery.
Proper selection of the voltage regulation set point may depend on many factors,
including the specific battery chemistry and design, sizes of the load and array with
respect to the battery, operating temperatures, and electrolyte loss considerations.
For flooded batteries the regulation voltage should be selected at a point that allows
the battery to achieve a minimal level of gassing. However, gassing should be
avoided for sealed, valve-regulated lead-acid (VRLA) batteries. Temperature
compensation of the voltage regulation set point is often incorporated in charge
controller design, and is highly recommended for VRLA batteries and if battery
temperatures exceed 5o C from normal ambient temperatures (25o C). A
discussion on voltage regulation set point selection and temperature compensation
is contained later in this chapter.
An important point to note about the voltage regulation set point is that the values
required for optimal battery performance in stand-alone PV systems are generally
much higher than the regulation or float voltages recommended by battery
manufacturers. This is because in a PV system, the battery must be recharged
within a limited time period (during sunlight hours), while battery manufacturers
generally allow for much longer recharge times when determining their optimal
regulation voltage limits. By using a higher regulation voltage in PV systems the
battery can be recharged in a shorter time period, however some degree over
overcharge and gassing will occur. The designer is faced selecting the optimal
voltage regulation set point that maintains the highest possible battery state of
charge without causing significant overcharge.

12-10
High Voltage Alarm

In some regulator designs a high voltage alarm is included to alert the system
user/operator of a dangerously high battery voltage condition. In the event of a
controller failure, or failure of a backup source regulator, the high voltage alarm
sends an audible or visible signal to the operator if the battery voltage exceeds a
pre-set level. The settings for the high voltage alarm should be slightly higher than
the maximum expected voltage regulation set point of the charge regulator (including
any increases for temperature compensation). For most applications typical high
voltage alarm settings range upward from about 14.8 volts.
Array Reconnect Voltage (ARV) Set Point

In interrupting (on-off) type regulators, once the array current is disconnected at the
voltage regulation set point, the battery voltage will begin to decrease. The rate at
which the battery voltage decreases depends on many factors, including the charge
rate prior to disconnect, and the discharge rate dictated by the electrical load. If the
charge and discharge rates are high, the battery voltage will decrease at a greater
rate than if these rates are lower. When the battery voltage decreases to a
predefined voltage, the array is again reconnected to the battery to resume charging.
This voltage at which the array is reconnected is defined as the array reconnect
voltage (ARV) set point.
If the array were to remain disconnected for the rest of day after the regulation
voltage was initially reached, the battery would not be fully recharged. By allowing
the array to reconnect after the battery voltage reduces to a set value, the array
current will cycle into the battery in an on-off manner, disconnecting at the
regulation voltage set point, and reconnecting at the array reconnect voltage set
point. In this way, the battery will be brought up to a higher state of charge by
pulsing the array current into the battery.
It is important to note that for some regulator designs, namely constant-voltage and
pulse-width-modulated (PWM) types, there is no clearly distinguishable difference
between the VR and ARV set points. In these designs the array current is not
regulated in a simple on-off or interrupting fashion, but is only limited as the battery
voltage is held at a relatively constant value through the remainder of the day. A
discussion on these types of regulators is included later in this chapter.

12-11

The voltage span or difference between the voltage regulation set point and the
array reconnect voltage is often called the voltage regulation hysteresis (VRH). The
VRH is a major factor that determines the effectiveness of battery recharging for
interrupting (on-off) type regulators. If the hysteresis is to great, the array current
remains disconnected for long periods, effectively lowering the array energy
utilization and making it very difficult to fully recharge the battery. If the regulation
hysteresis is too small, the array will cycle on and off rapidly, perhaps damaging
regulators which use electro-mechanical switching elements. The designer must
carefully determine the hysteresis values based on the system charge and discharge
rates and the charging requirements of the particular battery.
Most interrupting (on-off) type regulators have hysteresis values between 0.4 and
1.4 volts for nominal 12-volt systems. For example, for a regulator with a voltage
regulation set point of 14.5 volts and a regulation hysteresis of 1.0 volt, the array
reconnect voltage would be 13.5 volts. In general, a smaller regulation hysteresis is
required for PV systems that do not have a daytime load. This is because daytime
loads will pull the battery voltage down past the reconnect voltage and initiate array
charging during the day. If only nighttime loads are operated, then the amount of
time that on-off type regulators spend with the array connected will be less than if
loads were operating throughout the day.
Low Voltage Load Disconnect (LVD) Set Point

Overdischarging the battery can make it susceptible to freezing and shorten its
operating life. If battery voltage drops too low, due to prolonged bad weather for
example, certain non-essential loads can be disconnected from the battery to
prevent further discharge. This can be done using a low voltage load disconnect
(LVD) device connected between the battery and non-essential loads. The LVD is
either a relay or a solid-state switch that interrupts the current from the battery to the
load, and is included as part of most regulator designs. In some cases the low
voltage load disconnect unit may be a separate unit from the main charge regulator.
In regulators or controls incorporating a load disconnect feature the low voltage load
disconnect (LVD) set point is the voltage at which the load is disconnected from the
battery to prevent overdischarge. The LVD set point defines the actual allowable
maximum depth-of-discharge and available capacity of the battery operating in a PV
system. The available capacity must be carefully estimated in the PV system design
and sizing process using the actual depth of discharge dictated by the LVD set point.
In more sophisticated designs a hierarchy of load importance can be established,
and the more critical loads can be shed at progressively lower battery voltages.
Very critical loads can remain connected directly to the battery so their operation is
not interrupted.

12-12
The proper LVD set point will maintain a healthy battery while providing the
maximum battery capacity and load availability. To determine the proper load
disconnect voltage, the designer must consider the rate at which the battery is
discharged. Because the battery voltage is affected by the rate of discharge, a lower
load disconnect voltage set point is needed for high discharge rates to achieve the
same depth of discharge limit. In general, the low discharge rates in most small
stand-alone PV systems do not have a significant effect on the battery voltage.
Typical LVD values used are between 11.0 and 11.5 volts, which corresponds to
about 75-90% depth of discharge for most nominal 12 volt lead-acid batteries at
discharge rates lower than C/30.
A word of caution is in order when selecting the low voltage load disconnect set
point. Battery manufacturers rate discharge capacity to a specified cut-off voltage
that corresponds to 100% depth of discharge for the battery. For lead-acid batteries,
this cut-off voltage is typically 10.5 volts for a nominal 12-volt battery (1.75 volts per
cell). In PV systems, we never want to allow a battery to be completely discharged,
as this will shorten its service life. In general, the low voltage load disconnect set
point in PV systems is selected to discharge the battery to no greater than 75-80%
depth of discharge.
In cases where starting (SLI) batteries are used or it is otherwise desired to limit the
battery depth of discharge to prevent freezing or prolong cycle life, a higher LVD set
point may be desired. To protect the battery from freezing, the LVD set point may
be temperature compensated in some cases to increase the load disconnect voltage
automatically with decreasing battery temperature.
To properly specify the LVD set point in PV systems the designer must know how
the battery voltage is affected at different states of charge and discharge rates. In a
few designs, current compensation may be included in the LVD circuitry to lower the
LVD set point with increasing discharge rates to effectively keep a consistent depth
of discharge limit at which the LVD occurs.

12-13
Low Voltage Alarm

An optional regulator feature is a low voltage alarm (LVA) designed to alert the user
of a low battery state of charge condition. If the charge regulator were to fail, or the
array damaged, or during long periods of below average insolation, the battery state
of charge might drop below safe levels. A low-voltage alarm can be included in a
remote power system to monitor battery voltage and send an audible, visual or
electronic alarm if the battery voltage drops below some predetermined safe range.
Typical low voltage alarm settings for nominal 12-volt batteries are about 11.8-11.5
volts, corresponding to about 60-70 % depth of discharge.
One advanced feature of some controllers is current compensation of the LVA
voltage. When high load currents are flowing the battery voltage will be temporarily
lowered due to the internal voltage drop in the battery. This temporary low voltage
should not activate the LVA. It should only come on when a low battery voltage is
sustained for a long period of time. Current compensation circuitry monitors the load
current and lowers the LVA set point correspondingly, so that nuisance alarms are
avoided.
Load Reconnect Voltage (LRV) Set Point

The battery voltage at which a regulator allows the load to be reconnected to the
battery is called the load reconnect voltage (LRV). After the regulator disconnects
the load from the battery at the LVD set point, the battery voltage rises to its opencircuit voltage. When the array or a backup source provides additional charge the
battery voltage rises even more. At some point the controller senses that the battery
voltage and state of charge are high enough to reconnect the load, called the load
reconnect voltage set point.
The selection of the load reconnect set point should be high enough to ensure that
the battery has been somewhat recharged, while not to high as to sacrifice load
availability by allowing the loads to be disconnected too long. Many controller
designs effectively lock out loads until the next day or when the controller senses
that the array is again recharging the battery. Typically LVD set points used in small
PV systems are between 12.5 and 13.0 volts for most nominal 12-volt lead-acid
batteries. If the LRV set point is selected too low, the load may be reconnected
before the battery has been charged, possibly cycling the load on and off, keeping
the battery at low state of charge and shortening its lifetime.
As in the selection of the other regulator set points, the designer must consider the
charge rates for the loads and array and how these rates affect battery voltage at
different states of charge.

12-14
Low Voltage Load Disconnect Hysteresis (LVDH)

The voltage span or difference between the LVD set point and the load reconnect
voltage is called the low voltage disconnect hysteresis (LVDH). If the LVDH is too
small the load may cycle on and off rapidly at low battery state-of-charge (SOC),
possibly damaging the load or controller, and extending the time it takes to fully
charge the battery. If the LVDH is too large, the load may remain off for extended
periods until the array fully recharges the battery. With a large LVDH, battery health
may be improved due to reduced battery cycling but with a reduction in load
availability. The proper LVDH selection for a given system will depend on load
availability requirements, battery chemistry and size, and the PV and load currents.

12-15
Other Functions Associated With

Charge Regulation
Backup Energy Source Control
Another optional feature of some charge regulators is the control of auxiliary or
backup energy sources. Sometimes operating in conjunction with low voltage alarm
(LVA) circuits, the controller switches power to start a backup generator and
connects it to the battery when the state of charge and voltage reach a low level. An
audible alarm can be used for an occupied site to alert the user that it is time to
manually turn on the generator. Or the control unit can automatically switch on a
generator with electric start. Another control can be set to a higher voltage to turn
the generator off as the battery voltage reaches a predetermined higher state of
charge.
Equalizing Charge Capability

For some battery types, particularly for tall flooded lead-antimony types, periodic
equalization charges are required to maintain optimal battery performance. Under
most circumstances, this requires user intervention in system operation, by
bypassing the normal charge regulation circuit for the duration of the equalization
charge. To perform equalization charging automatically, some new regulators allow
their VR set point to be overridden periodically, either when the battery voltage has
dropped particularly low, or at regular time intervals (e.g. 15-20 days). The
frequency and extent to which equalization charges occur can be programmed into
the controller microprocessor circuitry.
Set Point Adjustability

While some charge controllers use fixed resistors to pre-set the controller set points
many have the ability to adjust or change the regulation and load disconnect set
points. Some also have provisions to adjust the hysteresis values as well.
Adjustments are typically made with potentiometers (single or multi-turn), DIP (dual
in-line package) switches, or circuit board jumpers. Although the system designer or
installer may need access to properly set the controller for the type of battery and
system configuration, user/operator access to regulator adjustments should be
discouraged.

12-16
Load Voltage Regulation

For some PV applications with critical loads such as telecommunications and
telemetry equipment, the load must operate at a specific voltage or within a
prescribed voltage range. While an external dc-dc converter between the battery
bank and load sometimes accomplishes this task, some advanced charge regulator
types can provide load voltage regulation. A DC-DC converter may be incorporated
in the controller circuitry, or a simple voltage regulator circuit may be used if the load
voltage requirements are less than the lowest expected battery voltage.
Regulation/Control Element Design

The regulation or switching control element in charge regulators can be either a
solid-state device or an electro-mechanical relay. Simple interrupting or on-off
regulators may use relays, however in most cases MOSFETs or power transistors
are used rather than relays because they have lower power requirements, are
smaller, and can operate for many more cycles. The regulator switching elements
must be properly rated for the application. The steady-state dc current ratings for
the regulator element should be at least 125% of the maximum PV array short-circuit
current. The peak or surge DC current ratings for the element should be at least
150% of maximum expected PV array short-circuit current.
The switching element that controls the low voltage load disconnect circuit is
generally an electro-mechanical relay, as the number of cycles this switching
element experiences is very limited compared to the regulation element. Generally,
relay contacts should be plated, or hermetically sealed and rated to handle at least
125% of the rated dc load current. Peak or surge dc current ratings for the load
switching element should also be able to handle expected high current surges from
the system loads.

12-17
Operational Limits
The environmental and mechanical design limits of charge regulators are an
important consideration for most PV applications. Generally, PV systems and
components are installed in remote areas, in unconditioned spaces, and subject to
the extremes of the weather. For these reasons most regulators have minimum and
maximum ratings for ambient temperature, battery temperature and relative
humidity. Where extreme environmental conditions exist the designer should
consider these specifications when selecting charge regulators.
The packaging and physical characteristics are another important characteristic of
charge regulators. In general, the regulator circuitry should be sealed from the
environment, either by conformably coating or potting the circuitry. A rigid case
should protect the regulator or the controller may be installed in a weatherproof
enclosure.
Terminations used to connect wiring to the charge regulator should be corrosion
resistant, and be large and sturdy enough to accept the conductor sizes that may be
used in the system.
Surge Protection and Grounding

Most regulators include some type of surge suppression devices on the array and
load circuits. Commonly, these devices are metal-oxide varistors (MOVs), which are
connected between the positive and negative terminals, and from this terminal to
ground. Under normal operating conditions MOVs are high impedance devices.
However under surge conditions, MOVs shunt energy to ground, bypassing sensitive
electrical circuits of the controller. While these devices do not always protect the
controller circuitry from harmful surges, they are strongly recommended for
regulators used in lightning prone areas and in dc/ac systems using inverters.

12-18
Service Disconnects and Overcurrent

Protection
Other features that are sometimes included system control centers are overcurrent
protection and disconnect devices. The requirements for overcurrent protection and
disconnect requirements are discussed in detail elsewhere in this manual. In
general, disconnects and overcurrent protection are needed between the PV array
source circuits and regulators, between the battery and regulators, and perhaps on
load circuits. When these components are included as part of the system control
center, the installation, operation and maintenance of the system are simplified.

12-19
Standard Configurations Of
Charge Regulation and Control
Systems
There are a number of ways that battery charge regulation and system controls are
configured photovoltaic systems. Depending on the type and size of the system, the
cost and availability of hardware, and designer choice, certain regulation and control
configurations may be more appropriate than others. Some of the more common
configurations of charge regulation and system control are described below. The
electronic design of charge regulators is presented in the next section.
Simple Series Path Configuration

The most common configuration is to install the charge regulator/controller in series
between the array and the battery. When the battery reaches the charge regulation
voltage set point of the controller, the regulator mechanism may either open-circuit
or short-circuit the array, or limit the current to lower than the peak array current.
The following figure shows two typical series regulator configurations used in PV
systems. The top diagram shows a simple design where the charge regulator is only
used to protect the battery from overcharge, but does not protect it from
overdischarge. In this configuration, the regulator is placed in series between the
PV array and battery, while the load is connected directly to the battery.
The bottom diagram in the figure shows a series configuration with battery
overdischarge protection. The charge regulator used in this design includes a low
voltage load disconnect circuit, and the load is connected to the charge regulator at
the load terminals rather than directly to the battery. Charge regulators of this type
are common and are designed to handle PV array currents up to 30 amps. Where
the load currents exceed the ratings of these controllers an external relay is often
driven by the charge regulator to control the load.

12-20
Simple Series Configurations

Without Overdischarge Protection
PV Array
Charge
Regulator
Battery
Load
With Overdischarge Protection

PV Array
Charge
Regulator
Load
Battery

Auxiliary Load Path Configuration

Photovoltaic systems are usually sized so that the array produces enough energy to
fully operate the loads during the winter months, in order to keep the battery fully
charged. When the solar insolation is high and the load usage is low the charge
regulator disconnects the array from the battery to prevent overcharge, wasting
valuable energy. Where a backup or non-essential load can be utilized the excess
array power can be diverted directly to the backup load or through the battery to the
backup load.

12-21
The figure below shows a basic configuration for auxiliary load connection in PV
systems with similar battery overdischarge protection as described for the series
path configuration. In this case when the battery reaches full state of charge, the
charge regulator diverts the array power to an auxiliary load while the primary load
continues to operate from the battery. In most cases these regulators connect the
array directly to the auxiliary load, and bypass the circuit to the battery. In this
design it should be noted that the array power and voltage delivered to the load will
not be regulated, so the load must be compatible with the array output. Electrical
loads that can be directly connected to the array output include auxiliary batteries,
DC motor appliances such a pumps or fans, or resistive elements used for water or
space heating. Whenever excess energy is available from the array, these types of
loads can absorb it. When the primary battery voltage drops and array power is
once again needed to recharge it the auxiliary loads power is cut off.
During the summer months charge regulation limits the array energy utilization by as
much as 50% in most small stand-alone PV applications. The use of an auxiliary
load can make full use of all the array energy throughout the year that would
otherwise be wasted. Array energy diversion to auxiliary loads is usually used in
remote homes or cabins, where the primary load usage is intermittent or infrequent,
and where a non-essential load can be utilized. Generally, commercial and
industrial PV applications have little need for the common types of auxiliary loads.
Auxiliary Load Configuration

Primary
Load
PV Array
Auxiliary load is OPTIONAL

to utilize excess array energy
Charge
Regulator
Battery
Auxiliary
Load

12-22
Parallel Path Configuration

A less common approach to battery charge regulation is a parallel path
configuration. In this configuration the auxiliary load is no longer an option - it is
necessary and required to protect the battery from overcharging.
The figure below shows a diagram of a parallel path configuration with
overdischarge protection. The charge regulator is not in series between the array
and battery, but is placed in parallel with the battery. The current from the array is
fed directly to the battery at all times. When the battery reaches full state of charge,
the regulator begins to pass excess array power to the auxiliary load. If the battery
is fully charged then all the power from the array flows across the battery terminals
and to the auxiliary load. The primary load will continue to operate.
The concern with this configuration is that the auxiliary load must always be
available and must be able to fully utilize the maximum expected array power. If not,
failure of the load or battery overcharge may occur. The advantage of this type of
configuration is that multiple sources of power can be connected to the battery.
Sources such a wind generators that produce large variations in current could
damage some regulators configured in series. But in this approach, the large battery
bank acts as a buffer, absorbing all the wind generator energy, and the regulator
must only gradually divert the energy away to the auxiliary load as the battery
becomes charged.

Primary
Load
PV Array
Battery
Auxiliary load is MANDATORY

to regulate battery overcharge
Charge
Regulator
Auxiliary
Load


12-23
Sub-Array Switching Configuration

For larger arrays, the commonly available charge regulators used for small standalone PV applications may alone not be rated to handle the higher array currents.
Charge regulation for arrays larger than 500 to 1000 peak watts is generally
accomplished by grouping the array into parallel sub-groups or sub-arrays, and by
regulating each of the sub-arrays independently. The common operational mode for
sub-array switching controllers is that each sub-array regulator is set to a slightly
different cut-off voltage. As the battery voltage rises, first one sub-array disconnects
at the lowest regulation voltage, and then the next is disconnected, gradually
reducing the charge current into the battery as the voltage increases to the highest
regulation voltage set point.

Sub-Array #1
Charge
Regulator
Sub-Array #2
Charge
Regulator
Sub-Array #3
Charge
Regulator
Battery
Optional LVD
Sub-Array #3
One sub-array may be connected directly to the
battery for finishing charge
Load


12-24
For this configuration, a common design approach is that several regulators are
used in parallel, each with its own sub-array connected. Another approach is one
that a master controller operates relays switching each sub-array, each set to a
slightly different voltage point. For example, a 12 volt array could have one subarray disconnected at 13.8 volts, a second at 14.0 volts, a third at 14.2 volts, and so
on. In this way the charge current is reduced in steps, and the battery gradually
approaches full charge. This approach allows for the use of simple on-off type
controls and yet results in the same gentle approach to full charge as complex multistage and constant-voltage regulators.
The switching elements used in sub-array switching control designs can either be
electro-mechanical relays or long-life mercury displacement relays. By separating
the control mechanism (relay) from the charge control circuitry greater flexibility in
design can be enjoyed, and very large array currents can be regulated with simple
single stage on-off type control elements (relays).
In systems with many sub-arrays one small sub-array may be left connected directly
to the battery (with appropriate fusing and disconnect switch for wire protection and
code compliance). This final sub-array current provides a gentle finishing charge at
the end of the day. The general requirement is that this one sub array contributes
only low charge rates (for example C/90 or less) to the battery to limit the battery
voltage.
The number of parallel sub-arrays determines the number of regulators or relays
required. To determine the number of regulators or sub-array relays required, divide
the maximum possible array current by the current capacity of one regulator or relay.
The maximum array current is the short circuit current of the entire array increased
by a safety factor for possible cloud and ground reflectance, usually a factor of 130%
is used.
Number of Sub-Arrays
and Regulators

Number of Parallel Modules X Isc X 1.3

Current Capacity of One Regulator
12-25
Example:
A large microwave repeater system requires 4 kW of solar array at 24

volts. If 35-watt modules are used, this means that a total of 114
modules are required.
4000 watts
35 watts/ module
114 modules
Since the system requires 24 volts, this means that two modules are
wired in series and 57 strings of two-in-series are wired in parallel. So
the number of parallel modules is 57.
114 modules total
2 in series
57 in parallel
A typical size charge regulator is rated at 30 amps (some are larger,

some smaller). Using the Isc of 2.15 amps for an 35 watt module, the
number of regulators and sub-arrays is given by
Number of Regulators

57 parallel X 2.15 amps X 1.30

30 amps/regulator
5.3 regulators, rounded up to 6
12-26
Exercise

!

"#
$%!

&'
((((((((((

)&'!!*&'

*#
##
#
+
, #

&!
'!
#%!
!
&!!

12-27
Electronic Designs for Charge

Regulation
.
The discussion in the preceding section show the different ways battery charge
regulation can be configured in photovoltaic systems. There are also a number of
variations in the function and electronic design of charge regulators.
Two basic methods exist for controlling or regulating the charging of a battery from a
PV module or array - shunt and series regulation. While both of these methods are
effectively used, each method may incorporate a number of variations that alter their
basic performance and applicability. Simple designs interrupt or disconnect the
array from the battery at regulation, while more sophisticated designs limit the
current to the battery in a linear manner that maintains a high battery voltage.
The algorithm or control strategy of a battery charge regulator determines the
effectiveness of battery charging and PV array utilization, and ultimately the ability of
the system to meet the electrical load demands. Most importantly, the controller
algorithm defines the way in which PV array power is applied to the battery in the
system. In general, interrupting on-off type controllers require a higher regulation set
point to bring batteries up to full state of charge than controllers that limit the array
current in a gradual manner.
Some of the more common design approaches for charge regulators are described
in this section. Typical daily charging profiles for a few of the common types of
regulators used in small PV lighting systems are presented in the next section.

12-28
Shunt Regulator Designs

Since photovoltaic cells are current-limited by design (unlike batteries) PV modules
and arrays can be short-circuited without any harm. The ability to short-circuit
modules or an array is the basis of operation for shunt regulators.
The figure shows an electrical design of a typical shunt type regulator. The shunt
controller regulates the charging of a battery from the PV array by short-circuiting the
array internal to the regulator. All shunt regulators must have a blocking diode in
series between the battery and the shunt element to prevent the battery from shortcircuiting when the array is regulating. Because there is some voltage drop between
the array and regulator and due to wiring and resistance of the shunt element, the
array is never entirely short-circuited, resulting in some power dissipation within the
controller. For this reason, most shunt controllers require a heat sink to dissipate
power, and are generally limited to use in PV systems with array currents less than
20 amps.
The regulation element in shunt regulators is typically a power transistor or MOSFET
depending on the specific design. There are three general variations of the shunt
regulator design. The first is a simple interrupting, or on-off type regulator design.
The second type limits the array current in a gradual manner, by increasing the
resistance of the shunt element as the battery reaches full state of charge. A third
approach is to pulse the current through the shunt element and vary the pulse time
or width depending on the battery state of charge (Pulse Width Modulation or PWM).
These variations of the shunt regulator are discussed next.

12-29
B a s ic S h u n t R e g u la to r D e s ig n
L o a d S w itc h in g E le m e n t
B lo c k in g D io d e
+
PV
A rra y
R e g u la tio n
C o n tro l
LVD
C o n tro l
DC
Load
B a tte ry
S h u n t E le m e n t

Shunt-Interrupting Design
The shunt-interrupting regulator completely disconnects the array current in an
interrupting or on-off fashion when the battery reaches the voltage regulation set
point. When the battery decreases to the array reconnect voltage, the regulator
connects the array to resume charging the battery. This cycling between the
regulation voltage and array reconnect voltage is why these regulators are often
called on-off or pulsing controllers. These are NOT to be confused with a true
pulsing or PWM type control, discussed later. Shunt-interrupting regulators are
widely available and are low cost, however they are generally limited to use in
systems with array currents less than 20 amps due to heat dissipation requirements.
In general, on-off shunt regulators consume less power than series type regulators
that use relays (discussed later), so they are best suited for small systems where
even minor parasitic losses become a significant part of the system load.
Shunt-interrupting charge regulators can be used on all battery types, however the
way in which they apply power to the battery may not be optimal for all battery
designs. In general, constant-voltage, PWM or linear regulator designs are
recommended by manufacturers of gelled and AGM lead-acid batteries. However,
shunt-interrupting regulators are simple, low cost and perform well in most small
stand-alone PV systems.

12-30
Shunt-Linear Design
Once a battery becomes nearly fully charged a shunt-linear regulator maintains the
battery at near a fixed voltage by gradually shunting the array through a
semiconductor regulation element. In some designs a comparator circuit in the
regulator senses the battery voltage, and makes corresponding adjustments to the
impedance of the shunt element, thus regulating the array current. In other designs
simple Zener power diodes are used, which are the limiting factor in the cost and
power ratings for these regulators. There is generally more heat dissipation in a
shunt-linear regulator than in shunt-interrupting types.
Shunt-linear regulators are popular for use with sealed VRLA batteries. This
algorithm applies power to the battery in a preferential method for these types of
batteries, by limiting the current while holding the battery at the regulation voltage.
Shunting Pulse Width Modulated (PWM) Design

This algorithm uses a semiconductor switching element across the array and battery
that is switched on/off at a high frequency with a variable duty. The time duration or
duty cycle of the pulses is varied to increase or decrease the time-average amount
of current passed on to the battery, and this maintains the battery at or very close to
the voltage regulation set point. Similar to the shunt-linear, constant-voltage
algorithm in performance, power dissipation within the regulator is considerably
lower in PWM design.
By electronically controlling the high speed switching or regulation element the PWM
regulator breaks the array current into pulses at some constant frequency, and
varies the width (time duration) of the pulses to regulate the amount of charge
flowing into the battery as shown in the figure. When the battery is at low state of
charge the current pulse width is practically fully on all the time. As the battery is
charged and the voltage rises, the pulse width is decreased, effectively reducing the
magnitude of the charge current.

12-31
The PWM design allows greater control over exactly how a battery approaches full
charge, and the high speed switching elements generate little heat. PWM type
regulators can be used with all battery types, however the controlled manner in
which power is applied to the battery makes them preferential for use with sealed
VRLA types batteries over on-off type controls. To limit overcharge and gassing, the
voltage regulation set points for PWM and constant-voltage controllers are generally
specified lower than those for on-off type regulators. For example, a PWM regulator
operating with a nominal 12 volt flooded lead-antimony battery might use a VR set
point of 14.4 to 14.6 volts at 25o C, while an on-off regulator used with the same
battery might require a VR set point of between 14.6 and 14.8 volts to fully recharge
the battery on a typical day.
P u ls e -W id th -M o d u la te d (P W M )
R e g u la to r D e s ig n
B a tte ry a t L o w S ta te o f C h a rg e
U n til th e b a tte ry is fu lly c h a rg e d ,
th e c u rre n t p u ls e s to th e b a tte ry
a re w id e .
B a tte ry a t H ig h S ta te o f C h a rg e
O n c e th e b a tte ry b e c o m e s fu lly
c h a rg e d , th e c u rre n t p u ls e s to
th e b a tte ry b e c o m e n a rro w e r.
!

12-32
Series Regulator Designs

As the name implies, this type of regulator works in series between the array and
battery, rather than in parallel as for the shunt regulator. There are several
variations to the series type regulator, all of which use some type of control or
regulation element in series between the array and the battery. While this type of
controller is used in small PV systems, it is the most practical choice for larger
systems due to the current limitations of shunt controllers. The figure shows an
electrical design of a typical series type regulator.
In a series regulator design array current is controlled in one of three ways.
A relay or solid-state switch can open the circuit between the array and the
battery to discontinue charging.
Array current is pulsed rapidly, and the pulse time or width is varied to keep the
battery voltage constant (PWM type).
A control circuit limits the current in a series-linear manner to hold the battery
voltage at a high value.
Because the series regulator open-circuits rather than short-circuits the array as in
shunt-controllers, no blocking diode is needed to prevent the battery from shortcircuiting when the controller regulates.
B as ic S eries R eg u lato r D e sig n

S e ries E le m e n t
L o a d S w itc h in g E le m e n t
+
PV
A rray
R e g u latio n
C o n tro l
LVD
C o n tro l
DC
Load
B a tte ry
"

12-33
Series-Interrupting 1-Step Design

The most simple series regulator is the series-interrupting type, involving a one-step
control, turning the array charging current either on or off. The charge regulator
constantly monitors battery voltage, and disconnects or open-circuits the array in
series once the battery reaches the regulation voltage set point. After a pre-set
period of time, or when battery voltage drops to the array reconnect voltage set
point, the array and battery are reconnected, and the cycle repeats. As the battery
becomes more fully charged, the time for the battery voltage to reach the regulation
voltage becomes shorter each cycle, so the amount of array current passed through
to the battery becomes less each time. In this way, full charge is approached
gradually in small steps or pulses, almost identical to the shunt-interrupting type
controller discussed earlier. The principle difference between the two is whether the
regulation is through shunting or series interrupting.
As mentioned with the shunt-interrupting type regulator, the 1-step seriesinterrupting type designs are best suited for use with flooded batteries and NOT for
sealed VRLA types, due to the fact that gassing results from the relatively long on
pulses.
Series-Interrupting, 2-Step, Constant-Current Design

This type of controller is similar to the 1-step series-interrupting type, however when
the voltage regulation set point is reached, instead of totally interrupting the array
current, a limited constant current remains applied to the battery. This trickle
charging continues either for a pre-set period of time, or until the voltage drops to
the array reconnect voltage due to load demand. Then full array current is once
again allowed to flow, and the cycle repeats. Full charge is approached in a
continuous fashion, instead of smaller steps as described above for the on-off type
regulators. Some two-stage controls increase array current immediately as a load
pulls down battery voltage. Others keep the current at the small trickle charge level
until the battery voltage has been pulled down below some intermediate value
(usually 12.5-12.8 volts) before they allow full array current to resume.
Series-Interrupting, 2-Step, Dual Set Point Design

This type of regulator operates similar to the series-interrupting type, however there
are two distinct voltage regulation set points. During the first charge cycle of the
day, the controller uses a higher regulation voltage to provide some equalization
charge to the battery. This may be applied on a preset frequency (e.g. every 15-20
days) or may be programmed to occur after the battery has been discharged to a
low state of charge.

12-34
Once the array is disconnected from the battery at the higher regulation set point,
the voltage drops to the array reconnect voltage and the array is again connected to
the battery. However, on the second and subsequent cycles of the day, a lower
regulation voltage set point is used to limit battery overcharge and gassing.
This type of regulation strategy can be effective at maintaining high battery state of
charge while minimizing battery gassing and water loss for flooded lead-acid types.
The designer must make sure that the dual regulation set points are properly
adjusted for the battery type used. For example, typical set point values (at 25o C)
for this type of controller used with a flooded lead-antimony battery might be 15.0 to
15.3 volts for the higher regulation voltage, and between 14.2 and 14.4 volts for the
lower regulation voltage.
Series-Linear, Constant-Voltage Design

In a series-linear, constant-voltage regulator design the regulator maintains the
battery voltage at the voltage regulation set point. The series regulation element
acts like a variable resistor controlled by the regulator battery voltage sensing circuit
of the regulator. The series element dissipates the balance of the power that is not
used to charge the battery, and generally requires heat sinking. The current is
inherently controlled by the series element and the voltage drop across it.
Series-linear, constant-voltage regulators can be used on all types of batteries.
Because they apply power to the battery in a controlled manner, they are generally
more effective at fully charging batteries than on-off type regulators. These designs,
along with PWM types are recommended over on-off type regulators for sealed
VRLA type batteries, because they gradually approach, but do not exceed, the
gassing voltage.
Series-Interrupting, Pulse Width Modulated (PWM) Design

This algorithm uses a semiconductor switching element between the array and
battery that is switched on/off at a high frequency with a variable duty cycle to
maintain the battery at or very close to the voltage regulation set point. Shunt-type
PWM designs (with the switching element across the array and battery) were
discussed earlier. Power dissipation within the series-interrupting PWM regulator is
considerably low, similar to the series-linear, constant-voltage algorithm.
As with all series type designs, there is no need for a blocking diode to prevent
battery shorting, unlike the shunt-type PWM designs.

12-35
Exercise

.

#
/

0
#

*

**1+2#

03 # #

4

* #

# #

12-36
Daily Operational Profiles for

Charge Regulators
The following sections present typical daily operational profiles for a few of the
different types of battery charge regulators commonly used in small stand-alone PV
systems. These daily profiles show how the different charge controller algorithms
regulate the current and voltage from the PV array to protect the battery from
overcharge.
About the Charge Regulator Daily

Profiles
The data presented in the graphs were measured during tests on operational PV
lighting systems at the Florida Solar Energy Center (FSEC) in February 1993.
Several identical systems were monitored, with the exception that each system used
a different battery charge controller. The data presented here are for a selected
clear day with no cloud cover, clearly showing the charge controller regulation
effects.
To properly understand the data presented in the graphs, it is helpful to know how
they were measured. The measured parameters included among others the solar
irradiance (Sun), battery voltage (Vbat) and current (Ibat), and PV array voltage
(Vpv) and current (Ipv). The designations in parenthesis are used in the legend key
for the daily profiles.
Each parameter was sampled every 10 seconds and averaged over a six-minute
period and recorded for a total of 240 data points daily. In addition, the minimum
and maximum of the battery voltage samples were recorded every six minutes.
These minimum and maximum voltages (based on 10 second samples) are key to
understanding how a battery charge controller operates.
In each of the following figures showing charge controller daily performance, there
are two graphs. The top graph shows the battery and PV array voltage versus time
for the clear day. Note that for clarity, the battery voltage is plotted on the left yaxis, while the PV array voltage is plotted with respect to the right y-axis on a
different scale. The bottom graph shows the battery and PV array currents over the
day, as well as the solar irradiance. Note that the currents are plotted on the left yaxis, and the irradiance is plotted on the right y-axis.

12-37
The sizing of the battery, PV array and load profile in the test systems was
configured to typify commercially available PV lighting systems. The different
charge controllers were selected from those commonly used in these type and sizes
of systems. The following table lists the nominal specifications for the FSEC test
systems.
Nominal System Specifications
Location:
Design Insolation:
PV Array:
Battery:
Load:
Florida Solar Energy Center (FSEC)

5 kWh/m2-day
Nominal 100 watt Pmp, 6 amps Imp
Flooded Lead Antimony, 12 volt, 100 Ah @ 20 hr rate
Nominal 3 amps, 8 hours nightly, 24 amp-hours per day
A final word of caution when examining the following daily operational profiles for the
different charge controllers. Since these were test systems designed to investigate
not only the behavior of the different controllers, but the effects the regulation set
points had on maintaining battery state of charge, the set points were not always
optimized for the specific system design. In some cases this was intentional while in
other cases was the result of the controller operating characteristics. The main point
to emphasize however is that the daily profiles presented here show how the charge
controllers typically operate in PV systems.

12-38
Shunt-Interrupting Charge Regulator

Daily Profile
A 24-hour daily profile for a small stand-alone PV lighting system operating with a
shunt-interrupting (on-off) type battery charge regulator is shown in Figure 12-9.
Beginning at the left of the two graphs (midnight), the constant lighting load is
operating (indicated by the approximate -3 amp load battery current), and battery
voltage decreases steadily from about 12.1 volts to 11.9 volts. At about 0400 hours
the lighting load is disconnected by the charge controller load regulation/timing
circuit (LVD cutoff voltage). At this point the battery current goes to zero, and there
is a sharp rise in the battery voltage as it approaches an open-circuit (no load)
voltage of about 12.35 volts. At sunrise (about 0700 hours), the battery voltage
begins to increase as the PV array current is fed into the battery. Until about
noontime (1200 hours), the PV array current and the battery voltage increase
steadily with increasing insolation as the battery is being recharged. Note that
during this period, the battery charge controller is not regulating and nearly all the PV
array current is fed into the battery.
At approximately Noon (1200 hours) the battery voltage reaches the regulation
voltage set point for the battery charge controller, and the controller begins to
regulate the PV array current. When this occurs, the battery current decreases in a
jagged manner characteristic of the 1-step interrupting (on-off) algorithm. The shunt
characteristic is demonstrated by the fact that once regulation begins, the array
current continues to follow the same profile as the solar irradiance, while the 6minute average PV array voltage decreases to an average of about 5 volts (time
averaging between zero and battery voltage). When this controller shunts, or shortcircuits the PV array at regulation, it causes the PV voltage to reduce to zero and
forces the array current to the short-circuit current value.
Up until regulation the minimum and maximum battery voltages closely match the
six-minute average battery voltage throughout the morning and during load
operation. With the onset of regulation the minimum and maximum battery voltages
are different from the 6-minute averaged voltages and indicate the approximate
controller set points. During regulation the maximum battery voltage is between 14.3
and 14.5 volts, corresponding to the voltage regulation set point for the battery
charge controller. The minimum battery voltage is consistently about 13.7 volts,
corresponding to the voltage at which the charge controller reconnects the array to
the battery to resume charging. The fact that the minimum voltage is consistent over
the regulation period indicates that the controller is regulating or cycling the battery
voltage between the voltage regulation and array reconnect set points at least once
every six minutes. The differences in the minimum and maximum battery voltages
during regulation demonstrate the operation of an interrupting or on-off type
controller algorithm. This voltage difference is often referred to as the controllers
hysteresis, or array regulation voltage span.

12-39
The hysteresis is an important specification for on-off controllers, and must be

selected properly to achieve good array energy utilization and proper battery
recharging.
Towards the end of the sunlight hours (1600-1700 hours) the PV array current
output reduces to a low enough value, in this case about 2.5 amps, wherein
regulation is not required to limit the battery voltage below the regulation set point of
the controller. Once the sun sets (about 1800 hours) the battery voltage begins a
gradual decrease to its open-circuit voltage.
Notice how the open-circuit voltage at the end of the day is higher than in the
morning, indicating that indeed the battery was charged and is now at a higher state
of charge. At about 2030 hours, the 3-amp load is reconnected and the battery
voltage begins to steadily decrease in transition to the next day.

12-40

Clear Day Operational Profile
Battery Voltage (V)
PV Array Voltage (V)
15
25
Vbat, max
20
14
Vbat, avg
15
13
Vbat, min
10
12
5
Vpv, avg
11
0
4
12
16
20
24
Time of Day (EST)

Battery & PV Array Current (A)
Irradiance (W/m2)
1000
10
Sun
7.5
Ipv, avg
800
5
600
2.5
400
0
Ibat, avg
-2.5
200
-5
0
4
12
16
20
24
Time of Day (EST)
#

12-41
1-Step Series-Interrupting Regulator

Daily Profile
A 24-hour daily profile for a small stand-alone PV lighting system operating with a 1step series-interrupting (on-off) type battery charge regulator is shown in Figure 1210. The night and early morning activity is similar to the first system described
earlier.
Until about Noon-time (1200 hours) the PV array current and the battery voltage
increase steadily with increasing insolation as the battery is being recharged. Note
that during this period the battery charge controller is not regulating and the PV array
current is approximately the same as the battery current. However, the minimum
battery voltage shows values slightly lower than the average and maximum battery
voltages during the morning charging period. This is a particular characteristic of the
charge controller in this test system, by which the array is periodically disconnected
from the battery to sense nighttime conditions.
At about Noon (1200 hours) the battery voltage reaches the regulation voltage of the
battery charge controller (about 14.1 volts), and the controller begins to regulate the
PV array current. When this occurs the battery current decreases to the jagged
characteristic of the 1-step interrupting (on-off) algorithm. The series characteristic
can be seen by the fact that once regulation begins, the average PV array current
decreases (time averaging between battery charging current and zero), while the
average PV array voltage approaches the array open-circuit voltage (time averaging
between battery charging voltage and Voc). When this controller open-circuits the
array during regulation, the result is zero PV current while operating the array at the
open-circuit voltage point.
With the onset of regulation the minimum and maximum battery voltages are
distinguished from the 6-minute average voltage, and show the approximate
controller set points. After regulation, the maximum battery voltage is about 14.1
volts. This maximum battery voltage corresponds to the voltage regulation set point
for the battery charge controller. The minimum battery voltage is between 13.2 and
13.4 volts, corresponding to the voltage at which the charge controller reconnects
the array to the battery to resume charging.
After sunset (~1800 hours) the battery voltage begins a gradual decrease to its
open-circuit voltage, which is again higher than it was in the morning, indicating
battery recharging has occurred. At about 2030 hours, the 3-amp load is
reconnected and the battery voltage begins to steadily decrease in transition to the
next day. In comparison with the shunt-interrupting controller discussed previously,
the regulation set point for this 1-step series-interrupting controller was considerably
lower, resulting in a lower battery state of charge. This is indicated by the lower
battery voltage just prior to the load being disconnected in the early morning.

12-42
Series-Interrupting Charge Regulator

Clear Day Profile in PV Lighting System
Battery Voltage (V)
15
25
Vpv, avg
20
Vbat, max
14
Vbat, avg
15
13
10
Vbat, min
12
11
0
4
12
16
20
24
Time of Day (EST)

Irradiance (W/m2)
1000
10
Sun
7.5
800
Ipv, avg
600
2.5
400
0
Ibat, avg
-2.5
200
-5
0
4
12
16
20
24
Time of Day (EST)
$

12-43
Modified Series Interrupting 2-Step

Constant Current Charge Regulator
Daily Profile
modified series interrupting 2-step constant current type battery charge regulator is
shown in Figure 12-11. The morning profiles are again similar to the previous
patterns.
At about Noon (1200 hours) the battery voltage reaches the regulation voltage set
point for the battery charge controller (about 14.9 volts), and the controller begins to
regulate the PV array current. In contrast to the series- and shunt-interrupting
controllers discussed previously, the battery current is not entirely disconnected from
the battery, but only limited to a lower value. When this occurs the battery current
decreases to below 2 amps, and remains in a current-limited mode through the
remainder of the day. The series characteristic is shown by the fact that once
regulation begins, the average PV array current also decreases while the average
PV array voltage approaches the open-circuit array voltage. This controller
regulates the array in a series-linear manner, by increasing the resistance between
the PV array and battery. The resistance is held at such a value that a limited
amount of current is allowed to flow from the PV array to battery after initial
regulation.
With the onset of regulation, the minimum and maximum battery voltages are
indistinguishable from the six-minute average voltage, indicating that the controller is
NOT an on-off interrupting type design. After the initial battery regulation at 14.9
volts, the voltage after regulation remains at about 14.1 volts through the remainder
of the day.
Once the sun sets (~ 1800 hours), the battery voltage begins a gradual decrease to
its open-circuit voltage. At about 2030 hours, the 3-amp load is again reconnected
and the battery voltage begins to steadily decrease as the battery is discharged.

12-44
Modified Series Charge Regulator

Battery Voltage (V)
15
25
Vbat, max
Vbat, avg
14
20
Vbat, min
15
13
10
Vpv, avg
12
11
0
4
12
16
20
24
Time of Day (EST)

Irradiance (W/m2)
1000
10
Sun
7.5
800
5
600
Ipv, avg
2.5
400
0
Ibat, avg
-2.5
200
-5
0
4
12
16
20
24
Time of Day (EST)


12-45
Daily Profile for Constant-Voltage

Series Charge Regulator
constant-voltage series type battery charge controller is shown in Figure 12-12.
Again the morning profiles are similar to previous cases.
regulate the PV array current. When this occurs the battery current gradually
decreases to about 1 amp by the end of the day. The series characteristic of this
controller is shown by the fact that once regulation begins, the average PV array
current also decreases while the average PV array voltage approaches the opencircuit array voltage. In principle this controller regulates the array in a series-linear
manner, by increasing the resistance between the PV array and battery through
semiconductor devices such as MOSFETs. The resistance is held at such a value
that limits amount of current that is allowed to flow from the PV array to battery after
initial regulation, while holding the array voltage at a constant value corresponding to
he controllers regulation voltage.
With the onset of regulation the minimum and maximum battery voltages are
indistinguishable from the six-minute average voltage, indicating that the controller in
not an on-off interrupting type design. After the initial regulation at 14.5 volts, the
voltage after regulation remains at this level through the remainder of the day.
Moving toward sunset (about 1800 hours), the array current is no longer high
enough to maintain the battery at the regulation voltage, and the battery voltage
begins a gradual decrease to its open-circuit voltage. At about 2030 hours, the 3
amp load is again reconnected and the battery voltage begins to steadily decrease
until the next day when charging resumes.

12-46
Constant-Voltage Series Charge Regulator

Battery Voltage (V)
15
25
Vbat, avg
Vbat, min
Vbat, max
14
Vbat, max
Vbat, avg
Vbat, min
20
Vpv, avg
15
13
Vpv, avg
10
12
5
11
0
4
12
16
20
24
Time of Day (EST)
Irradiance (W/m2)
1000
10
Sun
7.5
800
Ipv, avg
600
2.5
400
0
Ibat, avg
-2.5
200
-5
0
4
12
16
20
24
Time of Day (EST)


12-47
PWM Series Charge Regulator

Daily Profile
pulse-width-modulated (PWM) series type battery charge regulator is shown in
Figure 12-13.
regulate the PV array current. When this occurs the battery current decreases in a
jagged manner, and remains in a current-limited mode through the remainder of the
day. The series characteristic can be seen by the fact that once regulation begins,
the average PV array current also decreases while the average PV array voltage
approaches the open-circuit array voltage.
In the PWM design an oscillating signal operating at a frequency of several hundred
Hertz is used to regulate the array current. When the controller is not regulating, the
full array current is applied to the battery. When the regulation voltage is reached,
the current pulses are gradually reduced to hold the battery voltage at the regulation
set point.
In effect, the PWM design operates similar to the constant-voltage controller, with
the exception that there is a small hysteresis between the minimum and maximum
battery voltage after regulation. The PWM is essentially a high switching speed onoff type or interrupting type controller that does not allow the battery voltage to drop
significantly during regulation.
Once the sun sets (about 1800 hours), the battery voltage begins a gradual
decrease to its open-circuit voltage. At about 2030 hours, the 3-amp load is
reconnected and the battery voltage begins to steadily decrease in transition to the
next day.

12-48
Pulse-Width-Modulated Series Charge Regulator

Battery Voltage (V)
15
25
Vbat, max
Vbat, avg
Vbat, min
14
20
15
13
Vpv, avg
10
12
5
11
0
4
12
16
20
24
Time of Day (EST)

Irradiance (W/m2)
1000
10
Sun
7.5
Ipv, avg
800
5
600
2.5
400
0
Ibat, avg
-2.5
200
-5
0
4
12
16
20
24
Time of Day (EST)


12-49
Exercises

5# # #

#
"
6
&'&& #

# #
#
6
&'&'
&'&% #
7#

##*
##


12-50
Voltage Regulation Set Point

Selection
In stand-alone PV systems the ways in which a battery is charged are generally
much different from the charging methods battery manufacturers recommend. A
battery in a PV system must be fully recharged during the few daylight hours, a
much shorter time period than typically used for recharging standby batteries in UPS
applications (days or weeks) or industrial cycle applications (18-24 hours). For this
reason, the voltage regulation set point must be set higher than the usual float
voltage given by battery manufacturers, to permit full utilization of the array current,
but not to high as to excessively overcharge and gas the battery.
The optimal voltage regulation set point allows a maximum amount of Ah to be
replaced into a battery with minimum water loss and gassing. This is usually arrived
at empirically, and is different for different types of batteries and even different
manufacturers of the same type of battery.
The set points are probably more important than the particular type of regulator
design. A relatively simple charge regulator design with proper set points may work
better than a sophisticated controller which is not set properly for the type of battery
or the environment.
Battery Gassing is Key To Voltage

Regulation Set Point Selection
The onset of gassing in a lead-acid cell is determined by cell voltage, temperature,
and charge rate.
Under standard conditions, at a certain voltage (usually ~ 2.35 volts), the

chemical potential is high enough to separate water into hydrogen and oxygen,
causing water loss through gassing.
At higher battery temperatures, the internal chemical reactions occur faster and
the corresponding gassing voltage decreases. This is the primary reason why
temperature compensation is used for the voltage regulation set point.
At higher charge rates the internal voltage drop in the battery elevates the
voltage and brings on gassing sooner than would occur for a slower charge rate.

12-51
Some degree of gassing and overcharge is required for flooded lead-acid batteries,
but can be harmful to sealed VRLA captive electrolyte type batteries for which lost
electrolyte can not be replaced. In general, sealed maintenance free valveregulated batteries (using lead-calcium grids) require lower charge regulation voltage
set points than flooded deep cycling batteries (using lead-antimony grids). Battery
manufacturers should be consulted to determine the gassing voltages for specific
designs.
The figure shows the relationships between cell voltage, state of charge, charge rate
and temperature for a typical lead-acid cell with lead-antimony grids. At 27o C and at
a charge rate of C/20, the gassing voltage of about 2.35 volts per cell is reached at
about 90% state of charge (point A). At a charge rate five times faster (C/5) at 27o
C, the gassing voltage is reached at only 75% state of charge (point B). At a low
battery temperature of 0o C gassing does not occur until about 2.5 volt per cell, or 15
volts for a nominal 12-volt battery.
Lead-Acid Battery Charging Voltage as a

Function of State of Charge
3.0
2.9
Lead-Antimony Grids
Charge Rate
Cell Voltage (volts)
2.8
C/2.5
2.7
2.6
Gassing Voltage at 0
C/5
2.5
C/20
2.4
2.3
2.2
2.1
2.0
0
20
40
60
80
100

B

12-52
Suggested Voltage Regulation Set

Points
o
Some recommended ranges for charge regulation voltages at 25 C for different
battery types used in PV systems are presented in the table below. These values
are typical of voltage regulation set points for battery charge controllers used in
small PV systems. These recommendations are meant to be only general in nature,
and specific battery and regulator manufacturers should be consulted for their
suggested values.
Regulator Parameters
Battery Parameters
Regulator
Design Type
Charge
Regulation
Voltage at
25o C
Flooded
LeadAntimony
Flooded
LeadCalcium
Sealed,
Valve
Regulated
Lead-Acid
Flooded
Pocket Plate
NickelCadmium
On-Off,
Interrupting
Per 12 volt
battery
Per Cell
14.6 - 14.8
14.2 - 14.4
14.2 - 14.4
14.5 - 15.0
2.44 - 2.47
2.37 - 2.40
2.37 - 2.40
1.45 - 1.50
Per 12 volt
battery
14.4 - 14.6
14.0 - 14.2
14.0 - 14.2
14.5 - 15.0
Per Cell
2.40 - 2.44
2.33 - 2.37
2.33 - 2.37
1.45 - 1.50
ConstantVoltage,
PWM, Linear
%

The charge regulation voltage ranges presented in the table are much higher than
the typical values presented in manufacturers literature. This is because battery
manufacturers often speak of regulation voltage in terms of the float voltage, for
when batteries are float charged for extended periods (for example, in noninterruptible power supply (UPS) systems). In these and many other commercial
battery applications, batteries can be trickle or float charged for an extended
period, requiring a voltage low enough to limit gassing. Typical float voltages are
between 13.5 and 13.8 volts for a nominal 12-volt battery, or between 2.25 and 2.30
volts for a single cell.
In a PV system however, the battery must be recharged within a limited time (usually
during sunlight hours), requiring that the regulation voltage be much higher than the
manufacturers float voltage to ensure that the battery gets as much charge as
possible in a short period of time. If the charge regulation voltage in a typical PV
system were set at the manufacturers recommended float voltage, the batteries
would never be fully charged.

12-53
Temperature Compensation
As discussed previously, the electrochemical reaction and gassing in a battery is
highly dependent on temperature. Lower battery temperatures slow down the
reaction, reduce capacity and raise the voltage required for gassing. Conversely,
higher temperatures accelerate the reaction, increase grid corrosion, and lower the
gassing voltage.
For these reasons temperature compensation (TC) of the VR set point is
recommended in PV systems where battery temperature might vary more than 5 o C
from 25o C. Temperature compensation is also required for all type of sealed VRLA
captive electrolyte batteries where any gassing means permanent loss of electrolyte
and life. By using TC a battery can be fully charged during cold weather, and
protected from overcharge during hot weather.
A widely accepted value of temperature compensation for lead-acid batteries is -5
mV/ o C /cell. For a nominal 12 volt battery, this amounts to -30 mV /o C. Where
battery temperatures vary by as much as 30o C, temperature compensation may
result in the regulation set point varying by as much as 1.0 volt in a 12 volt system.
It is important to notice that the TC coefficient is negative, meaning that increases in
temperature require a reduction in the charge regulation voltage. If battery
temperatures are lower than the design condition, the regulation voltage is increased
to allow the battery to reach a moderate gassing level and fully recharge.
Conversely, the regulation set point is reduced if battery temperatures are greater
than design conditions.
Battery temperatures may be sensed with an external probe connected to the
regulator, or approximated with an on-board sensor in regulator circuitry.
Typical Temperature Compensation Coefficient

-.005 volts/o C
for one lead-acid cell
-.030 volts/o C
for a 12-volt lead-acid battery

12-54
If the electrolyte concentration has been adjusted for local ambient temperature
(increase in specific gravity for cold environments, decrease in specific gravity for
warm environments) and temperature variation of the batteries is minimal,
compensation may not be as critical. Typically, the LVD set point is not temperature
compensated unless the batteries operate below 0o C on a frequent basis.
A graphical representation of the change in the charge regulation voltage with
temperature for a typical 12-volt lead acid battery is shown in the figure below. At
25o C the charge regulation about 14.5 volts. As the battery temperature decreases,
the charge regulation voltage increases. Conversely, as the battery temperature
increases, the charge regulation voltage must be reduced.
Final Charging Voltage
for a 12 volt battery (volts)
16
15
14
13
10
30
20
40
Battery Temperature (deg.C)
50


12-55
Example:
A 12-volt battery has a nominal final charge voltage of 14.5 volts at

standard temperature (25o C).
If the battery operated in the winter at about 15o C., the final charge
voltage should not be 14.5 volts but should be increased to almost
14.8 volts.
Voltage Change
Final voltage
factor X temperature change
-.030 volts/o C X (15o C - 25o C)
+.3 volts
14.5 + .3
14.8 volts
And conversely, if the battery is in a hot climate, and the battery

temperature is 35o C., the final charge voltage should be reduced to
about 14.2 volts.
With temperature compensation built into the regulator, batteries in cold climates will
receive the full charging they require, and batteries in hot climates will not be
overcharged and gassed excessively. It is often considered an option, but should be
incorporated into every system. It should always be included when sealed batteries
are used, as they are especially sensitive to temperature effects. The battery is too
important and expensive a component to not include such a feature.

12-56
Exercises

*

#

8! 4
$ #
#

# )

9.0) #1+2#
((((((((((

:

0

#*
#

#+ #
# #
#* -

12-57
Oversizing Charge Regulators

The regulator must not only be able to handle typical or rated voltages and currents,
but must also be sized to handle expected peak or surge conditions from the PV
array or required by the electrical loads that may be connected to the regulator. It is
extremely important that the regulator be adequately oversized for the intended
application. If an undersized regulator is used and fails during operation, the costs
of service and replacement will be higher than what would have been spent on a
regulator that was initially oversized for the application.
Typically, we would expect that a PV module or array produces no more than its
rated maximum power current at 1000 W/m2 irradiance and 25o C module
temperature. However, due to possible reflections from clouds, water or snow, the
sunlight levels on the array may be enhanced up to 1.3 times the nominal 1000
W/m2 value used to rate PV module performance. The result is that peak array
current could be 1.3 times the nominal peak rated value if reflection conditions exist.
For this reason the peak array current ratings for charge regulators should be sized
for at least 130% of the nominal peak maximum power current ratings for the
module or array.
The figure below shows the irradiance profile for a typical clear day and for a cloudy
day with periods of enhanced irradiance due to cloud-reflectance. Note that during
the cloudy day irradiance levels peaked above 1300 w/m2!
&' ('$$$)*+

12-58
The size of a regulator is determined by multiplying the peak rated current from an
array times this enhancement safety factor. The total current from an array is
given by the number of modules or strings in parallel, multiplied by the module
current. To be conservative, the short-circuit current (Isc) is generally used instead
of the maximum power current (Imp). In this way, shunt type regulators that operate
the array at short-circuit current conditions are covered safely.
Regulator Size (amps)
Example:
# Modules or Strings in Parallel X Isc X 1.3
An array of 10 parallel 35-watt modules is needed for a 12-volt remote

home system.
Total array current =
Regulator size
10 modules X 2.15 amps Isc
21.5 amps array Isc
21.5 amps X 1.3 safety factor
28 amps
In the example, the array would produce 21 amps of current, but the regulator
should be sized to control up to 28 amps of current. A single regulator rated at 30
amps would be adequate, or two regulators rated at 15 or 20 amps could be used in
parallel. If two controllers were chosen the array would be split into two sub-arrays
of five modules each, and the output from the two regulators would be combined into
one battery. The two regulators could be set to slightly different final charge
voltages, and a two-step charging system would be the result.
Consult with regulator manufacturers to determine if they have already built a safety
factor into their rating value. Oversizing the regulator by 130% may not be
necessary if you are sure that the regulator design can handle any possible high
currents for short periods.

12-59
Exercises

)
### 3

;

&'+ ,#

-"#
#

##

#

'!

'

%!

%

)

' #

<1%=
###
*

&%!> # #

#
/

#

?

@

&!


12-60
Operating Without a Charge

Regulator
In most cases a charge regulator is an essential requirement in stand-alone PV
systems. However there are special circumstances where a charge regulator may
not be needed in small systems with well-defined loads. Beacons and aids to
navigation are a popular PV application, which operate without charge regulation.
By eliminating the need for the sensitive electronic charge regulator, the design is
simplified, at lower cost and with improved reliability.
The system design requirements and conditions for operating without a charge
regulator must be well understood because the system is operating without any
overcharge and overdischarge protection for the batteries. There are two cases
where battery charge regulation may not be required:
when a low voltage self-regulating module is used in the proper climate

when the battery is very large compared to the array.
Each of these cases is discussed next.

12-61
Using Low-Voltage Self-Regulating

Modules
The use of low-voltage or self-regulating PV modules is one approach used to
operate without battery charge regulation. This does not mean that the modules
have an electronic charge regulator built-in, but rather it refers to the low voltage
design of the PV modules. When a low voltage module, battery and load are
properly configured, the design is called a self-regulating system.
Typical silicon power modules used to charge nominal 12-volt batteries usually have
36 solar cells connected in series to produce and open-circuit voltage of greater than
21 volts and a maximum power voltage of about 17 volts. Why do we generally use
modules with a maximum power voltage of 17 volts when we are only charging a 12volt battery to maybe 14.5 volts? Because voltage drops in wiring, disconnects
overcurrent devices and controls, as well as higher array operating temperatures
tend to reduce the array voltage measured at the battery terminals in most systems.
By using a standard 36 cell PV module we are assured of operating to the left of the
knee on the array I-V curve, allowing the array to deliver its rated maximum power
current. Even when the array is operating at high temperature, the maximum power
voltage is still high enough to charge the battery. If the array were operated to the
right of the I-V curve knee, the peak array current would be reduced, possibly
resulting in the system not being able to meet the load demands.
In the case of using self-regulating modules without battery charge regulation we
want to take advantage of the fact that the array current falls off sharply as the
voltage increases above the maximum power point. In a self-regulating low voltage
PV module, there are generally only 28-30 silicon cells connected in series, resulting
in an open-circuit voltage of about 18 volts and a maximum power voltage of about
15 volts at 25o C. Under typical operating temperatures, the "knee" of the IV curve
falls within the range of typical battery voltages. As a battery becomes charged
during a typical day, its voltage rises and results in the array operating voltage
increasing towards the maximum power point or knee of the IV curve. In addition,
the module temperature increases, resulting in a reduction of the maximum power
voltage. At some point, the battery voltage high enough that the operating point on
the IV curve is to the right of the knee. In this region of the IV curve, the current
reduces sharply with any further increases in voltage, effectively reducing the charge
current and overcharge to the battery.

12-62
Self-Regulation Using
Low-Voltage Module
Phoenix in July at 3:00 p.m.
Current (amps)
2
M55
1
M65
8
10 12
Voltage (volts)
14
16
18
20
!
The figure shows a comparison of operating points in the afternoon in a hot climate
between a 36-cell (M55) module and a "self-regulating" 30-cell (M65) module. As
the battery voltage rises there is a more dramatic reduction in current from the 30cell module. In the afternoon, in this example, the battery voltage has risen to about
14.4 volts, and the current from the 30-cell module is almost one third that from the
36-cell module. The battery is reaching full charge, and the current should be
reduced. The reduction in charging current that would be accomplished by a charge
regulator for the 36-cell module is performed automatically with the low-voltage 30cell module.
Self-regulating modules are popularly used in the recreational vehicle (RV) market.
A single 30-cell module can be safely matched to a single large 120-150 Ah battery
without needing a regulator. The current from the module in the afternoon is low
enough so that the battery is not overcharging.

12-63
However, there are conditions that must be met, even with RVs, for a selfregulating design to work properly. If more than one module were connected to
only one 120-150 Ah battery, or if the battery were too small (less than 120 Ah), then
overcharging could occur each day, and a small charge regulator would be needed
in the system. Other factors must also be considered before using a self regulating
module. These are discussed below.
Special Considerations For Using Self-Regulating Module

Using a "self-regulating module" does not automatically assure that a photovoltaic
power system will be a self-regulating system. For self-regulation and no battery
overcharge to occur, the following three conditions must be met:
1. The load must be used daily. If not, then the module will continue to
overcharge a fully charged battery. Every day the battery will receive excessive
charge, even if the module is forced to operate beyond the "knee" at current
levels lower than its Imp. If the load is used daily, then the amp-hours produced
by the module are removed from the battery, and this energy can be safely
replaced the next day without overcharging the battery. So for a system to be
"self-regulating", the load must be consistent and predictable. This eliminates
applications where only occasional load use occurs, such as vacation cabins or
RVs that are left unused for weeks or months. In these cases, a charge
regulator should be included in the system to protect the battery.
2. The climate cannot be too cold. If the module stays very cool, the "knee" of
the IV curve will not move down in voltage enough, and the expected drop off in
current will not occur, even if the battery voltage rises as expected. Often "selfregulating modules" are used in arctic climates for lighting for remote cabins for
example, because they are the smallest and therefore least expensive of the
power modules, but they are combined with a charge regulator or voltage
dropping diodes to prevent battery overcharge.
3. The climate cannot be too hot. If the module heats up too much, then the drop
off in current will be too extreme, and the battery may never be properly
recharged. The battery will sulfate, and the loads will not be able to operate.
In summary, a self-regulating system design can simplify system design by

eliminating the need for a charge regulator, however these type of designs are only
appropriate for a very narrow range of applications and conditions. In most all
stand-alone PV system designs, a charge regulator is required.

12-64
Using a Large Battery or Small Array

A charge regulator may not be needed if the charge rates delivered by the array to
the battery are small enough to prevent the battery voltage from exceeding the
gassing voltage limit when the battery is fully charged and the full array current is
applied. In certain applications, a long autonomy period may be used, resulting in a
large amount of battery storage capacity. In these cases, the charge rates from the
array may be very low, and can be accepted by the battery at any time without
overcharging. These situations are common in critical application requiring large
battery storage, such as telecommunications repeaters in alpine conditions or
remote navigational aides. It might also be the case when a very small load and
array are combined with a large battery, as in remote telemetry systems.
In general a charging rate of C/100 or less is considered low enough to be tolerated
for long periods even when the battery is fully charged. This means that even during
the peak of the day, the array is charging the battery bank at the 100-hour rate or
slower, equivalent to the typical trickle charge rate that a regulator would produce
anyway.
To determine if a regulator is needed calculate the peak rate of charge expected
from the array and see if this rate is near or slower than the 100 hour rate. The
number of modules in parallel is multiplied by the module peak current (Imp) to give
the array peak current. Dividing the battery capacity by this current gives the hours
to fully charge the battery.
Peak Charge Rate (hours) =

Installed Battery Capacity (Ah)

Number of Parallel Modules X Imp
12-65
Example:
Four 75-watt modules are wired in parallel and connected to a battery

bank of four 120 Ah batteries in parallel.
Peak Charge Rate
120 Ah X 4 batteries in parallel

4 modules X 4.4 amps
480 Ah total
17.6 amps total
27 hours or the C/27 rate.
Because the 27-hour rate is faster than the general guideline of 100hour rate, this system would indeed need a charge regulator.
But if more autonomy were built into the battery bank, the ratio of charging current to
capacity would be reduced.
Example:
If five 360 Ah batteries were connected in parallel, the battery capacity

would be 1800 Ah, and the charge rate from four 75 watt modules
would be lower:
Peak Charge Rate
360 Ah X 5 batteries in parallel

4 modules X 4.4 amps
1800 Ah total
17.6 amps total
102 hours or the C/102 rate.
This example meets the general criteria presented, so this system

could operate safely without a charge regulator between the array and
battery bank. Proper fuses and disconnects would also be required for
safety.
The ratio of module charge rate to battery capacity depends on the local weather
data and the desired reserve or autonomy capacity built into the battery bank, so it
varies for each location and application. The 100-hour rate does not occur until
there is about 15 days of autonomy or more designed into the battery. This amount
of autonomy is unusual, and would occur only for systems that are designed to
operate in climates with extremely bad seasonal weather. This might be the case for
remote telecom repeaters in mountainous regions. In that type of application, the
cost savings of not including a regulator are minor compared to the cost of the
overall system. A charge regulator would probably be included anyway, for added
system control.
12-66
The preceding discussion illustrates that the conditions and circumstances

that allow photovoltaic systems to be designed without charge regulators are
quite special. In most applications, a charge regulator is required.
Exercises

<#

# :#
=

$

#
0

# "
*
,
3#
&!!

(((((((((()
; 3

##

#

#

)

;
#&!) A&';
3 '!

##

*
# ##
?!> ##
)
# #

#
*
,
'!B&!) AA?!

C' !)
##
)2).%
<<#4D

% ' "

4#
3#

3+
#
((((((((((
3#

12-67
Selecting Charge Regulators

The selection and sizing of charge regulators and system controls in PV systems
involves the consideration of several factors, depending on the complexity and
control options required. While the primary function is to prevent battery overcharge,
many other functions may also be used, including low voltage load disconnect, load
regulation and control, control of backup energy sources, diversion of energy to and
auxiliary load, and system monitoring. The designer must decide which options are
needed to satisfy the requirements of a specific application. In review of the material
presented in this chapter, the following table lists some of the basic considerations
when selecting charge regulators and controls.
Charge Regulator Selection Criteria
System voltage
PV array and load currents
Battery type and size
Regulation algorithm and switching element design
Regulation and load disconnect set points
Environmental operating conditions
Mechanical design and packaging
System indicators, alarms, and meters
Overcurrent, disconnects and surge protection devices
Costs, warranty and availability

12-68
Selected References
Stand-Alone Photovoltaic Systems - A Handbook of Recommended Design Practices, Sandia National
Laboratories, SAND87-7023, revised November 1991.
Naval Facilities Engineering Command, Maintenance and Operation of Photovoltaic Power Systems,
NAVFAC MO-405.1, December 1989.
Exide Management and Technology Company, Handbook of Secondary Storage Batteries and Charge
Regulators in Photovoltaic Systems - Final Report, for Sandia National Laboratories,
SAND81-7135, August 1981.
Bechtel National, Inc., Handbook for Battery Energy Storage in Photovoltaic Power Systems, Final Report,
SAND80-7022, February 1980.
S. Harrington and J. Dunlop, "Battery Charge Controller Characteristics in Photovoltaic Systems",
Proceedings of the 7th Annual Battery Conference on Advances and Applications, Long Beach,
California, January 21, 1992.

12-69
(End of Chapter)

12-70
CHAPTER TWELVE
CHARGE REGULATORS AND SYSTEM CONTROLS
12-1
Purpose of Charge Regulators and System Controls

Prevent Battery Overcharge
Prevent Battery Overdischarge
Provide Load Control Functions
Provide Status Information To Users
Interface and Control Backup Energy Sources
Divert PV Energy To Auxiliary Load
Serves As Wiring Center
12-2
12-3
12-4
12-5
12-5
12-6
12-6
12-7
Terminology
Nominal System Voltage
Nominal Load and PV Array Current
12-8
12-8
12-8
12-9
Other Functions Associated With Charge Regulation

Backup Energy Source Control
Equalizing Charge Capability
Set Point Adjustability
Load Voltage Regulation
Regulation/Control Element Design
Operational Limits
Surge Protection and Grounding
Service Disconnects and Overcurrent Protection
12-16
12-16
12-16
12-16
12-17
12-17
12-18
12-18
12-19
Standard Configurations Of Charge Regulation and

Control Systems
Simple Series Path Configuration
Auxiliary Load Path Configuration
12-20
12-20
12-21
12-23
12-24
Electronic Designs for Charge Regulation

Shunt Regulator Designs
Series Regulator Designs
12-28
12-29
12-33

12-71
Daily Operational Profiles for Charge Regulators

About the Charge Regulator Daily Profiles
Daily Profile
1-Step Series-Interrupting Regulator
Daily Profile
Modified Series Interrupting 2-Step Constant Current
Charge Regulator Daily Profile
Daily Profile for Constant-Voltage Series Charge Regulator
PWM Series Charge Regulator Daily Profile
12-37
12-37
12-39
12-39
12-42
12-42
Voltage Regulation Set Point Selection

Battery Gassing - Key to Voltage Regulation Set Point Selection
Suggested Voltage Regulation Set Points
12-51
12-51
12-53
12-54
Oversizing Charge Regulators
12-58
Operating Without a Charge Regulator

Using Low-Voltage Self-Regulating Modules
Using a Large Battery or Small Array
12-61
12-62
12-65
Selecting Charge Regulators
12-68

12-72
12-44
12-46
12-48
Chapter 12 Answers
Charge Regulators & System Controls

According to the data sheet, an SP75 module has a peak power of 75 Watts and an Isc
of 4.8 amps (when configured at 12 Volts).
Since the array is 750 Watts total, there must be 750 75 = 10 modules.
With a 12-volt system, the array will have all 10 modules in parallel. The maximum
amount of current that the array can produce is 10 times the Isc of one module. We
can therefore calculate the number of regulators required as
Number of Regulators
10 parallel X 4.8 amps X 1.30 (Safety factor)

30 amps/regulator
2.08 regulators, rounded up to 3

The array of 1200 Watts is split into two sub-arrays. Therefore, each sub-array is 600
Watts. Using 75-Watt modules, the number of modules in one array is
600 watts 75 watts/ module = 8 modules
At 12 volts, the array is configured as 8 modules in parallel. We use 4.8 amps as the
Isc for the SP75 modules. Since the sub-array is supposed to have a single regulator it
must handle the full amount of current for the 8 modules (including the safety factor).
Size of Regulator
8 parallel X 4.8 amps X 1.30 (Safety factor)
49.92 amps
This is greater than 30 amps, so we can not use a 30-amp regulator. However, we can
use a 50-amp regulator since the maximum expected current is slightly less than this
value. The correct answer is
d. 50 amps.

12-1
Charge Regulators and System Controls

The shaded area shows indicating greater total charge being put into the battery over
the period of the day.
Irradiance (W/m2)
1000
10
Constant
voltage
7.5
800
5
600
2.5
400
0
Modified
series
-2.5
200
-5
0
4
12
16
20
24
Time of Day (EST)

According to the Voltage Regulation Setpoint Table, the proper ending voltage for a
sealed VRLA battery is 2.33 - 2.37 volts when using a PWM controller. We will use the
average of this range, 2.35 volts, as our starting point. For a nominal 12-volt system,
the ending voltage will be equal to 6 times this (since there are 6 two-volt cells).
6 X 2.35 volts = 14.1 volts.
So, under normal circumstances (25 C), we would stop charging when the battery
reached 14.1 volts. However, this needs to be reduced because of the high
temperatures.
Voltage Change
Final voltage
factor X temperature change
-0.030 volts/C X (40 C - 25 C)
-0.45 volts
14.1 - 0.45 volts = 13.65 Volts
12-2


At 12 volts, all four modules will be in parallel. The maximum array current will be four
times the Isc of the module.
Total array current =
Regulator size
4 modules X 4.8 amps Isc
19.2 amps
19.2 amps X 1.3 safety factor
24.96 amps
The necessary size is greater than the smallest regulator size available (20 amps). We
need to use a regulator that is greater than 24.96 amps, so the correct answer is
b. 25-amp regulator

The shunt regulator is rated for 25 amps. However, this needs to include the 1.30
safety factor. The amount of array current that is allowed is:
Array Current
25 amps
1.30
19.23 amps
Using an Isc of 2.4 for the SP36 module, the maximum number of modules is
Maximum Modules
19.23
2.4
8.01
We round down and can use a maximum of 8 modules. The correct answer is:
b. 8 modules
12-3

According to the module literature, a 6-Watt module has a Imp of 0.39 amps. The peak
charge rate is determined as
Peak charge rate
Battery capacity
Array peak amps
Battery capacity
Peak charge rate X Array peak amps
100 hours X 0.39 amps
39 Ah
So,
Remember, if the battery size is greater than 39 Ah, the charge rate will be slower
than a C/100 peak charge rate. The decision to use a charge regulator would depend
on the available types of regulators, the pattern of load use and the environment. Since
the application is located in the desert (hot conditions) this might not be a good situation
to use a self-regulating module.

The required battery capacity is given as 250 Ah battery capacity. The peak charge
rate is:
Peak charge rate
250 Ah
2.5 amps
100 hours peak charge rate
This is just at the edge of the recommended peak charge rate for self-regulated system.
A regulator is probably not necessary from a charging standpoint. However, the
decision whether or not to use a regulator requires good judgement of the advantages
and disadvantages, as well as attention to overall system needs.
12-4
Chapter Thirteen
Energy Enhancement
Through Tracking
Technology
There are two methods for enhancing the output of a photovoltaic power system that
are commonly employed by system designers. Both of these approaches improve
the match between the potential output of solar modules and the load. Both
methods involve the word tracking, although one is electronic and the other is
mechanical. In this chapter we will discuss these methods for getting more out of
you solar generators.
The first method involves physically tracking the sun as it moves throughout the day
and the seasons. The second method involves electronically monitoring the
maximum power point of an array and transferring this maximum power to the loads.

13-1
Components Tracking
Tilting The Array To Match The

Load Profile
The most basic approach to tracking the sun position is to mount the array at a
fixed angle tilted up from horizontal so that the module surface faces the sun as it
passes through the sky each day. The mounting hardware can be ground mounted
or have the modules mounted at the top of a pole.
The sun path is high in the summer and low in the winter. The best system design
has the array tilted to an angle so that the variation or profile of insolation throughout
the year matches the load profile.
An example of the variation in insolation throughout a year is shown. The insolation
on a flat surface varies greatly, with the least during the winter and the greatest
during the summer. If the load demand is small in winter and large in summer, as
with air conditioning loads or water pumping for irrigation, then tilting the array near a
flat angle will give an insolation profile that best matches those load requirements.
If however, the load is relatively constant every month of the year, as might be the
case for a navigational aid or a constantly transmitting repeater, then a different
angle is better. Tilting the array up increases the insolation intercepted during the
winter months and sacrifices some during the summer months, with a resulting
profile that is more constant throughout the year. This more constant insolation
profile better matches the profile of a constant load.
Notice that as the tilt angle increases from horizontal, that a double peak profile
emerges. The insolation profile never gets perfectly flat. The spring and fall have
more insolation than the winter or the summer. This double-peak profile is typical for
latitudes above the Tropic of Cancer or Capricorn. Nearer to the equator, the suns
path actually passes overhead twice during the year. During the spring and fall, the
path will pass nearly straight overhead. During the winter and summer, the path will
actually be further down from straight above.
If a tilt angle is chosen that does not match the insolation profile to the load profile,
then the array will have to be quite large to produce enough output when insolation
is low and will produce wasted energy when insolation is high. By choosing a tilt
angle for the array that gives the best match, array size is minimized. Computer
models or repetitive hand calculations can predict the best angle to give a match
between load demand and insolation.

13-2
Components Tracking
New Delhi Insolation Profiles

Latitude 29 deg. N
600
500
400
300
200
100
0 deg.
15 deg.
30 deg.
Dec
Nov
Oct
Sep
Aug
Jul
Jun
May
Apr
Mar
Feb
Jan
45 deg.
60 deg.
Trivandrum Insolation Profiles

Latitude 8 deg. N
600
500
400
300
200
100
0 deg.
10 deg.
20 deg.
30 deg.
Dec
Nov
Oct
Sep
Aug
Jul
Jun
May
Apr
Mar
Feb
Jan
40 deg.


13-3
Components Tracking
Exercise

Trivandrum
Insolation Profile Data (Langleys)
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
-15 deg.
0 deg.
15 deg.
30 deg.
45 deg.
404
452
495
474
454
382
388
428
433
389
359
366
461
494
515
471
436
362
371
419
441
413
399
422
519
533
528
461
414
342
352
406
444
436
442
481
546
540
509
424
367
302
313
370
421
433
458
511
540
516
464
366
305
252
262
317
378
408
447
509
Leh
Insolation Profile Data (Langleys)
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Latitude: 8 deg. N
Latitude: 34 deg. N
0 deg.
15 deg.
30 deg.
45 deg.
60 deg.
75 deg.
207
262
332
438
476
503
494
457
422
340
262
209
268
317
375
464
484
502
497
476
467
404
336
278
316
355
394
461
462
471
470
465
481
445
393
335
343
369
388
434
419
420
423
431
467
456
423
369
349
363
364
388
361
356
361
380
430
442
427
381
334
336
323
327
293
284
290
314
373
404
406
369

13-4
Components Tracking
Trivandrum Insolation Profiles

600
500
400
300
200
100
Jan
Feb
Mar
Apr
May
Jun
July
Aug
Sep
Oct
Nov
Dec
Apr
May
Jun
July
Aug
Sep
Oct
Nov
Dec
Leh Insolation Profiles

600
500
400
300
200
100
Jan
Feb
Mar

13-5
Components Tracking
Sun Position Tracking

Although fixed angle mounting structures are simple and worry free, modules do not
get good solar exposure in the morning and evening. By mounting the modules on a
tracking structure, gains in total daily output of power of 30% or greater can be
achieved because the modules are facing the sun directly during the morning and
evening hours. Sun position trackers can move along one axis to capture most of
the suns energy daily, or can also adjust along a second axis for the altitude of the
sun and enhance seasonal performance.
The effect of sun position tracking is shown below for Leh at 34-deg. N latitude. The
module is tilted to 50 degrees, to give the best output during the winter. The daily
enhancement is shown for the summer and for the winter, as well as the
performance over a typical year.
During the middle of the day, the single axis tracking gives the same output as the
fixed array, because they are at the same angle at that time. But in the morning and
evening, notice that the single axis tracking gives more output. The array is tilted to
face the morning or afternoon sun directly, while the fixed array is still facing due
South in the Northern Hemisphere (and due North in the Southern Hemisphere).
The increased area under the curve for tracking indicates the increased energy
produced by the modules.
The winter performance in the same location is shown below. During the winter, the
sun is making a low path with only a few hours of exposure. The sun is out in front
of the fixed array during most of the arc of the day, so tracking and fixed receive
about the same amount of exposure. The enhancement during winter is lower than
in summer because there is less sun path to track!
Tracking adjustments by employing a second axis adds some output, depending on
the month and the latitude. The sun makes the same path in the sky during spring
and fall. It deviates up about 23.5 degrees in the summer and deviates down about
23.5 degrees in the winter, for a total angle change of about 47 degrees summer to
winter. Adjusting for this seasonal change adds mostly during the summer, when
the path in the sky is large and there are many hours of exposure.

13-6
Components Tracking
One S L-35 in Le h, June

2
Amps
1.5
1
0.5
0
5
11
13
15
17
19
Ho ur
Fix ed
Single A x is
Double A x is

One S L-35 in Le h, Ja nua ry

2
Amps
1.5
1
0.5
0
5
11
13
15
17
19
Ho u r
Fix ed
Single A x is
Double A x is

Ah/day
One S L-35 in Le h
18
16
14
12
10
8
6
4
2
0
Jan
Mar
Fix ed
May
Jul
Sep
Single A x is
Nov
Double A x is

13-7
Components Tracking
The enhancement of sun position tracking will be different depending on the latitude.
The performance curves of a module in Minicoy are shown below. The latitude of
the site is only 8 deg. north of the equator. The enhancement from tracking is more
uniform throughout the year, compared to the heavier gain only in the summer for
Leh at 34-deg. latitude. Winter and summer daily enhancements are about the
same, because the suns path is always a large arc in the sky.
One SL-35 in Minicoy, June

2
Amps
1.5
1
0.5
0
5
11
13
15
17
19
Hour
Fixed
Single Axis
Double Axis

Ah/day
One SL-35 in Minicoy

18
16
14
12
10
8
6
4
2
0
Jan
Fixed
Mar
May
Jul
Single Axis
Sep
Nov
Double Axis

13-8
Components Tracking
Exercise

!

"

#

$

%

$

%
%
&%

'

%
%

13-9
Components Tracking
Solar Powered (Passive) Sun Tracker

The typical commercial solar powered tracker design involves two tubes of Freon on
the east and west sides of the array. Each tube is partially shaded by a metal sheet
or cover, as shown in Figure 13-7. As the sun moves, one tube becomes more
exposed to the sun than the other. The Freon expands and either pushes a piston
or transfers oil to the other side, which causes the structure to move to follow the
sun.
Trackers can follow the sun along only one axis or can have dual axis tracking for
complete seasonal compensation. Most large utility scale photovoltaic systems
have the modules mounted on trackers, to maximize module output and minimize
acreage costs.
At very high latitudes, the tilt angle of the structure is great, and the difference in
weight between the two sides becomes less. Therefore, it may be more difficult for
solar powered trackers to adequately perform at high latitudes.
Shadow Masks
Freon Tubes

!


13-10
Components Tracking
Electrically Powered (Active) Sun

Tracker
Electrically powered trackers use motors to move the structure holding the modules.
One design uses photocells at the base of a tall vertical block. As the sun moves,
the shadow from the block moves across the photocells. Logic in the tracker
controller tells the motors to move the structure until the photocells see the same
amount of irradiance. This keeps the block, and the modules, pointed straight at the
sun all day long. The energy needed to operate the tracker is drawn from the
modules themselves, but is quite small, about 1/2 watt during daylight hours.
Two designs for active trackers are shown below. One design uses a tilt-and-roll
design where linear motors push the array around the azimuth and tilt the array up
or down to accommodate altitude tracking. A second approach uses a vertical axis
tracking design where a gear drive mechanism rotates the array about a vertical axis
while a linear motor tilts the array for altitude adjustments. Both designs use the
same form of electronic control unit mounted on the upper edge of the array.

'()%&

"#$%
&

13-11
Components Tracking

*+),-%./
Trackers can be designed to hold from 4-24 modules or even more. Wind loading of
large tracker structures should be a concern. But most designs can handle winds up
to 40 mph or greater.
Because the tracking is driven by electronics, it can operate equally well in all
latitudes. There is no greater difficulty tracking at high latitudes as might occur with
solar powered trackers.
Exercise

(

! )

*


13-12
Components Tracking
Electronic Power Tracking

A second method for enhancing the output of a solar array into a load involves the
use of an electronic maximum power point tracking circuit (MPT) between the
modules and the load. Such a device can be used for battery charging systems, but
has its greatest benefit when used with direct coupled water pumping systems.
Power Tracking With DC Motors

We have discussed how the load determines the operating voltage for an array. DC
motors (and batteries to some extent) will probably operate an array away from its
maximum power voltage (Vmp) throughout a typical day. This means that the
module-load matching is not optimized and that system efficiency is not maximized.
A maximum power tracking device (MPT) operating between the array and load will
force the array to operate at its maximum power voltage at all times. The circuitry
chops the DC from the array at high frequency, and rectifies it back to DC but at a
different voltage and current, keeping the total POWER (current X voltage) the
same. It translates array power into power at the optimum voltage for the load. So
the array and the load operate at different voltages and currents.
Pmax (array)
Imp (array) X Vmp (array)
Power (load)
I (load) X V (load)
An IV curve is shown with the point of maximum power indicated. If this power is
translated to a lower voltage, the current must rise to keep the value of power
constant. Thus a MPT effectively changes the shape of an IV curve from having a
flat region where current is almost constant into a descending geometric curve of
constant power. Thus the current available at voltages less than the Vmp of the
module actually increases. Dont think that energy or power has somehow been
magically created. The total power available stays the same. It is just the current
and voltage values that have been changed. Their product, the power, stays the
same.
This substantially increases efficiency for direct coupled DC motors and water
pumps and can even improve battery charging efficiency under certain
circumstances. The effect for a DC motor is shown below. During the middle of the
day, the module and motor should be matched in voltage by design. But in the early

13-13
Components Tracking
morning and late afternoon, or during overcast conditions, the module would
normally supply very little current directly to the DC motor. The motor would operate
the module at point A, well away from its maximum possible power output at point C.
By using the MPT, the motor now interacts with the power curve at point B,
substantially increasing the current and the voltage. The water pump begins to pump
earlier in the day, and continues to pump later in the afternoon than would be
possible without the MPT.
M a x im u m P o w e r T ra c k in g
Current (amps)
4
curve of constant power
3.5
3
2.5
maximum power point

of module
(V m p X Im p )
2
1.5
1
0.5
0
0
12
16
Voltage (volts)
20
24

O p tim ize d O p e ra tio n W ith

D C M o to rs
C u rre n t (a m p s)
4
W ith o u t M P T
m o to r cu rve
3 .5
A
3
2 .5
m o to r
m o d u le
1 .5
1
0 .5
0
W ith M P T
C
A
M 5 5 m o d u le cu rve
P h o e n ix, A rizo n a , 8 :0 0 a .m .
4
12
16
V o lta g e (vo lts)
C
m o d u le
20
B
m o to r
24


13-14
Components Tracking
Power Tracking With Batteries

The enhancement of MPT for battery charging is less dramatic. This is because
modules are designed to charge typical 12-volt batteries, and the number of solar
cells in series has been chosen deliberately to match the voltage requirements for
charging batteries. In very cold conditions however, there may be a slight mismatch
of the battery operating voltage and the module Vmp. The cold module temperature
will keep the voltage potential of the IV curve out beyond the typical charging voltage
of the battery.
An inherent efficiency of about 95% has been assigned to the MPT device. One of
the issues of using MPT with batteries is that the enhancement must be greater than
the internal efficiency of the device itself, or there will be no net gain at all.
An example of battery charging with MPT is shown below for a hot location, Raipur,
during the coolest and hottest months. A 36-cell SL-35 module is used. Even in the
coolest month in Raipur, September, the gain from MPT is so small that the inherent
inefficiency of the MPT actually penalizes the output into the battery, compared to
directly connecting the module to the battery. During the hottest month, May, the
input into the battery is even more reduced. The module output voltage potential is
already reduced due to high temperatures, and the MPT only subtracts its
inefficiency burden from the output.
The annual performance is shown below as well. In this hot climate, MPT does not
increase input into the battery at all during the year. The annual input without MPT
is approximately 57.3 kwh/year from one module, while the input with MPT at 95%
efficient is only 56.0 kwh/year.

13-15
Components Tracking
O ne S L-35 in Ra ipur in S e pte m be r

30
Watts
25
20
15
10
5
0
5
9 10 11 12 13 14 15 16 17 18 19
Ho u r
MPT @ 95% ef f ic ienc y
Battery

One S L-35 in Ra ipur in M a y

30
Watts
25
20
15
10
5
0
5
9 10 11 12 13 14 15 16 17 18 19
Ho u r
Battery


13-16
Components Tracking
O ne S L-35 in Ra ipur
200
Wh/day
150
100
50
0
Jan
Mar
May
Jul
Sep
Nov
Battery


13-17
Components Tracking
Another example is shown, this time in Leh. The temperatures here will be lower
than in the case of Raipur, and the difference in benefit of MPT with battery charging
will show more clearly. The enhancement is positive throughout the year, higher in
the cool month of January and less in the hot month of August. But here again, the
relative heat in all months makes the gain small. The annual total charge put into a
battery directly is 50.3 kwh/year while the input through a MPT is 53.6 kwh/year.
One Sl-35 in Leh in January

25
Watts
20
15
10
5
0
5
11
13
15
17
Hour
MPT @95% ef ficiency
19
Battery

One Sl-35 in Leh in August

25
Watts
20
15
10
5
0
5
11
13
15
17
19
Hour
MPT @ 95% efficiency
Battery

!
13-18
Components Tracking
The annual performance profile shows that there is a greater contribution from MPT
in a cold climate throughout the year, again more so in the winter than in the warmer
summer.
One SL-35 in Leh

200
Wh/day
150
100
50
0
Jan
Mar
May
Jul
MPT @ 95% efficiency
Sep
Nov
Battery

'
The negative trade-off that must be considered is the cost of the device, and the
increased risk of failure. Also the MPT must overcome its own energy inefficiency
(usually about 5%) to prove economical.

13-19
Components Tracking
Summary and Comparison of

Power Tracking and Sun Position
Tracking
By way of summary, we should compare the benefits of the two types of tracking
that have been described in this chapter. Both sun position tracking and maximum
power tracking enhance the transfer of power from a solar array to a load. They do
so in different ways, and are optimum in different environmental conditions.
Simply put, a sun position tracker maximizes the input of power into the system in
the first place, while an MPT device maximizes the output of the power to the load
(but does not affect or increase the power input to the modules).
A sun tracker would be beneficial in a climate of clear sky and not extremely high
winds. If the sky is overcast often, then the tracker will not yield any benefit. It
wouldnt matter where the module was facing
An MPT tracker would be of most benefit in the overcast situation. The modules
would capture what solar energy they could, and the MPT would help transfer that
energy on to the loads.
Both types of trackers can enhance output into batteries, although MPT matching is
less dramatic than sun position tracking. But both systems would be of greatest
benefit when coupled with DC motors, as in water pumping. There are thousands of
working systems worldwide that use sun position tracking or MPT to enhance water
pumping.
Exercise

(
)

!

!


13-20
Components Tracking
(End of Chapter)

13-21
Components Tracking
CHAPTER THIRTEEN
ENERGY ENHANCEMENT THROUGH TRACKING
TECHNOLOGY
13-1
Tilting The Array To Match The Load Profile
13-2

Solar Powered (Passive) Sun Tracker
Electrically Powered (Active) Sun Tracker
13-6
13-10
13-11
Electronic Power Tracking

Power Tracking With DC Motors
Power Tracking With Batteries
13-13
13-13
13-15
Summary and Comparison of Power Tracking and

13-20

13-22
Components Tracking
Chapter 13 Answers
Energy Enhancement through Tracking Technology

Location: Trivadrum
Trivandrum
600
500
-15
0
15
30
45
400
300
200
100
0
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
The total annual insolation is
Annual
-15 Deg 0 Deg 15 Deg 30 Deg 45 Deg

152,740 158,169 162,810 157,789 144,693
13-1
Energy Enhancement Through Tracking
Location: Leh
Leh
600
500
400
0
15
30
45
60
75
300
200
100
0
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
The total annual insolation is
Annual
0 Deg 15 Deg 30 Deg 45 Deg 60 Deg 75 Deg

134,051 148,188 153,617 150,351 139,972 123,245

Location: Leh, Latitude 34 N
Month
Largest June
Smallest Jan.
Fixed Axis
Ah/day
9
8
Double Axis
Ah/day
16
10
Percent
Increase
77%
25%
Fixed Axis
Ah/day
8
13
Double Axis
Ah/day
13
17
Percent
Increase
62%
31%
Location: Minicoy, Latitude 8 N

Month
Largest June
Smallest Jan.
The differences in enhancement are due to the different site latitudes.
13-2


13-3
Chapter Fourteen
Diodes in Photovoltaic
Systems
This chapter discusses the function of a very specific piece of equipment used in
photovoltaic systems, the solid state diode. This device performs two different
functions, depending on where it is installed, and those functions are often confused
and misunderstood. The purpose of this chapter is to clarify the differences in
performance and to make clear why diodes are used in photovoltaic systems.
A diode is a solid state device made of P-type and N-type silicon. In fact a silicon
solar cell is a diode. Solar cells use the internal electrostatic field of their P/N
junction to prevent electrons from flowing back into the cell after they have been
knocked loose by light. Diodes use their internal field to allow electric current to flow
one way in a circuit and prevent it from flowing back.
When diodes are installed in series with a string of modules they perform a blocking
function, preventing backflow down the module string. When diodes are installed in
parallel with modules they perform a bypass function allowing current to pass around
a shaded area of a module. These two functions will be explained next.

14-1
Components Diodes in PV Systems
Bypass Diodes For Shading

Shading Causes Mismatched Currents
When part of a photovoltaic module is shaded the shaded cells will not be able to
produce as much current as the unshaded cells. Since all the cells are connected in
series the same amount of current must flow through every cell. The unshaded cells
will force the shaded ones to pass more current than their new Isc. The only way
the shaded cells can operate at a current higher than their Isc is to operate in a
region of negative voltage, that is to cause a net voltage loss to the system. The
current times this negative voltage gives the negative power produced by the shaded
cells. In other words, the shaded cells will dissipate power as heat and cause "hot
spots". And the shaded cells will drag down the overall IV Curve of the group of
cells.
Mismatched Cells in Series

unshaded cell
25% shaded cell
series combination


14-2
Effect Depends On How Shaded

A module curve is shown with varying degrees of shading. Even with only one cell
out of 36 shaded by 50%, there is a significant loss of power in the voltage range of
battery charging. One cell completely shaded is even worse, but note that the
module is not "turned off" by one completely shaded cell.
For a module with three cells shaded the impact is of course worse still. But notice
that the effect of 25% shading on three cells is not as bad as 75% shading on one
cell, the same total area of shading. Having the shading spread over many cells is
not as severe as having all the shading located in one or a few cells.
Effect of Shading on Output

Potential
One cell shaded
Three cells shaded
same amount of shading

Current 3
Current 3
25%
25%
2
50%
50%
75%
-10
75%
100%
10
100%
20
-10
Voltage
10
20
Voltage


14-3
How Bypass Diode Reduces Loss

An example of a module partially shaded will illustrate how bypass diodes can help
to reduce the looses associated with shading. Lets look at two cases, one with a
module with three cells 50% shaded, and another with three cells 100 % shaded.
The exact amount of loss of power will depend on the amount of shading and on
where the shaded module operates on its new reduced IV curve. Where a shaded
module operates depends on the operating current of the other modules
connected to it.
In the example in the figure, the shaded module is connected in series with many
other modules. The other unshaded modules try to operate at a their normal peak
current, creating a high current level. If the new reduced Isc of the shaded module
remains above the operating current of the other series modules in the group, then
the shaded module will be able to operate in the forward part of its curve (at point A
in the figure), and will still be able to contribute voltage to the series string. The
voltage out of the module or cell string will be higher than going into the module, as
is the normal case.

14-4
Shaded Module In Series String

Isc for shaded
cells is below
the Isc for
unshaded cells.
Isc for unshaded cells

Level of current in array is
determined by other modules
A
3 cells 50% shaded
3 cells 100% shaded
-10
-5
Shaded module removes

voltage from system
10
15
20
Shaded module can still contribute voltage
(-)
Voltage
(+)
Current
Current
Voltage
(-)
(+)


14-5
However when the shading in a module is severe enough so that the new Isc of the
shaded module falls below the operating current of the other series modules in the
group, then the shaded module as a whole is forced to operate at a negative voltage
to pass that level of current (point B in the figure). The entire module subtracts
voltage from the group of modules. The voltage out of the module is actually lower
going into the module. It would be better if the entire module were taken out of the
circuit temporarily. It is also important to limit the amount of local heating that result
from this negative voltage to prevent damage to the plastic and stress to the cells.
This can be done by installing diodes in parallel with groups of cells in each module.
These are called bypass diodes, because they allow current to pass around
shaded cells and thereby reduce the voltage losses through the module.
A diode passes current when the voltage is higher on one side, but blocks any
current flow when the voltage is higher on the other side. In condition A, the diode
sees the same voltage polarity as the module, with the voltage going out of the
module higher than going in. Current that tries to flow from the high voltage side of
the module to the lower voltage side through the diode is blocked. The diode is
"reverse biased" and does not allow any current to backflow around the module. The
current produced by the shaded module passes on to the next module. It is as if the
diode were not in the circuit at all.
However under condition B, where enough cells become shaded so that the module
is forced to operate at a negative voltage, the final voltage going out of the module is
actually lower than the voltage going in. The voltage polarity shown is positive at
the beginning of the module and negative at the end, the opposite of the normal
case. The voltage across the diode becomes reversed ("forward biased") and it
conducts current through itself. All the current greater than the shaded cell's new Isc
is "bypassed" through the diode, thus reducing drastically the amount of local
heating at the shaded area. The diode also holds the entire shaded module or
group of cells to a small negative voltage of approximately -0.7 volts, thus limiting the
reduction in array output.

14-6
Severely shaded module is

limited to operating at only
-0.7 volts by the bypass
diode.
Shaded
Module
With
Bypass
Diode
-10
10
20
Current at
Current above
passes through
passes through diode
module
No current
passes
backwards
through bypass
diode
(-)
All current
passes through
module and
(+)
on to loads.
(+)
(-)


14-7
Placement Of Bypass Diode In Module

Cell Circuit
The simplest way to include bypass diodes in a module cell circuit is to connect the
negative of the diode to the negative starting point of a module, and the positive of
the diode to the positive end of the module. When there is severe shading, then the
bypass diode will bypass the entire module circuit.
To preserve more of the modules voltage, a diode can be placed around a portion
of the cells in the circuit. Two methods can be used due to different junction box
configurations. For modules with three strings of cells, as shown below, with the
circuit beginning at one end of the module and ending at the opposite end, one
bypass diode is installed in each of the two junction boxes at opposite ends of the
module. If the cell circuit is stretched out into a long string, you can see more clearly
how each diode is connected around 24 cells, and how the two diodes overlap in cell
coverage.
In the case of severe shading, current can flow through a group of 12 cells and then
through the bypass diode. Thus only 24 cells or 2/3 of the module circuit is
bypassed.
Dual Bypass Diode Placement

Overlapping Design


14-8
Another typical circuit design has four strings of cells, with the circuit beginning and
ending at the same end of the module. Only one junction box is used, and both
diodes are placed inside that box. Looking at the cell circuit stretched out, each
diode is connected around 18 cells and there is no overlap. In the case of severe
shading, current can pass through an unshaded group of 18 cells and then into a
bypass diode. Thus only 18 cells or 1/2 of the module is bypassed. This design
allows for even less voltage loss than the three-string configuration.
Dual Bypass Diode Placement

Adjacent Design


14-9
An Example Of Bypass Diode Reducing

Losses
An example serves to illustrate the effect of bypass diodes in (1) reducing the
amount of local heating and (2) preserving the array output to the load. Five 53-watt
36-cell modules (Siemens SM55) are connected in series with a nominal 60-volt
battery bank. The battery bank sets the operating voltage for the system at about
67.5 volts, which then determines the array current at 2.4 amps. Each module
therefore operates at 2.4 amps and each module contributes 13.5 volts (67.5 / 5
modules = 13.5 volts).
Array Of Five Modules

In Series With No Shading
Curve for
one module
-20
20
40
60
80
100
Battery determines voltage

for entire array.


14-10
For this example, assume that three cells in one module become 75% shaded. We
first show the result without bypass diodes, and then on the next page show the
effect with bypass diodes in place.
The entire array output potential is dragged down. The battery as always
determines the operating point for the array, and now results in an array current of
about 1.3 amps. The new Isc for the shaded module is only 1.1 amps, so the
shaded module is forced to operate at a negative voltage of -8 volts. The negative
voltage times the current gives the net power dissipated by the shaded module:
-8 volts X 1.3 amps = -10.4 watts
So about 10 watts of power is lost in shaded cells, in the form of heat. Remember
that an M55 module produces only 53 watts peak at standard conditions, and about
45 watts in field conditions when it is heated. So this potential power loss is quite
high.
Three Shaded Cells

Reduces Entire Array Output
Array current is
dragged down
by shaded cells
Shaded
module
is forced
to operate
at a
negative
voltage
-20
20
40
60
80
100
Voltage (volts)
Battery determines voltage
for entire array.


14-11
(For simplicity, we show the benefit of one single bypass diode around the entire
module). With a bypass diode installed in parallel with the shaded module, the
result is quite different. There are two regions to examine one below the Isc of the
shaded module and the other above it.
Looking at current levels below the new Isc of the shaded module, everything is left
as is. The bypass diode can do nothing here.
Looking above the new Isc of the shaded module however, the shaded module
cannot operate beyond approximately -0.7 volts, as this is the maximum negative
voltage allowed by the diode. So the output potential of the array is not dragged
down as severely (-0.7 volts instead of -8 volts!). For current levels above the new
Isc of the shaded module, the shaded module is completely bypassed, so the
composite curve for the entire array is that of only four unshaded modules instead of
five.
The battery voltage now operates the entire array at a current of almost 2.0 amps,
nearly back to the full-unshaded potential (instead of only 1.3 amps as before with
no diode).
The shaded module operates at a net power loss of only:
-0.7 volt X 2.0 amps = -1.4 watts
Thus the total dissipated power in shaded cells and bypass diode is only about
-1.4 watts instead of -10 watts as before. This can be broken down into the power
dropped in the shaded cells and the power loss through the diode. The module
passes slightly more than its Isc of 1.1 amps at -0.7 volts so it is dissipating only a
small amount of power itself:
-0.7 volts X 1.1 amps = -0.77 watts
And the current going through the diode is just the difference between the shaded
module current the array current, given as 2.0 amps - 1.1 amps = 0.9 amps. At a
voltage of -0.7 volts across the diode, this means that the power lost in the diode is
given by:
-0.7 volts X 0.9 amps = -0.63 watts
In summary, we have reduced the power lost from 10.4 watts to only 1.4 watts, with
0.77 watts dissipated in the cells and 0.63 watts dissipated across the diode. The
two main benefits of bypass diodes are quite evident: (1) array output is preserved;
and (2) local heating is minimized.

14-12
Bypass Diode Preserves

Array Output
Above this point, array output
is missing one module, but
is no longer dragged down
by shaded cells.
3
Voltage
loss
2
through
shaded
module is
1
limited to
-0.7 volts
-20
20
40
60
80
100


14-13
Bypass Diodes In High Voltage Arrays

Recall from the discussion in the chapter on Wiring, there are two general ways to
wire systems of voltage higher than 12 volts. The modules can be wired in parallel
first to create current, and then have the parallel wired groups connected in series to
get voltage. Or modules can be wired in series strings first to get voltage, and then
the strings can be wired in parallel to get current. The recommended way is the
latter, to wire in series groups. The discussion below illustrates another reason why
this is the preferred method.
Parallel Wired Groups

If modules are wired in parallel groups there is a need for a large external diode
around each parallel group, as shown below.
Bypass Diodes For

Parallel Wired Groups


14-14
If modules in a parallel group are shaded, then the group cannot produce the current
of the other groups in series with it. The unshaded modules in the group can
compensate by trying to produce more current. They do this by operating at lower
voltage, which moves their operating point up in current along their IV curve.
If the shading is bad enough, and the current compensation needs to be very large,
then the entire group will continue to compensate by operating further and further
down in voltage. The group can actually go past zero volts and operate in the region
of negative voltage to try to get the current level up to that of the other unshaded
groups in series with it. If this happens, all the bypass diodes in the separate
modules will begin to pass current. Now the current is not just that of one module,
but could be the current of the entire array. The small diodes for each module would
be overloaded and could fail.
The solution is to install an external diode large enough to handle the current of the
entire array. If a shaded group compensates by going into negative voltage, then
the current from the other groups bypasses the entire group through the large diode,
and no damage occurs. This can involve extra wiring and cost.
Current Compensation of
Parallel Wired Modules
current level of other modules
-Voltage
+Voltage


14-15
Series Wired Groups

In the case of modules wired in series groups the bypass diodes installed in each
module will be sufficient. No extra diode protection is needed. This is because the
current that can flow through a bypass diode is just one modules worth and no
more. The current flowing in a string is just the current of one module. If a module
becomes shaded, the current flows through the bypass diode and on to the other
modules in that string.
Bypass Diodes For

Series Wired Strings


14-16
Blocking Diodes Prevent Leakage

Diodes placed in series with cells or modules can perform another function that of
blocking reverse leakage current backward through the modules. There are two
situations where blocking diodes can help prevent this phenomenon.
Blocking reverse flow of current from battery through module at night.

In battery charging systems the module potential drops to zero at night, and the
battery could discharge all night backward through the module. This would not
be harmful to the module, but would result in loss of precious energy from the
battery bank. Diodes placed in the circuit between the module and the battery
can block any nighttime leakage flow.
Blocking reverse flow down damaged modules from parallel modules during day.
Blocking diodes placed at the head of separate series wired strings in high
voltage systems can perform yet another function during daylight conditions. If
one string becomes severely shaded, or if there is a short circuit in one of the
modules, the blocking diode prevents the other strings from loosing current
backwards down the shaded or damaged string. The shaded or damaged string
is "isolated" from the others, and more current is sent on the load. In this
configuration, the blocking diodes are sometimes called "isolation diodes".
Each of these situations will be discussed next.

14-17
Preventing Nighttime Leakage

The IV curves presented on the next page show how the module is operated during
the day and night by the battery. At night the IV curve passes through zero volts and
amps, and slowly slopes downward from there. The battery maintains its voltage
even at night, so the module is operated at about 12 volts. Since the entire curve is
below zero amps, the current passing through the module is "negative", or going the
wrong way. The amount of current leaked from the battery depends on the shape of
the IV curve at night. The fewer number of cells in series in the module, or the
poorer the curve shape, the greater the leakage current. Leakage currents for some
modules are only about 50 mA, so the power dissipated through the cells would be
only 12 volts X 0.050 amps = 0.6 watts. The power is spread evenly through all the
cells, so the power dissipated by any one cell would be only about 0.6 watts / 36
cells = 0.017 watts or 17 milliwatts.
Curve Shape Determ ines

Nighttim e Leakage
Daytime flow
of current
(+) Current
(+)
(+)
(-)
( -)
Leakage flow of
current at night
(-) Current
(+)
(+)
( -)
(-)


14-18
This can be prevented by installing a diode in series between the array and battery.
The diode allows current to flow into the battery, but "blocks" reverse flow at night.
So in this configuration, the diode is called a "blocking" diode. There is a small
penalty to pay in the form of a 0.6-1.0 volt drop across the diode.
12 Volt System
Blocking Diode
+
PV
Module
-
Battery
Load


14-19
If a single module or a 12-volt string of parallel modules is involved a single large

diode can be placed between the module or 12 volt array and the battery. Often
charge controllers have a blocking diode built in to the circuitry. If a higher voltage
array is made of groups of parallel wired modules then one large blocking diode can
be used for the entire array.
Parallel-Wired Strings

This is not the preferred way to connect modules in a higher voltage array though. If
one or more modules in a group are damaged, or if maintenance work must be done
on a group, the entire array will not be able to produce power at the required voltage.
It is recommended to connect modules in series strings first, and then connect those
groups in parallel. This is described next.

14-20
Preventing Daytime Leakage

The preferred method of wiring higher voltage arrays is to first connect modules in
series strings and then connect them in parallel. Each series string should have its
own blocking diode. Do not think that if there are four strings, as shown, that the
current from any string must flow through all four diodes. Current flows from each
string through only one diode and combines with the current from the other strings,
so there is no increase in voltage penalty.
Blocking diodes placed at the head of each parallel string can perform another
function during daylight operation. If one string becomes severely shaded, or if there
is a short circuit in one of the modules, the blocking diode prevents the other strings
from loosing current backwards down the shaded or damaged string. The shaded or
damaged string is "isolated" from the others, and more current is sent on the load.
In this configuration, the blocking diodes are sometimes called "isolation diodes".
Series-Wired Strings


14-21
Typically in this configuration all the blocking or isolation diodes are grouped
together in a field combiner box located at each structure. All the positive wires from
the series strings are fed separately into the box, and the diodes are mounted on
some type of wiring block. The separate outputs from the diodes are combined into
a large diameter wire, and all the current from the strings flows on to the controls
and battery. The negative wires are combined into a large diameter wire also.
Series-Wired Strings With

Combiner Box
Dual J-box
style modules
Single J-box
style modules
Field Combiner Box

Other functions that can be performed at the field combiner box include a disconnect
switch or circuit breaker and lightning protection devices. These could be
conveniently built in to the box and would allow further protection of the system from
short circuits and lightning strikes.

14-22
An example of a commercial field combiner box is shown below. This unit is

manufactured by Solar Electric Specialties in the US and includes blocking diodes,
separate pull fuses for each series string, a DC rated circuit breaker and a silicon
oxide varistor (SOC) lightning protection unit. The negative block combines the
negative wires from the strings and a ground wire from the lightning protection unit.
The block is then connected to a ground rod driven into the earth at the array. If
there is a lightning strike at the array, the dangerous current spike will be shunted to
ground at the array instead of passing through the wires all the way back to the
system controls.

14-23
Current and Voltage Rating of

Diodes
To insure the diodes are not stressed, the current rating of blocking and bypass
diodes should be twice the Isc that might flow through them, and the reverse voltage
rating should be twice the entire array Voc.
Current Rating of Diodes
Double Isc
Voltage Rating of Diodes
Double Voc
Most power modules produce current in the range of 3-4 amperes, so this means
that the current rating of bypass diodes should be at least 6-8 amps, a relatively
small value. Since the manufacturer does not know what the voltage of the final
array will be, the voltage rating of bypass diodes should be the highest reasonable
commercial value, usually 600 volts.
In the case of a single blocking diode for a 12-volt parallel string, the current rating
must be the Isc for the entire array. For example, if 8 35-watt modules were
connected in parallel for a 12-volt system, the single blocking diode needed should
have a current rating of at least
2 X 8 parallel modules X 2.15 amps (Isc) = 34.4 amps.
The voltage rating of the blocking or isolation diodes should be twice the full array
open circuit voltage (Voc). Do not use just the nominal system voltage. For the
simple 12-volt array with 8 modules in parallel, the number of series modules is only
one, so the voltage rating should be
2 X 1 series module X 20 volts (Voc) = 40 volts.
For a 48-volt nominal array, with 4 modules in series, the voltage should be at least
2 X 4 series modules X 20 volts (Voc) = 160 volts.

14-24
Exercise


14-25
(End of Chapter)

14-26
CHAPTER FOURTEEN
DIODES IN PHOTOVOLTAIC SYSTEMS
14-1
Bypass Diodes For Shading

Shading Causes Mismatched Currents
Effect Depends On How Shaded
How Bypass Diode Reduces Loss
Placement Of Bypass Diode In Module Cell Circuit
An Example Of Bypass Diode Reducing Losses
Bypass Diodes In High Voltage Arrays
14-2
14-2
14-3
14-4
14-8
14-10
14-14
Blocking Diodes Prevent Leakage

Preventing Nighttime Leakage
Preventing Daytime Leakage
14-17
14-18
14-21
Current and Voltage Rating of Diodes
14-24

14-27
Chapter 14 Answers
Diodes in Photovoltaic Systems

First Array:
Blocking
Bypass
Bypass
Bypass
Bypass
A diode should be rated for twice the expected Isc and twice the full array Voc. The
bypass diodes are connected to 4 modules in parallel, so their rating should be:
Current Rating 8 X Isc
Voltage Rating 8 X Voc
The blocking diode must handle all 4 parallel strings, so its rating should be:
14-1
Diodes
Second Array:
Blocking
Blocking
Bypass
Bypass
Bypass
Bypass
The bypass diodes are now connected to a single module, so the rating should be:
Each blocking diode is connected to a single parallel string, so each will be rated at:
14-2
Diodes
Chapter Fifteen
System Sizing
The purpose of photovoltaic system sizing is to calculate the number of solar
modules and batteries needed to reliably operate the load throughout a typical year.
This involves balancing the often-opposing goals of maximum reliability and
minimum cost.
This section presents simple methods for calculating array and battery size for
standalone systems. While reading this chapter keep in mind that not all choices in
system sizing are based on calculations. There are decisions that require judgment
on the part of the designer and the user. The mechanics of the calculations are
quite simple. It is the judgment of the designer about the efficiency and
appropriateness of the loads that makes a system well designed and cost effective.

15-1
System Design System Sizing
Basic Principles
The solar array is sized to replace the load on a daily basis based on average
weather conditions. The average is made up of below average days and above
average days so the array and battery must work together.
The array is NOT sized based on how quickly it can recharge the array after a
few days of below average weather. This would result in a large array most of
which is not needed or used during most of the year.
The battery has the job of supplying energy to the loads when the insolation is
below average. The array will replenish the battery during subsequent days of
above average insolation.
If there are concerns about quickly recovering from a storm or prolonged period of
below average weather then the designer should be considering a hybrid system
design. The backup generator (usually a diesel, gas or propane generator) would be
used to bring the batteries back to near full charge every few days during the winter
or periods of prolonged bad weather. The generator might not even be needed
during the summer months.
The sizing calculations presented in this chapter are based on designing a
standalone photovoltaic, non-hybrid and system. Hybrids are discussed in the
chapter on Hybrid Systems

15-2
Size Array For Worst Season

When first thinking about array sizing for standalone systems one could begin by
sizing so that the array output would be equal to the mathematical average of the
load during the entire year. In that way, on the average over the year, the array
would put back into the system as much energy as the load used. But this means
that for half the months of the year, every year, the battery would be partially
discharged. During such long periods of sustained partial charge sulfation of the
battery plates would occur. In standalone systems there is no generator available
on demand to recharge the batteries during below average weather. Battery
performance and life would be very poor, and the overall cost of the system over
time would be very high.
The proper approach to array sizing is to calculate the array needed during the worst
season of the year. This means that one would expect the battery to be fully
recharged even during the worst season and certainly during all the rest of the year.
This approach will reduce the sulfation that might occur on the battery plates and
lead to long system operating life and low maintenance costs over time.
If the worst season of the year is very far below the average for the year then sizing
for the worst month will give an array that is quite oversized for the rest of the year,
resulting in high initial costs. In that case it would make sense to consider a hybrid
system with a backup generator. But for small systems where a generator would be
too expensive, or in remote situations where the operation and maintenance of a
generator would be a burden, designing a standalone photovoltaic system will be the
most cost-effective approach.

15-3
Battery Sizing
Days of Autonomy or Reserve
The battery bank is sized to operate the loads during a long sequence of below
average insolation days. We can think of the battery as being full of charge.
During a below average day the array cannot supply all the amp-hours of charge
needed to replace what the load draws from the battery. So the battery ends up
being discharged at the end of the day. If the next day is again below average then
the battery again discharges some to operate the loads. This process can go on for
only so long before the battery is discharged to a point that is may become
damaged. The system designer must build into the battery capacity enough
equivalent days of charge to operate the loads autonomously, meaning without any
input of energy from the solar array. We refer to these equivalent days of reserve as
days of autonomy.
The usual rough values that are used in most system sizing calculations are about 35 days for non-critical applications and 7-14 days or even more for more critical
applications. Non-critical situations usually involve occupied systems, where the
users can adjust their load demand a little to accommodate the bad weather.
Critical situations might involve commercial or governmental systems for
communications or navigation, or important health needs such as hospitals and
clinics. And very remote systems must have a large capacity in their battery bank to
allow for the time it would take a maintenance crew to arrive at the site.
Rough Guide to Days of Reserve or Autonomy

Non-critical Applications
3-5 days
Typical 4 days
Critical Applications
7-14 days
Typical 10 days

15-4
Basic Battery Sizing Calculation

The basic formula for calculating battery size is presented below. It involves
multiplying the number of days of reserve or autonomy times the amount needed
daily for the load. This gives the first approximation to the size of the battery
capacity.
But we cannot allow all of the battery capacity to be discharged during the days of
autonomy. Manufacturers recommend that only 80% of even deep cycling
batteries, and only about 50% of shallow cycling batteries, be discharged. So we
must divide by the maximum percentage usable to give the amount of capacity to
install.
Maximum Percentage Usable

Deep Cycling Battery Type
Up to 80%
Shallow Cycling Battery Type
Up to 50%
The basic version of the formula is given below for calculating the battery capacity
that must be installed.
Battery Capacity =
Number of Days of Autonomy X Daily Load

Maximum % Usable
Each battery has a nominal voltage, depending on how many single cells are
connected by the manufacturer. Some large cells are only two volts; some smaller
units are six or twelve volts. The formula for the number of batteries to connect in
series to give the voltage for the loads is simply the nominal system voltage divided
by the nominal battery voltage.
Number of Series Batteries

15-5
Load Nominal Voltage

Battery Nominal Voltage
To illustrate these basic equations lets use the remote school system that was sized
on Page 11 in the Load Estimation chapter as an example.
Example:
The AC load demand for the school summed to 9104 Wh/day. If we use the
Trace 4024 sinewave inverter, and assume that the efficiency is about 90%
and the input voltage is 24 volts, then the DC load will be 421 Ah/day (9104
Wh 0.9 24 V = 421 Ah).
We shall consider this a non-critical system because the users can be a bit
flexible in their usage depending on the weather, so we will design for 5 days
of autonomy. We will assume that deep cycling industrial batteries are used.
The maximum % usable will therefore be 80%.
Battery Capacity
5 days X 421 Ah/day

0.8
2631 Ah of capacity
2105
0.8
If they were to use Trojan L16 deep cycling flooded batteries, rated at 350 Ah
and 6 volts then they would need about eight in parallel and four in series:
(2631 Ah 350 Ah = 7.5, round up to 8) and (24 volts 6 volts = 4).
As another example, look again at the small remote cabin system sized on Page 10
of the Load Estimation chapter. Assume that it will be used as only a weekend
cabin and that cost is a problem so low cost batteries will be used. Recall that it is a
pure DC system with no AC loads.
Example:
The remote weekend cabin load was calculated to be about 70 Ah/day at 12

volts. If low cost starting batteries are to be used, then we can use only 50%
of the capacity for days of reserve. If the cabin is to be used only for the
weekends, then only 2 days of reserve need to be built into the battery
capacity.
Battery Capacity
2 days X 70 Ah/day =
0.5
280 Ah of capacity
140
0 .5
If they were to use Delco 2000 shallow cycling batteries, rated at 105 Ah at
12 volts each, they would need three batteries in parallel (280 105 = 2.7)

15-6
Exercise

!

"#

$ %

&
"%#


15-7
Correcting For Cold Temperature

The basic formula presented must be modified to properly account for some
performance factors that affect battery capacity and depth of discharge. The first
factor that must be taken into account is the fact that batteries loose capacity when
they get cold. Recall from the discussion in the Battery Technology chapter that a
lead acid battery can deliver only a little more than 90% of its rated capacity at 0o C
and only about 80% at -20o C.
This reduction in available capacity with cold temperatures means that MORE
capacity must be installed than you would calculate at standard temperatures of
25o C. Install more than you need at 25o C so that when the battery gets cold, it will
still have the capacity that you need.
We do this by dividing by the Capacity Correction Factor given in manufacturers
literature. A typical reduction curve is shown below. Estimate the average
discharge rate using the method given on page 46 of the Battery Technology
chapter. The average rate of discharge for photovoltaic systems is given by the
formula:
Average Rate of Discharge (hours) = # Days of Autonomy X Load Operating Time
Maximum % Usable
where Load Operating Time can be estimated as follows:
use 24 hours
Continuous Loads:
Multiple Load Systems: Weighted Average Load Operating Time =
load time
loads
Percent of Rated Capacity
Temperature and Discharge Rate

Effects on Lead-Acid Battery Capacity
120
110
100
90
80
70
60
50
40
30
C/500
C/50
C/0.5
-30
-20
-10
10
C/120
C/5
20
30
40
50
Battery Operating Temperature - oC

15-8
Example:
A remote data acquisition system is installed on a mountain top where

the average 24 hour temperature can get as low as -10o C. The
system has 10 days of autonomy in the battery bank and uses deep
cycling batteries (maximum DOD of 80%).
The system has only one load of 2 amps operating 24 hours/day, so
the weighted average load time is just 24 hours.
Weighted Average Load Operating Time = 2 amps X 24 hours = 24 hrs
2 amps
The average discharge rate can be estimated to be:

10 days X 24 hours
0.8 max discharge
= 300 hour rate.
Reading between the 120 hour and 500 hour rate curves, the loss of
capacity factor reading up from -10 is about 0.91 or 91%. This means
that at this temperature, there would be only 91% of the capacity
available at 25o C. To have the capacity you really need for your
autonomy requirements, you would need to install 9% more.
For this example, the daily load of the data acquisition system is
2 amps X 24 hours/day = 48 Ah/day. Using the basic formula
presented earlier (not accounting for cold temperature) the capacity
would be given by:
Simple Capacity Calc
10 days X 48 Ah
0.8
600 Ah
But this is the capacity that we want, even during the coldest period
when it gets down to -10o C. To account for the loss of capacity at 10oC we need to divide by the correction factor to calculate the amount
to install when rated at 25o C:
Temperature Adjusted
Capacity Calc

600 Ah needed at low temp.

0.91 temperature factor
660 Ah at 25o C
15-9
Cold Temperature Data

Data indicating the lowest 24-hour average temperature are extracted from the
Siemens Solar database for various sites and presented in the Appendix section. It
is assumed that the battery is a large thermal mass and will not quickly cool off to
the coldest temperature in the night, but rather will slowly move between the heat of
the day and the cold of the night.

15-10
Specifying Battery Capacity At Proper

Discharge Rate
Another factor that needs to be considered when sizing batteries is that battery
capacity increases with slower discharge rates. Typically manufacturers list their
capacity at a standard rate of 8 or 10 hours of discharge. But in photovoltaic
systems, it is typical for systems to have discharge rates of 100-200 hours or slower,
due to all the extra capacity built into the battery for autonomy.
Most manufacturers give the capacity of their batteries at different rates of
discharge. You need to be able to specify the capacity needed for your system
at the proper discharge rate. Average rate of discharge was discussed in the
Battery Technology chapter, and is presented again here.
The average rate of discharge for photovoltaic systems is given by the formula:
Average Rate of Discharge (hours) = # Days of Autonomy X Load Operating Time

Maximum % Usable
The average load operating time is either 24 hours for a continuous load, or the
actual load operating time if a single load or the weighted average time if several
intermittent loads are present.
Load Operating Time:
Continuous Loads:
Multiple Load Systems:
use 24 hours
use weighted average load operating time
Weighted Average Load Operating Time =
load time
loads
Some examples of industrial battery manufacturers literature are shown on the next
page.

15-11
C&D Battery Company:

Cell Type
AH Capacity @ 25 deg. C.
1.75
vpc
Overall Dimensions
to 1.9 vpc
Length
Width
Weight
Height
8 h
72 h
120h
240h
480h
in.
mm
in.
mm
in.
mm
lbs.
kgs.
3DCPSA- 3
31
44
46
48
49
5.28
134
7.38
187
10.31
262
27.7
12.6
3DCPSA- 5
62
72
73
74
75
5.28
134
7.38
187
10.31
262
33.6
15.2
3DCPSA- 7
94
133
139
145
150
9.47
241
7.38
187
10.31
262
54.3
24.6
DCPSA-11
156
222
230
240
249
6.38
162
7.38
187
10.75
273
37.0
16.8
DCPSA-13
188
267
277
290
300
6.38
162
7.38
187
10.75
273
39.0
17.7
DCPSA-15
219
295
300
306
310
6.38
162
7.38
187
10.75
273
41.0
18.6
3KCPSA- 5
225
315
324
333
340
8.53
217
10.44
265
18.25
464
131.0
59.4
KCPSA- 7
337
384
389
394
400
3.62
92
10.44
265
18.25
464
58.0
26.3
KCPSA- 9
450
506
513
520
524
4.62
117
10.44
265
18.25
464
79.0
35.8
KCPSA-11
562
618
628
638
644
5.59
142
10.44
265
18.25
464
96.0
43.5
KCPSA-13
675
743
755
766
773
6.59
167
10.44
265
8.25
464
113.0
51.3
KCPSA-15
787
1011
1031
1047
1060
8.53
217
10.44
265
18.25
464
139.0
63.0

Varta Battery Inc.:

Type
Vb2306
Vb2308
Vb2310
Vb2312
Vb2407
Vb2408
Vb2409
Vb2410
Vb2412
Vb2414
Vb2416
Vb2418
Vb2420
Capacity (Ah) to 1.85 vpc final voltage

5 hour
10 hour
24 hour
48 hour
72 hour
120 hour
240 hour
270
360
450
540
630
720
810
900
1080
1260
1440
1620
1800
300
400
500
600
700
800
900
1000
1200
1400
1600
1800
2000
360
480
600
720
840
960
1080
1200
1440
1680
1920
2160
2400
408
544
680
816
952
1088
1224
1360
1632
1904
2176
2448
2720
432
576
720
864
1008
1152
1296
1440
1728
2016
2304
2592
2880
450
600
750
900
1050
1200
1350
1500
1800
2100
2400
2700
3000
468
624
780
936
1092
1248
1404
1560
1872
2184
2496
2808
3120


15-12
Example:
A remote telecom site will use deep cycling industrial batteries (maximum
DOD of 80%). The system will have 10 days of autonomy designed into the
battery. Two loads are present:
Load 1:
Load 2:
Microwave Repeater
Radio
Weighted Average
Load Operating Time
Average Rate of Discharge
10 amps
7 amps
24 hours/day
12 hours/day
= 10 amps X 24 hours + 7 amps X 12 hours

10 amps + 7 amps
=
19 hours
10 days X 19 hours/day
0.8 max. discharge
238 hour rate
The nearest rate used by either manufacturer shown on the previous page is
the 240 hour rate, so use values from this column for this system. If your
calculated average discharge rate is between rates used by a manufacturer,
use the next faster rate (shorter time) to be conservative.
--------------------------Compare the capacity available at the 240-hour rate to the nominal rates of
the manufacturers:
C&D Literature: use as an example the KCPSA-15 battery. Capacity at the
slow 240 hour photovoltaic rate is 1047 Ah, and capacity at their standard 8
hour rate is 787 Ah. So a typical C&D battery will deliver about
1047 Ah @ 240 hr / 787 Ah @ 8 hr = 0.33 or 33% more capacity at the slow
photovoltaic rate compared to their standard rate.
Varta literature: use as an example the Vb2410 battery. Capacity at the slow
240 hour photovoltaic rate is about 1560 Ah, and capacity at their standard
10 hour rate is 1000 Ah. So a typical Varta battery will deliver 1560 Ah @
240 hr / 1000 Ah @ 10 hr = 0.56 or 56% more capacity at the slow
photovoltaic rate compared to their standard rate.
Rule of Thumb: You will have about 30% more capacity available from a battery at
typically slow photovoltaic discharge rates (usually C/100 to C/300) compared to
manufacturers standard or nominal rates (usually C/8 or C/10). This is only a rough
rule of thumb to use when no other capacity vs. rate information is available.

15-13
Limiting Maximum Depth of Discharge To

Prevent Freezing
A final factor that must be included in battery sizing is the adjustment to the
maximum allowed depth of discharge to prevent freezing. As a battery discharges
the electrolyte turns into water, and the freezing point rises toward the freezing point
of pure water at 0o C. In cold climates, if the battery is allowed to discharge too
much, the electrolyte might freeze, damaging the battery.
Even if a deep cycling industrial battery is used in a system, the maximum depth of
discharge might have to be limited to less than the usual 80%. The adjustment can
be read from a chart as given below.
Correct for Limited Discharge

Due to Possible Freezing
Maximum D.O.D. (%)
80
60
40
20
0
-60
-40
-20
Lowest Battery Temperature (deg.C)

Using the lowest 24-hour average battery temperature read up to find the maximum
allowed depth of discharge. This concern applies only for temperatures below minus
8o C. For example, if the battery would get to -20o C, the maximum allowed depth of
discharge should only be about 53%, not the usual 80%.

15-14
Complete Battery Sizing Calculation

Putting all these correcting factors together, we develop a final formula for
calculating the capacity of a photovoltaic battery bank.
Battery Capacity
(@ specified rate)

Max % Usable X Temperature Derating Factor
(A) The Maximum Percent Usable is either the standard value of 50% for shallow
cycling batteries, 80% for deep cycling batteries, or a reduced value based on
freezing concerns. The designer can also reduce this value simply as a way to build
in more life to the system. For example, a designer might use a shallow cycling
battery, but size based on using only 30% of the capacity and not the usual 50%.
This might not impact the cost of the system too much, and would mean the battery
life would be extended.
(B) The Temperature Derating Factor is included to make sure that more capacity is
installed at 25o C so that when the battery gets cold and loses some capacity, there
will still be the required capacity present.
(C) The specified rate insures that you are taking into account the change of
capacity with slow discharge rate. Use manufacturers literature to choose a battery
that will give you the capacity you need at the average discharge rate of your
system.
An Array/Battery Sizing Form is presented next. Use this form for your battery
calculations. All the factors are indicated. An example is worked for your reference.

15-15
Example:
Calculate the capacity needed for the remote school example again, this time
with more information. The local meteorologist reports that it gets to an
o
average of -20 C in the winter at the site. This will mean that we
compensate for both loss of capacity with cold temperature, and limit the
maximum depth of discharge to prevent freezing.
(A) Using the freezing phenomenon chart in Figure 15-2, we determine that
the maximum allowable depth of discharge is about 50% or 0.5 , even though
we are using deep cycling type batteries that usually can be discharged to
80%.
(B) Calculate average rate of discharge. The weighted average load
operating time for the loads on p. 9-11 is given by:
Weighted Average
Load Operating Time =
9104 Wh/day
(8X40)+(2x11)+(2x200)+ 300+800+200 watts
4.5 hours
The number of days of autonomy was given as 5 and the maximum

discharge is limited to 50% due to freezing, so the average rate is given by
Avg. Rate of Discharge
5 days X 4.5 hours/day

0.5 max. discharge
= 45 hr rate
(C) Use Figure 15-1 to determine the Temperature Derating Factor. Use the
50-hour curve (closest to our 45-hour rate), and determine that the
o
Temperature Derating Factor at -20 C is 70% or 0.70.
(D) Calculate the capacity. We will use Varta batteries for this example.
Battery Capacity
5 days X 421 Ah/day load

0.5 X 0.70
6014 Ah
@ 45 hour rate
Looking at the Varta capacity Table 15-2, use the 48 hour column (closest to
our 45 hour rate) to select a particular model to use. We could choose 4 of
the Vb2412 units, as each gives about 1/4 of the capacity we need at that
rate.
Since they are 2-volt cells, we would need strings of 12 cells connected in
series to give the 24 volts for the inverter.

15-16
Array/Battery Sizing Form

Remote School
System Description:
Max. Daily Load
421 Ah
System Voltage
24 V
Average Discharge
Rate
Number of Days X Load Operating Time

Maximum Battery % Usage
Number of Reserve Days

Maximum Battery % Usage =
Coldest Avg Temperature. =
-20 C
Battery Capacity
(@ 45 hour rate)
Number of Days X Max. Daily Load

Maximum % Usage X Temp. Derating
Chosen Battery
Capacity =
5
0.5
] X [ 421 ]
] X [ 0.70 ]
[
[
Varta Vb 2412
Voltage
Battery Capacity
6014
0.5
45
(hours)
Temperature Derating =
1632 @ 48 hour rate
Parallel Batteries
0.85
6014 Ah
2-volt cell
= 3.7
round to
4
[ 1632 ]
Chosen Battery Cap.
System Voltage
= [ 24 ]
= 12
Chosen Battery V.
[ 2 ]
----------------------------------------------------------------------------------------------------------------Daily Load = ________ Module : _______ Module Output = ______
Series Batteries
Parallel Modules
Daily Load
.
Module X Derating X Charge
Output
Factor
Efficiency
[
[
Series Modules
= System Voltage
12

]
]X[
]X[
=
[________]
[
]
15-17
__________
__________
Exercise

'

()*

+

!

"

,

"-

!

"%#

)

!)*./.

()

0
&

,

$&

"%
. %
,

1.

&!#

#)

23

,2
3

"#


15-18
Battery Cell Size

When you calculate the capacity of your battery bank, you will have to decide how
many real battery units you will put together in parallel to give you your capacity. Do
you choose just one huge cell, or a couple in parallel, or three, or ten?
For example, if you calculated that you needed 1000 Ah of capacity, would you
choose a single 1000 Ah cell, or two 500 Ah cells, or four 250 Ah cells?
In general, you want to keep the number of parallel connected batteries to a
minimum. This will reduce the possibility of the batteries getting out of equalization
with each other. Some parallel connected batteries could become more or less
charged than others.
Rule of Thumb: As a general guideline, it is recommended that you connect no

more than four batteries in parallel. Use this as a general rule of thumb.
If you find that your choice of battery means you have to connect 20 batteries in
parallel, for example, you should look at choosing a larger capacity unit so you need
fewer in parallel!
Many system designers prefer to design their system with two parallel strings. In this
way, if there is a problem with a battery cell, part of the battery bank can be removed
and worked on, while the system still has nominal voltage and can operate.
Different designers will have different preferences.
Exercise

0

!4
-
"

"#
5

6
7

8

15-19
Array Sizing
In the previous discussion we showed how battery sizing was based on operating
the loads during a period of below average insolation days, or days of autonomy.
Now we look at how to size the array based on replacing the daily load on the
average every day.
Basic Array Sizing Calculation

The basic method to calculate the array size is to divide the average daily Ah load by
the number of Ah that one module will produce in a day. This will give the number of
modules needed in parallel to produce the current for the load. The number of
modules needed in series is given by dividing the nominal system voltage by the
nominal voltage of one module.
Number Parallel Modules=
Daily Load Demand (Ah)

Module Daily Output (Ah)
Number Series Modules =
Nominal System Voltage (volts)

Nominal Module Voltage (volts)

15-20
Modifications To Array Sizing Formula

Derate Module Output By 10%
In the real world, modules loose some of their output due to environmental factor.
Dirt, dust and gradual aging can reduce the module output. The standard practice is
to derate the calculated output of a module by 10% to account for these
unpredictable factors. Think of this as a normal engineering safety factor, required
in the case of photovoltaic system design because we are relying on the weather.
Variations are to be expected. So we build in a margin of safety to insure that the
system works year after year.
Increase Load By 10% For Battery Coulombic Efficiency

During the charging process, flooded lead acid batteries will gas. This means that
some of the charge pushed through the battery by the modules is not turning into
usable capacity but is escaping the system. Thus a little extra current must be
generated to overcome this loss. This is the coulombic efficiency of the battery.
Usual practice allows for about 5-10 % loss due to gassing. So we must increase
the array size by about 10% over the load demand to account for this inefficiency in
the battery.

15-21
Complete Array Sizing Calculation

We must therefore modify the simple array sizing formula to account for these
factors.
(A) Divide the Daily Load by the Battery Coulombic Efficiency. This effectively
increases the daily load and gives the true load that the array must replace.
(B) Multiply the Module Daily Output by the Derating Factor. This reduces the
expected output from a module due to environmental and aging losses, and gives a
more conservative estimate of what can be expected from a module in the real
world.
Together these factors give the complete array sizing formula.
Number Parallel Modules
Daily Load (Ah)

Coulombic X [ Module X Derating ]
Efficiency
Output
Factor
Number Series Modules
Nominal System Voltage (volts)

Nominal Module Voltage (volts)

15-22
Seasonal Variations Require Monthly Calculations

If the load profile is not constant it is wise to perform this calculation for the different
seasons or for each month. There is usually more output available in the summer,
spring and fall than in the winter. But the load may be greater in the summer as
well. The expected module output for each month or season should be divided into
the load for that month or season to calculate the number of modules needed for
each month. The required size would be the largest number needed during the
year.
For example, you may calculate that you need nine modules in the winter, but only
need seven modules in the summer. You would have to install the larger number to
meet your load demand throughout the year.

15-23
Choose A Module With The Proper Number

Of Cells In Series
The daily output of a module will depend greatly on the number of cells in series. All
cells and modules loose voltage with increasing temperature. This is a natural
phenomenon well understood by manufacturers. Modules are designed to operate
in different temperature climates for this reason.
36-Cell Modules For Hot Temperatures or MPT Applications

Manufactures usually produce a standard line of modules with 36 cells in series.
Modules with 36 cells in series are designed to operate near their Imp even in the
hottest of climates. The presence of 36 cells in series means that the Vmp at
standard temperature of 25 oC is about 17 volts, much more than needed to charge
a 12-volt battery. When these modules are placed in hot climates, they will loose
about two volts due to heat, and will have their Vmp at about 15 volts. This is
enough to charge all types of batteries, even in the hottest of climates. Modules with
36 cells therefore should be considered for all hot climates. They are also the
module of choice for systems with MPT, such as direct coupled water pumping
systems and utility interactive systems. All the potential power of the 36 cells will be
extracted by the MPT device, so no voltage is wasted.
33-Cell Modules For Moderate Temperatures

Solar module manufacturers may create modules with different number of cells in
series, so reduce module cost and to match different climates. The modules with 33
cells in series are designed to operate in moderate climates. The presence of 33
cells in series means that the Vmp at standard temperature of 25o C is about 16
volts, slightly more than needed to charge a 12-volt battery. When the module heats
up to about 40-45o C it will loose about one volt and will have its Vmp at about 15
volts. This is enough to successfully recharge any battery. But if the module is in a
very hot climate, then it will loose more voltage. If it heats up to 50o C or more, then
the voltage will drop to about 14 volts or less, and some current output reduction will
occur. This is not damaging to the module at all, but simply means that less than
optimal current will be generated. Thus 33 cell modules are best suited for
moderate climates and are not recommended for India.

15-24
Choosing A Module
36 cells:
Vmp @ 25oC = 17.4 volts

@ 50oC = 15.3 volts
hot climates
direct couple to DC motors
utility interconnection with MPT
33 cells:
Vmp @ 25oC = 15.9 volts

@ 50oC = 14.0 volts
moderate climates
general battery charging


15-25
Peak Hour Method Of Estimating Module

Daily Output
You can estimate module daily output by using weather data. This method involves
converting the actual measured insolation on a tilted surface into the equivalent
number of hours of peak hours of standard full sun irradiance at 1000 w/m2.
Multiplying the number of peak hours times the module peak output (Imp
measured at 1000 w/m2) gives an estimate for the number of Ah/day from a module.
Module Output Is Peak Hours X Peak Power

If the insolation data on a tilted surface is in units of kwh/m2 then the conversion to
peak hours is easy. For example, if the average insolation for a month is 6.6
kwh/m2, this can be rewritten as 6.6 hours X 1 kw/m2. But recall that 1 kwh/m2 is just
1000 watts/m2, and is just the standard irradiance used by manufacturers to rate
module output. It is as if the module was exposed to the standard peak irradiance
level of 1000 w/m2 for 6.6 hours. This of course did not happen, but it makes for a
simple calculation. The standard irradiance level of 1000 w/m2 is important because
module output at that condition is given in all manufacturers literature, so you can
simply use manufacturer's literature values. To calculate the Ah/day from a module
using peak hours, multiply the peak hours times the module Imp.
Estimating Module Output

Using Peak-Hour Method
Pretend that all real insolation was delivered at peak
irradiance level of 1000 W/m2
ex: 6.6 kwh/m2 =
1kw/m2
6.6 hours
peak
irradiance
peak
hours
Module Daily Output = Imp

(for 35 watt module)
= 2.0 amps
peak hours
6.6 hours
= 13.2 Ah


15-26
For example, the average daily irradiance profile is shown for a surface tilted to 20o
for Madras. The total insolation falling on that surface in March is 6.6 kw/m2. This
can be rewritten as 6.6 hours X 1 kw/m2. In the figure we show the equivalent
irradiance profile with the sun suddenly turning on to the peak irradiance level of
1000 w/m2 and staying at that level for 6.6 hours. Module output would be given by
multiplying this value for equivalent peak hours times the module Imp. For a typical
35 watt module with Imp of 2.0 amps, this would be 6.6 X 2 amps = 13.2 Ah/day.
For a typical 75-watt module with Imp of 4.4 amps, this would be 6.6 X 4.4 amps =
29 Ah/day.
If we look at the average daily irradiance profile for Madras in July, we see that the
total insolation that fell on a surface tilted to 20o would be only 4.5 kw/m2. This
would equate to 4.5 peak hours of irradiance.
Calculate array size based on the month with the lowest insolation for a load profile
that is constant throughout the year, or for each month and compare to the load
required for that month and choose the largest number of modules needed in any
month.
Peak Hours
o
Madras, Surface Tilted to 20 S
1200
Irradiance (w/m )
1000
800
6.6 kwh/m2
600
4.5 kwh/m2
400
200
0
4
10
12
14
16
18
20
Hour of Day
March
July


15-27
If insolation data is reported in units other than kwh/m2, such as Langleys

(calorie/cm2), megajoules/m2, or Btu/ft2, then multiply by the conversion factor
presented below to change into kwh/m2 and therefore peak hours.
Conversion Factors For Peak Hours
Unit of Insolation
Multiply by this factor to convert to peak hours...
kwh / m2 / day
1.0
Langley / day
0.01162
MJ / m2 / day
0.2777
Btu / ft2 / day
0.003155


15-28
Weaknesses in the Peak Hour Method
There are some simplifications in the peak hour approach that should be
understood. For example, the effect of temperature on module output is
neglected in the peak hour approach. As a means to compensate the module
Imp is used in calculations. Since battery loads usually operate the module at
voltages slightly lower than the Vmp and therefore at currents slightly higher than
the Imp, using Imp for module output is being conservative. The effect of
temperature on module output is greater for modules with fewer cells in series.
So the peak hour method is more accurate for 36 cell modules, and less
accurate for 33 cell modules, especially in hot climates. Predictions in cold
climates will be more accurate for all modules.
In the peak hour method, the total measured solar insolation is translated into
peak hours of operation. Actually, at the beginning and ending of each day,
there will be some time when the irradiance will be too low and the module or
array voltage will not be sufficient to charge a battery. This error is usually quite
small, but is more pronounced for modules with fewer cells in series. Predictions
of Ah/day/module will be more accurate with this method for 36 cell modules than
for 33 cell modules.
The peak hour method assumes that module output is completely linear with
irradiance. It assumes that all modules will convert solar irradiance into electrical
power the same. But this is not the case. For example, high efficiency single
crystal solar cells can convert at low light levels more efficiently than some other
technologies. So this peak hour approach of multiplying hours times rated
current can overestimate the output of certain technologies.
In general however, the peak hour approach is a useful method for quickly
approximating module output given local measured insolation data on a tilted
surface.

15-29
Exercise

.

69 :7

;<&

=

9

5$4

>!#

;?
;

@@@@@@@@@
5$4

>!

A@@@@@@@@@@
A@@@@@@@@@@
<?
;

@@@@@@@@@
5$4

>!


15-30
A@@@@@@@@@@
A@@@@@@@@@@
Leh
Latitude: 34 oN
Best Tilt: 50 oS
Jan
4.0
Feb
4.3
Mar
4.4
Apr
4.9
May
4.7
Jun
4.6
Jul
4.7
Aug
4.8
Sep
5.3
Oct
5.3
Nov
5.0
Dec
4.4
Jan Feb Mar Apr May Jun Jul Aug Sep

5.9
6.5
6.6
6.4
5.8
5.0
4.5
4.9
5.3

!
Oct
5.4
Nov
5.2
Dec
5.4
Madras
Latitude: 13 oN
Best Tilt: 15 oS

15-31
Two examples of how to use the data are presented next. The first example is the
remote school, and the second is the remote cabin.
Example:
The remote school 24-volt DC load demand was determined to be 421

o
Ah/day. Use the insolation data given for Leh. We would use the 50 tilt to
produce the flattest insolation profile for the year to match the expected
constant load for the school used throughout the year.
o
The lowest insolation during the year at 50 tilt angle is 4.0 kwh/m during
January. This translates into 4.0 peak hours. Calculating the output from a
typical 75 watt module:
Module Output
4.0 peak hours
17.6 Ah/day
4.4 amps Imp
To calculate the number of modules needed for this load, we assume a

conservative battery coulombic efficiency of 90%, and apply the usual 10%
derating of module output:
Number of Parallel Modules
Number of Series Modules
421 Ah/day daily load

0.9 X [ 17.6 Ah/day X 0.9 ]
29.5 , round up to 30 modules
24 volt system voltage

12 volt module voltage (nominal)
2 in series
Total number of modules = 2 series X 30 parallel = 60 modules
The work for this example is worked out on the next page, using the lower portion of
the Array/Battery Sizing Form.

15-32

System Description: Remote
School
(location: Leh worst
month: Jan)
Max. Daily Load
_____
_____
System Voltage
_____
_____
Average Discharge
Rate

Coldest Avg Temperature. =
_____
Battery Capacity
(@ ___ hour rate)

Chosen Battery
Capacity =
Temperature Derating
[_____ ] X [_____ ]
[
] X [
_____
_____ Ah
____________________________
_____@ _____hour rate
Parallel Batteries
= _____
(hours)
Voltage
Battery Capacity
=
Chosen Battery Cap.
[_____]
[
_____
_____
System Voltage
= [_____]
= _____
Chosen Battery V.
[
]
----------------------------------------------------------------------------------------------------------------Week-Averaged = 421 Module : 75 watt Module Output = 17.6 Ah/day
Daily Load
Series Batteries
Parallel Modules
Daily Load
.
Output
Factor
Efficiency
421 ]
[ 17.6 ] X [ 0.9 ] X [0.9 ]
Series Modules
= System Voltage
12

[24] =
12
15-33
29.5 -> 30
Example:
The weekend cabin system was determined to have a daily load demand of
69.7 Ah/day. But if the system is really only going to be for weekend use,
then the array can be sized based on the Week Averaged Load Demand .
The array can be sized to replace the 2-day weekend demand over the entire
7-day week. Recalculating the 69.4 Ah/day daily load spread over the week:
69.4 Ah/day X 2 days / 7 days = 19.8 Ah/day week averaged.
Now we size the array to replace this amount on a daily basis.
Let the location for this system be near Madras. In this system we will use a
35-watt module, with a typical Imp of 2.0 amps. The best tilt angle for a
o
constant load demand throughout the year is 15 S.
o
The lowest insolation during the year at 15 tilt angle is 4.5 kwh/m during
July. This translates into 4.5 peak hours. Calculating the output from a
typical 35 watt module:
Module Output
4.5 peak hours
9.0 Ah/day
2.0 Imp
Assume the same conservative battery coulombic efficiency of 90%, and the
usual 10% module output derating factor.
The number of 35 watt modules needed for this application is given by
19.8 Ah/day avg daily load

0.9 X [ 9 Ah/day X 0.9 ]
2.7, round up to 3 modules

12 volt module voltage (nominal)
1 module in series
The work for this example is again worked out on the next page, using the lower
portion of the Array/Battery Sizing Form.

15-34

System Description:
Weekend Cabin (Madras
worst month: July)
Max. Daily Load
_____
_____
System Voltage
_____
_____
Average Discharge
Rate

_____
Coldest Avg. Temperature. =
Battery Capacity
(@ ___ hour rate)
Chosen Battery
Capacity =
_____
_____

[_____ ] X [_____ ]
[
] X [
_____ Ah
_____
_____
_____
_____@ _____hour rate
Parallel Batteries
(hours)
Voltage
Battery Capacity
=
Chosen Battery Cap.
[_____]
[
System Voltage
= [_____]
= _____
Chosen Battery V.
[
]
----------------------------------------------------------------------------------------------------------------Week-Averaged = 19.8 Module : 35 watt Module Output = 9.0 Ah/day
Daily Load
Series Batteries
Parallel Modules
Daily Load
.
Output
Factor
Efficiency
[ 19.8 ]
[9.0 ] X [ 0.9 ] X [0.9 ]
Series Modules
= System Voltage
12

[_12_]
12
15-35
2.7
-> 3
1
Siemens Solar Module Output Tables

Siemens Solar has developed a variety of computer sizing programs that assist in
designing photovoltaic power systems for a wide variety of applications. The
advantage of using the output of a computer program is that all the important
variables, such as the effect of temperature on module performance, is taken into
account. Greater precision is possible when accounting for variables such as
battery efficiency and sun angle. The results of these computer calculations are
made available to you in the Appendix section to allow for more precise calculations
than is possible with peak hour estimations.
The Siemens Solar Module Output Tables contain module output values
calculated from these programs for over a hundred sites in the form of
Ah/day/module for each month of a typical year. Use these values in the complete
array sizing formula to calculate the number of modules needed.
Assumptions
For each location, the data includes the latitude and longitude, the Siemens Solar
map code identifier number, and then twelve output values, one for each month.
The values are calculated by the program PVSYSTEM, and are based on the
following assumptions:
(1) Constant load every month. This simplifies the automatic calculation process
and is applicable to a wide variety of situations.
(2) Best tilt angle for constant load. The computer chooses the array tilt angle that
best matches a constant load. This tilt angle is presented at the end of each
locations data.
(3) Module operating voltage of 14.3 volts. This is based on an assumed average
battery voltage of about 13.7 volts plus about .6 volts added for system losses
(4) Ground reflectance of 0.20 from dry grass.
Multiply SM55 Data By 1.44 To Create SP75 Data

There is a table of data for the 36-cell SM55 and for the 33-cell SM50-H. The values
for the SM55 can be used for the SP75 as well. Simply multiply the SM55 value by
1.44 to account for the larger cell used in the SP75.

15-36
The first page of the North American Region data table is presented to illustrate the
information presented. There are five regions of data: North America; Latin and
South America; Europe; Asia; and Africa. The most appropriate regional data for
you has been included.
NORTH AMERICAN
REGION
Module = SM55
Ah/day/module
Location
Map
Code
Lat
Long :
:
:
UNITED STATES
Alabama
BIRMINGHAM
MOBILE
MONTGOMERY
Alaska
ADAK
ANNETTE
BARROW
BETHEL
BETTLES
BIG DELTA
FAIRBANKS
GULKANA
HOMER
JUNEAU
KING SALMON
KODIAK
KOTZEBUE
MATANUSKA
MCGRATH
NOME
SUMMIT
YAKUTAT
Arizona
PHOENIX
PRESCOTT
TUCSON
WINSLOW
YUMA
Arkansas
FORT SMITH
LITTLE ROCK
California
ALPINE
ARCATA
ARROWHEAD
BAKERSFIELD
BLYTHE
BUTLER VALLEY RANCH
CARLSBAD
CARRISA PLAIN
CHULA VISTA
DAGGETT
DAVIS
867
868
224
34
31
32
87
88
86
900
901
902
136
904
905
158
907
908
909
910
911
912
185
913
914
915
916
52
55
71
61
67
64
65
62
60
58
59
58
67
62
63
65
63
60
177
132
157
162
152
146
148
145
152
135
157
152
163
149
156
165
149
140
197
871
220
873
874
33
35
32
35
33
112
112
111
111
115
875
180
35
35
94
92
692
881
693
882
695
1115
696
1223
697
884
152
33
41
34
35
34
41
33
35
33
35
39
117
124
117
119
115
124
117
120
117
117
122
:
:
:
:
:
:
:
:
:
:
:
:
:
:
:
:
:
:
:
:
:
:
:
:
:
:
:
:
:
:
:
:
:
:
:
:
:
:
:
:
:
:
:
:
:
:
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
DEC
Best
Tilt
11.38 13.12 14.14 14.75 13.98 13.39 13.14 14.20 14.64 15.48 13.41 11.34
12.73 14.37 14.82 14.63 13.55 12.63 12.10 13.09 13.98 15.86 14.14 12.36
11.84 13.49 14.39 15.00 14.02 13.49 13.13 14.14 14.50 15.82 14.01 12.12
55
55
55
5.49
4.67
0.00
3.45
0.02
1.61
0.79
2.67
4.17
3.24
4.95
4.57
0.01
3.62
2.04
0.62
2.02
2.94
8.03
7.46
2.24
8.09
5.56
7.40
6.79
7.78
8.07
5.89
9.04
7.96
5.25
9.12
7.15
6.66
7.07
5.74
9.48
10.07
11.21
12.27
12.58
13.34
12.88
13.41
12.21
9.04
12.60
12.24
12.14
16.20
12.31
11.58
12.62
9.66
10.03
11.49
15.56
13.00
15.49
14.65
14.22
14.83
13.26
11.02
12.45
12.88
15.32
14.72
13.52
13.97
14.34
11.35
9.26
11.55
12.09
11.67
15.30
14.32
13.89
13.37
12.54
10.55
11.51
11.20
15.58
13.17
12.41
13.52
13.81
10.50
8.54
10.33
15.13
10.80
14.47
13.35
13.26
12.82
12.26
10.42
10.62
11.21
15.23
12.30
11.67
13.22
12.08
10.03
8.39
10.67
15.22
9.61
12.89
12.84
12.37
12.47
11.85
9.85
10.04
10.83
13.35
11.50
10.68
11.21
11.01
9.39
8.27
10.41
10.36
8.28
11.22
12.46
11.30
12.25
10.96
9.10
9.27
10.95
11.34
11.05
9.76
9.73
10.11
8.81
8.74
9.88
6.47
9.20
10.62
11.46
10.61
11.31
10.41
7.93
9.91
10.30
10.36
9.92
9.72
9.77
9.98
8.10
8.70
7.10
2.79
7.50
6.34
7.56
6.79
8.78
8.99
5.53
9.59
9.75
6.53
7.75
6.77
7.11
7.86
6.46
6.96
5.21
0.03
4.10
1.28
3.29
2.66
3.83
5.51
3.72
6.33
5.93
0.71
4.76
3.41
1.99
3.72
3.56
4.90
3.34
0.00
1.62
0.00
0.00
0.00
0.72
2.18
1.63
3.46
3.26
0.00
1.75
0.33
0.00
0.20
1.37
70
75
55
80
80
80
80
80
80
75
80
75
75
80
80
80
80
75
17.62
17.95
18.38
17.46
18.56
19.61
19.60
20.12
19.62
20.43
20.31
20.84
21.00
21.00
21.69
20.48
21.31
21.20
21.47
21.59
19.07
20.72
20.05
20.51
20.37
17.59
19.95
18.72
19.64
19.16
16.73
17.58
17.02
17.81
17.43
18.01
18.21
18.14
18.60
18.81
19.81
20.72
19.98
20.53
20.35
20.32
20.84
20.53
20.55
20.74
18.76
19.14
19.14
18.93
19.42
17.05
17.31
17.60
16.76
17.92
55
50
50
50
50
12.76 14.02 14.34 13.87 13.73 13.67 14.00 14.87 14.98 15.62 13.93 12.45
12.28 13.88 14.20 13.71 13.72 13.64 13.71 14.64 15.02 15.83 13.60 12.04
60
60
13.55
10.07
13.08
13.28
17.02
9.59
13.34
16.38
14.65
17.08
11.62
55
65
55
60
50
65
55
60
55
60
60
17.49
12.19
16.73
15.79
19.47
8.90
17.29
16.23
18.11
18.53
13.20
15.57
13.11
16.96
17.86
20.59
10.91
17.90
15.93
17.63
19.78
16.69
16.90
14.06
18.50
18.03
20.40
13.02
16.85
16.45
17.36
19.28
17.51
14.40
13.40
18.38
17.46
19.01
12.72
14.55
16.89
14.86
17.72
17.18
16.62
13.03
18.69
17.09
18.19
13.51
14.59
15.95
14.56
16.88
16.82
17.05
12.59
18.64
17.26
17.20
14.98
14.86
16.17
15.37
16.53
17.50
17.65
12.93
19.69
18.64
18.69
15.89
15.59
17.29
16.02
18.04
18.64
19.18
14.12
21.39
19.84
19.70
16.43
16.58
17.58
17.66
19.60
19.56
19.56
13.07
21.18
19.34
20.45
11.57
15.67
18.96
17.43
19.86
17.98
19.03
10.66
19.76
15.77
19.71
9.17
17.84
14.60
18.00
18.27
14.57
16.76
9.55
17.06
12.40
17.83
8.71
15.92
14.04
16.02
16.63
11.18

"# $$ %$&

15-37
Please find the Data Tables appropriate for your region in the Appendix section.

15-38
Two examples of how to use the data are presented next. The first example is the
remote school, and the second is the remote cabin.

15-39
Example:
The remote school DC load demand was determined to be 421

Ah/day. Assume the location is near Yuma, Arizona. Looking at the
data for Yuma, we find that the lowest output during the year for an
SM55 module is 17.43 Ah/day during July. Assume a conservative
battery coulombic efficiency of 90%. Apply the usual 10% derating of
module output.
The number of SM55 modules needed for this applications is given by

.9 X [ 17.43 Ah/day X .9 ]
29.8
round up to
30 modules in parallel

12 volt module voltage (nom.)
2 in series
The work for this example is worked out on the next page, using the lower portion of
the Array/Battery Sizing Form.

15-40

System Description
Remote School
Max. Daily Load
_____
_____
System Voltage
_____
_____
Average Discharge
Rate


Battery Capacity
(@ ___ hour rate)
_____
_____
(hours)
_____

[_____ ] X [_____ ]
_____
Ah
[
Chosen Battery
Capacity =
] X [
_____________
_____@ _____hour rate
Parallel Batteries
Voltage
Battery Capacity
=
Chosen Battery Cap.
[_____]
[
_____
=
_____
System Voltage
= [_____]
= _____
Chosen Battery V.
[
]
----------------------------------------------------------------------------------------------------------------Week-Averaged = 421 Module : SM55 Module Output = 17.43 Ah/day
Daily Load
Series Batteries
Parallel Modules
Daily Load
.
Output
Factor
Efficiency
[ 421 ]
[17.43 ] X [ 0.9 ] X [0.9 ]
Series Modules
= System Voltage
12

[_24_]
[ 12 ]
15-41
29.8 -> 30
=
Example:
The weekend cabin system was determined to have a daily load

demand of 69.7 Ah/day. But if the system is really only going to be for
weekend use, then the array can be sized based on the Week
Averaged Load Demand . The array can be sized to replace the 2day weekend demand over the entire 7-day week. Recalculating the
69.4 Ah/day daily load spread over the week: 69.4 Ah/day X 2 days /
7 days = 19.8 Ah/day week averaged. Now we size the array to
replace this amount on a daily basis.
Assume the location is in the mountains near Arrowhead, California.
Looking at the data for Arrowhead, we find that the lowest output for an
SM55 module is 13.08 Ah/day during January. We have decided to
use SP75 modules in this application. So the output from a SP75
would be adjusted from this value for an SM55 by multiplying by the
factor of 1.44.
SP75 output in January
=
=
=
SM55 output X 1.44

13.08 Ah/dayX 1.44
18.84 Ah/day
Assume the same conservative battery coulombic efficiency of 90%,

and the usual 10% module output derating factor.
The number of SP75 modules needed for this applications is given by:
19.8 Ah/day avg daily load

.9 X [ 18.84 Ah/day X .9 ]
1.3 modules
round up to
2 modules in parallel

12 volt module voltage (nom)
1 in series
The work for this example is again worked out on the next page, using the lower
portion of the Array/Battery Sizing Form.

15-42

System Description:
Weekend Cabin
Max. Daily Load
_____
_____
System Voltage
_____
_____
Average Discharge
Rate


Battery Capacity
(@ ___ hour rate)
_____
_____
(hours)
_____

[_____ ] X [_____ ]
_____
Ah
[
Chosen Battery
Capacity =
] X [
_____________
_____@ _____hour rate
Parallel Batteries
Voltage
Battery Capacity
=
Chosen Battery Cap.
[_____]
[
_____
=
_____
System Voltage
= [_____]
= _____
Chosen Battery V.
[
]
----------------------------------------------------------------------------------------------------------------Week-Averaged = 19.8 Module : SP75 Module Output = 18.84
Series Batteries
Ah/day
Daily Load
Parallel Modules
Series Modules
Daily Load
.
Output
Factor
Efficiency
[ 19.8 ]
[ 18.84 ] X [ 0.9 ] X [0.9 ]
= System Voltage
12

[ 12 ]
[12 ]
15-43
1.3
-> 2
Exercise

5<!!

,

<

,
$>

5

$
!4 B

5
5>!0

5<!!

5>!
%%
0
"%


15-44
Checking Battery Sizing

Checking Depth of Discharge
There are a couple of checks that you can perform on your array and battery
calculations to learn more about how the system might perform. The first is to
calculate the average daily depth of discharge for the installed battery. This check
can be interesting because you may have chosen to install a deep cycling battery,
but if the number of days of autonomy is large, the actual daily depth of discharge
can be quite shallow.
Daily Depth of
Discharge
Example:
Daily Load (Ah)

Total Installed Capacity
Daily Load (Ah)
Number Parallel Batteries X Battery Capacity
The remote school problem worked on pages 16 and 32 has been

designed with 3175 Ah of installed capacity. The daily load was 421
Ah at 24 volts. A check on the actual average depth of discharge
shows:
Daily DOD
421 Ah daily load

3175 Ah battery capacity
.13 or about 13% daily

15-45
Checking Maximum Rate of Charge

Another check on your calculations is the maximum rate of charge of the battery
bank by the installed array. The battery should not be charged too fast during peak
irradiance periods or damage could occur. Battery manufacturers will tell you the
maximum rate of charge for specific models of batteries. If no specifics are
available, refer to this chart of typical maximum allowable charge rates for different
types of batteries.
Typical Maximum Charge Rates

Pure lead grids, stationary battery
Lead calcium grids, shallow cycling battery
Lead antimony grids, deep cycling battery
C/25
C/15
C/5
The formula for checking the maximum charge rate of your installed system is given
below. You simply divide the total peak current that can come from the installed
parallel modules into the total installed battery capacity.
Maximum Rate
Of Charge
Example:
Total Installed Capacity (Ah)

Total Array Peak Current (Amps)
Number Parallel Batteries X Battery Capacity

Number Parallel Modules X Imp
The remote school problem worked on pages 16 and 32 has been designed
with 30 75-watt modules in parallel by 2 modules in series for 24 volt
operation. There will be 3175 Ah of installed capacity. A check on the
maximum rate of charge shows:
Max. Rate
3175 Ah battery capacity

30 parallel modules X 4.4 amps Imp (75 watt module)
24 hours
This is slower than the typical maximum of five hours for deep cycling
batteries, so the array is not too big for the battery.

15-46
Calculating Tilted Insolation

The insolation data used must be incident on a tilted surface. Raw weather data is
usually measured on a flat (horizontal) surface, and it will not represent the amount
of insolation available when the modules are mounted on a tilted surface. Data must
either be measured on a tilted surface, or some mathematical work must be done to
the raw horizontal data to predict insolation on a tilted surface.
General Concepts in Translating From Horizontal to Tilted

Insolation
For many regions of the world, the only available solar radiation data is in the form of
monthly averages for daily global horizontal insolation. Photovoltaic system
designers require insolation data for tilted surfaces to properly size systems. For this
reason, methods are needed to translate actual measured horizontal data to
reasonably accurate values for tilted surfaces. Presented next is a number of
empirical and geometric relationships required for this estimation.
The total solar insolation or radiation intercepted on a tilted surface is made up of
three contributors: (1) direct beam radiation coming from the actual disk of the sun;
(2) diffuse whole-sky radiation coming from the entire canopy of the sky; and (3)
reflected radiation bouncing from the ground or nearby surfaces.
Total Insolation on a Tilted Surface
=
+
+
Direct (beam) Insolation

Diffuse (whole-sky) Insolation
Reflected Insolation
When measurements of total (global) insolation on a horizontal surface are

recorded, and then averaged, these components are lost, and all that remains is a
total bucket of insolation with no information about how much was direct and how
much was diffuse.
We need to separate these two components from the total, because they contribute
to the total insolation on a TILTED surface in different ways. The diffuse component
is just a function of how much of the total canopy of the sky the tilted surface sees.
The direct component needs to be added up as the angle of the sun changes during
the day, and is a function of the exact path of the sun on that day at that latitude.
The reflected component just depends on the tilt angle of the surface as it
accumulates the same fraction of both the direct and diffuse.

15-47
Global Horizontal Insolation

Sun
Diffuse whole-sky radiation
Direct
beam
radiation
Horizontal
terrestrial
surface

'( $) *+
Tilted Insolation
Sun
Direct
beam
Diffuse whole-sky radiation
Reflected
radiation
Tilted
terrestrial
surface

,( $

15-48
Summary of Steps
Summary of steps to calculate insolation on a fixed tilted surface using monthly
average insolation data measured on a horizontal surface:
Step
Procedure
Variable
Establish the desired tilt angle of the modules and the

latitude of the site.
2
Read from literature the measured average terrestrial
insolation on a horizontal surface for each month.
3
Determine the declination and sunset hour angle for a
representative day for each month. These are geometric
variables that depend only on day of year and latitude.
4
Determine the extraterrestrial horizontal insolation. This
is the radiation outside the atmosphere, a variable due
the non-circular orbit of the Earth around the Sun.
5
Calculate the clearness index, the ratio of actual
terrestrial to extraterrestrial insolation. This indicates
how much of the theoretical radiation actually got through
to the surface.
6
Determine the ratio of the diffuse to total insolation. This
comes from a model of the sky, and separates the
diffuse or indirect radiation from the total. (The remainder
would be the beam or direct radiation coming from the
disk of the sun).
7
Determine the sunset angle for a tilted surface (if
different from the angle for a horizontal surface).
8
Determine the beam radiation tilt factor. This is a
geometric factor that adjusts for the surface area
presented at a tilt angle compared to horizontal.
9
Select the ground cover and the associated ground
reflectivity factor.
10
Calculate the final total insolation on a tilted surface by
adding up the contribution for direct, diffuse, and
reflected insolation.

'
-($

15-49
s, L
H
, hs
Ho
Kt
H diffuse
h' s
Rb
g
H tilted
Equations for Conversion from Horizontal to Tilted Insolation

First, the declination, , must be calculated for a representative day for each month
of the year:
360
= 23.45 cos
(284 + n)
365
Eqn. 2-1, p.12 (Hsieh)
Where n = julian day number for the representative day in each month.
The following table gives the recommended average day for each month of the
year:
Date
January 17
February 16
March 16
April 15
May 15
June 11
July 17
August 16
September 15
October 15
November 14
December 10
Julian Day number

17
47
75
105
135
162
198
228
258
288
318
344
Source: S.A. Klien, Calculation of Monthly Average Insolation on Tilted Surfaces,

Solar Energy, vol. 19, p.325, Permagon Press Ltd., 1977.

,.$/-
Next, the sunset hour angle, hs, is determined for the representative day for each
month:
hs = cos 1 ( tan L tan )
Eqn. 2-10, p.15 (Hsieh)
Where L = site latitude.

15-50
The monthly average daily solar insolation on an extraterrestrial horizontal surface,

H o , is then calculated. The over bar designates a monthly average value.
Ho =
24
360n
2hs
Hsc 1 + 0.033 cos
cos L cos sin hs +
sin L sin
365
360
Eqn. 3-38, p.44 (Hsieh)
Where Hsc = solar constant = 1353 W/m2

Next, the monthly average clearness index, K T , is defined by the ratio of the
monthly average daily insolation on a terrestrial horizontal surface to the radiation on
an extraterrestrial horizontal surface:
KT =
H
Ho
Eqn. 3-42, p.52 (Hsieh)
Then, the diffuse to total radiation ratio, H diffuse H , is given by Collares-Pereira &
Rabl:
] (
H diffuse
= 0.775 + 0.00653( hs 90) 0.505 + 0.00455(hs 90) cos 115 K T 103
H
Eqn. 3-51, p.61 (Hsieh)
The sunset hour angle, h's, on tilted surfaces facing the equator is determined by:
hs = min hs , cos 1 [ tan( L s) tan ]
Eqn. 3-55, p.62 (Hsieh)
Where s = surface tilt angle.

15-51
The monthly average beam radiation tilt factor, R B , is in general a complicated

function of atmospheric transmittance. However, Liu and Jordan have estimated R B
as the ratio of extraterrestrial radiation on the tilted surface to that on a horizontal
surface for the month. For surfaces facing the equator, R B is estimated by:

cos( L s) cos sin hs +
h sin( L s ) sin
180 s
RB =

cos L cos sin hs +
h sin L sin
180 s
Eqn. 3-53, p.61 (Hsieh)
The monthly mean total radiation tilt factor, R , is calculated by:

R=
H tilted
H diffuse
H diffuse 1 + cos s
1 cos s
= 1
RB +
+ g
2
2
H
H
H
Eqn. 3-52, p.61 (Hsieh)
Where g = ground reflectivity, generally use 0.2 as default value. A table of ground
reflectivity factors is presented below for various common surfaces:
Ground Condition
Ground Reflectivity factor

g
Earth roads
Water surfaces from above
Coniferous forest (winter)
Weathered blacktop
Gravel, bituminous roof
Soils (clay, loam)
Dry grass
Crushed rock
Weathered concrete
Green grass
Forest in autumn, ripe field
Dark building surfaces
Dead leaves
Light building surfaces
Snow
.04
.07
.07
.10
.13
.14
.20
.20
.22
.26
.26
.27
.30
.60
.75

0)

15-52
Finally, the monthly average daily total solar radiation , H t , on tilted surfaces is
approximated by:
H tilted = R H
Eqn. 3-52, p.61 (Hsieh)
This final equation can be re-written in another form to more clearly show the three
components adding up to give the total insolation:
We can rewrite
H H diffuse as the direct component
equation can look like the following:
H direct . Then the
Total Insolation on tilted surface = direct + diffuse component + reflected

1 + cos s
1 cos s
H tilted = H direct R b + H diffuse
+ Hrg
2
2

15-53
Tracking Surfaces
For tracking surfaces, the monthly average insolation must be estimated using longterm hourly radiation data. First, a more comprehensive form of the equation for the
monthly average beam radiation tilt factor, R B , must be used to account for the daily
variation in surface tilt (s) and azimuth angles () for the tracking surface. An
equation for R B for surfaces of any orientation has been developed by Klien:
R B = [ A B + C + D + E ] 2 cos L cos sin hs +

hs sin L sin Eqn. 3-56, p.62 (Hsieh)
180

Where,
(hs hr )
180
B = (sin cos L sin s cos )

(hs hr )
180
C = (cos L cos cos s)(sin hs sin hr )
A = (cos s sin sin L)
D = (cos cos sin L sin s)(sin hs sin hr )
E = (cos sin s sin )(cos hs cos hr )

hr = sunrise hour angle on tilted surface, deg.
Charts used for rapid graphical solutions of R B are presented in Hsieh.

15-54
The hourly average radiation on a tilted surface can be approximated from monthly
average daily insolation by the following analytical expression suggested by
Collares-Pereira and Rabl:
r=
cos h cos hs
( a + b cos h)
24
sin hs (2hs 360) cos hs
Eqn. 3-44, p.58 (Hsieh)
Where r = ratio of hourly total radiation to the daily total radiation

h = hour angle in degrees at mid point of hour
a = 0.409 + 0.5016 sin(h s 60)
b = 0.6609 0.4767 sin(h s 60)
The hourly totals for insolation on the tracking surface must then be integrated over
the entire day, over the range of surface tilt and azimuth angles to determine the
daily total radiation on tilted surfaces.
In practice, insolation enhancement factors are generally applied to the insolation on

equator facing latitude tilt surfaces to approximate the insolation on tracking
surfaces. For single-axis tracking, insolation enhancements of between 15 and 20
percent are typical. For two-axis trackers, enhancement values of between 25 and
30 percent are common. In general, passive trackers and areas with higher diffuse
to total radiation ratios will have lower insolation enhancements due to tracking.
References
Hsieh, J.S., Solar Energy Engineering, Prentice-Hall, 1986.

15-55
(End of Chapter)

15-56
CHAPTER FIFTEEN
System Sizing
15-1
Basic Principles
15-2
Size Array For Worst Season
15-3
Battery Sizing
Days of Autonomy or Reserve
Basic Battery Sizing Calculation
Correcting For Cold Temperature
Cold Temperature Data
Specifying Battery Capacity At Proper Discharge Rate
Limiting Maximum Depth of Discharge To Prevent Freezing
Complete Battery Sizing Calculation
Battery Cell Size
15-4
15-4
15-5
15-8
15-10
15-11
15-14
15-15
15-19
Array Sizing
Basic Array Sizing Calculation
Modifications To Array Sizing Formula
Complete Array Sizing Calculation
Choose A Module With The Proper Number Of Cells In Series
Peak Hour Method Of Estimating Module Daily Output
Siemens Solar Module Output Tables
Checking Battery Sizing
Calculating Tilted Insolation
15-20
15-20
15-21
15-22
15-24
15-26
15-36
15-45
15-47

15-57
Chapter 15 Answers
System Sizing

A photovoltaic system needs to be sized for the worst case to insure that it will operate
reliably year round. If the system were only sized to meet the average requirements,
then the system would be undersized for half the time. This could result in a lack of
sufficient energy to support the load during the periods of poor weather. Even if the
system could support the load in these conditions, the battery would remain at a low
state of charge for long periods of time. This would damage the battery and reduce its
lifetime. Sizing for the worst case results in a PV system that will operate reliably
throughout the year and extend the lifetime of the battery.

The load demand calculated in Problem 9-1 is 230 Ah/day. Since low cost batteries will
be used, the maximum % useable capacity will be 50%.
Battery capacity
5 days X 230 Ah/day

0.50
2,300 Ah
Since the system is 12 volts, we will need a 12-volt battery of this capacity or sufficient
smaller voltage cells in series.

The load demand calculated in Problem 9-3 is 146.8 Ah/day. Using deep cycle
industrial batteries, we have 80% useable capacity.
Battery capacity
14 days X 146.8 Ah/day

0.80
2,569 Ah
The overall battery bank voltage must be configured at 24 volts. This could be two 12volt batteries in series, four 6-volt batteries, or twelve 2-volt cells.
15-1
System Sizing

Lead-acid cells should not be completely discharged, even for high quality, deep cycle
batteries. Discharging lead-acid batteries beyond the 80% DOD limit can damage the
plates, cause the plates to break apart, or reverse the cell polarity. Any of these could
significantly reduce the battery lifetime.

The only capacity information we have for the batteries is the "rated capacity" of 80 Ah.
We do not know how this varies with rate. To get an assumption of the capacity at a
reasonable discharge rate, we will use the "rule of thumb" indicated in the text. We will
multiply the rated capacity by 130%. So the capacity will be:
80 Ah X 130%
80 X 1.30
104 Ah
The load demand is 30 Watts / 12 volts = 2.5 amps. The total daily Ah load is 2.5 amps
X 12 hours = 30 Ah/day. The battery autonomy is specified to be 5 days. Low cost
batteries will result in a maximum % useable capacity of 50%. The load operating time
is 12.
The average rate of discharge is
Avg Rate of Discharge
5 days X 12 hours
50% useable capacity
120 hour rate
The coldest 24-hour temperature is 0 C, which results in a temperature derating of

92% (using the C/120 curve from Figure 15-1). There is no limit on the depth of
discharge due to freezing concerns, so the maximum % useable capacity is 50%.
Battery Capacity
(@ specified rate)

Max % Usable X Temperature Derating Factor
Battery Capacity
(@ 77 hour rate)
5 days X 30 Ah
.50 X .92
326 Ah
The number of parallel batteries is:

Parallel Batteries
326 Ah
104 Ah per battery
3.13 batteries
15-2
rounded to 4 batteries.
System Sizing
Since this is a 12 V system the number of batteries in series is 1. Note that we required
4 batteries in parallel for a very small load. A larger load would have required more
than 4 batteries in parallel, but this is not good practice. This illustrates one of the
limitations of using automotive batteries for solar applications.

We calculated the load demand in Problem 9-3 to be 146.8 Ah/day. The number of
days of autonomy is given as 14. The deep cycling batteries will allow a maximum
charge of 80%. However, we have an additional limitation due to concerns about
freezing, which limits the maximum % capacity to 75% (refer to Figure 15-2).
For the weighted average load operating time, we should only take the peak loads that
will be on at the same time (the transmit loads for both transmitter #1 and #2):
Weighted operating time
8 X 12 + 5 X 8
8+5
136 / 13
10.5 hours
The average discharge rate will be:

Average Discharge rate
Days of Autonomy X Operating Time

Maximum Discharge %
14 X 10.5
0.75
196 hours
At - 30 C, we need to derate the battery capacity 88% for temperature (using the
C/120 curve of Figure 15-1). The battery capacity is then:
Battery Capacity
@ C/196
14 days X 146.8 /day

.75 Max discharge X .89 temperature factor
3079 Ahrs at the 196 hour rate
Looking at Table 15-1, we see columns for 120 hours and 240 hours. To be
conservative, we will use the 120-hour column (since this gives the smaller rating for the
same cell). There are no cells with a capacity of 3114 Ah, so we will need to use
multiple cells in parallel. The capacity of KCPSA-15 cells is 1031. Three of these cells
in parallel give 3093 Ah, slightly more than we need. To get a 24-volt string of 2-volt
cells, we need 12 cells in series. Multiplying by 3 cells in parallel, we will need 36 cells
total.
15-3
System Sizing

Our requirement from Problem 15-6 is for 3079 Ah at a discharge rate of C/120. There
are a number of possible configurations:
Cells in
Paralle
l
2
3
4
Cell
Model
Cell
Capacity Total Ah
Vb2412
Vb2407
Vb2312
1800
1050
900
3600
3150
3600
Although 5 strings of the Vb2310 (750 Ah) would also give sufficient capacity, we rule
this out because it exceeds the Rule of Thumb for using no more than 4 parallel strings.
Since the Vb2407 is the smallest size that meets our requirements, it would probably be
the best choice. However, we would need to look at overall price, cell size and weight,
battery layout and similar details.

Location: Leh
The month with the lowest insolation is January, with 4 kWh/m2/day. To convert to
peak hours, we multiply by the conversion factor (in this case just 1).
4 kWh/m2/day X 1 = 4 peak sun hours
Looking on the data sheet, we find the following information:
Module
SP36
SP75
Pmax
36
75
Imp
2.1 Amps
4.4 Amps
The output for each module is the Imp multiplied by the peak sun hours
SP36:
2.1 amps X 4 peak sun hrs = 8.4 Ah/day
SP75:
4.4 amps X 4 peak sun hrs = 17.6 Ah/day
15-4
System Sizing
Location: Madras
The month with the lowest insolation is July, with 4.5 kWh/m2/day. The peak sun hours
are:
4.5 kWh/m2/day X 1 = 4.5 peak sun hours
Again, we take the Imp and multiply by the peak sun hours:
SP36:
2.1 amps X 4.5 peak sun hrs = 9.5 Ah/day
SP75:
4.4 amps X 4.5 peak sun hrs = 19.8 Ah/day

The load demand in Problem 9-1 is 230 Ah/day. We assume a nominal 12-volt system.
Looking at the data for Montgomery, Alabama, we find that the lowest output during the
year for an SM55 module is 11.84 Ah / day in January. We use the same derating
factors in the example: a battery coulombic efficiency of 90% and a module output
derating of 10%.
Total number of modules

0.9 X 11.84 Ah/day X (1 - 0.10)
230
9.59
24 modules (rounded up)

12 volt nominal module
1 in series
1 series X 24 parallel = 24 modules.
15-5
23.98 modules
System Sizing

The load demand in Problem 9-3 is 146.8 Ah/day. According to the data for Yuma,
Arizona, the lowest output during the year for an SM55 module is 17.43 Ah / day in July.
Since we are using the SP75 module, we adjust the output by a factor of 1.44.
SP75 output
SM55 Output X 1.44
17.43 Ah /day X 1.44
25.1 Ah/day
Assume the same conservative battery coulombic efficiency of 90% and a module
output derating of 10%.
Total number of modules
146.8 Ah/day daily load

0.9 X 25.1 Ah/day X (1 - 0.10)
146.8
20.33
8 modules (rounded up)

12 volt nominal module
2 in series
2 series X 8 parallel = 16 modules.
15-6
7.22 modules
System Sizing
Chapter Sixteen
System Wiring
All the precise calculations in the world can be sabotaged by improper wiring of the
components. Poor choice of wire size can restrict battery charging and eventually
cause system failure even with an adequately sized array. Imagine someone working
on the electrical system. Anticipate problems or dangers and how you could eliminate
the dangers with safety disconnect switches and fuses. Thinking about safety and
wiring requires that you consider what can possibly go wrong not assuming that
everything will be safe and that people will not do stupid things. Design the system as if
you were living with it and consider how you would want yourself and your family
protected from the danger of fire and system failure.
Electrical wiring is a specialized field and this chapter cannot possibly cover all the
issues completely. In the United States the National Electric Code (NEC) covers all
issues of wiring in depth and specifically covers solar photovoltaic systems in Article
690. A professional electrician or electrical contractor should be involved in the wiring
of photovoltaic power systems.
The entire photovoltaic system should be diagrammed before costing or installation.
This should include the wiring of modules in the array or sub-arrays, controllers, AC and
DC load centers, batteries, inverter if present, grounding and circuit protection. The
wiring process is the time to consider safety features such as fuses, circuit breakers,
ground fault interrupters, and grounding rods and wires. Also efficiency and cost
concerns will affect many specific choices such as the sizes and types of wires to use.
In this chapter we will look at three aspects of photovoltaic system wiring: (1) proper
wire sizing and insulation; (2) array wiring in series and parallel groups; and (3) safety
equipment wiring.
Proper wiring and the design of safe, user-friendly photovoltaic systems are the
most overlooked aspects of PV system design. Adherence to the National
Electric Code and safe practice will result in reduced hazards associated with
electrical installations. Careful design of the wiring subsystem will result in
efficient and reliable PV systems that are safely and easily serviced.

16-1
System Design - System Wiring
The National Electric Code and

the UL Listing
The National Electric Code (NEC), first developed in 1897, is a code of standards for
the design of safe electrical power systems in the US. Article 690 of the Code
addresses safety standards (such as overcurrent protection) for the installation of PV
systems. Most local electrical inspectors in the US use the NEC as the code of practice
(sometimes with slight modifications) in their jurisdiction. The NEC states that
adherence to the recommendations made will reduce the hazards associated with
electrical installations.
The NEC suggests and many inspectors require that an approved testing laboratory
test electrical equipment used. The most common of these national testing
organizations is the Underwriters Laboratories or UL. Most building inspectors expect
to see UL on electrical products used in electrical systems. This presents a problem for
the PV industry as many products have low production rates that do not justify the costs
for the testing. As a result it is still difficult to design a PV system that has used all UL
listed components. Proper documentation on the use of the non-UL listed equipment
will assist in gaining the approval of the local inspector. Another helpful source of
information on the NEC as it applies to PV is Photovoltaic Power Systems and the
National Electric Code, Suggested Practices. This publication is available from the
Sandia National Laboratories, US Dept. of Energy, (505) 844-3698.
An important concept to understand is that the code is not designed to create the most
cost effective or efficient systems. Its focus is safety. There may be a tension between
a system designers desire to make a system low cost and the code requirements that
lead to added components or larger size choices. Safety, reliability, efficiency and cost
effectiveness must all be considered by the designer and by the user.
Exercise

!
"
# #
# $!

16-2
Proper Wire Type and Insulation

Wire can be solid or stranded. Often AC house wiring is done with solid wire. In
photovoltaic systems, however, the DC currents are higher than typical for AC
situations, and therefore the wire size must be larger. Solid wire becomes stiff and
difficult to work with in large sizes and photovoltaic installers often use stranded wire.
Wire can be made of aluminum or copper. Aluminum wire can be considered for very
large wire runs, needing 4/0 for example because it costs less than equivalent copper
wire. For most wiring applications, however, copper wire should be used.
Most wire is insulated with a type of plastic called thermoplastic. Thermoplastic is
designed to withstand considerable heat. When current flows through a wire its
resistance causes power to be dissipated. If the current and/or resistance is high the
power dissipated can be significant, causing the wire to get warm or even hot. For this
reason the insulation must be able to withstand heat. While most wire is made of
thermoplastic, rubber insulation is also available.
Thermoplastic and rubber insulation are designated by a letter code that indicates the
type of environment that is most appropriate for its use. The most common type is T,
which means thermoplastic insulation for use in dry areas where temperature will not
exceed 60 C. Type TW insulation should be used for wet (or dry) areas. Some
common types of wire and their U.S. designations are listed in the table.
Letter Designation
Type of Insulation
T
TW
THW
Thermoplastic
Thermoplastic Wet
Thermoplastic, High
temperature, Wet
Thermoplastic, High,
High temperature, Wet
Thermoplastic High,
High temperature,
Nylon jacket
Rubber insulation
Rubber, High
temperature, Wet
Rubber, High, High
temperature
Rubber, High, High
temperature, Wet
THHW
THHN
R
RHW
RHH
RHW-2
For Use in:

Dry Areas where temp will not exceed 60 C
Wet (or Dry) Areas where temp will not exceed 60C
Dry Areas where temp will not exceed 90C.
Dry Areas where temp will not exceed 60C

Dry Areas where temp will not exceed 90C
Wet or Dry locations where temp will not exceed
90C.


16-3
Color Coding of Wire

The color-coding of wire makes wiring easier and is used to designate its function. It
also minimizes the possibility that incorrect connections will be made. For AC house
wiring in the US white or gray is always used for neutral or main system grounds (as
opposed to equipment ground that is commonly green or bare wire). The hot wires can
be black, red, blue or yellow. Black is the most common; but, in cables with two hot
wires, black and red are used.
In PV systems the NEC specifies that in a DC circuit the system grounded conductor be
white. There is no convention designating the color of the ungrounded conductor but
typically red or black are used. Green or green with yellow-stripped is used for the
equipment ground.
Wire
AC(Below 600 Volts)
DC (Below 600 Volts)
Neutral or Ground
White or Gray
White
Hot (or high side)
Black, Red, Blue, Yellow
No NEC Requirement, but

typically black or red.
Equipment Ground or
Grounding
Bare, Green, or Green with

Yellow Stripe
Bare, Green, or Green with

Yellow Stripe

!
Conductor Types
A cable is a connection of two or more wires combined in a common outer covering. It
is pre-assembled with wires of various sizes and insulation covers. All of the wires are
protected with an outer sheath made of plastic, rubber, or metal. Most of the wire you
will be working with comes in cable form, which is easier to work with than multiple
individual wires.
The most commonly used cable for electrical wiring of AC systems is nonmetallic
sheathed cable or type NM. This is commonly referred to as Romex, which is used in
house and building wiring in dry locations. Non-metallic sheathed cable is available in
different versions for a variety of other applications including areas of moisture ( NMC)
and wet areas buried in the ground (UF) commonly referred to as underground feeder
cable.
16-4
Wires and cables used in PV systems are typically designed for outdoor situations with
moisture and sunlight resistance or are placed in conduit to protect against weather.
The following cable types are the most commonly used in PV systems.
Wire/Cable Type
USE or USE-2
(Underground Service
Entrance)
or
Design Use
When made to UL
Standard, it has a 90C
rating and is sunlight
resistant. Moisture
proof, resistant to acids
and chemicals.
Applications in
Photovoltaic
Systems
Design Issues for

Photovoltaic
Systems
Best cable for module

interconnects
The -2 designation
means that the
insulation is rated for
90C. wet or dry, and is
especially suited for the
hot temperatures that
module wiring would
endure.
TC is sunlight resistant
and generally marked as
such.
Not recommended
within battery
enclosures, Should
specify sunlight resistant
coating if used outside.
SE (Service Entrance)
TC (Tray Cable)
UF (Underground
Feeder)
THHN (Dry Locations)
Used for module

interconnects
Designed for use in wet
areas especially outside
where it may be
exposed to weather,
also may be burred
underground.
Must be used in
conduit
For use in array wiring,

and between array
wiring and the PV
disconnect device
Multi-stranded flexible
UL listed cable available
in large sizes
Multi-stranded flexible
cable available in large
sizes
Interior house wiring, dry
locations
Used for battery /

Inverter wiring
and
Used between the array

wiring and the PV
disconnect device
THWN (Wet Locations)
USE/RHH/RHW
Welding Cable
NM (Romex)
Used for battery

interconnects and for
battery/inverter wiring
Limited. Used in
conduit to connect array
to PV disconnect device
and for short DC wiring
in the house.
Must be used in conduit

with insulation rated at
90 C. In larger sizes, is
difficult to bend.
Conduit sizing depends
upon number and size
of conductors
Not UL Listed
Conduit sizing depends

upon number and size
of conductors
"#$%&

16-5
Consider Both Ampacity and

Voltage Drop in Wire Sizing
There are two major considerations in sizing wire for a photovoltaic system:
Ampacity
Voltage drop
Ampacity is the current carrying capacity of the wire, or its ability to carry current safely
without overheating. Ampacity is based upon the wire size or cross-sectional area, the
type of materials its constructed of (e.g., copper, thermoplastic insulation), its length,
and its temperature. For these different conditions, the National Electric Code has
determined a maximum ampacity for each wire type (Romex, USE, etc.) for various
temperatures and lengths. This is the same standard for which PV systems need to be
designed.
The second consideration in wiring is voltage drop. Unlike higher voltage AC systems
where a small voltage drop will not have measurable impact on the efficiency and
deliverability of energy, PV systems are very sensitive to voltage changes. These
voltage drops must be considered in the design of a photovoltaic system.

16-6
Wire Sizing Based on Ampacity

Ampacities for common types' wire used in photovoltaic systems are presented in the
tables on the next page. Wire sizes are given in both the American Wire Gauge (AWG)
and the metric method of cross sectional sizing in square millimeters. The value in the
regular font is for wires in conduit while the italic numbers are the ampacity values of
single conductors in free air. These values represent the maximum current that the
conductor can safely carry continuously while the ambient air temperature is at 30 C.
If the ambient temperature is higher then the maximum current must be reduced. The
reduction for temperature is discussed later in this section.
We can identify general areas in photovoltaic system designs where wires are sized
based on ampacity: (1) short connections of DC wiring made indoors from batteries to
switches, controls and other components; (2) AC wiring from the inverter to a
distribution panel and on to the AC loads; (3) battery-to-inverter wiring; and (4)
conductors carrying array current.
Short DC Wiring From Battery

Wires to connect switches, controls and other components are put within conduit. The
runs are usually so short that voltage drop is not a problem and the wires are sized
based on ampacity. Usual safety practice is to oversize the wire by 125% above the
continuous current that the wire might handle. (This does not include wire carrying
current flow from the array into the battery. There is an additional safety factor that is
applied. This is explained later.)
AC Load Wire Sizing

For AC load wiring usual electrical practices apply. Circuits are designed to carry
typically 15 or 20 amps and usually AWG 14 or 12 wire is sufficient. Wires are usually
carried within conduit. Again, the usual safety factor for ampacity recommended by the
NEC is 125% of the maximum current that might flow continuously.

16-7
Wire Sizing (AWG)

Ampacity (amps)
Wire Size
(AWG)
UF
USE,RHW,
THHW,THWN
60 o
75 o
THHN,THHW
RHW-2,USE-2
90 o
14
12
20 25
25 30
20 30
25 35
25 35
30 40
10
8
30 40
40 60
35 50
50 70
40 55
55 80
6
4
55 80
70 105
65 95
85 125
75 105
95 140
2
1
95 140
110 165
115 170
130 195
130 190
150 220
125 195
150 230
170 260
00
000
145 225
165 260
175 265
200 310
195 300
225 350
0000
195 300
230 360
260 405
'
Regular Font = NEC 1993, Table 310-16. Ampacities of Insulated Conductors Rated 0-2000 Volts.
Not More Than Three Conductors in Raceway, Cable, or Earth (Directly Buried), Ambient Temp. of 30 oC.
Italic Font = NEC 1993, Table 310-17. Ampacities of Single Insulated Conductors Rated 0-2000 Volts
in Free Air Based on Ambient Temp. of 30 oC.
W ire Sizing (m etric)

W ire Size
(m m 2 )
Am pacity
(am ps)
2.5
4
6
10
16
25
35
50
70
95
120
150
32
42
54
73
98
129
158
198
245
292
344
391
W ire am pacity rated at am bient tem perature of 30 deg.C., m axim um con ductor tem perature of 60
deg.C ., a nd assu m ing cable touches surfaces of w alls, etc. V oltage loss fa ctor rated at 20 deg.C.
(
16-8
Inverter Wire Sizing

For the case of battery-to-inverter wiring the DC currents can be very high and care
must be taken to choose a wire adequate for the job. Inverters used for large
residential or commercial applications are often rated to produce 2000-4000 watts
continuously. If the input voltage to the inverter is 12 volts this will mean DC currents of
180-400 amps. Even if the inverter input voltage is 24 volts the current can be 100-200
amps. Very large wires are needed to handle such currents.
The distance between the battery and the inverter should be kept as short as possible
to minimize voltage drop at such high currents. Usually no more than about six feet or
two meters is recommended as the maximum length for inverter cables.
The current that should be used for inverter wire sizing should correspond to the full
load continuous output of the inverter, not just the load demand being sized for today.
A system designer may often choose to install a larger inverter than is needed, to
accommodate future growth. The wire size installed, however, should be designed
around the full continuous demand of the inverter. This will allow future growth in the
power to the loads without requiring costly re-wiring from small wires perhaps needed
today to larger wires needed in the future.
The formula for calculating the current that should be used as the basis for inverter
cable wiring is presented next. The continuous power rating of the inverter is divided by
the full load efficiency. This gives the input current needed to overcome internal
inverter inefficiency. This value is divided by the lowest DC input voltage to the inverter
(usually about 11 volts for 12 volt systems and about 22 volts for 24 volt systems) not
by just the nominal voltage (12, 24 or 48 volts). This is to account for the fact that at a
lower battery voltage the inverter will draw higher currents to meet load power
demands.
The NEC safety margin of 125% should be applied on top of this real current value for
purposes of choosing wire with adequate ampacity. In other words the calculated
current might actually be flowing for extended periods of time, and the wire chosen
should have an ampacity of at least 125% more than this value.

16-9
Ampacity of
Inverter Wire
Maximum Continuous Input Current X 1.25
Maximum
Continuous
Input Current
Continuous Inverter Output Rating (watts)

Full Load Efficiency X Lowest Input Voltage
Nominal Voltage
12
24
48
Example:
Lowest Input Voltage

11
22
44
The total load demand for a residential system comes to about

2.5 kW continuous. The system designer chooses to install the
Trace 4048 sinewave inverter, with 48-volt input and rated for
continuous output at 4000 watts. Efficiency at 4000 watts is about
87%.
Max. Inverter Input Current=
4000 watts
.87 X 44 volts
=
104 amps
Applying the 125% safety factor, the wire chosen should have an
ampacity of at least:
Ampacity value
104 X 125% =
130 amps
104 X 1.25
So in this example, a recommendation might be to use type USE/RHH/RHW cable with

an ampacity rating of at least 130 amps.

16-10
Conductors Carrying Array Current

Wires carrying current from the modules or array are to be sized based on their short
circuit current (Isc), rather than just on their maximum power current (Imp). The usual
NEC code practice is to choose wire with ampacity rating of at least 125% higher than
the array Isc.
Modules listed to Underwriters Laboratory Standards have an additional factor applied.
Module Isc is to be increased by 125% before any other calculations are considered, to
account for possible environmental enhancements of output such as bright clouds or
reflective ground cover. This factor is to apply even before the NEC safety factor of
125% is applied.
In other words, apply two factors to wires carrying array current:
Derate the ampacity of the wire by 20% (apply as 125% factor) to insure that the
wire is not operated continuously at more than 80% of its rated value.
Derate the expected continuous output current of a solar module by 125% to

account for expected continuous enhancements to output.
The combination of the two factors means that a factor of 156% should be applied to
the published Isc value of a module for ampacity sizing purposes.
Array wiring ampacity safety factor (UL and NEC)
156%
This factor should be applied to all short lengths of wire interconnecting array power to
switches, controls and other equipment.

16-11
Adjustment To Ampacity For Temperature

The ampacity of a conductor is based on the ability of the insulation to adequately
dissipate the heat generated by the current flow through the copper. Most ampacity
ratings are based on the ambient temperature around the wire being 30 C.
If the ambient temperature is greater than the rated value of 30 C. then the insulation
will not be able to dissipate heat as well. The ampacity rating of the conductor should
be decreased to reflect the increased ambient temperature.
The temperature derating factors recommended by the NEC for various conductors are
presented below. These must be applied against the standard values presented earlier
in Tables 16-4 and 16-5. The same factors apply for conductors in conduit or in free
air.
Ampacity Temperature Correction Factors

Ambient
Temperature (oC)
Standard Temperature Rating

60 oC
75 oC
90 oC
1.00
.94
.88
.82
.75
.67
.58
.33
---
1.00
.96
.91
.87
.82
.76
.71
.58
.41
26-30
1.00
31-35
.91
36-40
.82
41-45
.71
46-50
.58
51-55
.41
56-60
--61-70
--71-80
--


16-12
Example:
SM55 modules are mounted on a fixed structure in the desert. The

back of the module is exposed to free air. Ambient temperatures
can get up to 45 oC. Type USE-2 (rated for 90oC) AWG #10 is
used to interconnect the modules.
The temperature derating for 90oC conductors at 45oC is .87.
The standard ampacity of USE-2 wire AWG #10 in free air
is 55 amps. [Refer to the italic numbers in Table 16-4].
The combined derating for the ampacity is given by:
Ampacity
= 55 amps rating X .87 1.25 NEC 1.25 UL

= 30 amps
Example:
SP75 modules are mounted on a roof. The conductors are in

conduit, and the back of the module is mounted very close to the
roof surface where temperatures can get up to 75 oC. Type USE-2
(rated for 90oC) AWG #8 is used to interconnect the modules.
The temperature derating for 90oC conductors at 75oC is .41.
The standard ampacity of USE-2 wire AWG#8 in conduit
is 55 amps. [Refer to the regular font numbers in Table 16-4].
The combined derating for the ampacity is given by:
Ampacity
= 55 amps rating X .41 1.25 NEC 1.25 UL

= 15 amps
Temperature can lead to a substantial reduction in the ampacity of conductors,

especially in photovoltaic systems where wires interconnecting modules and connecting
an array to the main system components are usually exposed to high ambient
conditions.
Array wiring from the array in the field is usually sized even larger than this factor would
dictate. The array in the field is typically a long distance from all the other equipment in
the power system. The long distance of this run of wire requires a different approach to
wire sizing one based on voltage drop not on ampacity. This is discussed later in this
chapter.

16-13
Exercise

"%&'

%&
()*!+

,-
#.

!
//////////

" 0"+12.&

344+56)7
&'
.%
8!3

&!+
,
,
"

9
:
!
//////////

;<
" =

=!3
#6)
.% !3
>
&!&?&@.6!
>
&!(?&@.!%
!3
= =

&!
+

,//////////

>6) ,//////////


16-14
Wire Sizing Based on Voltage Drop

The second factor in wire sizing is voltage drop. This is often a small consideration in
common utility connected high voltage residential and commercial systems but
becomes critical in photovoltaic system design. Sizing based on voltage drop is
especially important for the wire run between the array and the battery. This distance is
usually quite large and voltage drop can be a big problem.
All conductors have some small resistance that causes a loss of voltage depending on
the diameter and length of the wire. The primary effect of small wire between a
photovoltaic array and a battery bank is to reduce the amount of current that flows into
the battery. This is because the resistance of the wire causes the voltage potential of
the array to be reduced at all current levels. More voltage is loss at higher array
currents. This results in a "dragging in" of the IV curve of the array, as shown below.
Effect of Voltage Drop

on Module Output
Phoenix
50 feet to battery
4 parallel SM50-H modules
12
10
no wire loss
AWG #4
Current 8
(amps)
6
AWG #10
AWG #14
4
2
-4
0
4
Voltage (volts)
12
16
Range of
battery
voltage
)

16-15
The batteries operate in a narrow range of voltage close to the "knee" of the array IV
curve if the knee is brought in due to small wire. The operating point on the curve
then drops down the IV curve and less current flows to the battery. The net effect is to
operate the array at a lower current level. The figure shows that if there was no
voltage loss in the wire and the batteries were operating at about 13.5 volts, then the
current from the array would be about 11 amps. With the losses expected using AWG
#10 wire the battery would get only about 8 amps. By using common residential wire
size of AWG #14 the current would be only about 6 amps, close to half of what is
expected. The current would not be what the designer anticipated and over time the
battery would never be fully recharged and the system would fail.
Size For 2% Maximum Loss

The National Electric Code recommends that there be no more than 5% voltage drop in
any circuit between a source of power (typically a main service entrance in a building)
and the furthest load. In stand-alone photovoltaic systems we can apply this rule to
mean there should be no more than 5% voltage loss from the modules to the DC loads
or into the inverter for AC systems.
An even more stringent recommendation is to not have more than a 2% voltage drop in
any feeder circuit or any branch circuit. A feeder circuit is any circuit from a source of
power (for example, a photovoltaic array or a battery bank) to a distribution center
where circuit protection is available (for example, a central wiring center or DC load
center). A branch circuit is any circuit from a distribution center to a load (for example,
from a DC load center to the loads). Some photovoltaic system designers use the 2%
factor from the modules to the battery and then from the battery to the loads to
preserve voltage and improve performance. This results in larger wires and slightly
greater costs but preserves more of the expensive energy generated in the first place.

16-16
Voltage Drop Factors

The voltage drop in a wire is calculated using the Voltage Loss Factors listed on the
next page (ohms/foot or volts/ampere/foot). The current and the total length of wire
(going to and returning from) are multiplied by the Factor to give the volts lost.
Voltage Drop in Wire = Current x Total Wire Length x Voltage Factor
The voltage loss in high voltage AC circuits is usually quite small whereas the losses in
low voltage DC systems can be quite high. This is not because the current is DC or AC,
but rather because the voltages in AC systems are usually higher (typically 120 or 240
volts) while the voltages in DC systems are usually low (typically 12 or 24 volts). Some
examples will illustrate this.
Example:
A 1000 watt load (ex: heater, microwave oven, hair dryer) is located 25
feet from a load center. The total length of wire for the circuit is therefore
50 feet. The amps of current to the load would be given by
Current
Power Volts
8.3 amps
1000 watts 120 volts
According to the ampacity charts, AWG #14 wire would be more than
sufficient to handle 8 amps.
If the wire used was standard household AWG #14 (voltage loss factor of
.002823 volts/amp/foot), the voltage loss in the circuit would be:
Voltage Drop =
(#14 wire)
(120 volts)
1000 watts X 50 feet X 0.002823 v/a/ft

120 volts
1.18 volts
The recommended maximum of 2% voltage loss for a 120-volt circuit would be a drop
of 2.4 volts. The 1.18-volt drop in this circuit is less than 2.4, so in this case the wire
chosen based on ampacity is also acceptable in terms of voltage loss.

16-17
Voltage Loss Factors (AWG)

Wire Size
(AWG)
Voltage Loss Factor

(volt/amp/foot)
14
12
10
8
6
4
2
1
0
00
000
0000
.002823
.001775
.001117
.0007023
.0004416
.0002778
.0001747
.0001385
.0001099
.0000871
.0000691
.0000548
*
Voltage Loss Factors (Metric)

Wire Size (mm2)
2.5
4
6
10
16
25
35
50
70
95
120
150
Voltage Loss Factor

(volt/amp/meter)
0.0084864
0.0052881
0.0034434
0.0021112
0.0013206
0.0008371
0.0005614
0.0004183
0.0003151
0.0002406
0.0001907
0.0001518
+

16-18
If the same load were operating at nominal 12 volts, however, the current would be
much higher to produce the same power and the wire size needed would be much
greater.
Example:
The same 1000 watt load is operating at 12 volts nominal, and is

located 25 feet from a DC load center. The total length of wire for
the circuit is therefore 50 feet. The amps of current to the load
would be given by
Current
volts
Power Volts
83.3 amps
1000 watts 12
(ten times higher than before)
According to the ampacity charts, the AWG #14 wire used in the
previous example would be insufficient to handle this high current.
We would need at least AWG #4 wire if we used 75 C rated
conductor.

16-19
Another example illustrates that using wire with enough ampacity does not necessarily
give the acceptable voltage drop of 2%.
Example:
An array of four SP75 modules is wired in parallel at 12 volts. The

array is located 30 feet from the load center and other components
of the power system.
The nominal Isc from the array will be given by
Array Isc
4 X 4.8 amps Isc
19.2 amps
The NEC and UL safety factors apply as this is module current.

Conductor ampacity must be overrated by 125% and current must
be overrated by 125%. The current that should be used for
ampacity is given by
Array Current for Ampacity
19.2 X 1.25 X 1.25
30 amps
Assume the conductors are in buried conduit with ambient of about

30 C. so no temperature derating will apply at this time. If we use
THHN (90 C rated) conductor, then AWG #12 wire would appear
adequate to handle the current.
But now we look at the voltage drop that would occur using
# 12 wire. The total length of conductor is 2 x 30 feet = 60 feet.
Using the nominal array Isc value of 19.2 and the voltage loss
factor for #12 wire we find that the voltage drop would be
Voltage Drop =
=
19.2 amps X 60 feet X .001775 v/a/ft

2.05 volts
The nominal system voltage is only 12 volts. This voltage drop of 2.05 volts is more
than 17% of that nominal voltage! The acceptable voltage drop for a 12-volt system is
only 2% of 12, or only .24 volts.

16-20
This illustrates that choosing wire based on ampacity for the long run from the array into
the main system components is unacceptable. We must size that run of wire based on
voltage drop. The wire chosen will be much larger than dictated by ampacity alone. To
size for acceptable voltage drop, we must use the equation on Page 14 in reverse.
Insert the wire length, the array current, and the acceptable voltage drop and solve for
the voltage loss factor. An example will illustrate.
Example:
Using the above array again, the array current is 19.2 amps, the
total wire length is 60 feet. The acceptable voltage drop is 2% of
12 volts or .02 X 12 = .24 volts.
Voltage Loss Factor
Acceptable voltage drop

Current X Total Length
.24 volts
19.2 amps X 60 feet
.000208
Referring to the Voltage Loss Factor chart of Table 16-7, the factor for #4 wire
(.0002778 volt/amp/foot) is not small enough, so go to the next larger wire. This means
that #2 wire would be the size needed to keep the voltage drop less than 2%. Compare
this to the #12 that was acceptable with regard to ampacity only.

16-21
Voltage Drop Tables

A set of tables has been created to allow easy wire sizing based on voltage drop.
Tables have been created for nominal 12, 24, 36, 48 and 120-volt systems.
Distances indicated are for the one-way length of the conductor. The total round-trip
length of conductor has been taken into account in the calculations.
American Wire Gauge (AWG) tables indicate lengths in feet and metric (mm2) tables
indicate in meters.
Procedure for using the tables:

1. Determine the current that might be flowing in the wire ( use Isc for array wire
calculations);
2. Read across the row to find the one-way distance (if between values go to the next
larger number);
3. Read up to find the acceptable wire size for 2% voltage drop.
If your current is greater than indicated in the tables then use the formula on Page 17
and the Voltage Loss Factors on Page 18 to determine the acceptable wire size. When
using the formula be sure to use the total length of the conductor, not just the one-way
distance that is used in the tables.
(Ampacity limits have not been indicated in these charts. Be sure to check that the
ampacity of the wire is not exceeded especially in the case of high current values).

16-22
Voltage Drop Table -- 12 Volt Nominal

Maximum one-way distance for 2% voltage loss in 12 volt systems
Wire Size (AWG)
Amps
1
2
3
4
5
6
7
8
9
10
15
20
25
30
35
40
45
50
60
70
80
90
100
12
10
00
000
0000
43
21
14
11
9
7
6
5
5
4
3
2
2
1
1
1
1
1
1
1
1
0
0
68
34
23
17
14
11
10
8
8
7
5
3
3
2
2
2
2
1
1
1
1
1
1
107
54
36
27
21
18
15
13
12
11
7
5
4
4
3
3
2
2
2
2
1
1
1
171
85
57
43
34
28
24
21
19
17
11
9
7
6
5
4
4
3
3
2
2
2
2
272
136
91
68
54
45
39
34
30
27
18
14
11
9
8
7
6
5
5
4
3
3
3
432
216
144
108
86
72
62
54
48
43
29
22
17
14
12
11
10
9
7
6
5
5
4
687
343
229
172
137
114
98
86
76
69
46
34
27
23
20
17
15
14
11
10
9
8
7
866
433
289
217
173
144
124
108
96
87
58
43
35
29
25
22
19
17
14
12
11
10
9
1092
546
364
273
218
182
156
136
121
109
73
55
44
36
31
27
24
22
18
16
14
12
11
1377
689
459
344
275
230
197
172
153
138
92
69
55
46
39
34
31
28
23
20
17
15
14
1737
868
579
434
347
289
248
217
193
174
116
87
69
58
50
43
39
35
29
25
22
19
17
2190
1095
730
548
438
365
313
274
243
219
146
110
88
73
63
55
49
44
37
31
27
24
22
Wire Size (mm2)

Amps
1
2
3
4
5
6
7
8
9
10
15
20
25
30
35
40
45
50
60
70
80
90
100
[distances in feet]
14
[distances in meters]
2.5
10
16
25
35
50
70
95
120
150
15
7
5
4
3
2
2
2
2
1
1
1
1
0
0
0
0
0
0
0
0
0
0
24
12
8
6
5
4
3
3
3
2
2
1
1
1
1
1
1
0
0
0
0
0
0
35
18
12
9
7
6
5
4
4
4
2
2
1
1
1
1
1
1
1
1
0
0
0
63
31
21
16
13
10
9
8
7
6
4
3
3
2
2
2
1
1
1
1
1
1
1
99
50
33
25
20
17
14
12
11
10
7
5
4
3
3
2
2
2
2
1
1
1
1
154
77
51
38
31
26
22
19
17
15
10
8
6
5
4
4
3
3
3
2
2
2
2
217
108
72
54
43
36
31
27
24
22
14
11
9
7
6
5
5
4
4
3
3
2
2
311
155
104
78
62
52
44
39
35
31
21
16
12
10
9
8
7
6
5
4
4
3
3
441
221
147
110
88
74
63
55
49
44
29
22
18
15
13
11
10
9
7
6
6
5
4
583
291
194
146
117
97
83
73
65
58
39
29
23
19
17
15
13
12
10
8
7
6
6
745
373
248
186
149
124
106
93
83
75
50
37
30
25
21
19
17
15
12
11
9
8
7
930
465
310
233
186
155
133
116
103
93
62
47
37
31
27
23
21
19
16
13
12
10
9
,

16-23

Wire Size (AWG)
(amps)
1
2
3
4
5
6
7
8
9
10
15
20
25
30
35
40
45
50
60
70
80
90
100
12
10
00
000
0000
85
43
28
21
17
14
12
11
9
9
6
4
3
3
2
2
2
2
1
1
1
1
1
135
68
45
34
27
23
19
17
15
14
9
7
5
5
4
3
3
3
2
2
2
2
1
215
107
72
54
43
36
31
27
24
21
14
11
9
7
6
5
5
4
4
3
3
2
2
342
171
114
85
68
57
49
43
38
34
23
17
14
11
10
9
8
7
6
5
4
4
3
543
272
181
136
109
91
78
68
60
54
36
27
22
18
16
14
12
11
9
8
7
6
5
864
432
288
216
173
144
123
108
96
86
58
43
35
29
25
22
19
17
14
12
11
10
9
1374
687
458
343
275
229
196
172
153
137
92
69
55
46
39
34
31
27
23
20
17
15
14
1733
866
578
433
347
289
248
217
193
173
116
87
69
58
50
43
39
35
29
25
22
19
17
2184
1092
728
546
437
364
312
273
243
218
146
109
87
73
62
55
49
44
36
31
27
24
22
2755
1377
918
689
551
459
394
344
306
275
184
138
110
92
79
69
61
55
46
39
34
31
28
3474
1737
1158
868
695
579
496
434
386
347
232
174
139
116
99
87
77
69
58
50
43
39
35
4380
2190
1460
1095
876
730
626
548
487
438
292
219
175
146
125
110
97
88
73
63
55
49
44
Wire Size (mm2)

Amps
1
2
3
4
5
6
7
8
9
10
15
20
25
30
35
40
45
50
60
70
80
90
100
[distances in feet]
14
2.5
10
16
25
35
50
70
95
120
150
29
15
10
7
6
5
4
4
3
3
2
1
1
1
-
47
24
16
12
9
8
7
6
5
5
3
2
2
2
1
1
-
71
35
24
18
14
12
10
9
8
7
5
4
3
2
2
2
2
1
1
1
1
1
1
126
63
42
31
25
21
18
16
14
13
8
6
5
4
4
3
3
3
2
2
2
1
1
198
99
66
50
40
33
28
25
22
20
13
10
8
7
6
5
4
4
3
3
2
2
2
308
154
103
77
62
51
44
38
34
31
21
15
12
10
9
8
7
6
5
4
4
3
3
433
217
144
108
87
72
62
54
48
43
29
22
17
14
12
11
10
9
7
6
5
5
4
622
311
207
155
124
104
89
78
69
62
41
31
25
21
18
16
14
12
10
9
8
7
6
882
441
294
221
176
147
126
110
98
88
59
44
35
29
25
22
20
18
15
13
11
10
9
1165
583
388
291
233
194
166
146
129
117
78
58
47
39
33
29
26
23
19
17
15
13
12
1491
745
497
373
298
248
213
186
166
149
99
75
60
50
43
37
33
30
25
21
19
17
15
1860
930
620
465
372
310
266
233
207
186
124
93
74
62
53
47
41
37
31
27
23
21
19
-

16-24

Wire Size (AWG)
Amps
1
2
3
4
5
6
7
8
9
10
15
20
25
30
35
40
45
50
60
70
80
90
100
12
10
00
000
0000
128
64
43
32
26
21
18
16
14
13
9
6
5
4
4
3
3
3
2
2
2
1
1
203
101
68
51
41
34
29
25
23
20
14
10
8
7
6
5
5
4
3
3
3
2
2
322
161
107
81
64
54
46
40
36
32
21
16
13
11
9
8
7
6
5
5
4
4
3
513
256
171
128
103
85
73
64
57
51
34
26
21
17
15
13
11
10
9
7
6
6
5
815
408
272
204
163
136
116
102
91
82
54
41
33
27
23
20
18
16
14
12
10
9
8
1296
648
432
324
259
216
185
162
144
130
86
65
52
43
37
32
29
26
22
19
16
14
13
2061
1030
687
515
412
343
294
258
229
206
137
103
82
69
59
52
46
41
34
29
26
23
21
2599
1300
866
650
520
433
371
325
289
260
173
130
104
87
74
65
58
52
43
37
32
29
26
3276
1638
1092
819
655
546
468
409
364
328
218
164
131
109
94
82
73
66
55
47
41
36
33
4132
2066
1377
1033
826
689
590
517
459
413
275
207
165
138
118
103
92
83
69
59
52
46
41
5211
2605
1737
1303
1042
868
744
651
579
521
347
261
208
174
149
130
116
104
87
74
65
58
52
6571
3285
2190
1643
1314
1095
939
821
730
657
438
329
263
219
188
164
146
131
110
94
82
73
66
Wire Size (mm2)

Amps
1
2
3
4
5
6
7
8
9
10
15
20
25
30
35
40
45
50
60
70
80
90
100
[distances in feet]
14
2.5
10
16
25
35
50
70
95
120
150
44
22
15
11
9
7
6
5
5
4
3
2
2
1
1
1
1
1
1
1
1
0
0
71
35
24
18
14
12
10
9
8
7
5
4
3
2
2
2
2
1
1
1
1
1
1
106
53
35
27
21
18
15
13
12
11
7
5
4
4
3
3
2
2
2
2
1
1
1
188
94
63
47
38
31
27
24
21
19
13
9
8
6
5
5
4
4
3
3
2
2
2
298
149
99
74
60
50
43
37
33
30
20
15
12
10
9
7
7
6
5
4
4
3
3
462
231
154
115
92
77
66
58
51
46
31
23
18
15
13
12
10
9
8
7
6
5
5
650
325
217
162
130
108
93
81
72
65
43
32
26
22
19
16
14
13
11
9
8
7
6
933
466
311
233
187
155
133
117
104
93
62
47
37
31
27
23
21
19
16
13
12
10
9
1324
662
441
331
265
221
189
165
147
132
88
66
53
44
38
33
29
26
22
19
17
15
13
1748
874
583
437
350
291
250
218
194
175
117
87
70
58
50
44
39
35
29
25
22
19
17
2236
1118
745
559
447
373
319
280
248
224
149
112
89
75
64
56
50
45
37
32
28
25
22
2791
1395
930
698
558
465
399
349
310
279
186
140
112
93
80
70
62
56
47
40
35
31
28


16-25

Wire Size (AWG)
Amps
1
2
3
4
5
6
7
8
9
10
15
20
25
30
35
40
45
50
60
70
80
90
100
12
10
00
000
0000
170
85
57
43
34
28
24
21
19
17
11
9
7
6
5
4
4
3
3
2
2
2
2
270
135
90
68
54
45
39
34
30
27
18
14
11
9
8
7
6
5
5
4
3
3
3
430
215
143
107
86
72
61
54
48
43
29
21
17
14
12
11
10
9
7
6
5
5
4
683
342
228
171
137
114
98
85
76
68
46
34
27
23
20
17
15
14
11
10
9
8
7
1087
543
362
272
217
181
155
136
121
109
72
54
43
36
31
27
24
22
18
16
14
12
11
1728
864
576
432
346
288
247
216
192
173
115
86
69
58
49
43
38
35
29
25
22
19
17
2748
1374
916
687
550
458
393
343
305
275
183
137
110
92
79
69
61
55
46
39
34
31
27
3466
1733
1155
866
693
578
495
433
385
347
231
173
139
116
99
87
77
69
58
50
43
39
35
4368
2184
1456
1092
874
728
624
546
485
437
291
218
175
146
125
109
97
87
73
62
55
49
44
5510
2755
1837
1377
1102
918
787
689
612
551
367
275
220
184
157
138
122
110
92
79
69
61
55
6947
3474
2316
1737
1389
1158
992
868
772
695
463
347
278
232
198
174
154
139
116
99
87
77
69
8761
4380
2920
2190
1752
1460
1252
1095
973
876
584
438
350
292
250
219
195
175
146
125
110
97
88
Wire Size (mm2)

Amps
1
2
3
4
5
6
7
8
9
10
15
20
25
30
35
40
45
50
60
70
80
90
100
[distances in feet]
14
2.5
10
16
25
35
50
70
95
120
150
58
29
19
15
12
10
8
7
6
6
4
3
2
2
2
1
1
1
1
1
1
1
1
94
47
31
24
19
16
13
12
10
9
6
5
4
3
3
2
2
2
2
1
1
1
1
142
71
47
35
28
24
20
18
16
14
9
7
6
5
4
4
3
3
2
2
2
2
1
251
126
84
63
50
42
36
31
28
25
17
13
10
8
7
6
6
5
4
4
3
3
3
397
198
132
99
79
66
57
50
44
40
26
20
16
13
11
10
9
8
7
6
5
4
4
615
308
205
154
123
103
88
77
68
62
41
31
25
21
18
15
14
12
10
9
8
7
6
866
433
289
217
173
144
124
108
96
87
58
43
35
29
25
22
19
17
14
12
11
10
9
1244
622
415
311
249
207
178
155
138
124
83
62
50
41
36
31
28
25
21
18
16
14
12
1765
882
588
441
353
294
252
221
196
176
118
88
71
59
50
44
39
35
29
25
22
20
18
2330
1165
777
583
466
388
333
291
259
233
155
117
93
78
67
58
52
47
39
33
29
26
23
2981
1491
994
745
596
497
426
373
331
298
199
149
119
99
85
75
66
60
50
43
37
33
30
3721
1860
1240
930
744
620
532
465
413
372
248
186
149
124
106
93
83
74
62
53
47
41
37


16-26

Wire Size (AWG)
Amps
1
2
3
4
5
6
7
8
9
10
15
20
25
30
35
40
45
50
60
70
80
90
100
12
10
00
000
0000
425
213
142
106
85
71
61
53
47
43
28
21
17
14
12
11
9
9
7
6
5
5
4
676
338
225
169
135
113
97
85
75
68
45
34
27
23
19
17
15
14
11
10
8
8
7
1074
537
358
269
215
179
153
134
119
107
72
54
43
36
31
27
24
21
18
15
13
12
11
1709
854
570
427
342
285
244
214
190
171
114
85
68
57
49
43
38
34
28
24
21
19
17
2717
1359
906
679
543
453
388
340
302
272
181
136
109
91
78
68
60
54
45
39
34
30
27
4320
2160
1440
1080
864
720
617
540
480
432
288
216
173
144
123
108
96
86
72
62
54
48
43
6869
3434
2290
1717
1374
1145
981
859
763
687
458
343
275
229
196
172
153
137
114
98
86
76
69
8664
4332
2888
2166
1733
1444
1238
1083
963
866
578
433
347
289
248
217
193
173
144
124
108
96
87
10919
5460
3640
2730
2184
1820
1560
1365
1213
1092
728
546
437
364
312
273
243
218
182
156
136
121
109
13774
6887
4591
3444
2755
2296
1968
1722
1530
1377
918
689
551
459
394
344
306
275
230
197
172
153
138
17369
8684
5790
4342
3474
2895
2481
2171
1930
1737
1158
868
695
579
496
434
386
347
289
248
217
193
174
21902
10951
7301
5475
4380
3650
3129
2738
2434
2190
1460
1095
876
730
626
548
487
438
365
313
274
243
219
Wire Size (mm2)

Amps
1
2
3
4
5
6
7
8
9
10
15
20
25
30
35
40
45
50
60
70
80
90
100
[distances in feet]
14
2.5
10
16
25
35
50
70
95
120
150
146
73
49
37
29
24
21
18
16
15
10
7
6
5
4
4
3
3
2
2
2
2
1
236
118
79
59
47
39
34
29
26
24
16
12
9
8
7
6
5
5
4
3
3
3
2
354
177
118
88
71
59
51
44
39
35
24
18
14
12
10
9
8
7
6
5
4
4
4
628
314
209
157
126
105
90
79
70
63
42
31
25
21
18
16
14
13
10
9
8
7
6
992
496
331
248
198
165
142
124
110
99
66
50
40
33
28
25
22
20
17
14
12
11
10
1538
769
513
385
308
256
220
192
171
154
103
77
62
51
44
38
34
31
26
22
19
17
15
2166
1083
722
542
433
361
309
271
241
217
144
108
87
72
62
54
48
43
36
31
27
24
22
3109
1554
1036
777
622
518
444
389
345
311
207
155
124
104
89
78
69
62
52
44
39
35
31
4412
2206
1471
1103
882
735
630
551
490
441
294
221
176
147
126
110
98
88
74
63
55
49
44
5825
2913
1942
1456
1165
971
832
728
647
583
388
291
233
194
166
146
129
117
97
83
73
65
58
7453
3727
2484
1863
1491
1242
1065
932
828
745
497
373
298
248
213
186
166
149
124
106
93
83
75
9302
4651
3101
2326
1860
1550
1329
1163
1034
930
620
465
372
310
266
233
207
186
155
133
116
103
93
"

16-27
Exercise

" %

$!3
.%
!+ 0
,
//////////"+1

+ 0

,
>
!"3445!7
!
//////////"+1

"
&'
%&'
A) %&' $!+
0

,
//////////"+1

+ 0

,
>
!"3445!7
!
//////////"+1

" B))
%&' !3

)
C
!+ 0

,
//////////"+1

+ 0

,
>
!"3445!7
!D -E !
//////////"+1


16-28
Series/Parallel Array Wiring

We now look at how to arrange modules into an array. Modules can be connected in
parallel, plus-to-plus, to add current. Or they can be connected in series, plus-to-minus,
to add voltage. And arrays can be built using both wiring methods.
We recommend drawing array circuits in two steps, first for clarity as a theoretical
schematic, and then for actual details as they would be installed.
Schematic Drawing
We first draw the arrangement of modules in a schematic form with boxes representing
modules placed above one another for series connections, and placed side by side for
parallel connections. This is very clear visually. We can think of higher voltage as
higher stacks of series connected modules and increased current coming from
modules pouring or flowing together side-by-side in parallel.
Installed Drawing
Then we show the modules as they would be installed in actual field arrangement.
There are generally two junction box arrangements used in solar modules: separate
positive and negative boxes on opposite ends of a module; or single junction boxes
containing both positive and negative terminals.
For the single junction box arrangement all the junction boxes are lined up on the same
side, so all installed drawings look similar except for the details of where the wires
connect.
For the separate J-box arrangement installed 12-volt systems look similar to their
schematic drawing, but series wired systems of higher voltage (typically 24 and 48
volts) involve flip-flopping the modules so that the plus and minus connections are
close together.
We demonstrate recommended wiring arrangements for both single junction box and
dual J-box arrangements for 12, 24 and 48-volt systems next. For example, the SM55
has separate positive and negative terminal junction boxes, while the SP75 module has
a single J-box.

16-29
12-Volt Array Wiring

The schematic arrangement of a 12-volt array involves placing the modules side by
side, with all the positive terminals connected in parallel, and all the negative terminals
also connected in parallel. The current from each module adds to give the total array
current while the voltage of the entire array is the same as the voltage of one module.
12 Volt A rray
S chem atic
)

16-30
The actual installed wiring of a 12-volt array using dual J-box modules would also
have all the positive connections along one side and all the negative ones along the
other.
12 V o lt A rra y
A s In stalle d
D u al J-b o x
) "
The installed wiring of a 12-volt array for single J-box modules would also have all the
positives connected together in parallel, and all the negatives connected together.
12 Volt Array
As Installed
Single J-box
) '
16-31
Higher Voltage Array

When we move on to systems with voltages greater than 12 volts there are two
approaches that can be taken. We can connect the modules in parallel groups first, to
produce the current that is needed, and then connect these groups in series to give the
required voltage. This is shown on the left in the figure below. In actual field conditions
however, it may be necessary to remove modules from an operating array for repair or
other maintenance work. If the array is wired as shown on the left, in parallel groups,
then removing a group leaves the rest of the array without enough voltage to properly
operate. The whole array must be shut down.
The second method involves connecting modules in series groups first to produce the
required voltage, and then connecting similar series strings in parallel to produce the
required current. This is the recommended approach. If a series group is removed
from the circuit, the rest of the strings still operate at the nominal system voltage, and
the array continues to charge the batteries or operate the load.
Parallel Groups or
Series Groups
Buss bar
Buss bar
) (
16-32

For a 24-volt array we need strings of two modules wired in series. Any number of such
series strings can then be connected in parallel to give the required current.
A schematic view of a 24-volt array is shown below. Strings of two modules are
stacked vertically, indicating higher voltage. Current flows up through the lower
module, gaining nominal 12 volts. The current from the lower module in each string
then flows on up through the second module in the string where it gains another 12
volts. The current out of each top module combines to flow on to the battery or load.
The total array current returns from the battery or load to the bottom of the array.
There the current splits into equal parts and each string draws one-modules-worth of
current.
2 4 V o lt A r r a y W i r i n g
S c h e m a t ic
)

16-33
The installed wiring for 24 volts using modules with separate J-boxes is shown below.
The two-module strings would be flip-flopped alternating positive and negative Jboxes. The intermediate wires for each series string, connecting the positive of the
bottom module to the negative of the top module, are shown to the right of the array.
The positive outputs from the two-module strings would all combine, and the negative
inputs would also combine. Typical ground mount structures hold up to eight modules,
so four 24-volt strings could be mounted on one structure.
2 4 V o lt A r r a y
A s In s t a lle d
D u a l J -b o x
) *
The installed wiring for single J-box modules is shown again with all the J-boxes aligned
along one side. The intermediate connection within each string, from the negative of
one module to the positive of the second module, is shown.
2 4 V o lt A r r a y
A s In s ta lle d
S in g le J -b o x
) +
16-34

In the case of 48-volt systems, we need four modules connected in series. Strings of
four modules can be combined in parallel to give the final array current.
48 V o lt A rra y
S c h e m a tic
+
) ,

16-35
Installed field wiring for a 48-volt system with dual J-box modules would again have the
modules flipped. Typical ground mount structures hold up to eight modules, so two
48-volt strings could fit on one structure.
48 Volt Array
As Installed
Dual J-box
) Installed wiring for single J-box modules would also have groups of four modules
connected in series.
48 Volt Array
As Installed
Single J-box
)
16-36
Exercise

3

!
B))
A

.% !"

!3 F0 G
!
3
!>
!
#
H
I
//////////

//////////
//////////

//////////
//////////

//////////
&
//////////

//////////
//////////

//////////
//////////

//////////
//////////

//////////
//////////

//////////
//////////

//////////
.
//////////

//////////
..
//////////

//////////

16-37
10
11
Load
6
zero volts
3
) ./...0.

16-38
Wiring of Safety Equipment

Battery Bank Wiring
Limit the Number of Parallel Batteries
When connecting batteries we can increase the total capacity by connecting batteries in
parallel. This connects plates of the same type of material (lead or lead-dioxide) and
effectively makes a bigger battery at the same voltage. It is recommended that not
more than four parallel connections be made in a bank. Parallel strings of batteries are
more susceptible to shorter battery life due to uneven charging of different strings. The
greater the number of parallel connections the greater the chance for advanced aging
of the battery. If more than four parallel strings are needed this would indicate that
perhaps a battery cell with too small of capacity is being used. Choose a larger
capacity cell so that fewer parallel strings are needed.
Battery Bank Wiring

A
(+)
+
(+)
(+)
(+)
+
+
+
-
(-)
-
Max 6-8 in parallel
(-)
(-)
Connect at opposite corners
(-)
) "
16-39
Series and Parallel Connections

To operate at a higher voltage connect batteries in series. For example, the second
battery bank shown (B) could be four 12-volt batteries connected in series to give a final
voltage of 48 volts nominal.
If both large capacity and high voltage are needed then connect in series and parallel.
For example, the battery bank on the far right (D) has four batteries connected in series
for 48 volts and then three strings connected in parallel to give more capacity at 48
volts.
Tap Opposite Corners

If series and parallel connections are made it is good practice to tap off the battery at
opposite corners, as shown. This helps to equalize the path of the current through the
batteries and helps to insure equal charging and discharging.
Use Solid Buss Bars

Another good practice is to avoid individual wires connecting parallel strings. Instead
use a large copper bar drilled with holes for the parallel connected terminals. The solid
copper bar gives excellent conductivity and reduces the possibility of corrosion or loose
connections at the terminals.

16-40
Overcurrent Protection Devices

Whenever there are voltage sources such as batteries, photovoltaic systems, or
generators, there is the potential for dangerously large currents. A battery bank in a
short circuit condition can discharge hundreds or thousands of amps causing damage
to equipment and even physical harm or death. Care must be taken to protect
equipment from such high current conditions. As a result NEC requires that every
ungrounded conductor must be protected by an overcurrent device, which breaks the
current if too much current flows. This is done by using fuses or circuit breakers in the
circuits. A circuit breaker is a switch that opens a circuit if excessive current flows. A
fuse consists of a wire or metal strap that will melt (open) if excessive current flows.
Most fuses and circuit breakers have been developed for AC use and as a result are
unrated for DC use. Those that are rated for DC current are in many cases undersized
for the necessary currents when short circuits occur. As a result it is very important to
use fuses and circuit breakers rated for their intended operation when used in DC
circuits. Since most PV systems may have transients surges (lightning and motor
starting), inverse-time circuit breakers (the standard type) or time-delay fuses should be
used in most cases. Presented below are some important considerations for
overcurrent protection, as specified by the US National Electric Code.
Overcurrent Protection
Each source circuit must be protected by an overcurrent device on the ungrounded

conductor.
Use DC rated circuit breakers and fuses in order to meet NEC Code.
Batteries must have current limiting fuses for overcurrent protection.
All PV equipment (e.g. controllers, inverters, etc.) must have safety disconnects to
disconnect it from all sources of power.
Specifying Overcurrent Protection
Overcurrent devices must be able to handle the rated current that will open circuit.
All fuses and circuit breakers should be in enclosures.
Circuit breakers or fuses should not be paralleled in the same circuit.
Protection ratings of the overcurrent protection device must never exceed the
current ratings of the wire.
Grounded conductors should never be fused or switched.

16-41
Types of Fuses and Circuit Breakers

There are a number of fuses and circuit breakers that have been UL listed for DC
operation that are commonly used in PV systems.
Manufacturer
Type
Voltage
(volts)
AIR
(Amps)
Comments
5,000-25,000
These are not current

limiting and hence can not
be used to protect other
fuses and circuit breakers.
Can be used between the
battery and inverter.
Heinemann
Circuit Breakers
65-250
Continuous
Rating
(Amps)
15-700
Square D
Circuit Breakers
48
15 - 70
5,000
These are not current

limiting.
Bussmann
Gould
Littlefuse
Class T Fuse Fast Acting

Current Limiting
125 - 300
1 - 1200
20,000
It is a fast acting current

limiting fuse but seems to
produce good results
where time delay fuses are
specified.
Bussmann
Gould
Littlefuse
RK-5 Time Delay

Current Limiting
125-600
0.1 -600
20,000
Most common fuse used

for protecting from battery
short circuit. Has time
delay and current limiting
feature with high AIR.
'
1)
..
Continuous Ratings Of Overcurrent Devices

Fuses and circuit breakers have two current ratings:
Continuous
Maximum interrupt level or AIR (Amperes Interrupting Rating).
The continuous rating is the current that the device will allow to pass continuously. For
example, a 100-amp fuse will allow 100 amps to pass continuously. The fuse will blow
when the current exceeds this continuous level. How long the device lets the current
pass before it breaks the circuit depends on the exact design of the device and its time
delay. Many devices will allow 125% of the continuous rating pass without breaking.
Currents exceeding 200% may blow the device in a few minutes. Levels exceeding
500% may blow the device in a few seconds and so on. The exact performance of a
fuse or circuit breaker is shown in a graph of current vs. time. An example of a time
graph for a circuit breaker and a fuses is shown next.

16-42
Fuse Example
Presented below is the current vs. time characteristics of FLNR series time-delay
current-limiting fuses, DC rated, Class RK5. The time-delay models are really
designed to allow surge currents to flow through to loads such as motors and
compressors. They will have a slower response time curve than fast acting fuses
designed for sensitive electronic equipment. The continuous current range is shown
from 15 amps to 600 amps. For each value of fuse the amount of current that it will
pass is shown compared to the time before the fuse will interrupt the current.
Example:
The 200-amp rated fuse curve is shown (as the dark line). The
curve indicates that the fuse will pass 600 amps for about 60
seconds, and 1000 amps for about 10 seconds and so on. It will
allow 2000 amps for only about 0.2 seconds, and 4000 amps for
only about 0.02 seconds. So at higher currents, the time to
interrupt is quite short.
) '
16-43
Circuit Breaker Example

The current vs. time curves presented below are for Heinemann Series GJ circuit
breakers. Three ranges are shown with the top range indicating standard time delay,
the middle range representing medium time delay, and the lower range illustrating short
time delay. The current values are indicated as percent of model continuous rating.
The current range for GJ type breakers is from 100 amps to 250 amps. DC interrupt
ratings range from 10,000 amps DC for the 125 VDC version to as high as 25,000
amps for the 65 VDC designation.
Example:
Examine the top standard time delay range. If we were to specify

a 200 amp rated breaker, then 500% corresponds to 1000 amps.
The curve indicates that the breaker will allow 500% (about 1000
amps for a 200 amp rated breaker) to flow for between 2-10
seconds. 2000 amps would correspond to 1000% for our 200 amp
breaker, and the current allowed at 1000% is between 0.01-0.4
seconds.
) (

16-44
Interrupt Rating of Overcurrent Devices

The second current rating is the interrupt rating known as AIR, Amperes Interrupting
Rating. This is a measure of how high above the continuous rating the device can
safely break current flow without being damaged by short circuit conditions. For
example, a common AC breaker for homes might have a continuous rating of 60 amps
and an interrupt rating of 10,000 amps of AC current.
Batteries Are The Major Concern

Because photovoltaic modules are inherently current limited high short-circuit currents
from the solar array are not a problem. However, a single 220 amp-hour 6-volt deep
discharge battery may produce short-circuit currents as high as 8000 amps for a
fraction of a second. Such high currents can generate excessive thermal and magnetic
force that can cause an underrated device to burn apart. As a result the interrupt rating
is very important when designing protection for battery circuits.
Derate AC Ratings By 90%

Overcurrent protection devices such as fuses and circuit breakers, that are designed to
interrupt currents in AC systems expect the current to fall to zero every half-cycle, or
about every 8 milliseconds. The arc that is created by large currents can be easily
extinguished at that point. But with DC circuits the current does not go to zero.
Overcurrent devices with interrupt ratings for AC operation must be derated
usually by at least 90% for DC operation! For example, this means that a fuse or
circuit breaker rated to interrupt 10,000 amps AC should be expected to safely interrupt
only 1000 amps of DC current.
Current-Limiting Devices
Batteries can produce thousands of DC amps under short circuit conditions. A 100-Ah
battery could release up to 2000 amps of short circuit current, while a 1000-Ah cell
might be able to produce up to 20,000 amps! This level of current can overwhelm and
damage downstream overcurrent protection. This downstream overcurrent equipment
needs protection too! Specially designed fuses, called current-limiting fuses, can
interrupt large currents. These are generally classified as RK5 or RK1 (DC rated Class
J and T fuses can be used as well).

16-45
A typical high interrupt rating for a current limiting fuse is 200,000 amps of AC current.
Under DC conditions this is derated to 20,000 amps. So a current limiting fuse should
be connected to each battery string up to about 600-1000 Ah capacity. In this way
downstream overcurrent equipment is protected from excessively high currents. This
means that designers should plan to break very large battery banks into units of
maximum 600-1000 Ah. Single banks of greater than 1000 Ah will need specially
designed overcurrent protection systems.
Current Limiting Fuses

(+)
(-)
)

16-46
Exercise

D '

HED

!.
,
//////////

D 41J$
.
$ =
&

,
///// /////

> ,
///// /////

"%'
$

!+

A* ) ,5 7!
<B<

16-47
Safety Disconnects
Safety disconnects or switches are placed into power systems to allow equipment to be
safely installed and maintained. The NEC requires that all source circuits or voltage
sources in photovoltaic systems such as photovoltaic arrays, battery banks, and
generators, must be able to be isolated in case there is a problem. Circuit breakers and
fused disconnect switches provide a means to disconnect safely, as well as providing
for overcurrent protection. The code limits the total number of switches or circuit
breakers in a PV system to six. They are further required to be grouped and marked
and readily accessible to be pulled.
All safety disconnects in photovoltaic systems must be DC rated for voltage as well as
current. This may involve specifying a more expensive or heavy duty version of an AC
rated model. DC rated switches are marked to indicate added protection against
arcing.
Typically there are four locations where disconnect devices are needed in photovoltaic
systems. They are: (1) between the array and the charge regulator; (2) between the
regulator and the battery; (3) between the battery and any DC loads or load center; and
(4) between the battery and the inverter. Disconnect means may be built into small
charge regulators or system controllers. For example, the load terminals on charge
regulators may have circuit breaker protection built -in.
Safety Disconnects
Modules
Regulator
Battery
DC Loads
Inverter & AC Loads
) *

16-48
Safety Disconnect Design Using Circuit Breakers

There are many ways to design safety disconnect protection into a photovoltaic power
system. One approach that is quite flexible and compact is to use DC rated circuit
breakers and pull-fuses. These components can be sized for different currents and yet
housed in the same box for convenience of service and ease of installation. All
components must be DC rated to be able to break the arcs associated with DC
currents. A general schematic of a possible approach is shown below.
Circuit Breaker Method

Of Safety Disconnect Design
array ground
system ground
RegArrayBattery-
Inv-
DC LoadA
Array+
Battery+
Reg Array+
Reg Battery+
Inv+
DC Load +
) +
The charging circuit passes current from the Array + through circuit breaker A and on to
the charge regulator (terminal designated as Reg Array +). Current returns back from
the regulator (from terminal designated as Reg Battery +) and through circuit breaker
B, and then on to the Battery +. The negative current from the battery enters from the
Battery -, and passes through the central negative wiring block E, where it continues on
to the Array -. This completes the charging circuit.

16-49
The load circuit is adjacent and uses common components. The current begins at the
Battery +, flows through a jumper to circuit breaker C and then on to the Inverter +
terminal. The current can also flow to one of possibly many DC load breakers D where
it then flows on to the DC Load + terminal of various DC loads. Current would return
from the Inverter -, or from the DC Load -, and connect at the central negative wiring
block E, where it would then flow back to the Battery -. This would complete the load
circuit.
The flexibility of this method is that the charging circuit breakers and the load breakers
do not have to be of the same size. For example, the array might need only a 30-amp
breaker while the inverter might need a 200-amp breaker and the DC loads might use a
100-amp breaker.
The dashed lines represent the equipment grounding wires. All equipment grounds are
collected at one central point, the central negative wiring block E. This includes the
grounding wire from the array ground. This grounding wire must come all the way in
from the metal mounting structure and the module frames in the field. All negative
electrical conductors also connect at E. Thus this point becomes the one central point
in the circuit where electrical and equipment grounds are bonded. A main system
ground wire then connects from this central connection point to a ground rod that
makes the firm ground to earth.

16-50
Example of Circuit Breaker Method:

Power centers are now commercially available that have all necessary means for
array, battery, and load circuit protection built-in. An example of the schematic wiring of
a Powercenter from Ananda Power Technologies Inc. is shown. The system uses large
current limiting fuses in a dual pull-out design as the disconnecting means between the
array and battery (top fuse in Pull Out), and between the battery and inverter and DC
load breakers (bottom fuse in Pull Out). This is a compact and low cost method for
including both overcurrent protection (with current limiting strength) and disconnecting
means in one space-saving unit.
The Array Mercury Contactors are controlled by charge regulator circuit boards. The
relays open the array charging circuit to prevent battery overcharge, and stay open at
night to prevent leakage from the battery through the array.
) ,

16-51
Safety Disconnect Design Using Fused Disconnect Switches

Another approach to safety disconnect design is to use fused switches. These are
large industrial grade knife switches with integrated fuse holders. The fuse acts as the
overcurrent protection and the knife switch acts as the disconnect means. Again these
must be DC rated heavy-duty models to handle the array, battery and load DC currents
of photovoltaic systems.
The difficulty with this approach is that switches of different current levels may not be
able to fit in the same box. Separate disconnect boxes may be needed for the array
circuit and the load circuit, as shown above.
The charging circuit passes current from the Array + through a fused switch and on to
the charge regulator (terminal designated as Reg Array +). Current returns back from
the regulator (from terminal designated as Reg Battery +) and through another fused
switch, and then on to the Battery +. The negative current from the battery enters from
the Battery -, and connects to a negative wiring block A, where it continues on to the
Array -. This completes the charging circuit.
A separate box services the load circuit. The current begins at the Battery +, flows into
a fused switch and then on to the Inverter + terminal. The current can also flow
through a short jumper wire to an adjacent fused switch where it then flows on to the
DC Load + terminal of the DC load center. Current would return from the Inverter -, or
from the DC Load -, and connect to another negative wiring block B, where it would
then flow back to the Battery -. This would complete the load circuit.
All the ground wires from the equipment connect in to one of the negative wiring blocks,
shown at block A in this drawing. A final system ground wire connects from the block
A to a ground rod driven into the earth.

16-52
Fused Switch Method

array ground
ArrayBattery Array+
Battery+
inverter
ground
A
Reg Array+
system
ground
regulator
ground
DC
load
ground
Inv+
DC Load+
Reg Battery+
Reg-
Inv-
DC Load-
) -

16-53
Example of Fused Switch Method:

An example of a 24-volt AC and DC photovoltaic system using the fused disconnect
switch approach is shown on the next page. The charging currents from parallel 24-volt
strings flow into a field junction box at the base of each sub-array. Each string flows
through a blocking diode and combines into a larger wire. From the field junction boxes
the current from the sub-arrays travels the long distance to the central equipment of the
system.
The currents are wired into separate fused switches and then on to separate charge
regulators. The currents from the regulators combine at positive block A and negative
block B. The positive current flows on to a fused switch and then on to the positive
terminals of the battery bank. Thus the charge regulators are protected from both the
array and the battery.
The battery is shown with three parallel strings each with a current limiting fuse to
protect the downline fuses and disconnects from possible battery short circuit currents.
The negative wire runs from the opposite corner of the battery bank and connects to the
negative block B. Current flows back through the charge controllers and back to the
sub-arrays.
There are two load circuits. One circuit has current flowing from the battery positive
terminal through a fused switch and then on to the DC distribution box. The other load
circuit has current flowing through a larger fused switch and then on to the positive
terminal of the inverter. The inverter AC output is connected to the AC distribution box.
Negative conductors from the two load types returns to the negative block B and then
on to the negative of the battery bank.

16-54
24-Volt AC/DC system diagram illustrating fused switch approach
(+)
Controller #1
Controller #2
(-)
DC
load
center
AC
load
center
Inverter
(-)
(+)
)
16-55
Exercise

H # !

H.'.(!
H "K$
5 7 B 5
7
>
BH5 7
$!
H E
BH5 7 8>

8E K$ 8
8 !
3 "
"

!3
"

$
!3
1 D # !

H

H.'%.!3 !

!
= 8 "

!

16-56
System And Equipment Grounding For

Photovoltaic Systems
One of the most important safety features of electrical wiring is the ground. A ground is
the electrical connection made between one side of an electrical circuit and the ground
or earth. This provides a low resistance path for current flow that in many cases will
prevent shock for certain types of accidents and equipment failures. There are two
types of grounding in any circuit: (1) the system ground; and (2) the equipment ground.
The system ground is where one leg of a circuit is grounded. In the case of an AC
circuit this is the neutral leg. In the case of a PV DC circuit it is usually the negative leg.
By grounding the overall system current surges from lightning and distribution accidents
are shunted to the earth, thus protecting people and equipment from electric shock and
damage.
The equipment ground, on the other hand, connects all exposed, non-current-carrying
metal surfaces to ground. A short circuit, also known as a fault condition, occurs when
an electrical conductor makes contact to any metal structural component (wiring box,
appliance, conduit, etc.), or to another conductor. Current can flow through this new
path, which might include a person's heart if they are touching the faulty component.
Dangerous currents can be made to avoid traveling through a person touching faulty
equipment by having an alternative highly conductive path connecting the equipment to
the earth, in the form of a grounding wire. A large current can flow through the wire,
large enough to cause the overcurrent devices (fuses or circuit breakers) to open the
circuit and stop the flow. The equipment ground wire is not a part of the power circuit.
It is a normally a non-current carrying wire and is on "stand-by" in case the normal
circuit becomes unsafe.
If any lightning protection system is included in the system a very good connection to
ground should be made. A lightning rod or other conductive target that might be hit by
lightning should have a direct, very low resistance path to a ground rod, because if
struck by lightning it might have to conduct tens of thousands of amps of current for a
few moments. Any resistance or bends in the wire will cause the enormous currents to
jump to a nearby object, and control of the current is lost.
All exposed metal surfaces, including the module frames, array mounting structure, any
metal housings of controllers or load centers, and all loads should be grounded. All
grounds, including equipment, electrical and lightning protection should come together
and be connected to one ground rod. This is often called the final "system ground".

16-57
Grounding Details
System and equipment grounding conductors share the same electrode.

There should be only one single point where the system ground and the equipment
ground are bonded together.
DC and AC subsystem should share the same electrode.
Equipment grounding wire should have the same ampacity as the overcurrent
device protecting the particular circuit.
For lightning protection at the array use system of grounding electrodes bonded
together.
Grounding Electrode (Rod)
Locate as close to system grounded conductor as possible.

Connect the grounding conductor to the system on the battery side of the charge
controller at a point to the system.
Grounding electrode conductor needs to be as large as the largest conductor in the
system is and be at least No. 8 AWG.
NEC requires that the grounding electrode be 5/8 diameter with at least 8 feet
driven into permanently moist soil.
If the solar array will be a great distance away from the central ground rod for the
system a separate ground rod for the array structure can be driven close to the array.
However there should be a bonding wire that connects this remote ground rod to the
main system ground rod. This bonding wire keeps both ground rods at the same
voltage potential. If the two rods were not connected a nearby lightning strike would
result in a momentary surge of current into the earth that would cause large differences
in ground potential due to the resistance of the earth. The two ground rods would not
both be a "zero" potential and large induced currents will be created in the ground
system. By having the wire connecting the two rods the resistance of the earth is
overcome, and the two rods stay at the same potential and no induced current result.

16-58
System and Equipment

Grounding
Electrical equipment
System ground
Array disconnect
Bonding wire
Equipment
ground
System
grounding
electrode
Array frame
grounding
electrode
) 2$

16-59
Disconnect Arrangement for

Ungrounded and Grounded Systems
The U.S. NEC code stipulates that for systems above 50 volts (Voc for photovoltaic
type systems), one DC conductor shall be grounded. This means that system
designers can choose an ungrounded configuration for 12 and 24-volt nominal systems.
For ungrounded systems disconnects are required on both the positive and negative
conductors leading to and from the battery. For grounded systems disconnect and
overcurrent protection are only required on the ungrounded conductor. The figure
below shows the overcurrent and disconnect configurations for grounded and
ungrounded PV systems.
Battery Overcurrrent and Disconnect Requirements

from charging
circuits
to load
circuits
from charging
circuits
to load
circuits
current limiting
fuses
disconnect
switches
Battery
Battery
Grounded Systems
Ungrounded Systems
) "

16-60
Lightning and Surge Protection

Solar modules mounted on the tops of buildings or high can attract lightning strikes.
Besides grounding equipment such as the array frames, system designers must provide
appropriate means to deal with lightning induced surges coming into the system
conductors. Metal oxide varistors (MOV) and silicon oxide surge arrestors (SOV) are
used as surge suppression devices on PV systems. Both have a breakdown voltage at
which they dissipate energy to ground. SOVs are preferred due to the fact that they do
not draw current when they are off and if they fail, they fail in the open circuit condition.
They are rated for surge currents of up to 100,000 amps at 300 volts or higher. SOVs
are commonly installed either at the array in the array field junction box or as part of the
system controller or power center.
A simple schematic of the concept and an actual drawing of a field junction box with
SOV surge protection are shown. In the schematic notice that both the positive and
negative conductors from the array are connected through SOVs to ground. Normally
the SOV would be an open circuit and no connection would be made to ground. But in
case of a large voltage surge the SOV would short circuit and the dangerous voltage
spike would be shunted to ground at the array, thereby protecting the downline
equipment.
Simple Surge Protection

Diagram
To controls
and battery
From array
SOV's
Bonding wire
to system electrode
From array frame
Array grounding electrode
) '$./. $.

16-61
An example of a field junction box with surge protection is presented below. Allowance
for up to eight separate array positive wires is made. Each positive connection is made
to a fuse and blocking diode and then combined into one positive buss bar. Current
flow through a field disconnect circuit breaker and then on to the positive output
terminal block. The array negative(s) would connect to the negative terminal block.
The SOV surge arrestor has connections to both the positive buss bar and to the
negative terminal block. It is also connected to the ground block. This is where the
high lightning induced surge current would flow.
The array grounding electrode would be connected to the ground block as would the
bonding wire to the system grounding electrode.
) ( 0$3.04/ $.

16-62
Fully Worked Example

To help illustrate the application of the rules and concepts presented in this chapter a
complete wiring exercise is presented below.
Example: Remote Cabin, DC-Only System
Array Size:
Batteries:
Load:
Six 12 volt 75-watt SP75 Modules

700 amp hours at 12 volts (4 Trojan L-16)
75-Watt Peak DC
Description:
The modules are mounted on a rack on a hill behind the house, 50

feet from the batteries and control center. The modules are wired
in parallel. Non-metallic conduit is used to run the cables from the
module rack to the control panel. A disconnect and control panel
is mounted on the back porch. The batteries are in an insulated
box under the porch within a few feet of the control panel. All the
loads are DC with a peak combined power of 75 watts at 12 volts,
primarily due to a pressure pump on the gravity-fed water supply.
The battery bank consists of four Trojan L-16 (350-amp-hour,
6-volt) deep-cycle batteries wired in series and parallel.
Module Interconnects:
Determine Short Circuit Current

The first step in sizing the wiring is to determine the short circuit
current of the array and then adjust for the NEC and UL safety
factors and temperature.
The array short circuit current can be determined by multiplying the
number of SP75 modules (six modules) by their rated short circuit
current (Isc = 4.8 amps):
Array Isc

=
=
=
number of parallel modules X rated Isc

6 X 4.8 amps
28.8 amps
16-63
Module Interconnects (continued):
Adjust for UL / NEC and Temperature

To be conservative, use the highest temperature rated insulation,
90 C. Assume that the back of the modules will get up to 65 C.
The temperature derating for 90 C conductors at 65 C is 0.58.
Applying the NEC oversizing factor of 125% and the UL oversizing
factor of 125% and the temperature derating of 0.58 to determine
the required ampacity rating at standard conditions:
Ampacity
Required
= 28.8 amps 0.58 X 1.25 NEC X 1.25 UL

= 77.6 amps
Type USE-2 (rated for 90C) will be used to interconnect the

modules. The standard ampacity of USE -2 # 8 AWG in free air is
80 amps. This is sufficient to meet the above requirement.
Wire From Field Junction Box to Control Panel:

Because wire runs are typically a relatively long distance between
the module field junction box and control panel, voltage drop
becomes the main concern. However, it is good design to first
look at ampacity before calculating voltage drop.
Check Ampacity for the Chosen Wire

Cable in conduit exposed to 40 C temperatures is chosen for the
module to control panel runs. If RHW-2 cable with 90 C rating is
selected, then the derating factor is 0.91.
Ampacity
Required
28.8 amps 0.91 X 1.25 NEC X 1.25 UL
49.5 amps
This is safely covered by the chosen wire size of #8 AWG, which

has a rated ampacity in conduit of 55 amps at 30 C.

16-64
Wire From Field Junction Box to Control Panel (continued):
Size Wire for Voltage Drop

Using the 2% maximum voltage drop rule, we can refer to
Table 16-2 (Voltage Drop Table--12 Volt Nominal) for the 50 feet
run to determine wire size. The voltage drop should be sized for
the array current at peak power, not the short circuit condition.
The array current at peak power can be determined by multiplying
the number of modules (six modules) by their Imp (Imp = 4.4
amps):
Array Imp
number of panels x rated Imp
6 X 4.4 amps
26.4 amps
Using the Voltage Drop Table we see that for 25 Amps, 1/0 AWG
is required for a 2% drop with a maximum one-way distance of 55
feet. So we must use 1/0 size wire for this part of the circuit.
Connecting 1/0 AWG to the #8 AWG from the modules will require
a power distribution block in the field junction box.
Wiring in the Control Center and to the Batteries:

These wires are short (typically a few feet) and hence voltage drop
will not be a concern. However, they should be sufficient size to
handle the ampacity requirements. Cable type THHN size # 6
AWG is used in the control center. We already have determined
(see above) that #8 AWG THHN has sufficient ampacity. The 1/0
cable from the negative DC input is used to the point where the
grounding electrode conductor is attached in order to meet code.
The same #6 AWG is used to connect from the control center to
the batteries. If this distance is greater than a few feet, one should
consider fusing right at the battery bank and increasing the size of
the conductors from the battery to the control center.

16-65
Grounding Conductors:
Equipment grounding conductor sizing requires a number of
calculations and tables. (See Section 250 in NEC). #6 AWG is a
sufficient size to handle the grounding for this system. The
grounding electrode conductor should be # 1/0 AWG, as it is
required to be as large as the largest conductor in the system is.
Overcurrent Protection and Disconnect:
Each source circuit must be protected by overcurrent devices. A
current-limiting fuse in a fused disconnect switch is required by
code to protect the load and the array circuit from the batteries.
Array Circuit
The design current of the array circuit fuse must be greater than
the array short circuit current:
28.8 amps X 1.25 NEC X 1.25 UL = 44 amps
A 50 amp RK-5 type fuse is chosen to meet this requirement.
Each component must be able to be disconnected from all sources
of power. The charge control is between the array and the battery.
Hence a 50-amp single-pole circuit breaker (cheaper than a fused
disconnect switch) is specified to isolate the controller from the
array.
Load Circuit
To determine the protection on the load circuit, the maximum load
power of 75 watts at 12 volts is converted to
75 12 = 6.25 amps. The protection rating in this circuit must only
meet the standard NEC overrating factor:
6.25 amps X 1.25 NEC = 7.8 amps.
A 30-amp fuse will meet this requirement easily.
Both fuses should have an amperes interrupting rating (AIR) of
around 20,000 amps sufficient to withstand the short circuit
currents from the battery under fault conditions.

16-66
Voltage Ratings:
Disconnects, fuses, and circuit breakers must have a voltage rating
of at least 1.25 times the system open circuit voltage. Standard
module Voc is 22 volts.
22 volts X 1.25 NEC = 27.5 volts.
) 5./0$.

16-67
Wire Sizing Worksheet

An example of a worksheet for collecting information about the wiring design for a
photovoltaic system is presented on the next page.
Charge controller capacity can be calculated, including the environmental safety factor
of 1.3 for array current enhancement from reflections.
Wiring between equipment can be listed and sized based on voltage drop or ampacity
considerations.
Size inverter wire and most small connections based on ampacity. Remember to
oversize any conductors of array current by the NEC and UL safety factors. Apply
temperature corrections as appropriate.
Size the run from array to central wiring, and any long runs to DC loads using voltage
drop considerations. Size based on only 2% voltage drop.

16-68
Controller and Wire Sizing Worksheet

Charge Control :
Total Charge Controller Capacity
# Parallel Modules X Isc X 1.3
= [
Charge Controller Model :
] X [
] x 1.3
______________ amps
__________________________________________________________________________
Charge Controller Capacity : ________________ amps
Number of Controllers
and Sub-arrays
________________
Sub-Array :
Number of Parallel Modules =
amps
__________________
Nominal Sub-Array Current
________________
Wiring :
Wire Segments Endpoints
Voltage
% loss Current
One-way Wire
Distance Gauge
Wire
Type
Ampacity Fuse
or CB
Rating
________________ to ________________: _______
_______ _______ _______ _______ _______ _______ _______
________________ to ________________: _______
_______ _______ _______ _______ _______ _______ _______
________________ to ________________: _______
_______ _______ _______ _______ _______ _______ _______
________________ to ________________: _______
_______ _______ _______ _______ _______ _______ _______
________________ to ________________: _______
_______ _______ _______ _______ _______ _______ _______
________________ to ________________: _______
_______ _______ _______ _______ _______ _______ _______
________________ to ________________: _______
_______ _______ _______ _______ _______ _______ _______
________________ to ________________: _______
_______ _______ _______ _______ _______ _______ _______
________________ to ________________: _______
_______ _______ _______ _______ _______ _______ _______
________________ to ________________: _______
_______ _______ _______ _______ _______ _______ _______
________________ to ________________: _______
_______ _______ _______ _______ _______ _______ _______

16-69
(End of Chapter)

16-70
CHAPTER SIXTEEN
SYSTEM WIRING
16-1
The National Electric Code and the UL Listing
16-2
Proper Wire Type and Insulation

Color Coding of Wire
Conductor Types
16-3
16-4
16-4
Consider Both Ampacity and Voltage Drop in Wire Sizing
16-6
Wire Sizing Based on Ampacity

Short DC Wiring From Battery
AC Load Wire Sizing
Inverter Wire Sizing
Conductors Carrying Array Current
Adjustment To Ampacity For Temperature
16-7
16-7
16-7
16-9
16-11
16-12
Wire Sizing Based on Voltage Drop

Size For 2% Maximum Loss
Voltage Drop Factors
16-15
16-16
16-17
Voltage Drop Tables

16-22
16-23
16-24
16-25
16-26
16-27
Series/Parallel Array Wiring

12 Volt Array Wiring
Higher Voltage Array
16-29
16-30
16-32
16-33
16-35

16-71
Wiring of Safety Equipment

Battery Bank Wiring
Overcurrent Protection Devices
Safety Disconnects
System And Equipment Grounding For Photovoltaic Systems
Disconnect Arrangement for Ungrounded & Grounded Systems
Lightning and Surge Protection
16-39
16-39
16-41
16-48
16-57
16-60
16-61
Fully Worked Example
16-63
Wire Sizing Worksheet
16-68

16-72
Chapter 16 Answers
System Wiring

Refer to the National Electric Code or the Photovoltaic Power Systems and the National
Electric Code, Suggested Practices.

The inverter has a continuous output of 2400 watts. A nominal voltage of 24 volts
means that the input voltage could be as low as 22 volts (see table in the text). Given
an efficiency of 85%, we calculate the maximum input current as:
Max Inverter Input Current
2400 Watts
.85 X 22 volts
128.3 amps
Using the 125% safety factor from the NEC, our wire needs to have an ampacity of at
least:
Ampacity
128.3 X 125%
160.4 amps
128.3 X 1.25

Per Table 16-4, a conductor made of 14 AWG wire and type THHN (75 C) insulation is
rated for not more than 20 amps. At an ambient temperature of 40 C, this value needs
to be derated by 0.88 as shown in Table 16-6. We also need to divide by the NEC
safety factor:
Derated Ampacity
Ampacity at 25 C X Temperature Correction Factor

NEC Safety Factor
20 X 0.88
=
1.25 (NEC)
16-1
14.1 amps
Wiring
The current flow required by the loads is equal to:

Load Current
Qty X Watts
Volts
13.3 amps
4 X 40
12
Since the ampacity under these conditions is greater than the load current, the
conductor can be used to carry the load current. The answer is then "Yes."

The temperature derate for 90 C conductors at 40 C is .91. Using the array Isc, we
calculate the minimum ampacity of the wire as:
Minimum Ampacity =
19.2 amps X 1.25 NEC X 1.25 UL

0.91 Temperature Factor
32.9 amps
With a 75 C conductor, the temperature derate for a 40 C ambient is .88. The

minimum ampacity of the wire is:
Minimum Ampacity =
19.2 amps X 1.25 NEC X 1.25 UL

0.88 Temperature Factor
34 amps

We first evaluate the voltage drop only, using a current of 6 amps and a total distance
of 20 X 2 = 40 feet. Based on a 12-volt system, the acceptable voltage drop is 2% of
12 volts or .02 X 12 = .24 volts.
Voltage Loss Factor

.24 volts
6 amps X 40 feet
0.001
Referring to Table 16-7, we see that the first conductor size with a voltage loss factor
less than this number is 8 AWG.
16-2
Wiring
Next looking at the ampacity, we calculate the minimum required ampacity as:
Minimum Ampacity
6 amps X 1.25 NEC
7.5 amps
Using THHN (90 C) in conduit, we find using Table 16-4 that 14 AWG has an ampacity
of 25 amps, easily accommodating this current. This is a case where the wire size will
be determined by the voltage drop rather than just the ampacity.

We will first evaluate the voltage drop. The current is 2 X 40 Watts 24 volts = 3.3
amps. The total distance is 35 X 2 = 70 feet. Based on a 24-volt system the
acceptable voltage drop is 2% of 24 volts or .02 X 24 = .48 volts.
Voltage Loss Factor

.48 volts
3.3 amps X 70 feet
0.0021
Per Table 16-7, we see that the first conductor size with a voltage loss factor less than
this number is 12 AWG.
For the minimum ampacity required, we calculate it as follows (there are no
temperature derates):
Minimum Ampacity
3.3 amps X 1.25 NEC
4.1 amps
Referring to the values of THHN (90 C) in conduit in Table 16-4, we see that 14 AWG
has an ampacity of 25 amps, easily accommodating this current. Again, the voltage
drop in this case will determine the necessary conductor size.
16-3
Wiring

We will first evaluate the voltage drop for the array. Using 5 modules in parallel the Isc
for the array will be 5 X 3.45 = 17.25 amps. The total distance is 50 X 2 = 100 feet.
Based on a 24-volt system the acceptable voltage drop is 2% of 24 volts or .02 X 24 =
.48 volts.
Voltage Loss Factor

.48 volts
17.25 amps X 100 feet
0.00028
Per Table 16-7, we see that the first conductor size with a voltage loss factor less than
this number is 4 AWG.
We next look at the minimum ampacity required. Remember, for an array circuit, we
need to use both the NEC and UL safety factors:
Minimum Ampacity
17.25 amps X 1.25 NEC X 1.25 UL = 26.9 amps
Referring to the values of THHN (90 C) in conduit in Table 16-4, we see that 12 AWG
has an ampacity of 30 amps.

Point
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
Current
Flowing
12 amps
3 amps
6 amps
3 amps
3 amps
3 amps
3 amps
3 amps
3 amps
9 amps
12 amps
16-4
Voltage
0 volts
0 volts
0 volts
0 volts
12 volts
24 volts
36 volts
48 volts
48 volts
48 volts
48 volts
Wiring

Referring to Figure 16-14, we first identify the curve for a 60A fuse. This is the third
curve from the top. We next find the time 0.1 seconds across the top and move down
to where this intersects the 60A fuse curve. Reading across horizontally, we find this is
about 900A of overcurrent.

An overcurrent of 400 amps for a 100-amp breaker represents a 400% load. Referring
to Figure 16-15, we find the 400 percent load on the bottom and then move up to
standard time delay range (top range shown). This indicates a time between 4-11
seconds to interrupt.
If we were to use a short time delay breaker (bottom range), the time is .03 and .2
seconds to interrupt.

Again looking at Figure 16-15, we find the 300% load at the bottom and move up until
we reach the 5 second mark. This occurs within the medium time delay range.

Refer to the appropriate diagram.

Refer to the appropriate diagram.
16-5
Wiring
Chapter Seventeen
Water Delivery Systems
Water pumping is a fast growing and wonderful application for photovoltaic power
systems. There is a natural match between the availability of sunlight and the need for
water. Most photovoltaic pumping systems do not use batteries thus avoiding a costly
and high maintenance component and increasing the reliability greatly. In rural
undeveloped areas there is a critical need for fresh underground water to help prevent
diseases spread by using surface waters. A low maintenance photovoltaic powered
water system can bring health and prosperity to remote villages, without the burdens of
paying for maintenance, spare parts or fuel. In this chapter we discuss issues
associated with photovoltaic powered water delivery systems including performance
characteristics of different types of pumps and sizing concerns.
The focus of this discussion is on direct coupled solar powered water pumping
systems. These systems output water in proportion to available solar insolation. They
pump more water during summer months, typically when more water is needed, and
avoid the need for batteries and regulators. Such direct coupled systems may not be
as efficient as battery coupled designs in the sense that they cannot pump during any
weather or at night. But the advantages of reliability, simplicity and low cost make
these types of systems very attractive for a wide range of applications and users around
the world.

17-1
System Design Water Delivery Systems
Why Solar Water Pumping

Advantages of Solar Water Pumping
Photovoltaics offer many advantages to traditional water pumping technologies such as
a diesel engine generator or a windmill. A properly specified PV pumping system will
require little or no maintenance, has no fuel requirements, and will have a long life
expectancy particularly for the solar panels. The major disadvantage of PV is the initial
cost.
Advantages
Solar
Pumping
Diesel
Engine
Generator
Low initial capital cost

Familiar Technology
Disadvantages
No fuel required
Little maintenance
Environmentally Benign
Panel life is 20-30 years
Cost effective for small
power demands
High initial capital cost

Uses unfamiliar technology
Parts may be hard to obtain
Requires regular
maintenance and fuel
brought to site
Requires dependable
operator and service
support
Environmentally harmful
Expensive when
considering life cycle cost


17-2
Unique Characteristics of Solar Water

Pumping
PV water pumping technology (typically DC) has been developed around a variable
power source, the sun, as compared to traditional pumping (AC) which relies on a
constant power supply such as the grid or an engine generator. As a result PV water
pumping systems have the following unique characteristics.
1. Premium on Efficiency: Because sunlight is expensive to convert to electricity,
there is a high premium to use efficient motors and properly sized pump heads in
order to keep the power requirements and hence the initial cost down for the PV
system.
2. Use of DC Motors: DC motors are commonly used with PV for pumping to avoid
the loss of efficiency and complexity when converting DC power to AC. DC motors
work well at varying voltage and speed which is ideal for operating from a constantly
changing power source such as solar energy.
3. Time of Delivery: Direct coupled PV pumps delivery water only when the sun is
out. This may require some type of water storage in order to satisfy the need when
the sun is not out.
4. Range of Effectiveness: PV pumps are only available for a set range of conditions
of head and daily pumped volume. The upper limits for a PV pump are 200 meters
of head and 400 m3 of daily pumped volume.
5. Portability: PV pump systems can be made to be portable (trailer mounted for
example). This can offer many advantages. For example, in the western USA
livestock is often moved seasonally to different pastures. Instead of having
permanent pumps one PV pump can be moved to different well sites.

17-3
Markets and Applications

The number of solar pumping installations is estimated to be over 10,000 worldwide in
1990. Cumulative pump sales is believed to be over 5000 per year (1994) increasing at
a rate of 30 to 40% per annum in the developing world. PV pumping systems are most
economical for small power demands typically less than 1000 Wp. They can vary from
one or two modules (about 80 Wp) for livestock watering or household use to several
kilowatts for village water pumping. This range is well matched to pumping demands
commonly found in the developing world.
The reasons for increasing pump sales are: (1) falling real prices of PV pump systems
due to decreasing cost of modules and pumps; (2) more reliable pump technology (the
first pumping systems on the market in the early 80s had a high failure rate); and (3)
increasing awareness and acceptance by users such as NGOs (non-government
organizations) and governments in the developing world.
There are four main markets for photovoltaic powered water pumping systems with a
growing variety of applications serving people around the world.
Village Water Supply

Solar pumping is best suited for villages because of the small quantities of water
involved and the high value of delivering clean water for drinking. Hand pumps might
already be used, or villagers may be simply gathering surface water and carrying it back
to their village. But surface water is easily contaminated and disease control becomes
a big problem. Sub-surface water pumping using photovoltaics brings both a reliable
and a clean water supply to people in remote areas.
Solar pumping becomes cost competitive with diesel for village water supply when
average daily water requirements are less than a volume-head product of 800 m 4 (e.g.,
40 m3/day through a 20 m head) and where average solar irradiation is greater than 2.8
kWh/m2/day. (Source: World Bank Technical Paper #168).
If explicit water purification is needed, then solar power systems can operate them as
well. Systems that operate on the basis of reverse osmosis, UV irradiance, or chlorine
can all be operated directly by photovoltaic power systems.

17-4
Irrigation
Water can be pumped from nearby streams or canals or lakes to irrigate fields. Or it
can be draw up from even hundreds of feet. Photovoltaic systems can be permanently
mounted in the fields, be portable so they can serve multiple farmers, or be moved to
different fields as needed.
Drip irrigation is more efficient than flood irrigation and is especially suited for hot
climates. However, it involves more equipment and is more expensive. Photovoltaic
power systems can operate the pumps as well as the pressurizing systems to pass
water through filters that are often needed for drip systems.
In general solar water pumping for irrigation is not competitive when compared to diesel
pumping. An irrigation system usually requires a large amount of power for a short
seasonal period, whereas solar water pumping has its greatest advantages with small
scale year-round use.
However, there are niches for PV pumping with such applications as micro irrigation for
vegetable gardens and plots of less than 1 hectare. As a general rule solar pumping
systems for irrigation are most cost effective compared to diesel pumps where the peak
daily water requirements (volume-head product) are less than about 250 m4 and where
insolation is greater than 4 kWh/m2/day.
Consider solar water pumping if the following conditions are met:
Volume-Head
Product
(m4)
Average Daily
Solar Insolation
(kWh/m2/day)
Less than 800
Greater than 2.8
Irrigation
Less than 250
Greater than 4.0
Source: Solar Pumping: World Bank Technical Paper #168


17-5
Livestock Watering
Livestock watering using photovoltaics has been a high value application in developed
countries. The cost of maintaining utility lines to pumps in remote areas is often greater
than the total revenue for the power to run the pump. As a result there is a movement
in the US utility industry to offer photovoltaic pumps as a substitute for extending grid
power to such small and infrequently used loads.
Systems can be installed in remote ranch areas to provide a reliable water supply with
little maintenance and noise. Water can be pumped day and night with the use of
batteries. Or systems can be designed to work without batteries and accumulate the
water pumped during daylight in storage tanks, which then can deliver water day or
night.
Interesting applications such as pond aeration are also possible. Modules operate
pumps that bubble water into ponds to mix up temperature gradients and prevent or
postpone freezing.
Residential Needs
Homes have a variety of pumping needs including submersible pumps for wells,
pumping water to pressure lines, and water for irrigation and landscape maintenance.
A variety of specialized DC pumps for use with photovoltaics have been developed for
this purpose.
PV pumping for homes is typically a part of a larger residential solar electric power
system. In many cases standard AC pumps are used because they use the AC power
available from the inverter in the system.

17-6
Typical Usage Requirements

The following are some suggestions for water requirements in developing and
developed regions of the world. They are meant only as a rough guideline to help give
a sense of what might be required for different applications.
Typical Human Water Requirements in Developing Areas (per person per day)
Region
Africa
Southeast Asia
Western Pacific
Eastern Mediterranean
Europe(Algeria, Morocco,
Turkey)
Latin America, Caribbean
World Average

Gallons
(Liters)
4 - 10
8 - 20
8 - 25
10 - 22
5 - 17
15 - 35
30 - 70
30 - 95
40 - 85
20 - 65
20 - 50
70 - 190
10 - 24
35 - 90
Typical Animal Water Requirements (per animal per day)

Animal
Horse
Cow (dry)
Cow (milking)
Beef Cattle
Hog
Sheep
Turkeys(per 100)
Chickens(per 100)
Hens(per 100)
Rule of thumb


Gallons
Liters
12
15
40
12
4
2
13
4
8
45
57
150
45
15
10
50
15
30
10 gpd per 1000 lbs.
37 lpd per 450 kg.
17-7
Breakdown of Daily Use (per day)

Use
Gallons
Liters
Developing Areas
Survival
Public Hydrant
Single Home Faucet
Multiple Connections
w/bath, toilet, sink
0.6
11
11-16
45-65
2
40
40-60
80-240
Industrialized Areas
Personal (including
cooking)
Shower
Bath
Sink
Washing Machine
Dish Washer


10
38
25-60
30-35
1-2
30-50
10-20
95-225
115-135
4-8
115-190
38-75
17-8
Battery Powered Or
Direct Couple To Array
There are basically two types of PV powered water delivery systems: (1) pump and
motor directly connected to an array (or using a maximum power tracking circuit); and
(2) motor connected to battery bank charged by a solar array.
The advantage of direct coupled water delivery systems is the underlying simplicity of
the design. There is more water pumped during summer periods of high insolation,
which is usually when more water is required, especially for livestock watering and
irrigation. During periods of low insolation there is less water pumped, but there is less
evaporation as well. The system has few components to fail and the initial cost can be
kept low. As photovoltaic and maximum power tracking technologies have become
more efficient and reliable over time the trend worldwide favors installing direct coupled
pumping systems over systems using batteries.
Battery Power or
Direct Coupled to Array?
Advantages
Disadvantages
Direct
Coupled
to Array
Simplicity
Reliability
Low maintenance
Low initial cost
Match to season
Low efficiency if no
MPT installed
Requires storage tanks
for 24-hour availability
Battery
Powered
Predictable supply
Higher efficiency
Supply starting surge
current
Maintenance
Complexity
High costs over time
Charge control failure


17-9
One of the disadvantages of direct coupled systems is the inefficiency of the match
between a DC motor and a PV array during periods of low irradiance. The array can be
designed to have the motor operating at the array Vmp during the peak hours of the
day, as discussed previously in the chapter on Output Curves. This means that each
module operates near 15 volts instead of being held back to around 12 for battery
loads. But during the morning and evening hours or days of low insolation the array will
operate the motor far below its optimum voltage and away from the array Vmp. The
array may be producing power, but it is not being efficiently passed to the load. The
usual solution for this problem is to include a maximum power tracking device between
the array and DC motor. It operates the array at its maximum power voltage and
converts the power into the most useful voltage and current for the motor. Of course it
adds somewhat to the initial cost, has some inherent inefficiency (usually 95%), and
adds some complexity back to the system. But batteries and charge controllers are still
avoided. And the overall efficiency of batteries (about 80% overall) is still lower than
the efficiency passed through a typical maximum power tracker.
Another disadvantage of direct coupled systems is the need for some form of water
storage if water is required at night or during low insolation days. This means
effectively higher initial costs, and perhaps some maintenance costs associated with
maintaining the tank and its support structure. But it can be argued that it is more
effective to store water than to store electricity in batteries.
The advantage of battery coupled systems is just the opposite. All the array energy is
passed to the batteries, and can be extracted during day or night. The motor can be
operated at its optimum voltage all the time, not just the peak hours of the day, and
even on overcast days, thus prolonging motor life and improving system efficiency. And
another key feature of battery coupled systems is the ability to supply a high surge
current to start difficult motors or counter a high torque portion of a pumping cycle.
The key disadvantage of battery coupled systems is that batteries are involved at all. If
they are treated well they can last for many trouble-free years. However they are
adversely affected by high temperatures, will require cleaning of the terminals and
water level maintenance, age with time, and can sulfate, corrode, and fail. Some form
of charge control is required to prevent overcharge, and a motor controller is necessary
to tell the motor when to operate and when to shut off. This all adds up to a more
complex system, with an associated higher probability of some sort of failure.
Which type of system is best for a particular situation depends on comparisons of initial
cost, costs over time, reliability, efficiency and availability.

17-10
Terminology
The most critical pieces of load information for doing water system sizing are the total
vertical lift, or head, and the volume of water per day required. Other details will help
to fine tune the design, but these two pieces of information are critical.
Flow: The rate at which water is delivered by the pump, usually measured in
gallons/minute or liters/second.
Volume: The total amount of water needed daily. Usually given in gallons/day or cubic
meters/day.
Suction Head: Vertical distance from surface of water to center of pump when pump is
located above water. There is no suction head for a submerged pump.
Discharge Head: Vertical distance from center of pump to surface of storage tank
water or point of free discharge.
Static Head: Vertical distance from surface of water to surface of storage tank water or
point of free discharge.
Static Head
Suction Head
Discharge Head
Pressure Head: If a final discharge pressure is desired, the pump must be able to
supply the flow with the needed energy. The Pressure Head is the equivalent feet of
head the pump needs to be able to pump to supply this final pressure.
Pressure Head (feet)
Pressure (psi)
2.31
Friction Head: There is a loss of energy as water moves through a pipe. The smaller
the pipe diameter and the faster the flow, the greater the loss. The pump must be able
to supply enough energy to overcome these losses.

17-11
The Friction Head is the equivalent distance the pump must be able to push water to
have enough energy to overcome these losses. Tables of friction loss coefficients for
different size pipes, different types of pipe, and for fittings, are presented in the
Appendix. Factors are available for fittings, elbows, as well as for different pipe types
(steel, plastic, copper) and for different pipe diameters.
Draw Down: When the pump draws water from the well, the level of the surface may
drop depending on the ability of the surrounding earth to replenish the well. The draw
down is the distance from the surface of the water when it is being pumped to the
surface level when it is static. This can amount to many tens of feet, and depends on
rate of pumping. The higher the rate, the greater the draw down.
Pumping Head Components

Hf
friction head
Hp
pressure head
(2.3 X psi.)
surface
mounted
pump/motor
Hd
discharge head
Hs
suction head
Hdr
drawdown

Total Dynamic Head: The final total head the pump must be able to deliver at the
desired rate of pumping, including all the previous heads.
Total Dynamic Head

Suction Head + Discharge Head +

Pressure Head + Friction Head + Draw Down
17-12
Calculating Total Head

The purely vertical displacements of suction head and discharge head and drawdown
are relatively simple to determine by measurement. The other adjustments to total
head, such as friction head and pressure head must be calculated.
Friction Head Calculation

Water flowing through straight length of pipe experiences friction, and this reduces the
effective output of a pump. The amount of loss depends on the diameter of the pipe
and the flow rate through the pipe. It is as if the pump has to overcome more vertical
feet of lift, even though the water may be moving horizontally. The friction losses have
been translated into effective vertical feet of lift or head in various tables presented
next. The losses are presented as equivalent vertical feet per 100 feet of actual
pipe length.
A component of friction loss is the turbulence and friction caused by fittings, elbow
joints and other connections. Their energy reducing effect has also been converted into
equivalent vertical feet of head in various tables presented next. The losses are
presented as equivalent vertical feet.
Friction head is computed in two steps:

(1) First the equivalent feet of head due to fittings is calculated. Then this length is
added to the actual length of pipe, to give the total equivalent length of pipe.
(2) The friction loss equivalent feet of head is then computed for this total length, and
added to the true vertical displacements of suction and discharge head and drawdown
to give the final total head of the system.

17-13
Size of Fitting, inches
1/2
3/4
1 1/4
1 1/2
90o Ell
1.5
2.0
2.7
3.5
4.3
5.5
40o Ell
0.8
1.0
1.3
1.7
2.0
2.5
Long Sweep Ell
1.0
1.4
1.7
2.3
2.7
3.5
Close Return Bend
3.6
5.0
6.0
8.3
10.0
13.0
Tee--Straight Run
3.3
4.5
5.7
7.6
9.0
12.0
Glove Valve Open
17.0
22.0
27.0
36.0
43.0
55.0
Angle Valve Open
8.4
12.0
15.0
18.0
22.0
28.0
Gate Valve--Open
0.4
0.5
0.6
0.8
1.0
1.2
Check Valve (Swing)
11
13
Check Valve (Spring)
12
14
19
Tee--Side Inlet / Outlet
!"#$% %
GPM
3/8
1/2
3/4
1
4.25
1.38
.356
.11
2
15.13
4.83
1.21
.38
3
31.97
9.96
2.51
.77
4
54.97
17.07
4.21
1.30
5
84.41
25.76
6.33
1.92
6
36.34
8.83
2.69
8
63.71
15.18
4.58
10
97.52
25.98
6.88
15
49.68
14.63
20
86.94
25.07
25
38.41
30
35
40
45
50
(Equivalent vertical feet/100 feet of length)
&

17-14
1 1/4
.10
.21
.35
.51
.71
1.19
1.78
3.75
6.39
9.71
13.62
18.17
23.55
29.44
1 1/2
.10
.16
.24
.33
.55
.83
1.74
2.94
4.44
6.26
8.37
10.70
13.46
16.45
.10
.17
.25
.52
.86
1.29
1.81
2.42
3.11
3.84
4.67
GPM
3/8
1/2
3/4
1
4.30
1.86
.26
2
15.00
4.78
1.21
.38
3
31.80
10.00
2.50
.77
4
54.90
17.10
4.21
1.30
5
83.50
25.80
6.32
1.93
6
36.50
8.87
2.68
7
48.70
11.80
3.56
8
62.70
15.00
4.54
9
18.80
5.65
10
23.00
6.86
12
32.60
9.62
15
49.70
14.70
20
86.10
25.10
25
38.60
30
54.60
35
73.40
40
95.00
45
' &
GPM
3/8
1/2
3/4
1
6.2
1.8
.39
2
19.6
6.0
1.2
5
30
5.8
1.6
7
53
11.0
3.2
10
19.6
5.3
15
37.0
9.9
18
55.4
16.1
20
18.5
25
27.1
30
39.3
35
48.5
40
45
50
( &

17-15
1 1/4
1 1/2
.34
.51
.70
.93
1.18
1.46
1.77
2.48
3.74
6.34
9.65
13.60
18.20
23.50
30.70
.24
.33
.44
.56
.69
.83
1.16
1.75
2.94
4.48
6.26
8.37
10.79
13.45
.10
.13
.17
.21
.25
.34
.52
.87
1.30
1.82
2.42
3.10
3.85
1 1/4
1 1/2
2.1
3.2
3.9
5.3
7.6
10.2
13.2
16.2
19.4
1.5
2.1
2.8
3.5
4.2
5.1
2.2
3.9
6.2
6.9
10.4
14.3
18.7
25.4
30.0
39.3
Example:
A water pumping system is proposed to supply water to a storage tank
near a village. The piping used throughout is to be 1-inch diameter and the flow rate
will be about 10 gpm. It will draw from a nearby stream, so assume no drawdown.
The system uses a surface mounted centrifugal pump, so water is drawn up to the
pump and discharged above the pump. The total vertical lift will be 150 feet, from the
top of the stream to the top of the water in the tank (5 feet suction to pump, 145 feet
discharge from pump to water in tank).
There will be the following components:
370 feet of 1 inch plastic pipe; a check valve (spring type); three 90 deg. elbow joints;
two 45 deg. elbow joints/
Calculating the equivalent feet for the fittings (using Table 17-6):
90 deg. elbows:
3 X 2.7 feet
45 deg. elbows:
2 X 1.3 feet
spring check valve:
=
=
Equivalent length for fittings:

Length for friction calculation:
8.1 feet
2.6 feet
8 feet
18.7 feet
370 feet + 18.7 feet

=
388.7 feet
Friction Head (using Table 17-7):

388.7 feet X 6.88 feet/100 feet
100
=
26.7 feet
So the total dynamic head will be

suction head + discharge head
5

145
17-16
+ friction head
+
26.7
176.7 feet
Types of Motors For

PV Pumping Systems
There are three types of motor types used in photovoltaic powered pumping systems:
(1) brushed (Permanent Magnet) DC motors; (2) brushless (Permanent Magnet) DC
motors; and (3) AC motors.
The most important distinction is between AC and DC motors. DC motors can be
coupled directly to the array producing the most efficient and simplest systems. AC
motors require an inverter, thereby increasing complexity and adding inefficiency.
However, AC motors are more available, cheaper and more familiar to pump people
around the world. AC motors are typically used in larger photovoltaic pumping systems
while DC motors are best used in smaller systems.
The DC pumps used for PV applications are generally of the permanent magnet type.
In a conventional DC motor a magnetic field is produced electromagnetically by the field
windings. In a permanent magnet motor, the magnet produces the magnetic field.
Thus no power is consumed in the windings leading to higher efficiencies, which is
always the preference in PV systems.
No one type of motor dominates the marketplace, as each one has certain advantages
and disadvantages, as presented in the next table.

17-17
Motor Type
Brushed DC
The traditional DC
motor in which
brushes conduct
current into the
spinning portion of
the motor.
Permanent magnets
produce a magnetic
field inside the motor
shell
Brushless DC
Advantages
High technology
motor using a
complex electronic
system to precisely
alternate the current
causing the rotor to
spin
Disadvantages
Simplest and
most efficient
motor for use
with PV systems
No complex
control circuitry.
Motors start
without a great
current surge and
will run slowly but
not overheat with
reduced voltage.
Efficient
No maintenance
required
AC Motors
Inductive motors
No brushes to
replace
Can use existing
AC motor/pump
technology, which
is cheaper and
available
worldwide. Can
handler larger
pumping
requirements.
Special Features
Brushes need
replacing
periodically
(Typical
replacement
interval, 2000 to
4000 hr. or 2
years)
Electronic
computation adds
extra expenses,
complexity,
increasing risk of
failure
In most cases, oil
cooled, cant be
submerged as
deep as water
cooled AC
Requires an
inverter to
change DC into
AC adding cost
and complexity
Less efficient
than DC units
Prone to
overheating if
current is not
sufficient to start
motor, or if
voltage is too low
Requires
maximum power
tracker for
optimum
performance
Only available in
small motor sizes
Increasing
current (by
paralleling
modules)
increases torque,
Increasing
voltage (by series
modules)
increases speed
Growing trend
among PV pump
manufacturers to
use brushless
DC, primarily for
centrifugal type
submersibles
Can be single or
three phase
Inverters are
designed to vary
frequency to
maximize power
to the motor in
response to
changing light
levels
)*$ +


17-18
Types of Pumps For Photovoltaic

Pumping Systems
There are two general classes of pumps: (1) centrifugal; and (2) positive displacement.
Both centrifugal and positive displacement pumps can be further classified into those
with motors that are (a) surface mounted and those, which are (b), placed into the water
(submersible).
Centrifugal pumps have blades or impellers that rotate at high speed, creating a
pressure and forcing water to flow. Positive displacement pumps move a volume of
water through a distance by means of a plunger or cavity and then displace that water
with another quantity of water behind it, and so on. Pumps in both classes have
advantages and disadvantages, which are discussed next.
Pump Types
Centrifugal
Surface
Centrifugal
Submersible
Straight or
Self-Priming
Submersible
Jet Pump
Positive
Displacement
Surface
Positive
Displacement
Submersible
Helical
Cavity
Diaphragm
Jack Pump
Vertical
Turbine

, ,

17-19
Centrifugal Pumps - Surface

In general centrifugal pumps are good for moving large volumes of water. The high
speed of the impellers makes moving gritty water a problem though. The performance
of a centrifugal pump connected directly to a photovoltaic array is very sensitive to the
peak irradiance. As the irradiance goes down the current from the array goes down
and the motor rotates more slowly. A change in the speed will be directly proportional
to a change in the flow rate, but will have a geometric effect on head capability. That is
a 20% reduction in irradiance will result in 80% of original current and 80% of original
flow rate capability, but will result in (.80) x (.80) = .64 or 64% of original head capability.
Small changes in irradiance will result in large changes in the output of the pump, and
may lead to the pump not meeting the head requirement and effectively "shutting off"
until higher irradiance returns.
Straight Centrifugal
Installed above the water, it has a single impeller. Suction head is limited by
atmospheric pressure to approximately 20 feet, but can push above pump (discharge
head) hundreds of feet. Must be "primed" (filled with water) before each start-up.
Commonly used for surface water pumping, from streams, lakes, or to move water
along the land through pipelines.
"-


17-20
Self-Priming Centrifugal
Has a chamber above impeller which holds water after shut-down so the pump is
"primed" before each start-up.
Jet Pump
Centrifugal with a portion of the flow returning to a venturi on the input side. This
increases suction head to 150 feet but with a reduction in net flow. The venturi can be
placed just in front of the impeller chamber, or all the way down at the input of the
suction pipe. Low cost solution for relatively low flow rates and heads. Low efficiency
however.
Vertical Turbine
Series of impellers in one long narrow cylindrical casing that is submerged below water
level, connected to the motor on the surface by a long drive shaft. Allows for deep
pumping at high rates by using huge motors above ground. Heads limited by shaft
length. Efficiency is reduced due to twisting, friction, vibration, and weight of shaft and
bearings. Typically used for large-scale irrigation with large AC or diesel motors.
"-

17-21
Centrifugal Pumps - Submersible

Submersible: Similar to vertical turbine but has waterproof DC or AC motor connected
directly to pump, submerged below water level. Pump and motor assembly is
suspended by the pipe that carries water to surface. No drive shaft means can pump
from great depths (1000 feet).
!"-


17-22
Positive Displacement Pumps - Surface

In general positive displacement or volumetric pumps are good for meeting large head
requirements with small or moderate volumes. The rates are not as high as for
centrifugals, but the use of total available solar energy is generally better. Small
changes in the irradiance on an array will slow the motor but will not reduce its ability to
meet the head requirement as much as for centrifugals. This is because it is still
pushing one small unit of water after the next, just a bit slower. So once a positive
displacement type pump has pumped enough to meet the head requirement, it will
continue pumping all day long. Generally more of the solar energy over the day will be
directly translated into volume of water pumped.
Helical Cavity: Helical rod rotates inside a cavity of slightly different pitch, creating a
vertically moving cavity. Submerged under water and connected to motor above
ground by long drive shaft. Motor power is lost in friction, weight, vibration and twisting
of drive shaft and bearings. Requires high torque, may necessitate batteries or
maximum power tracker to supply surge to overcome tremendous friction. Can move
very gritty water, whereas high-speed centrifugals would be eroded quickly. Suitable for
low flow rates and moderate depths.

17-23
Jack Pump: Motor pulls on oscillating jack above ground which pulls on a long drive
shaft connected to a plunger with flapper valve below water level. Each cycle of the
jack moves a volume of water into the drop pipe, and displaces water previously
pushed into the pipe. Low flow rates achieved but great depths possible. Flapper valve
needs periodic replacement. Above ground mounting means can use wide variety of
either AC or DC motors.
A variation of the jack pump design has motor and small jack mechanism packaged in
narrow cylindrical casing like submersible. Pump and motor can be submersed below
water level.
"- +$

17-24
Positive Displacement Pumps

Submersible
Diaphragm Pump: The motor rotates an uneven cam, which opens and closes valves.
They produce low to medium flow rates and operate to medium to high heads (few
hundred feet). They can be built as a surface mounted design or in a waterproof
submersible design. Many commercially available small diaphragm type pumps for
personal water supply operate at 24 volts needing only 2-4 modules.
'"- $

17-25
Pump
Self-Priming
Surface
Centrifugal
Use
Surface mounted
motor and pump for
high volume, low
head applications
commonly used for
irrigation from rivers
Advantage
Disadvantage
Ease of installation,
easily serviced, wide
range capabilities,
many pumps use
water as a bearing
lubricant
Limited head: < 7 m
Vertical
Turbine
Down hole, medium

volume, mediumhead pump
Head: < 80 feet
Jet
Medium-head,
medium -volume
pump
Limited Head: < 30

m
Submersible
Centrifugal
Down-hole, mediumvolume, high-head,

integrated
motor/pump unit
Head: < 200 m
Ease of
maintenance with
surface mounted
pump
Self priming
Wide range of
capabilities
Good sand/tilt
tolerance
Low equipment and
maintenance costs
Can be used beyond
suction limit
Very reliable
Easy access to
motor and pump
Least expensive
intermediate-head
pump
Adaptable to very
small wells (50 mm)
Motor is directly
coupled to impellers
Easily cooled
because it is
submersed
Multistage impellers
accommodate a
wide range of heads
No noise, no
pumphouse
Use water as
bearing
Can pump from
great depth
Limited 5 meter
suction head
Relatively inefficient
compared to
centrifugal s which
have flooded inlets
Have a narrow
efficiency range with
respect to head
In abrasive
conditions, impeller
wear can lead to
reduced efficiency
Shaft losses reduce
efficiency compared
to submersibles
Shaft and borehole
alignment are critical
to successful
operation
Hard to install and
maintain
Relatively inefficient
compared to other
types of pumps
Sand/grit causes
impeller wear and
decreasing efficiency
Saline conditions will
corrode metal
housing
High capital cost,
highly expensive to
repair (must remove
pump from well)
Can be damaged by
running dry
.$$

17-26
Pump
Surface
Positive
Displacement
- Rotary or
Mono pumps
(Helical
Cavity)
Surface
PositiveDisplacement
Reciprocation
Piston
(Jack or
Nodding
Donkey
Use
Medium to highhead, medium-flow,

Mono or Moyno
pumps
Head: < 150 m
High-head, low-flow,
down hole piston
and cylinder driven
by sucker rod from
surface
Advantage
Head: < 150 m
Disadvantage
Very robust
Efficient over wide
range of head except
for under 20 meters
Simple construction
Self priming
No back-flow valve
required
Efficient over a wide

range of motor
speed
Can pump from
great depths
Simple design, easy
to repair
Submersible
Positive
Displacement
(Diaphragm)
Medium head, low

flow applications
Head: < 80m
Few moving parts

High efficiency over
the range of head
Low internal friction
Tolerant of sand and
other particles
Serviceable with
hand tools
Small size makes
them easy to remove
from wells
Can be used in
portable
applications.
Sand or very hard

water can cause
premature degradation
of rubber stators
Requires gearing
Can overload motors if
downstream valves
are closed
Installation is difficult
Requires batteries or
power-conditioning to
supply high starting
torque (current)
Maintenance requires
periodic replacement
of leathers and
cylinder, must remove
piston and cylinder
from well to service
seals
More expensive than
centrifugals of same
size
Cannot tolerate sand
or sediment
Efficiency decreases
as piston seals wear
Requires batteries or
power-conditioning to
supply high starting
torque (current)
Low capacity
Not appropriate for
deep wells
Diaphragms must be
replace every 1-2
years
Most designs use
brushed DC motor
requiring regular brush
replacement
Limited to heads
below 100 meters
Not all models can be
rebuilt
Adapted from Pump Selection, Water and Sanitation for Health Project, US AID p. 42
.$$ $

17-27
Optimal Operation Ranges for

Different Types of Pumps
100
Total Head (m)
Positive Displacement
Submersible
10
Surface Suction
Centrifugal
Handpump
1
1
10
100
Daily Volume (m3/day)
(/ / 0

17-28
Pump Type Comparison

Centrifugal
Positive Displacement
High speed impellers
Volumetric movement
Large volumes
Lower volumes
Moderate head
High head
Loss of flow rate with higher head
Flow rate less affected by head
Low irradiance reduces ability to

achieve head
Low irradiance has little affect on

achieving head
Possible grit abrasion
Unaffected by grit
),

$

17-29
Exercise

Model
Type

Head
17-30
Flow Rate
Power Needed
Primary Requirements
For Choosing A Pump
The first requirement that any pump must satisfy is the total head. If the head cannot
be reached then no water is pumped at all. If possible the total dynamic head should
be know including friction losses in the pipes and any drawdown that will occur. Often
all this information is not available. In that case certainly at least the suction and
discharge head must be known.
The second requirement is that the pump be able to deliver the required volume of
water in a typical day to the specified head. Direct coupled solar pumping systems will
operate only during daylight, and the output will vary with the daily insolation. A simple
approach to choosing a pump begins with estimating the amount of peak sun hours
available at the site, and dividing this into the total daily volume needed. This gives the
rate of pumping during peak sun hours that the pump must perform. Choose a pump
that can operate at the required head and that can output at this rate. The power
needed by the pump to operate at this rate and this head will then determine the size of
the solar array.
(1)
Required Total Dynamic Head (including friction and drawdown)
(2)
Required Pump Flow Rate
Daily Volume Required

Peak Sun Hours
Factors for converting insolation values into peak sun hours:
Unit of Daily Insolation
Multiply measured insolation by this

factor to convert to peak hours...
kwh / m2
1.0
Langley (cal/cm2)
0.01162
MJ / m
0.2777
Btu / ft2

0.003155
17-31
If one pump is not adequate to deliver the total volume perhaps the manufacturer offers
a way to combine pumps together. An example of a system by Grundfos is shown on
the next page that allows up to three pumps to be operated in the same borehole,
coupled either horizontally for large boreholes or coupled vertically for smaller borehole
diameters.
+*

1%
Exercise

!" #$
&""" #$
!"(""" #$

% ##$
&" '$
&& % ##$

17-32
Supplementary Considerations
For Choosing A Pump
Site, Well, and Water Considerations
The diameter of the borehole must be known because it can restrict your choices for
pump models. Typically submersible pumps are available in 4 inch or 6 inch diameters
or in larger sizes. A system designer who chooses a large pump on the basis of
meeting the volume demand must also confirm that the pump will indeed fit in the well
in the field!
The flow capacity of the well should be known so that the pump does not over pump
the well. If the rate of pumping far excess the recovery rate of the well from the
surrounding ground then the water level will continually drop during pumping, and the
pump could be exposed to air and pump in dry conditions. This might harm the pump.
If more volume is needed than can be delivered by a direct coupled pumping system
operating quickly during sunlight hours, then perhaps a battery coupled system should
be considered. In such a design a slower pump rate could be sustained 24 hours a day
perhaps staying under the natural recovery rate of the well.
Abrasive particles in the water are another consideration in choosing a pump type.
Very gritty water can cause high-speed impellers in centrifugal types to erode quickly.
Positive displacement type pumps should be considered because they are less affected
by grit. Or perhaps manufacturers of centrifugal pumps could supply an appropriate
filter to reduce the grit effect. Such filters will probably reduce output performance to
some degree.

17-33
Electrical Considerations
One electrical consideration is whether the system should be direct coupled to modules
or battery coupled. The usual trade-offs are simplicity, reliability, low maintenance, and
lower cost for a direct coupled system against 24-hour availability of water and higher
motor operating efficiency with batteries.
Another choice is whether a fixed array mount should be used or a sun position tracker.
Fixed structures are less expensive and can withstand extremely high winds. But
trackers can increase water output by 50-60% during summer months. The reduction in
the number of modules needed could equal or exceed the cost of the tracker.
If a pump model is chosen that uses an AC motor then an inverter will be required.
Most common commercial inverters require a battery bank for stable and narrow
voltage window. (Grundfos by way of contrast makes a special inverter designed to
operate directly connected to the solar array).
If a pump with a DC motor is chosen then the consideration is whether to use a brushtype commutated motor or a brushless electronically commutated design. Brushes will
have to be serviced periodically while electronic commutation circuitry could increase
the complexity and decrease the reliability of the overall system. Details of the motor
and control design must be considered.

17-34
Pump Output Curves

Just as solar modules have output curves so do water pumps. Solar module IV curves
compare current against voltage given conditions of light irradiance and temperature.
Water pump curves compare flow rate against head given conditions of motor power to
the pump.
If a pump is paired with different motors or the motor is operated at different voltages,
the output of the pump will be different. In AC systems often manufacturers present a
pump performance diagram when it is matched to different horsepower motors. Pumps
to be operated by DC motors (typical for solar pumps) often present performance
curves for the same motor operated at different DC voltages.
Pump curves show the flow rate possible from a pump at a given head. Typically, head
is listed along the vertical axis (in meters or feet), and flow rate is listed along the
horizontal axis (in gallons per minute or liters per second).
To read a typical head-flow curve begin with the head requirement and read across to
the curve and then down to find the flow rate that would be produced at that head with
that power. As an example, performance curves for the A.Y. McDonald Solar Sub DC
water pumps are presented on the next pages, showing performance at various
voltages.
Derate Array Power To 80%

When looking at photovoltaic coupled pump curves it is important to realize that the
power available from a solar array in field conditions will be less than the standard rated
power of the array, due to heat and dust. Typically a derating of 80% is applied to the
standard peak power of the array to give the expected actual power available to the
pump in field conditions. Modules will operate typically at about 50-60 oC. so their
power is derated by about 12-15% (recall 1/2% power loss per oC above 25 oC.).
Added to this is expected loss due to dirt and dust in agricultural settings of about 510%, giving a total derating of about 20%. In other words power values indicated in
solar coupled water pumping literature should be increased by about 20% to give the
size of array at standard conditions needed for the performance indicated.

17-35
Example:
An A.Y. McDonald model 211012DK pump is being used to pump

water 100 feet above the surface of a well. If the pump is coupled
to a nominal 45 volt array (three modules in series operating near
Vmp of about 15 volts), read across from 100 feet to the 48 volt
curve (assumes each module operating at 16 volts) and then down
to find a rate of 350 gallons per hour (1.3 m3/hr). If a nominal 60
volt array were used (four modules in series), read across from 100
feet to the 66 volt curve (assumes each module at 16.5 volts) and
then down to find a rate of about 630 gph (2.4 m3/hr).
500
A.Y. McDonald
211012DK
450
400
Power needed
at pump
Head (feet)
350
300
90 V, 1190 W
250
75 V, 880 W
200
66 V, 680 W
150
48 V, 350 W
100
34 V, 170 W
50
0
0
100
200
300 400
Flow (gph)
500
600
700
.2+)3

$

17-36
A.Y.McDonald
211009DP
350
300
Power needed
at pump
Head (feet)
250
200
90 V, 1190 W
75 V, 880 W
150
66 V, 680 W
100
48 V, 350 W
50
34 V, 160 W
0
0
200 400
600 800 1000 1200 1400 1600 1800

Flow (gph)
.2+))(

$
The 211009DP model of pump has similar shaped curves but different values. This
model of pump cannot achieve the same total head, but can pump more volume.
Centrifugal pump curves always curve slowly downward, as seen in the A.Y. McDonald
curves. At lower head the same pump/motor will deliver much higher flow. Or
conversely, as head is increased the flow rate decreases drastically. This effect of
decreasing flow with increasing head is partially because at higher head. More of the
spinning impeller's energy is used to just hold up the column of water above it. Less
energy is available for moving the column. So flow rate goes down as head goes up.

17-37
Positive displacement type pumps show little decrease in flow rate with increasing head
compared to centrifugal pumps. This is because the column of water is held up by a
mechanical design of the pump (a flapper valve, diaphragm, or progressing screw, for
example) so little more energy is needed to hold up a large vertical column of water
than a small one. But positive displacement types tend to be able to pump less volume
than centrifugals types.
An interesting shape of curve occurs for a jack-type positive displacement pump. At
high head the nature of the curve is similar to positive displacement pumps. That is,
there is little effect of increasing or decreasing head on flow rate. But at low heads the
jack-type pump seems to be able to increase the flow more than the progressing cavity
types. The curves for the Solarjack SJA series of jack type pumps are shown as an
example. The power values given are power required at the pump.
Pump Jack Output Curves
900
800
700
Head (feet
600
500
400
300
200
100
0
0
2.5
7.5
10
12.5
15
Flow (gpm )
470 w atts
660 w atts
454.45

$
17-38
Pump Efficiency
Manufacturers may offer a pump model that can deliver to a required head at different
flow rates depending on the voltage or array power delivered. After a designer has
selected various pumps that can deliver to the specified head they must choose the
best pump among the choices. Perhaps the primary factor affecting this choice is the
efficiency of the pump at that particular head. Choosing a pump that operates at a
higher efficiency will mean that fewer modules will be required, so initial cost is lower.
The efficiency of a pump (including the motor) has a maximum at some value of head
and flow and is less anywhere away from that point. The efficiency is determined by
dividing the output power of the flow by the input electrical power to the motor.
Efficiency of Pump/Motor =
Output power of flow

Input electrical power from array
Ideal efficiencies of between 25% and 60% are possible although real world
performances will probably be limited to 40-50%. Added losses are due to heating in
the motor and friction and twisting losses in shaft type systems.
The formula for calculating efficiency divides the theoretical power needed to move
water by the actual power specified for the pump by the manufacturer.
Efficiency of Pump/Motor
Fc
Head x Flow x Fc
Electrical Power
conversion factor for units ( from MJ to kWh)

2.7
163.5
if units are:
0.003
0.189
meters and m3 per hour

meters and m3 per minute
feet and gallons per hour
feet and gallons per minute
You may have determined that a number of manufacturers pumps will reach your
required head and also pump enough water. Use this formula to calculate the
efficiency at the required head for each of the pumps. Compare the efficiencies and
the flow rates or volumes, and choose the pump that has the highest efficiency, unless
other factors override this choice. You will then be delivering water to your head
requirement with the fewest modules per volume.

17-39
Example:
Examine the 211012DK pump from A.Y. McDonald, and calculate

the efficiency for various voltage and power conditions. Pick
points all along the curves so as to see where the efficiency
reaches a maximum.
First examine efficiency along the 48 volt curve:
Efficiency @ 150 ft. =
=
=
150 ft X 160 gph X 0.003 factor

350 watts
.206 or about 21% efficient
350 watts
.30 or 30% efficient
350 watts
Next examine efficiency along the 66 volt curve:

=
=
=

17-40

680 watts
680 watts
680 watts
680 watts
The 211009DP pump cannot achieve as high of head as the DK

pump (320 feet maximum head Vs 490 feet), but it can pump at
greater flow rates.
Examine this pump at various points along its 66 volt curve and
compare to the DK pump:
=
=

680 watts
680 watts
680 watts
So the DP choice is more efficient than the DK pump, and both pumps show the
characteristic peaking of efficiency in the midrange of the pump head range.
The are both about the same efficiency at 150 feet, showing that different pumps may
be similar in efficiency at particular operating points, but that over their range one model
is more efficient than another.
It is this type of analysis that needs to be done to decide which pump is the best choice
for a particular head and flow rate. Of course other must also be weighed, and pump
efficiency may be overridden if these other factors indicate a more cost-effective
solution.

17-41
Exercise

$

) *+ ,- !!"!-.
/ 0% /1* 2""
)
3"4
-. "4

/1*

,- !!"!-.
5555555555
5555555555

5555555555 6
$
7 / 0% /1*
5555555555
5555555555

5555555555 6
$
8

$
9 $ : $9

!""

#
$%""

&'()*
)+,

+-!. /%"
&'(0 1
1
2222222222 3
4
2222222222 34
2222222222 345%"

22222222223 4 22222222226

17-42
Considerations and Calculations:

Designing a Solar Pumping System
Choosing the right solar pumping system for a project is not an easy task. It requires a
background in photovoltaics and water pumping and experience working with
photovoltaic pumps in the field. Up to a few years ago there were only a few PV pumps
on the market representing a narrow range of head and daily volume.
Today there are more diverse PV pumps on the market representing a broader range of
pump types and applications. Photovoltaic pumping systems are able to pump up to a
depth of 200 meters and can provide up to 400 m3 per day under optimum conditions.
The following steps are a guide to getting the right pump chosen for your specific
application.

17-43
Step
Action
Visit the site; determine the rough layout of the system and the
likely water source.
2
Obtain information on the well and water resources (i.e. well
diameters, depths, yields, drawdowns, seasonal river levels, and
monthly groundwater information). Consult with local authorities
and private welldriller who have knowledge of the area. (See
Supplementary Considerations for Choosing a Pump).
3
Estimate the likely daily demand for water for the applications
for each month. Be sure to consider water supply needs in the
future in for example a growing village population.
4
Gather monthly data on the solar insolation or peak hours
(convert from kWh/m2/day, MJ/m2, Langleys, etc.)
5
Calculate an estimate for peak flow rate, based on the average
daily insolation available.
6
Determine the total dynamic head, including drawdown and
friction losses if possible.
7
Choose a pump type (Positive Displacement, Submersible,
Floating or Surface Suction) that best fits the head, volume
requirements. (Recall the discussion previously on Optimum
Operation Ranges for Different Types of Pumps).
8
Choose motor/pump systems that will deliver to your required
head, and that can meet or exceed your flow requirement. (Use
manufacturers literature and help from your pump supplier).
9
Calculate the efficiency of the pump(s) at your required head.
Choose the pump with the highest efficiency that can deliver your
required flow rate.
10
Estimate the array size for each month, using the Array Size
formula presented next for this calculation, based on pump
efficiency and available peak hours of insolation.
11
Choose the largest array size as calculated above.
12
Compare manufacturers best case performance with your array
estimation. Adjust system size as necessary. Be conservative.
13
Reevaluate your decision. If you havent used the pump before,
get feedback from other users. Make sure that it fits other
considerations (i.e. computability with the water quality, ease of
maintenance, availability of support, etc.)
14
Complete the system design. (See Component Design
Considerations.
15
Procure and install system. (See System Installation).
,

17-44
Solar Pumping Array Sizing Formulas

Array Size
Use the following formula to calculate the size of solar array need to pump a given
amount of water to a given head. (This is a slightly different version of the efficiency
formula presented earlier).
Array Size (Wp)
H x V X Fc
P.H. x Fm x Ft x eff.
where:
H
V
Fc
=
=
=
P.H.
Fm
Ft
eff.
total pumping head in meters or feet

desired daily water volume in m3/day or gallons/day
conversion factor for units ( from MJ to kWh)
2.7
if units are: meters and m3 per hour
0.003
if units are: feet and gallons per hour
equivalent peak hours of daily insolation
(global solar radiation in the array plane (kWh/m2/day))
array/load matching factor
0.95 for pumps with maximum power trackers
0.9
for centrifugal
0.8
for others
temperature derating factor for array power loss due to heat
0.8
for warm climates
0.9
for cool
pump/motor efficiency at required head (wire to water)
Forming Array into Series and Parallel Modules:
# Series Modules
# Parallel Modules =

Pump nominal voltage

15 volts/module
Array Wp
# Series Modules X Module Wp
17-45
Peak Flow Rate

Use the following formula for estimating the peak flow rate that the pump must deliver
at the specified head to meet the daily volume requirement. (This was presented in the
earlier discussion of Primary Requirements for Choosing a Pump).
Flow Rate
Daily Volume Requirement

Peak Sun Hours
V
P.H.
Factors for converting insolation values into peak sun hours:

Unit of Daily Insolation
P.H. = Insolation times this factor...
kWh / m2
Langley (cal/cm2)
MJ / m2
Btu / ft2
1.0
0.01162
0.2777
0.003155
Pump/Motor Efficiency
Use the following formula for estimating the efficiency of the pump at the required head.
Efficiency
H x Flow Rate x Fc
Pump Electrical Power Requirement
Reality Check : Manufacturer literature usually quotes the efficiency as measured

under ideal conditions, as high as 50 to 60%. Field measurements of pump efficiency
are rarely ever above 40%. A pump/motor (wire to water) efficiency of 30 to 40% is a
realistic estimate based on field testing. Surface centrifugal are less efficient hence we
recommend using a 30% efficiency, centrifugal submersibles are in mid-range (35%
efficiency) and positive displacement (surface and submersibles) have the highest
efficiency (40%). Higher head may drive the efficiency of centrifugals even lower.

17-46
Example:
A water pumping system is proposed for a remote rural community.

The projected population for the town in five year is 300 people.
Step (2)
We are given the following information:

Water is in a well, and needs to be pumped a long way through
existing piping to a elevated storage tank for the village.
Vertical distance, static water to tank inlet pipe
Drawdown estimate
Total length of pipe
Diameter of pipe
30 m
7m
500 m
1.5 inch
PVC pipe
Step (3)
If we assume typical consumption of 55 ltrs/person/day, then the

requirement will be 300 people X 55 l/day = 16,500 liters/day, or
16.5 m3/day, or about 4340 gallons/day.
Step (4)
Local weather data is provided. The best month is June with 9.1
kwh/m2, and the lowest month is January with 5.7 kwh/m2. Use the
lowest month to be conservative. Convert to 5.7 peak hours.
Step (5)
Required peak flow rate
Step (6)
Total vertical head =

=
=
Friction head:
Volume / Peak Hours
16.5 m3 / 5.7 peak hours
2.89 m3/hour or 764 gph
drawdown + discharge
7 m + 30 m
37 m
assume at least one 90 elbow, which

adds 4.3 feet or 1.3 m to the total
equivalent length of pipe.
Equivalent length
=
=
500 m + 1.3 m
501.3 m
Use friction loss factor based on 800 gph

and 1.5 inch pipe, given as 1.74 feet/100 ft
or 1.74 m/ 100 m.

17-47
Friction Head
Total Dynamic Head
(501.3 m / 100) X 1.74

=
8.7 m friction head
=
=
=
total vertical + friction heads

37 m + 8.7 m
45.7 m
Round to 46 m.
Step (7)
A DC centrifugal submersible with maximum power tracking is

proposed for the project.
Step (8)
An AY McDonald 211009DP pump (with built-in maximum power

tracking) is selected to be used for the project. Our total dynamic
head of 46 m converts to about 150 feet. Reading across from 150
feet, the 75-volt configuration yields about 800 gph (or 3 m3/hr),
which exceeds our requirement of 764 gph. So we will work with
this curve. The manufacturer indicates that the pump requires 880
watts in this configuration.
Step (9)
Efficiency
Step (10) Array Wp
150 feet X 800 gph X 0.003 as Fc

880 watts pump power
0.41 or 41%
150 feet X 4340 gpd X 0.003 Fc

.
5.7 hours X 0.95 Fm X 0.8 Ft X 0.41
1100 watts
We choose to use a 75 watt module for this project.

# Series Modules
75 nominal volts
15 volts/module
5 modules in series
Array design:

1100 watts
5 series modules X 75 Wp
2.9 modules
Round up to 3 modules.
5 series X 3 parallel = 15 total
17-48
Exercise

*
$ ;
""" #$ !!"
;
& % #$ <
$ (

) =&4
$
) ,- -. -7 )
8

( $ -

/

:

* (

$
$( $

$

$

5555555555

) -.
5555555555
5555555555
5555555555
5555555555
5555555555 $
5555555555
) -7
5555555555
5555555555
5555555555
5555555555
5555555555 $
5555555555
17-49
Component Design Considerations

A typical pump installation (direct coupled) includes the following major components:
Array, Pump Controller, pump cable, pipe, grounding wire and safety rope.
Array
An important choice is whether a fixed array mount should be used or a sun position
tracker. Fixed structures are less expensive and can withstand extremely high winds.
But trackers can increase water output by 30 to 40% during summer months. The
amount gained depends upon your insolation levels, the time year, and your latitude.
Using a tracker could reduce in the number of modules needed, which could then equal
or exceed the cost of the tracker. Trackers, though generally very reliable, do add
complexity to the system and may need to be serviced overtime.
Pump Controllers or Power Conditioning Units for Pumps

Most PV pump manufacturers include now with their system some type of impedance
matching device so that the systems will operate at optimum power (max. power point),
matching the electrical characteristics of the load and the array. In these conditions
both the motor and the array can function close to their maximum efficiency over a
range of conditions and light levels. For instance, high current may be produced so that
the motor/pump will start in low solar irradiance levels. If the pump does not come with
a maximum power tracker or linear current booster it is advisable to consider one
particularly when using positive displacement pumps, which have to work against the
full pumping head on start-up.
Inverters
There are a variety of PV pump manufacturers who have developed PV pumping
systems using inverters to drive single or three phase AC motors in direct coupled
systems. These inverters differ from those used for battery systems in that they not
only convert DC to AC but include circuitry that involves some form of impedance
matching as well. Typically these are variable frequency inverters that control the
frequency and hence voltage of an AC pump to best match the output of a PV system.
At this stage they have proved very reliable and efficient, though they add additional
cost to the overall systems as compared to a direct coupled DC motor/pump.
If a standard AC off-the-shelf pump is specified, then it is necessary to have a standard
commercial inverter with a battery bank. Though less efficient, because of their fixed
voltage operation, designer can optimize the motor/pump subsystem for maximum
efficiency.
17-50
Water Level Sensors, Float Switches, and Wire

External water level sensors are used in wells with uncertain or marginal capacities
when the pump/controller does not provide dry running protection. If the output
capacity of the well is close to the maximum pumping capacity, sensors should be
installed to prevent damage to the pump. Some pump manufacturers, however, claim
that their pumps are not damaged by running dry. Float switches keep water storage
tanks from overflowing.
Wire used to connect the pump to the array must be sized properly to minimize line
losses. This is especially true for low voltage systems (12 and 24 Volts) where a small
voltage drop can cause significant power losses. Wire type should be selected for the
application (i.e. water submersion, sunlight, et al). The pump cable should be carefully
spliced and sealed when attached to the pump leads to avoid corrosion.
Disconnects, Grounding, Fuses

Pumping systems, like other PV systems, need to be safe. In the US this means
pumping systems must be designed to meet National Electric Code. The major
requirements for PV pumping systems are:
1.
2.
3.
4.
Proper grounding of conductors and equipment

Disconnects on the array and all equipment
Fuses or circuit breakers on all ungrounded conductors
Properly sized and insulated wires.

17-51
Water Pumping System Installation

Each pump type (e.g. surface centrifugal, pump jack, diaphragm) can require a wide
range of different tools, equipment, etc. to install. As a general rule surface pumps are
easier to install and require few tools. Setting pumps down deep wells (50 meters or
more) may require specialized pump setting equipment only available from the local
well driller. The majority of PV pumping systems failures is cause by controller or pump
problems. Proper installation will minimize the occurrence of these problems.
The most common installation problems are the following:
Varying Water Levels: The water level in a well may vary seasonally as well as hourly.
Most pumps will fail if run dry. Pumps must be mounted to keep the inlet below the
water level at all times. If the replenishment rate of a well is less than the maximum
possible pump rate then a level switch should be included to keep the pump from
running dry.
Protect the Pump Inlet: Sand is a primary cause of pump failure. If a pump is
located where dirt and sand may be pulled into the pump, a sand screen should be
used. Consult the manufacturers for recommended methods.
Ground the Equipment: Water pumps attract lightning because of the excellent
ground they provide. Avoid locating arrays on high spots. Consider erecting lightning
rods on high ground around the pump to attract lightning away from the pump. Ground
all equipment. Do not ground to the pipe because the ground would be interrupted
during maintenance.
Avoid Long Pipe Runs: Friction can significantly increase the head and thus the size
of the PV array.
Protect the Control Equipment: All electronic control equipment should be housed in
water-resistant boxes. All wire should be approve for outdoor use or installed in
conduit. Any cables submersed in water should be appropriate for those applications.
Protect the Well and Array: The solar array, well and pump (surface mounted) needs
to be protected from people and animals.
Protect the Pump from Freezing: Surface pumps and pipes need to be automatically
drained in order to prevent against freezing. A frozen pipe can cause a pump to burn
out in winter conditions.

17-52
Matching Array To Pump

Direct Coupled Arrays
It is not enough to connect an array with enough power to a water pump to achieve
peak performance. The array must be configured so that the peak power voltage and
current are closely matched to the motor load curve during the day. Presented below is
a comparison of the performance of a surface centrifugal pump connected to the same
size (power) array but configured in two ways: 3 in series by 10 in parallel (high voltage,
low current); and 2 in series by 15 in parallel (lower voltage, higher current). This will
illustrate the effect of array / pump mismatch. Later we present these two array
configurations connected to the same pump using maximum power tracking and how
this helps eliminate the problem.
This simulation uses solar data for Madras, India during the month of March. The array
is made of thirty 35-watt modules (peak array power of 1050 Wp) with the array tilted to
15 deg. facing south. The surface centrifugal used is the A.Y. McDonald 820309DS1.
Each simulation shows the array IV curve at 8:00 am, 9:00 am and Noon, and the total
daily water pumped is indicated. The pump/motor load curve is also indicated to show
how the pump motor operates the array.
In the first configuration, 3 series X 10 parallel, the array IV curve intersects the
pump/motor load curve at about 23 volts at 8:00 and about 31 volts at 9:00 and only at
about 32 volts at Noon. The maximum power points of the array are indicated by
crosses and are all at about 40-50 volts during these time periods. Therefore, we see
that this configuration of array is not being operated near any of its maximum power
points by the pump. The total output from the pump is 6836 gallons per day in March
with this configuration of the 1050 Wp array.
In the second configuration, 2 series X 15 parallel, the array IV curve intersects the
pump load curve at 30 volts at 8:00, about 33 volts at 9:00, and about 34 volts at Noon.
These voltages correspond more closely to the indicated maximum power points of the
array during these times and the pump is able to deliver 9322 gallons per day an
increase of 36%!
The second configuration of array, with more modules in parallel for more current, is a
better match to this particular pump/motor. At Noon the 3 X 10 and the 2 X 15 arrays
are delivering about the same power to the pump (660 and 673 watts respectively). But
early in the morning (and symmetrically in the late afternoon as well) the 2 X 15 array is
delivering more current to the pump than the 3 X 10 array (12 amps vs. 9 amps). The
2x15 array is operating the pump up its load curve higher and sooner than the other
configuration.

17-53
Direct Coupled to 3 Series X 10 Parallel Array

35
30
Current
25
noon 660 watts
20
Direct
Couple
6836 gpd
15
9am 429 watts
10
8am 210 watts
5
0
0
10
20
30
40
50
60
70
Voltage
56,* *
Direct Coupled to 2 Series X 15 Parallel Array

35
25
Current
Direct
Couple
9322 gpd
Noon
+ 673 watts
30
20
9am 589 watts
15
8am 356 watts
10
5
0
0
10
20
30
40
50
60
70
Voltage
!56,**

17-54
Maximum Power Tracking

The introduction of maximum power tracking (MPT) to this water pumping system
removes the mismatch problem. The peak power available from the 1050 Wp array is
translated into current and voltage that best operates the pump, independent of
whether the array is wired at 3 series X 10 parallel or 2 series X 15 parallel.
Both configurations are able to deliver 348 watts at 8:00, 561 watts at 9:00, and 760
watts at Noon. Both configurations allow the pump to produce 10,055 gallons per day
in March. Notice that this volume is greater than the amount delivered by the better
matched direct coupled system presented above. That system configured as 2 series X
15 parallel delivered 9322 gpd. The MPT is able to better match even that
configuration to the pump and deliver in this case about 8% more water in March. (The
more mismatched configuration of 3 series X 10 parallel is increased by 47% using
MPT!)
This illustrates the desirability of having MPT in solar water pumping systems. The
configuration of the array is not such a critical factor with MPT. In fact, the array can be
configured for maximum voltage, and therefore minimum current and voltage losses
and smaller wire size and subsequent lower cost.
This is one major reason why solar pump manufacturers typically supply a MPT unit
with their pumps, or even build the MPT circuitry into the motor. The solar water
pumping market is one area where a bit of added complexity is worth the price.

17-55

3 Series X 10 Parallel Array
35
30
25
Current
MPT
10,055 gpd
noon
760 watts
20
9am 561 watts
15
8am 348 watts
10
5
0
0
10
20
30
40
50
60
70
Voltage


35
noon
760 watts
30
Current
25
9am 561 watts
MPT
10,055 gpd
20
15
8am 348 watts
10
5
0
0
10
20
30
40
50
60
70
Voltage
'

17-56

We can also look at the effect of tracking the array to face the sun and examine its
effect on pump output. Again both array configurations are shown directly coupled to
the centrifugal pump with with no MPT. (If we examined the combined effect of MPT
and sun position tracking we would again see that the final output of the two
configurations would be equal. The MPT would translate the available power from both
arrays, each at 1050 Wp, to the pump equally well.)
Both systems show significant improvement in volume output with sun position tracking.
The more mismatched configuration (3 X 10) shows a 54% increase, from 6836 to
10,534 gpd, and the better matched system (2 X 15) changes from 9322 to 13,745 gpd
for a 47% increase. These increases are slightly different due to the way the array IV
curve intersects with the pump load curve.
The 3 X 10 configuration profits a bit more from sun position tracking because all the
effect of tracking translates into more current into the pump. This system is getting
more power at 8:00 with sun position tracking than it had even after 9:00 as a fixed
angle system.
The 2 X 15 configuration also profits from sun position tracking but the gain is less
because the pump is operating the array beyond its maximum power voltage. The gain
from 8:00 to Noon is small.
The decision of whether to use sun position tracking or MPT with solar water pumping
systems may depend on a number of factors, including the remoteness and harshness
of the location, the clarity of the sky (sun position tracking wont show a large gain with
overcast skies), and the cost of the two methods. Reliability and simplicity should still
be the guiding factors in deciding how best to enhance the output of your solar array.

17-57
Direct Coupled and Sun Tracking

35
30
Current
25
20
15
9am 600 watts
10
noon 660 watts

+
+
+
Tracking
Direct
Couple
10,534 gpd
8am 507 watts
5
0
0
10
20
30
40
50
60
70
Voltage
(
Direct Coupled and Sun Tracking

35
30
+ Noon
+ 673 watts
+
25
Current
Tracking
Direct
Couple
13,745 gpd
20
9am 700 watts
15
8am 667 watts
10
5
0
0
10
20
30
40
50
60
70
Voltage
)

17-58
(End of Chapter)

17-59
CHAPTER SEVENTEEN
WATER DELIVERY SYSTEMS
17-1
Why Solar Water Pumping

Advantages of Solar Water Pumping
Unique Characteristics of Solar Water Pumping
17-2
17-2
17-3
Markets and Applications

Irrigation
Livestock Watering
Residential Needs
Typical Usage Requirements
17-4
17-4
17-5
17-6
17-6
17-7
Battery Powered Or Direct Couple To Array
17-9
Terminology
Calculating Total Head
17-11
17-13
Types of Motors For PV Pumping Systems
17-17
Types of Pumps For Photovoltaic Pumping Systems

Centrifugal Pumps - Surface
Centrifugal Pumps - Submersible
Positive Displacement Pumps - Surface
Positive Displacement Pumps Submersible
17-19
17-20
17-22
17-23
17-25
Primary Requirements For Choosing A Pump
17-31
Supplementary Considerations For Choosing A Pump

Site, Well, and Water Considerations
17-33
17-33
17-34
Pump Output Curves

Derate Array Power To 80%
17-35
17-35

17-60
Pump Efficiency
17-39
Considerations and Calculations:

Designing a Solar Pumping System
Solar Pumping Array Sizing Formulas
17-43
17-45
Component Design Considerations
17-50
Water Pumping System Installation
17-52
Matching Array To Pump

Direct Coupled Arrays
17-53
17-53
17-55
17-57

17-61
Chapter 17 Answers

Refer to manufacturer's literature.

a. Required volume = 10 m3/day. Insolation = 6 kwh/m2/day.
Peak sun hours = Insolation X Conversion Factor = 6 kwh/m2/day X 1 = 6 peak hours.
Required Flow Rate =
Convert to gallons:
Daily volume
Peak Sun Hours
10 m3/day
6 peak sun hours
1.67 m3/hour X 264.2 gallons/m3 =
1.67 m3/hour
441.2 gallons/hour
b. Required volume = 5000 gallons/day. Insolation = 450 Langleys.

Peak sun hours = 450 Langleys X 0.01162 = 5.23 peak hours.
5000 gallons/day
5.23 peak sun hours
956 gallons/hour
956 gallons/hour
264.2 gallons/m3
3.62 m3/hour
Convert to m3/hour:
c. Required volume = 10,000 gallons/day. Insolation = 5.5 kwh/m2/day.

Peak sun hours = Insolation X Conversion Factor = 6 kwh/m2/day X 1 = 5.5 peak hours.
10,000 gallons/day
5.5 peak sun hours
1,818 gallons/hour
Convert to m3/hour:
1,818 gallons/hour
264.2 gallons/m3
6.88 m3/hour
17-1

Referring to the information for the A.Y. McDonald pump in Figure 17-12, we find the
curve for 680 Watts. At a head of 300 feet, the output is about 50 gph. This is equal to
50 gph 60 = .83 gallons per minute (gpm). We can calculate the efficiency as:
Efficiency of Pump Motor =
Head X Flow X Fc
Electrical Power
300 X .83 gpm X 0.189

680 Watts
.069 = 6.9%
Looking at figure 17-14 for the Solarjack SJA pump, we look at the curve for 660 Watts
at 300 feet of head. The output is about 6.25 gpm. This is equal to 6.25 gpm X 60 =
375 gph. The efficiency is then:
300 X 6.25 gpm X 0.189

660 Watts
.537 = 53.7%
There is a sizable difference in efficiency. The McDonald pump is a centrifugal type

and does not have as wide of range of efficiency. The Solarjack pump is operating
near its peak efficiency.

According to the data sheet for the Shurflo 9300, the flow rate at 100 feet is 103 gallons
per hour. This is equal to 103 60 = 1.72 gallons per minute. The motor current at
100 feet is given as 2.6 amps. Using a system voltage of 30 volts, the power is
Power
2.6 amps X 30 volts
78 Watts.
The efficiency is:

Head X Flow X Fc
Electrical Power
100 X 1.72 gpm X 0.189

78 Watts
.417 = 41.7%
17-2

The estimated insolation is 6.5 peak sun hours. The farmer's required flow rate is then:
4000 gallons/day
6.5 peak sun hrs
615 gallons/hr
a. Using the McDonald DK Pump:

We look at Figure 17-12. At 110 feet, it looks like the output is just about 615
gallons/hour using the 66V curve. However, to be conservative, we will use the 75V
curve. Using the 75V curve, the output is about 700 gallons/hour, and the pump
requires 880 Watts.
Efficiency
Array Wp

0.26 or 26%
110 feet X 4000 gpd X 0.003 Fc

6.5 hours X 0.95 Fm X 0.8 Ft X 0.26
1028 Watts required
We choose to use a 75-watt module for this project.

# Series Modules
75 nominal volts
15 volts/module
5 modules in series
=
Total array:
1028 watts
2.74 modules
rounded to 3 modules
The actual flow rate will be 700 gph X 6.5 peak hours = 4550 gallons/day. The output
per module is:
Output per module =
4550 gallons
15 modules
17-3
303 gallons/module
b. Using the McDonald DP Pump:

We look at Figure 17-13, we see that at 110 feet of head, we get about 850
gallons/hour using the 680-watt curve (66 V). This is sufficient for our daily
requirements.
Efficiency
Array Wp

0.41 or 41%
110 feet X 4000 gpd X 0.003 Fc

6.5 hours X 0.95 Fm X 0.8 Ft X 0.41
652 Watts required
Again, using a 75 Watt module# Series Modules
66 nominal volts
15 volts/module
4.4 modules in series
=
Total array:
652 watts
1.73 modules
The actual flow rate will be 850 gph X 6.5 peak hours = 5525 gallons/day. The output
per module is:
Output per module =
5525 gallons
10 modules
553 gallons/module
Clearly, the DP pump is much better suited for this application. At the required head /
flow combination, the efficiency of this pump is much higher than the DK pump. This
shows the importance of selecting the correct pump for the specific project
requirements.
17-4
Chapter Eighteen
Hybrid Systems
Multiple Energy Sources - Flexible
Approaches
Hybrid power systems include a combination of energy sources. The energy
sources might be solar arrays, wind turbines, diesel generators, biomass plants, or
others. Together, these sources provide the total energy required by the load.
These training materials consider mainly photovoltaic - diesel hybrids, although other
types of systems are briefly discussed.
The general purpose of the hybrid system is to get the best benefits of each
technology while reducing some of the disadvantages. A pure stand-alone PV
system, for example, requires very little maintenance but depends on the amount of
sun light for all its energy. On the other hand, a diesel generator can provide a
dependable amount of energy whenever it is required, but it requires fuel and
frequent maintenance. A hybrid system using both a PV array and diesel engine
can provide energy regardless of changing weather patterns with less maintenance
and fuel use than a dedicated prime-power engine. This illustrates how the features
and benefits of each energy source are blended in a hybrid system.
Hybrid systems can take many forms, and there is no single design that is best for
every situation. Instead, the hybrid designer must consider the different
technologies and components available and then choose a combination that best fit
the particular requirements. Good hybrid design consists of finding a balance
between many factors.

18-1
Hybrid Systems
Fundamentals of Hybrid Power

Systems
On-Demand Power Generation
One of the key distinctions of most hybrid power systems is the ability to produce
on-demand power generation. On-demand means that the system itself can
directly control the production of energy. In a simple stand-alone photovoltaic
system, there is no on-demand power source. The only power source is the sun,
and energy is only produced when the sun is shining. This energy is stored in a
battery, but if the battery runs low on energy, then the system must wait until the sun
rises and begins to produce additional energy. If a diesel engine and battery
charger are connected to this system, the engine and charger represent the ondemand source. If more energy is needed, then a control system starts the engine
and produces additional energy.
Examples of variable energy sources
Solar
Wind
Small scale hydro-power
Examples of on-demand sources
Utility (if reliable)
Diesel generators
Gasoline generators
Thermoelectric generators
Biomass plants
Most hybrid systems include an on-demand power source. This provides much
better control of energy production in the system and allows the system greater
flexibility to deal with changes in either the load or in the variable energy resource.

18-2
Hybrid Systems
Better Use of Renewable Energy

Production
Stand-alone systems using renewable energy are often designed around a worstcase situation. Because the renewable energy is variable, the system must be
designed to accommodate the period of minimum energy production (the worst-case
design period).
Example:
The table below shows horizontal insolation for Denver, Colorado. It

also shows the results of tilting the array (at 65) to optimize the worst
case month, as well as tilting the array (at 35) to optimize the annual
average. The system designed for the worst case month will be based
on 4.77 sun-hours each month, while the system designed for the
annual average will be based on the average of 5.73 sun-hours.
Month
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Average
Horizontal
Insolation
Tilt to Optimize
Worst Case
2.65
3.55
4.82
5.92
6.73
7.41
7.16
6.44
5.44
4.10
2.78
2.31
4.94
5.30
5.65
5.55
5.18
4.85
4.85
4.85
5.07
5.47
5.51
4.87
4.77
5.16
Tilt to Optimize
Annual
Average
4.69
5.36
5.93
6.23
6.41
6.75
6.62
6.42
6.19
5.54
4.47
4.18
5.73
Since stand-alone systems are designed with the worst case in mind, they are
oversized for the rest of the time. This results in extra installed capacity. The
additional energy in the peak periods can not be used and is wasted. In effect, the
whole system performance suffers from the worst month. If the difference is severe
between the worst case and the other months, this can result in wasted energy
equal to 100% or more of the design load.
Continuing the example from above, the system designed for the worst
case month has a worst case month (Dec.) of 4.77 sun-hours. The
best month is 5.65 sun-hours (February). So during February, the
system produces 5.65 / 4.77 = 118% of the energy necessary. If all
the energy over 4.77 sun-hours is excess, then an additional 18% of
the load energy is wasted.
18-3
Hybrid Systems
A properly designed hybrid system can make better use of the renewable energy.
This is because the hybrid system can be designed around the average case, not
the worst case period. In general, this results in a smaller renewable energy
component and less wasted power. It is important to remember that the renewable
energy source in a hybrid system is actively supported by another power source.
Exercise

Month
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Average
Horizontal
Insolation
Tilt to Optimize
Worst Case
6.50
5.71
4.49
3.47
2.43
1.84
2.10
2.85
3.71
4.92
6.08
6.34
4.20
4.85
4.94
4.43
4.19
3.55
2.84
3.22
3.83
3.98
4.43
4.76
4.63
4.14
Tilt to Optimize
Annual
Average
6.20
5.91
4.83
4.13
3.21
2.50
2.84
3.59
4.14
5.07
5.91
5.99
4.53

!
"
#
$
b)

!
"
#
$

!
!

"
%

!

"
#
$
18-4
&
Hybrid Systems
Load Matching
Another advantage of hybrid power systems is a better match between power
production and the load. This can be as simple as being able to turn on a power
source for peak loads. A carpentry shop may use a small amount of power during
normal operation, but need more power to run a large lumber saw. It makes sense
to provide the baseline amount of smaller power for most of the day and then run an
on-demand source specifically to power the saw. Similarly, a remote village may
have a large peak demand for lighting and entertainment in the evening hours.
Load Size (kW)
Step Load
12
16
20
24
Time

Sometimes the size of the load indicates that a hybrid might be useful. Large
applications require large current flows and higher voltages. These can be costly to
provide using only solar, for example.
For village or facility power systems, there is often a collection of different types of
load equipment. Different types of equipment may be suited to different sources of
power. Rather than try to supply all the loads with a single power source, a hybrid
may match different resources to each individual load. PV, for example, is very well
suited for water pumping for humans and agriculture. The source of power (the sun)
tends to be most available during hot, dry periods. Other loads may be better suited
for a diesel generator. Equipment that has high start-up surges for example may run
well with a generator. For example, an ice-making unit, which has a large
compressor, has high starting currents that can be supplied by a generator. 3-phase
loads are also typically run from a generator. While such loads can be run from
photovoltaics and appropriate power-conditioning units, it may be simpler to provide
an engine for a specific load.

18-5
Hybrid Systems
Applications of Hybrid Systems

Hybrid systems are generally used for larger loads than pure stand-alone systems.
Generators represent a compact source of high power output. Since the generator
offsets some of the photovoltaic power, the same size array can serve a larger load
in a hybrid system. For smaller systems, the cost of generators and additional
controls tends to outweigh any real economic advantage to a hybrid system.
Because of the inherent flexibility in a hybrid approach, hybrid systems are also used
when there is variability in the load. Any time there is a base load with a limited
duration of higher loads, there may be a good match for a hybrid power system. If
the system must serve a mix of loads, with different characteristics, it may also be a
good application for a hybrid system.
Typical Size Ranges Hybrid systems can be cost effective in the range of 10 kWh
to 220 kWh. Note that some of the other advantages of hybrid system may result in
their use outside of this range. Conversely, if the load is within this range, it does
not necessarily imply that a hybrid system is the correct solution. Other factors such
as alternative sources of power, availability of technical support, renewable
resources, and customer preference must be taken into account. If the utility grid is
reliable and easily accessible, for instance, there may be very little justification for a
hybrid based only on the load size.
Telecomm Power Certain types of telecommunications equipment can require a
high level of power (1-4 kW or more). While diesel generators power many
telecommunications loads, this becomes a problem in very remote areas.
Maintenance at such sites is very difficult and expensive. Some sites may only be
accessible by helicopter, making it impractical to bring in large amounts of fuel and
oil. While some of these sites become stand-alone systems, for others a hybrid
system is a good compromise between initial cost and on-going maintenance.
Village Power Small rural communities may be candidates for hybrid power
systems. In some areas of the world, these villages are difficult to electrify by
traditional grid-extension. The size of the loads and the variability caused by many
households can justify a hybrid system. In addition, the reduced maintenance
requirements in comparison to a diesel-only system are a benefit when skilled
technicians may not be available.
Facility Power Some sites are dedicated facilities for a specific use. This could
be for commercial use such as an ice making plant for fishing, for governmental use
such as clinic or a recreational use such as a hotel. These types of facilities often
have higher load requirements than a simple stand-alone system can provide. In
addition, there can be sufficient load diversity to make use of different power
sources. There may also be additional reasons to consider a hybrid power system
over a diesel. A hotel promoting the ecological beauty of its setting, for example,
would not want to have a diesel running 24 hours per day.

18-6
Hybrid Systems
Advantages of PV/Diesel Hybrid Systems

The purpose of designing a hybrid system is to utilize the best advantages of the
various technologies. Hybrid systems have several advantages over either standalone PV systems or prime power diesel systems:
Better Use of Renewable Energy Compared to a stand-alone renewable energy
system, a hybrid can make better use of the renewable energy received. Since
hybrids can be designed for the average case, instead of the worst case, they can
more evenly use the energy through the year.
Higher System Availability In a normal stand-alone system, there is always some
chance that there will not be enough energy to support the design load. This is
because the incoming energy follows a statistical pattern, and there is always a
small chance of a long period of very poor sun. The ability of a stand-alone system
to support its load under these variable conditions is called availability. It is
important to recognize that availability is different from reliability. Typical
commercial PV systems have an availability of approximately 99.7% or higher.
However, to get higher availability even more PV must be added. A hybrid system
with an on-demand power source avoids this problem, since the on-demand source
can cover the small amount of time of poor weather.
Lower Diesel Maintenance than Prime Power Maintenance on a diesel engine
is directly related to the amount of run time. Therefore an engine in a hybrid system
that runs half the hours compared to a prime power engine will need only half the
maintenance. In addition, poor engine loading can also result in maintenance
problems. Prime power diesels are often loaded at some fraction of their full rated
capacity. Ideally, an engine should be loaded between 85-95%. However, because
of peak loads, available equipment or uncertainty about loads, some engines
operate at a much lower loading, 75%, 50% or even less. This results in a number
of problems, including coking, glazing, slobbering, which can result in higher
maintenance costs.
Higher Fuel Efficiency In addition to the maintenance problems presented by low
loading, running a diesel at significantly less than full load also results in poor fuel
efficiency. While a fully loaded engine uses 100% of the rated fuel consumption, an
engine that is 10% loaded may still require as much as 50% of the full consumption.
Therefore, the amount of energy produced per gallon or liter of fuel will be much
lower. In a hybrid system, the loading of the generator can be controlled for higher
fuel efficiency.
Reduced Capital Costs Standalone PV systems require very little maintenance in
the field, but they have a higher initial cost. Prime power diesels on the other hand,
may be very cheap initially, but cost more in terms of continued fuel and
maintenance costs through the life of the system. A hybrid system may cost less
than the equivalent stand-alone system and use less fuel than a simple diesel-only
system.
18-7
Hybrid Systems
Better Flexibility for Loads Hybrid systems offer great flexibility in their ability to
meet changing loads. Since the generator can provide additional energy by simply
running longer, it is easier for a hybrid to accept an increase in the loads. Stand
alone PV systems can not handle a significant increase in the load above the design
load, and the only way to allow for growth is to over-size the system for the initial
load. Similarly, a prime power diesel can not easily accept load growth and must
either sized for the actual load or run poorly at low load levels initially.

18-8
Hybrid Systems
Disadvantages of PV/Diesel Hybrid

Systems
Like any other technology, hybrid systems have disadvantages. To properly specify
and design hybrid power systems, these must be understood as well. Compared to
stand-alone PV systems and/or prime power diesel systems, hybrids can have the
following disadvantages:
Higher Control Complexity The consequence of using multiple power systems
together and utilizing more sophisticated operation is increased control complexity.
In addition to being able to control and monitor each separate power sub-system,
the controls must also allow for the interactions between sub-systems. However, the
introduction of microprocessor controls increases both the functionality and reliability
of the control system.
Greater Initial Engineering Hybrid systems are relatively more complex than
either stand-alone systems or prime power diesels to design and engineer. The
initial engineering work to create and manufacture a hybrid requires a broad
knowledge of various components, as well as the ability to perform system
integration. It can often be difficult to determine what solutions are acceptable and
which will cause further problems. Experienced firms with the right capabilities can
address the many tradeoffs in hybrid design to assure that the system is designed
correctly.
More Frequent Maintenance than Stand-Alone Systems Although hybrid
systems can require significantly less maintenance than a prime power system,
hybrid system do require more maintenance than a stand-alone PV system. This
maintenance is essential for the operation of the hybrid. Since batteries are cycled
harder, cell watering and check-ups are very important. The addition of a diesel also
increases the number of maintenance tasks as well as the expertise required for the
technician.
Higher Technical Expertise Needed for Troubleshooting Problems with hybrid
operation can require higher technical competence to identify and solve. This may
be a problem for locations where there is a shortage of skilled technicians. It can
also present difficulties if the end-user of the system does not have access to those
skills (in a remote village for example). This means that the repair and service of the
final system should be addressed before installation.
Increased Pollution and Noise A photovoltaic power system uses clean, silent
power -- the sun. Introducing a diesel into such a system also brings unwanted
pollution and noise. Diesel engines produce carbon dioxide (CO2) and other
undesirable emissions. The noise of the engine operating may disturb nearby
people or bring unwanted attention to the system. Compared to a prime power
diesel, however, both noise and pollution can be much reduced, and the system can
operate more efficiently.
18-9
Hybrid Systems
Comparisons of Various System Types
Initial Cost
Recurring
Costs
Maintenance
Fuel Use
Load
Flexibility
System
Complexity
Use of
Renewables
Pollution
Noise
Hybrid
System
High
LowModerate
Moderate
LowModerate
High
Stand-Alone
PV
High
Low
Cycle
Charge
Moderate
Moderate
Prime Power
(Diesel only)
Utility
Low
Low - High
High
Low
Low
N/A
Moderate
Moderate
High
High
Low
N/A
Low
Moderate
Low-High
High
High
Low
Moderate
Low
Low
High
ModerateHigh
Low
Low
N/A
N/A
N/A
High
ModerateHigh
High
High
Low
Low
Moderate
LowModerate


18-10
Hybrid Systems
Hybrid Configuration and Operation

Power Sources
As described previously, the key characteristic of a hybrid system is that it includes a
mix of power sources. This combination of power sources gives the system more
flexibility than if it relied on only a single source. Furthermore, using more than one
source may make the system less expensive than if it had to rely on a scarce
renewable resource during part of the year. This section describes some of the
various power sources available in hybrid systems.
Photovoltaic Arrays
PV is well suited for use in remote areas, since it is extremely reliable and requires
very little maintenance. Many telecommunications loads use PV as the sole power
source, and hybrid systems are a logical extension of this application of PV. The
difficulties with photovoltaics are the initial cost and the requirement for unobstructed
land to place the array. Nevertheless, the advantages of PV make it one of the
more common renewable energy sources used for hybrid systems.
PV Arrays
Advantages
Very low
maintenance
High reliability
Long product life
Disadvantages
High initial cost
Requires land with
no shading
Seasonal resource
Engine Generators
Engine generators represent the single most common on-demand power source.
They are very common, compact in size and inexpensive to purchase. Their
disadvantages are that they require on-going maintenance and fuel, and produce
noise and pollution.
Engine Generators
Advantages
Disadvantages
Widespread use and Require on-going
support
maintenance and
fuel
Small size
Produce noise and
Low initial cost
pollution

18-11
Hybrid Systems
Wind Turbine Generators

Wind Turbine Generators (WTGs) convert wind energy into electrical energy via a
rotating electric generator. WTGs are commercially available in sizes ranging from
50 watts to many hundreds of kilowatts. The larger units (> 100 kW) are typically
used in utility-scale wind farms. They are not suitable for use in remote power
systems, although there are beginning to be exceptions to this. The medium size
units (5 kW to 50 kW) are often suitable for remote power use, and the small ones
(50W - 5 kW) are typically configured specifically for charging batteries.
The WTG may be connected to the hybrid system on either the AC side or on the
DC side. If the unit is connected on the AC bus, it feeds AC directly into the grid,
which can cause dramatic fluctuations in voltage and frequency. Unless a very
sophisticated control system is used, the wind contribution in this type of system is
limited to a small percentage of the overall energy.
Alternately, the AC power delivered out of the WTG can be converted to DC and
used to charge batteries. The batteries exert a dramatic stabilizing effect on system
performance, although, they too can be overwhelmed with drastic fluctuations. In a
system with multiple generating sources, it is important to coordinate charging
between the various sources. For example, if a WTG is charging the batteries at
their maximum rate during a windy period, and the sun comes out bright and strong,
the net effect could be to overheat the batteries due to the excessive current.
Similar coordination is needed during diesel chargingthe generator cannot be
stopped and started continually just to compensate for fluctuating wind speeds.
There are a number of specific points to note when considering WTG hybrids:
First, wind energy is much more sensitive to wind speed than solar energy is to
irradiance. In fact, WTG output theoretically varies as the cube of the wind speed
(Power V3). This has benefits and disadvantages. The disadvantage is that if you
had estimated the wind speed to be somewhat higher than it really is, the net energy
output of the WTG would be dramatically less than predicted. On the other hand, in
a good wind climate, the output of the turbine is so large that other generating
sources cannot even compete economically. In general, if the average wind speed
is greater than 4 meters per second, a WTG will be less expensive in a hybrid
system than a PV array, even in a good solar climate. However, because of the
wide variability of wind speed on a daily, seasonal and annual basis, it may still
make sense to have both solar and wind generators in a hybrid system.
Second, WTGs are much more sensitive to location than PV arrays. It is not too
inaccurate to assume that the insolation at a particular site will be relatively similar to
that actually recorded for nearby sites (airports, city centers, etc.). However, the
wind speed is very dependent on specific and highly localized factors, such as the
location of nearby hills and trees. The wind speed is also dependent on the height
of the turbine, so an anemometer at 3 meters will not accurately represent the

18-12
Hybrid Systems
energy available to a turbine on a 20-meter tower. The best advice is to monitor the
wind speed at a site for at least one year before making final decisions on WTG
installation. There are a number of excellent wind monitoring kits which can help
you perform this task.
Third, because of the extra energy farther off the ground, it pays to install as tall a
tower as possible. Care must be taken not to shade the PV array with this tower.
Summary of Energy Considerations with Wind Turbine Generators
1. Wind Turbine output depends very heavily on wind speed
2. Wind speeds can vary significantly over small areas
3. Height increases the amount of energy available
In summary, wind turbines can form an excellent generating source in a hybrid
system. However, since the output is so sensitive to the wind speed, and since the
wind speed is so sensitive to the particular climate, it pays to make a careful
resource assessment before committing large amounts of investment in wind
turbines.
Wind Turbine Generators

Advantages
Disadvantages
Does not require any Requires additional
fuel
site work to install
Potentially lowest
Requires additional
cost source of
maintenance
renewable power
Usually requires
extended period of
data monitoring
Thermo-Electric Generators
Thermo-electric generators (TEGs) operate based on the principle that electricity
can be produced from the temperature difference between dissimilar metals. A fuel,
usually natural gas, is burned in the presence of a pile or stack of dissimilar metals,
producing a continuous DC current and voltage. Commercially available units have
been widely used in the gas pipeline industry, where there is an abundant supply of
cheap gas for fuel.
The basic principles of TEG's result in a number of specific advantages and
disadvantages. TEG's have no moving parts and there is no mechanical wear as
compared to an engine generator. TEG's can be started after a long shutdown,
which makes them suited for seasonal use. The thermoelectric affect works better
and is more efficient in colder weather. A TEG does, however, require fuel and fuel
storage. The efficiency is very low compared to other types of generators, typically
4% or less. The advantage of working well in cold climates becomes a
disadvantage in a hot climate. Finally, typical commercial units are only in the range
of 10 - 1,000 Watt output, limiting their practical use to smaller systems.
18-13
Hybrid Systems
Thermoelectric Generators
Advantages
Disadvantages
No moving parts
Requires fuel
Ability to start after
Low efficiency
long shutdowns
Do not work well in
Excellent
hot climates
performance in cold
Limited size range
weather
TEGs can be used in hybrid systems with small PV arrays. The best use for these
hybrids is in climates where the insolation varies dramatically over the course of the
yearsuch as far northern climates. Rather than oversizing the PV array by a factor
of three or more, a TEG can be used to supply some or all of the load during the low
insolation months. Since these months also tend to be the coldest, the TEGs are
most efficient. Finally, the waste heat from the TEGs can be used to supply space
heating for batteries and telecommunications equipment.
Micro-Hydro
Hydroelectric power is generated from the energy of falling water. A number of
small, commercial hydro plants are available in the 3 kW to 50 kW range, which can
be incorporated into a hybrid system. Since the output of a hydro plant is relatively
continuous, it lends itself to an AC bus type of system, with renewable or diesel
generators making up the extra energy required by fluctuating loads or seasonal
changes.
Power produced by a micro-hydro site depends directly on the height difference
(head) from the supply to the turbine, as well as the amount of flow. Since the
amount of water can vary seasonally, proper design of a micro-hydro system often
involves at least a year of data on the water flow.
The economics of this class of system are widely variable and depend mostly on the
annual availability of a stream, and on the ease of building and maintaining the
turbine. Clearing of debris and insuring maintenance are also concerns of
hydroelectric generators.

18-14
Hybrid Systems
Micro-Hydro
Advantages
Disadvantages
Requires no fuel
Depends on specific
sites
No pollution
Requires additional
May provide
civil works at site
economic power
Typically requires
some resource
monitoring
Maintenance
concerns
Utility / Grid
Although systems with a utility supply are sometimes not considered to be true
hybrids, there are times when the overall system behavior resembles a typical hybrid
system. These can be instances when there is utility power present, but it is not very
reliable. In such circumstances the utility is treated more like a variable energy
source rather than an on-demand power source.
Exercises

'

()*
!

* ()

*


18-15
Hybrid Systems
Bus Configurations
One of the basic distinctions in hybrid design is the bus configuration. Broadly
speaking, this is the general architecture of the system. Hybrids are usually
categorized as DC bus systems or AC bus systems.
DC Bus Hybrids
In a DC bus system, a generator / rectifier combination charges the batteries. The
center of the system is considered to be the common DC bus where the battery,
rectifier and PV array are all connected together. This is the most common hybrid
system configuration. This system was developed from the idea of cycle charged
diesel generators, and it is ideal for telecom hybrid systems, since the loads are
typically DC. Note that a DC bus system can supply AC loads by adding a separate
inverter to provide AC power.
The PV array charges the battery based on the available solar energy. The DEG is
normally only started when the batteries reach a low voltage state and need to be
recharged. The DEG is connected to a dedicated battery charger that supplies DC
current. Sometimes there may be an emergency transfer switch so that the
generator can power an AC load directly. Under normal conditions, however, the
engine only runs to charge the battery and to provide the additional energy needed
for restore the battery to full charge.
Pictorial Representations of DC bus configurations
DEG
Rectifier
AC
PV
Array
DC
(+)
(-)
Battery
Bank
DC
Loads


18-16
Hybrid Systems
DEG
Rectifier
AC
PV
Array
Transfer
Switch
DC
DC
AC
(+)
(-)
Battery
Bank
Inverter
AC
Loads
DC
Loads

! "
AC Bus Hybrids
AC bus hybrids utilize the generator to power some or all of the loads. The batteries
are charged using a bi-directional power conditioning unit that serves as both
inverter and rectifier. The focus of the system is considered to be the common AC
bus where the generator, inverter and AC Loads are all connected together. Since
the generator takes a much more active role in the system, these are also
sometimes called interactive hybrids. An AC bus hybrid may have some DC loads,
but it is more common to have only AC loads.
The PV array still supplies energy to the battery depending on the solar insolation.
The generator however, can now charge the batteries and supply power to the load.
The bi-directional inverter can take power from the battery bank to supply AC power,
or can draw AC power from the generator to charge the batteries. Some bidirectional units are capable of operating in synchronized or parallel mode where
both the inverter and generator can supply the load together. This allows for great
flexibility in handling peak loads.

18-17
Hybrid Systems
DEG
AC
Loads
Inverter/
Rectifier
AC
PV
Array
DC
(+)
(-)
Battery
Bank
DC
Loads

"
Exercises

#
+

,'-!'-

'-!'-.-
/-!-
0%

1
2

"

$
0%

- -
'

18-18
Hybrid Systems
Hybrid Operation
Daily Operation
A hybrid system will go through shorter and more intense cycles than a stand-alone
system. While a stand-alone system may remain at almost a full state of charge for
long periods of time, most hybrid systems will cycle through the battery every couple
of days. Rapid changes in the battery result caused by a sequence of array and
rectifier charging.
The figure below illustrates a sample hybrid system's performance over 4 days. The
PV array and rectifier both contribute about half the total energy in the system. The
battery bank provides about 1.5 days of storage. This results in a system with rapid
cycles and heavy utilization of the battery. Three lines are shown, one depicting the
average battery state of charge, one line showing the energy provided by the PV
array, and one line showing the rectifier contribution. For this example, the load is
assumed to be constant.
D ay 1
D ay 2
D ay 3
D ay 4
B a tte ry
P V A rra y
R e c tife r

%&'()
The figure shows the battery initially at nearly full charge. During the first day, the
battery state of charge falls as the load draws power. The PV array replaces some,
but not all of the energy used by the load. On the Day 2, the load again draws down
the battery, with some of the loss being replaced by the PV. At the end of the
second day the battery is discharged sufficiently to bring the rectifier on line. The
rectifier provides a large amount of current in a short interval of time, bringing the
battery back nearly to full charge. As the rectifier shuts down additional PV energy

18-19
Hybrid Systems
finishes the charging. At the end of the third day, the load keeps drawing power
from the battery. On the fourth day, the load continues to draw down the battery
with some replacement from the PV array. This is very similar to Day 1 and we can
see that the system will continue to run in cycles with the battery eventually reaching
a low point, causing the generator to start.
Performance Factors
The pattern of charge and discharge cycles shown above is typical of hybrid
systems. Obviously, the cycles may not be identical each day because of several
changing factors:
1. The load may change from day to day.
2. The battery may be at a different state of charge at the start of the day.
3. The weather may change, resulting in more input from the PV array.
4. The generator may be programmed to run differently, such as providing an
equalize cycle.
The annual performance of a hybrid system is typically measured by the following:
Total Energy Production
Contribution (%) by each source
Number of battery cycles
Generator run hours
Number of generator starts
Generator fuel consumption
Each of these factors is determined by the size of the various system components.
Some of the relationships may not be obvious at first. However, if you stop to think
about what is actually happening within the system, then the explanation should be
apparent. Below are some general guidelines for most hybrid systems. Note that
computer programs must be used for detailed performance results.
Battery Size The size of the battery affects how many cycles the battery goes
through. The smaller the battery the more frequently the battery will be charged
and discharged.
Generator / Rectifier Size The sizes of the generator and rectifier mainly
determine how long the generator will run for each charge cycle. They do not affect
how much of the total energy comes from the generator. So, to keep the annual run
hours low choose a relatively large engine.
Array Size The size of the array does affect how much of the annual energy is
supplied by PV. Obviously the bigger the array the more energy it will produce.
Inverter Size In a DC bus system the size of the inverter does not affect the
annual performance directly. In an AC bus system, the size of a bi-directional
inverter affects the generator run-time similar to the rectifier size.

18-20
Hybrid Systems
Different Roles of Generators

Although generators are frequently included in a hybrid system, there are several
different ways that a generator can be used. The operation of the generator can be
simplified into the following types:
Prime Power The generator runs whenever the load requires power. Some prime
power applications have an engine running continuously throughout the
entire year.
Back-up Power Another power source is the primary means of supplying the load.
The generator runs only rarely, and it is only intended to handle
excessive demand.
Cycle Charge A pure cycle charge system consists of only a generator, battery
charger and battery bank. The generator provides all the power, but it
runs only when the battery requires recharging.
Emergency Power The generator is not intended to run under normal
circumstances. In some systems the emergency generator is activated
by manual operator control.
In actuality there can be a range of energy contributions and run-times, so that

hybrid systems will tend to fit somewhere in between these types. While each
approach may be suited to a particular set of circumstances, it is important to
confirm the users expectation for how the system will operate.
Cycle
Charge
Prime Power
Emergency
Back-Up

*+

18-21
Hybrid Systems
Exercises

,
0%

-
# !

#3-

4
/5

,66&
7

#3-

895

2

2

,66& 7

"

"

"

18-22
Hybrid Systems
Component Selection
PV Array
The photovoltaic array for a hybrid system must meet the same strict requirements
as a stand-alone application:
Component
Selection
36 cells for
adequate
voltage
Rugged
design
High energy
output
15+ year
warranty
1. Sufficient Voltage for Battery Charging A solar module should

have at least 36 cells to ensure sufficient voltage for battery
charging. Modules with fewer cells may not reach a high enough
voltage in hot climates.
2. Rugged Design Like stand alone systems, hybrid systems are
often in very remote locations, with harsh environments and
severe weather.
3. High Energy Production Consider a high output module, like a
single crystal cell design that produces high amounts of energy
even under low light levels. That way the investment in the array
produces the most benefit.
4. Warranty Modules from major manufacturers now include

warranties of 15 years or more. Since a hybrid system is an
investment for a long period of time. The modules should be covered by a
manufacturer's guarantee for a significant portion of that time.
Test
certification
5. Test Certification Most modules are now tested to a variety of national and
international standards. In the US, the Underwriters Laboratory has a test for PV
modules. In Europe, the Commission of European Communities (CEC) and
International European Community (IEC) have standards for module testing.
By choosing a reputable supplier you will ensure the high quality of the product as
well as getting strong support.

18-23
Hybrid Systems
Battery
The battery bank is a key element in a hybrid system. The battery is literally at the
center of the system, and the quality of the battery can influence the entire system
performance. A well-designed system makes good use of the battery storage. A
poorly designed system suffers when the battery does not perform. The battery
bank also represents a significant cost in the system and is worthy of extra attention.
Component
Selection
Choose a
deep cycle
battery
Don't use SLI

or float type of
batteries
Take care with
VRLA
batteries
Choose a Deep Cycle Battery

Because the battery in a hybrid system is used much more heavily
than in a standalone PV system, it is important to select the proper
battery. A heavy duty, cycling type battery is most appropriate, with
typical specifications requiring a cycle life of 1500 cycles to 80%
DOD. This will allow the battery to reach 7-10 years of life, even
with cycles every 2-3 days. Forklift batteries are appropriate, even
though some high antimony types may only have five-year design
lives. Higher quality tubular plate low antimony telecom batteries
can last up to twice as long, but can also be significantly more
expensive.
Batteries designed for float applications and UPS operation are definitely not
appropriate, and neither are SLI (auto) batteries. Marine type cycling batteries are
a marginal choice, while wheel chair and golf-cart batteries may be used in a careful
design. Care should be taken with sealed (gelled or absorbed electrolyte) VRLA
type batteries, since most of these were developed for low cycling applications.
They are also much less forgiving of the variable rates of charge and discharge
experienced in a hybrid system.

18-24
Hybrid Systems
Engine / Generator
The engine generator is a critical component of a true hybrid system, since it
supplies a significant part of the energyit is not just a backup. There are a wide
range of choices for construction, fuel and accessories that must be considered.
Component
Selection
Prime Power vs. Stand-By

Use an engine
rated for
prime power
Be sure the
engine is
designed for
remote starts
Use air-cooled
engines for
small systems
Choose fuel
type based on
application
Because the engine in a hybrid system may be running for many

hundreds or thousands of hours per year, it should be specified as a
prime-duty engine rather than a standby-duty engine. Prime power
engines are rated for continuous duty environments. Stand-by
engines are only expected to run on an infrequent basis. Using a
stand-by engine in a hybrid application may result in excessive wear
and early failure. Note that some manufacturers will rate a specific
engine for both prime duty and stand-by, with the prime duty rating
being lower than the stand-by rating. In a standard hybrid system,
the engine should last the life of the system without replacement.
Air-Cooled vs. Water-Cooled
Add proper
accessories
for the system
Generators have two types of cooling systems: air-cooled and water

cooled designs. Air-cooled engines have the advantage of simplicity
and ruggedness. Water-cooled engines are smaller and quieter for a given power
output. Air-cooled engines require proper ventilation for both intake and exhaust
airflow. Water-cooled engines require the availability of clean water and proper
technician training. Because reliability is typically more important than size and
noise, air cooled engines are preferred for remote power systems, at least in the
small sizes.
Choice of Fuel
The choice of fuel for the generator is of primary importance. Generators using
diesel fuel are known as diesel electric generators (DEGs) and are very common
around the world. Diesel fuel is commonly available (although the quality can vary
significantly). Service and trained diesel technicians are also comparatively easy to
find, making diesel generators a good choice for many types of systems. Diesel
gensets are often rated for 20,000 to 50,000 hours of operation. In cold weather
(approaching 32 F or 0 C) diesels can have starting problems although these can
be addressed through the addition of glowplugs and appropriate grades of fuel.

18-25
Hybrid Systems
Diesel Fuel
Advantages
Disadvantages
Readily Available
More expensive than
gasoline engines in
Difficulty with cold
small sizes
weather starting
Noise and pollution
Fuel can be stored
for several months
Gasoline/petrol generators are very lightweight and compact for the amount of
power they produce, making them ideal for automotive applications and light duty
contractor generators. They also tend to be the least expensive generators
available. Their two major limitations are fuel storage and design life.
Gasoline/petrol is much more difficult to store in large quantities because it is more
volatile (meaning it burns easier) than other fuel options. This makes it prone to
explosions if not stored properly, or if there is a leak in the system. In addition these
engines are typically rated to provide only about 1,000-2,000 hours of engine life
(compared to 20,000-50,000 hours for diesel gensets) making them suited only for
small backup systems such as cabin power systems.
Gasoline (Petrol)
Advantages
Disadvantages
Readily Available
Hazardous to store
large quantities
Usually technicians
are available
Fuel can not be
stored for long
Engines are small
periods
and inexpensive
Engines typically not
rated for long life
Another common fuel choice for engines in commercial hybrid systems is liquefied
petroleum gas (LPG). These are sometimes modified DEGs. While LPG is less
efficient in terms of fuel consumption compared to gasoline, it is much cleaner and
requires less engine maintenance interval, thus justifying its use in many situations.
Certain environmentally sensitive locations such as National Parks will only allow
LPG generators because of concerns regarding pollution. LPG is volatile and
requires specialized fuel feed systems.

18-26
Hybrid Systems
Liquefied Petroleum Gas

Advantages
Disadvantages
Very clean burning
Not always easily
available
Less engine
maintenance
Specialized
technicians
Low pollution
Requires special fuel
systems for very cold
climates
Natural gas is occasionally used for generators, particularly on a pipeline or other

location where there is essentially a free supply. Similar to LPG, natural gas can
be very clean burning. The reliability of natural gas engines can also be quite good.
However, unless there is a supply of fuel nearby, this fuel is very seldom used for
remote hybrid applications.
Natural Gas
Advantages
Disadvantages
Very clean burning
Typically available
only in specific
Less engine
applications
maintenance
Specialized
Low pollution
technicians
Remote Starting
It is very important to ensure the engine selected has remote start capability (unless
an operator will be on the site at all times). The engine start circuit must be
specifically designed to allow remote starting, otherwise it will not operate properly in
a hybrid system. An engine for a hybrid system must also include an automatic
choke that will automatically adjust the air mixture between starting and normal
operation.
Even if the engine can be remotely started, it is also important to determine the type
of signal(s) needed to start and stop the engine. The simplest system is a two-wire
start system in which there is a single pair of wires. When the wires are connected
the engine will start and continue to run. When the wires are disconnected, the
engine will stop. Other types of start systems take 3 or more wires and have
separate signals to start the ignition and to operate the engine. These type of
systems are more complicated, since the control system must now determine how
long to close the starter circuit, whether to allow repeated crank attempts, etc.

18-27
Hybrid Systems
Whether or not the engine has a two-wire start system, the important factors are that
the engine is designed for remote starting and that the type of start system is closely
coordinated with the system controls.
Generator Accessories
Engine-generators in hybrid systems have special duties, compared to a traditional
prime diesel or standby diesel engine. Hybrid engines work in a demanding
environment, with frequent starts, extended hours of operation and typically
infrequent maintenance. For these reasons, a number of the following accessories
may be justified.
Long run lube systems

Because maintenance intervals are such a large cost factor in hybrid systems, it
makes sense to reduce maintenance as much as possible. The best way of doing
this is to reduce the frequency of the most common maintenancechanging the
engine oil. Typical standby diesel engines have an oil change interval of only 100
hours, since nobody expects them to be run very much. In continuous (prime)
operation this represents just over four days of use, meaning that the oil would need
to be changed almost twice per week! Even in a hybrid system with 1000 hours of
diesel operation per year a crew would need to visit the site every month.
Diesel engines designed for prime service typically have oil change intervals ranging
from 250 hour to 500 hours, and they often have available options to extend the oil
change interval to 1500 hours or more. With standard engines the maintenance
interval for a hybrid is thus approximately every 3 months with the potential to reduce
that to semi-annual or even annual visits.
Heavy duty start systems

A remote site diesel engine is only useful if it starts when it is needed. For this
reason alone, it is usually worth the investment in some sort of enhanced starter
system such as a 24V heavy-duty starter. In addition if the start reliability can be
increased a smaller battery can be used, and the system can have significant
savings. It also pays to use a multiple-attempt, crank-limited start controller to
prevent damage to the engine during start attempts.
Solar assisted start batteries

One of the primary causes for start failures is a weak battery. In a hybrid system,
where the engine may not be called on to start for a month or more, the starter
batteries may self-discharge to the point where they wont support a start when
requested (especially in hot seasons, when the diesel is used least). A small PV
module dedicated to charging the diesel start battery is often a worthwhile
18-28
Hybrid Systems
investment. The system should be checked, however, to ensure that it does not
overcharge the batteries.
Heavy duty air filters

Keeping the engine air clean is just common sense. If you keep the engine clean, it
will last longer. Dry type filters are preferable to oil bath filters in remote
environments.
Dual fuel filters and/or fuel-water separators

Water and other contaminants in the fuel are one of the primary causes of engine
failure at remote sites. A heavy-duty fuel filter, including a fuel-water separator
(supplied by RAYCOR and others) is a wise investment.
Reduced vibration mounts

Since the engine is a hybrid system is often installed near expensive batteries and
electronic equipment, it makes sense to specify shock absorbent mounting of the
engine. This is a relatively cheap option that will pay for itself in subtle ways.
Critical zone muffler system

Engines will typically ship with an industrial grade muffler, which is intended for a
loud, industrial setting. Residential grade mufflers are available which can reduce
the noise to about 95 dBA. If noise is a particular issue (for a village system, for
example), then a critical grade (hospital zone) muffler is the solution. These
mufflers reduce the noise to about 85 dBA. No one will even notice that the diesel is
operating!

18-29
Hybrid Systems
Control Systems
Component
Selection
Identify all
necessary
control
functions
Hybrid systems require significantly more control than either

standalone PV or prime diesel power systems. The control system
is a central part of a hybrid system and must be integrated with the
overall operation and control. There are several choices in selecting
the proper control system, including control functions, data
monitoring, and types of control systems. A good system designer
needs to understand the range of choices available.
Design the
control
systems as
part of the
total system
Hybrid Control Functions
In addition to charge regulation and load management, the controller

must provide control of the generator and rectifier, as well as monitoring the
additional currents, voltages and equipment associated with a hybrid system.
Typical functions provided by the control system include the following items.
Control Functions
Starting Engine
A hybrid control system needs to be able to start the generator under the
appropriate circumstances. The type of relays and control points required
depend on the start circuit of the engine. In addition, the controller needs the
proper data to determine when the engine should be started, which may
include battery voltage, current and amp-hour readings, time of day or other
parameters.
Stopping Engine
The engine must be stopped at the conclusion of a cycle.
Regulating Array
The power from the array into the battery must be regulated. The relays used
to disconnect the array must be sized for the maximum array current
(including edge of cloud effects).
High Voltage Disconnect
A high voltage disconnect protects the battery from over charging by
disconnecting the battery from all charging sources. The relays must be sized
according the maximum output for each source.
Low Voltage Load Disconnect
A low voltage load disconnect protects the battery from over discharge. The
relays used to disconnect the load must be sized for the maximum load
current.
Exercise Engine
Engines that sit unused for a long period of time may experience problems
starting. To avoid this, hybrid systems sometimes include an engine exercise
function, which periodically starts and runs the engine for a short time.

18-30
Hybrid Systems
Equalize Battery
In addition to normal battery charge cycles, the system may include an
equalize cycle to reduce any voltage differences between battery cells.
System Alarms
Engine Alarms
One of the most important functions of a hybrid system controller is to monitor
the engine. The control system must be able to recognize potential
dangerous conditions and shut the engine down before physical damage
occurs. There are certain specific indications that should be monitored as a
minimum. While the control system does not need to have a separate alarm
signal for each of the following situations, it should be able to respond to any
of them.
- Low oil pressure
- Engine over-temperature
- Engine over-speed
- Engine over-crank (engine has failed to start)
Battery Voltage
An essential feature of a controller is to protect the battery bank from overcharge and over-discharge. The actual control functions for high voltage
disconnect and load voltage disconnect are explained above. In addition, it is
often useful to sound an alarm so that the user is aware that the voltage is out
of normal limits.
Data Monitoring
System Voltages
System voltages can provide a quick check on the state of the system. A
number of readings through the course of a day also depict array
performance, load use and approximate battery state of charge.
- Array Voltage
- Load Voltage
- Battery Voltage
Current
The current in each of the main circuits is also an important variable. From
the current, one can calculate array output, load demand and charging
efficiency.
- Array Charging Current
- Load Current
- Battery Current (Charge and Discharge)
Temperature
Temperature information is used to adjust the battery charging algorithm and
may also be important to monitor for the load equipment.
- Battery Cell Temperature
- Load Enclosure Temperature
- Ambient Temperature

18-31
Hybrid Systems
Miscellaneous Parameters
Depending on the system, there may be additional items that should be
recorded. For example:
- Fuel level
- Wind speed
- Site insolation
Types of Control Systems

There are many different classes of control systems available. This section
discusses the three major classes, starting with the simplest.
Manual Controls
In this control scheme an operator is always available. When the battery voltage
falls to a certain point the operator turns the diesel generator on and supervises
charging of the battery. When the battery is fully charged the operator shuts the
engine off manually.
While this is the simplest technique in terms of control hardware it is also very prone
to errors in judgment or attention. It can also be very costly both in terms of operator
labor and in consequences of a simple lapse in attention. This type of labor is often
classified as long hours, low effort and poses a special challenge to maintaining
skilled labor.
Separate Subsystem Controls

In this type of system each subsystem has its own controls. For instance the PV
array has its own controller while the generator is controlled via a separate voltage
sense circuit. If a wind turbine is present its control is completely separate from the
other controllers and likewise for the inverter.
Obviously this technique only works for fairly simple DC bus systems and does not
allow tight system integration which holds the potential for future cost reductions.
System Level Controls

In this type of control system a single central controller or a network of distributed
controllers provides coordinated control of all system functions. This type of control
system is usually implemented with microprocessors and can include long-term data
logging and remote monitoring capabilities. Some modern inverters include some or
all of the control functions listed below. The hybrid system design must decide the
proper approach based on all the parts of the system.

18-32
Hybrid Systems
Battery Chargers / Rectifiers

The battery charger is the interface between the AC generator and the DC energy
storage system (battery), so it is a critical, but often overlooked component of a
hybrid system. Proper selection of a battery charger is essential to good hybrid
system design. Three phase chargers are preferred over single-phase units
because of reduced AC current, increased efficiency, and less DC ripple. However,
for smaller systems only single-phase chargers may be available.
Types of Chargers
Component
Selection
Choose the
right
combination of
factors:
- Cost
- Efficiency
- Temperature
rating
- Power
Quality
Most battery chargers are meant to be operated off of utility-grade

AC mains power and special attention is needed for use with diesel
generators. The process of charging batteries has two
componentschanging the AC to DC (rectification) and control of
the charging current. There are a number of different types of
chargers commercially available, each of which has advantages and
disadvantages. These are discussed below.
Diode Bridges (half bridge, full bridge and three phase bridge)
These units are the most basic type of battery chargers and tend to be very cheap.
They offer very little control of the end of charge. Single-phase half bridge chargers
have very significant ripple even when filtered heavily. Single phase, full bridge
chargers have less than 5% ripple and are very efficient; however, they have no
inherent control.
SCR Chargers
This class of chargers is based on silicon-controlled-rectifiers (SCRs). In essence
these are diodes that can be turned off thus offering a method of regulation. SCR
rectifiers tend to be inexpensive and moderately efficient but also have a reputation
for being extremely noisy in electrical terms. This affects both the AC input circuit
and the DC output. The high frequency spikes caused by switching the SCRs off
tend to cause problems with the alternators and voltage control circuits used with
engine generators. Most manufacturers require 20-40% oversizing to ensure proper
operation.

18-33
Hybrid Systems
Hybrid Bridge / SCR Chargers

This class of rectifier uses a diode bridge to provide the lower portion of the charge
potential (e.g. 48V in a 48V system) with SCRs regulating the last few volts (e.g.,
the 48-56V range in a 48V charger). This improves efficiency and reduces the noise
problem significantly; however, it adds significant cost to the charger since two
power stages and two control circuits are required. They are available in an
extremely wide range of voltages and currents.
Ferroresonant chargers
These chargers produce very clean power using ferroresonant transformers and
diode bridges. Their drawback is size, weight and low efficiency, but they are
available in a wide range of voltages (up to 260 V) and currents (up to 800 amps).
Saturated Amplifier (Mag-Amp) Chargers

These chargers use a specialty circuit that takes advantage of specialized behavior
of certain types of amplifier circuits. The 3-phase versions have moderate
performance, low cost, and moderate weight, and they tend to be extremely simple
and therefore rugged. (They are commonly used in mines and other harsh
environments.) They are available in a wide range of voltages (up to 260 V) and
currents (up to 800 amps). The single-phase versions are horrible (70% efficiency
and power factors as low as 0.55).
PWM (high frequency) chargers

The latest technology makes use of high frequency switching circuits similar to those
used in some PV controllers and sine wave inverters. These units are lightweight,
efficient, and highly controllable. They tend to be more expensive than other options
and are usually only available in smaller sizes (50 amps and less, typically 48V
max).
Bi-directional inverters as chargers

Some inverters operate in fully bi-directional mode; that is, they can operate either
as DC to AC converters (inverters) or AC to DC converters (rectifiers). They can be
efficient and cost effective, since the same circuit supplies two different functions.
Ferroresonant, Saturated Amplifier and PWM rectifiers are most commonly

used in hybrid systems although other types find uses as well. Check to ensure that
the selected charger is compatible with other components in the system especially
voltage regulators on engines.

18-34
Hybrid Systems
Efficiency and Power Factor

In general one should choose a battery charger to be as high as efficiency as
possible. Typical specifications for chargers are:
Three phase charger
Single phase chargers 75-85% efficient
85-92% efficient
0.8 to 0.9 power factor

0.7 to 0.85 power factor
Power Quality and Other Considerations

Power quality for a battery charger affects two areas the DC current into the
battery and the AC power used by the charger. The quality of the DC output will
affect any DC loads on the system. The amount of disturbance is typically rated as
an amount of ripple. Specify ripple specifications for the DC leg (2% ripple absolute
maximum, typically less). AC power quality will affect the generating source (and
may affect the AC loads on an AC bus system). Check the AC noise specifications
along with power factor.
In some models of battery chargers acoustic noise can also be a problem. Find out
what the unit produces under worst case situations.
Finally confirm that the rectifier is designed for a rugged environment (e.g. wide
temperature variations, high humidity, dust or marine atmosphere) not an airconditioned computer room.

18-35
Hybrid Systems
Inverters
Inverter technology can be critical to hybrids that are specified to supply AC power.
There are three main classes of DC-AC converterssquare wave, modified square
wave, and sine wave. The technology follows the same basic classes as battery
chargers discussed above. The best choice is a single or 3-phase, transistorized
PWM, sine wave inverter provided the cost is appropriate.
Power Quality
Component
Selection
Choose the
right combination of
factors:
- Cost
- Efficiency
- Temperature
rating
- Power
Quality
Inverter power quality is very important to AC loads. The most basic

evaluation of power quality is the waveform. The quality of power is
the worst with a square wave and best with a sine wave. Some
devices may not operate with square wave power or may experience
overheating.
Wave Form Power Quality
The line below indicates the extremes of power quality. Square

wave is the worst and pure sine wave is the best. However, costs
usually increase with power quality, too.
Square Wave
Modified Square Wave
(Worst Quality)
Sine Wave
(Best Quality)
Voltage and Frequency Tolerances

The next general evaluation of power quality is the voltage and frequency tolerances
of the output power. Voltage and frequency tolerances are usually specified as
percentages of the nominal value.
Total Harmonic Distortion

The most specific measure of wave quality is the total harmonic distortion. As
explained in the section on AC Power, total harmonic distortion measures how far
the waveform deviates from a pure sine wave. Modern sine wave inverters can
achieve a total harmonic distortion of less than 5%.
DC Charge Current
If the inverter is bi-directional, then the DC charging current power quality should
also be examined. Refer to the section above on battery chargers for this
discussion.
18-36
Hybrid Systems
Structures and Enclosures

Hybrid components require some method of mounting and protection from the direct
effects of the weather. This may be accomplished by placing each group of
components in a separate enclosure or there may be one shelter that contains all
the components. Generally smaller systems use the first approach while larger
hybrids will have a building or even several buildings dedicated to the system.
Prevention of vandalism is another reason for providing the proper enclosures for
the hybrid system.
Component
Selection
Choose
structures /
enclosures
based on:
- Weather
- Battery
location
- Corrosion
- Vandalism
protection
Often it is desirable to separate the batteries from any electronic

equipment such as inverters, rectifiers or load equipment. Flooded
lead acid batteries produce hydrogen and trace amounts of acid
vapor when under heavy charge. The acid vapor can accelerate
corrosion in sensitive electronics. For this reason the batteries
should be placed in a separate room or enclosure.
A system enclosure both solves certain problems and creates

additional ones. An equipment shelter may protect the components
from very low outside temperatures and from wind-blown dust and
debris. However, the temperature and ventilation within the shelter
must also be considered. Diesel engines produce very large amounts of heat and
so do power electronics to a lesser extent. Proper ventilation must provide cool,
clean air for these types of equipment.
Fuel storage represents an additional concern in a hybrid system. Diesel engines
often use a smaller day-tank together with a main fuel tank. The National Electrical
Code limits the amount of diesel fuel that can be stored inside a building to 660
gallons (2500 liters). Any outside fuel storage should be secured with a lock, since
diesel fuel is a valuable commodity, especially in remote areas!
Any structure or components exposed to the ambient conditions should be
engineered for the proper corrosion resistance. Sand, wind, rain, sun and salt-spray
can rapidly work to produce rust and corrosion in all but the most durable materials.
Stainless steel, quality hot-dipped galvanized steel and aluminum are the preferred
choices.

18-37
Hybrid Systems
PV / Diesel Hybrid Sizing and Design

The flexibility of hybrid power systems is matched by the possible variations in
systems. There are many different approaches to designing a hybrid system. It is
not important whether the system follows a particular design approach, but whether
the system functions properly and supports its design loads.
The hybrid designer must determine several parameters to size a system:
Design Loads and Load Profiles

Type of Bus Configuration
Battery Bus Voltage
Battery Sizing
Generator and Battery Charger Sizing
Generator Operation
PV and other Renewable Sizing
Some system-level design is necessary to make the proper choices during the basic
sizing. It is very important to keep in mind the overall system operation during this
process. There are many interactions between components in a hybrid system and
the designer must ensure that all the overlapping requirements are met.
This section addresses some of the fundamentals of system sizing and design
specifically for PV / Diesel hybrids. Although this particular combination of power
sources is given emphasis many of the design principles are valid for other types of
systems.
Trade Offs in Hybrid System Design
Initial Cost
Customization
Flexibility
Design Limits
Simplicity
Standardization
Operating Cost

)

18-38
Hybrid Systems

Just as defining the design load is a crucial step in stand-alone PV system design,
the overall load for a hybrid system is equally important. For AC loads, additional
information about the frequency, number of phases and power factor is necessary.
Design
Guidelines
Questions to ask to determine system loads:

1. What are the different types of loads to be used with the system?
Get complete,
accurate load
information.
For DC and
AC loads:
- voltage
- power
- duty cycle
- peak power
For AC loads:
- frequency
- 1 ph or 3 ph
- power factor
2. For all loads (both AC and DC):

What is the voltage?
What are the voltage limits (max and min)?
What is the power (or current)?
Does the power or current change according to voltage?
When does the load operate (duty cycle)?
What is the peak power drawn by the load?
3. For each AC load:
What is the frequency?
Is the load single phase or 3 phase?
What is the power factor?
Not only the size of the load is important, but also the hourly load profile. The
system must be able to accommodate any peaks. The load profile might also
suggest certain control strategies. For example, in an AC bus design load peaks
should be coordinated with the generator since this improves system efficiency.
Example:
DC Load Calculation
For a certain telecommunications application, the load is given as:
225 Watts at 24 Volts for 14 hours per day
100 Watts at 24 Volts for 10 hours per day
So
225 W x 14 hours = 3,150 watt-hrs/day
100 W x 10 hours = 1,000 watt-hrs/day
Total load = 4,150 watt-hrs.
4150 / 24 hrs = 173 W continuous. 225 W peak

18-39
Hybrid Systems
Example:
AC Load Calculation
In a certain rural community, the community leaders would like to allow
the following equipment per household. There are 40 households total
in the community:
2 x 50 watt fluorescent lights (12 hours per day), power factor = 0.6
1 x 25 watt radio (6 hrs per day), power factor = 1.0
For one household
2 x 50 Watts x 12 hours = 1200 Watt hrs/day
1 x 25 Watts x 6 hours = 150 Watt hrs/day
Total load = 1,350 Watt hrs /day. 1350 / 24 hrs = 56.3 Watts
continuous
Peak power = 2 x 50 Watts + 25 Watts = 125 Watts
The apparent power is:
2 x 50 Watts / 0.6 power factor = 167 Volt-Amperes (VA)
1 x 25 Watts / 1.0 power factor = 25 VA
Total apparent power = 192 VA
For 40 households
Total load = 40 x 1,350 Watts = 54 kWhrs/day.
Total real power = 40 x125 Watts = 5000 Watts
Total apparent power = 40 x 192 VA = 7680 VA
Total power factor = 5000 W / 7680 VA = 0.65

18-40
Hybrid Systems
Bus Configurations
Determining the Correct Configuration
The choice of whether to use a DC bus or AC bus is determined by the loads and
the overall system operation. If all the loads are DC power, then a DC bus makes
the most sense. If the majority of the loads are AC then an AC bus system might
make sense. AC bus hybrids are particular useful when the generator is required to
support part of the load.
Design
Guidelines
For a DC-only
system, use a
DC bus.
For an AConly system,
use an AC
bus.
In general an AC bus design requires more sophisticated controls

and is more complicated to operate. However, they can also be
more efficient, since energy from the generator can go directly to the
loads. In the DC bus system the generator energy must go through
an AC/DC converter (rectifier) and then back out to a DC/AC
converter (inverter) to reach the load. Depending on the efficiency
of these components this can add significant losses. For example, if
the rectifier efficiency is 80% and the inverter efficiency is 85%, then
any energy going through the pair of components is only at a 0.8 x
.85 = .68 = 68% efficiency.
For a mixed
system,
evaluate both
options.
The hybrid designer must be careful making such comparisons.

These efficiency gains for the AC bus system only apply to the
energy that is supplied directly by the generator. In addition an AC
bus system may need to use single-phase alternators and rectifiers
based on the load requirements. Since the generator and rectifier pair for a DC bus
system can be independent of the load some efficiency can be gained by using
three phase alternators and rectifiers, which are more efficient than single-phase
units. This can result in the net efficiency of a small, single phase AC bus hybrid
being less than an equivalent DC bus hybrid using a 3-phase generator and battery
charger.

18-41
Hybrid Systems
Bus Voltage Selection

The battery bus voltage in a hybrid system can make an important difference on
both the cost and the efficiency of the system. In general the battery voltage should
be as high as possible within the constraints of equipment availability and local
safety codes. This is because the higher battery voltage allows lower operating
currents resulting in lower losses and higher overall efficiency. (Remember that
wattage losses are proportional to the square of the current.) Since the cost of
wires, fuses, circuit breakers and other auxiliary equipment is also a function of
current capacity a higher voltage will result in savings for those components as well.
Comparison of Losses,
Assuming 120W Load with 0.5V Voltage Drop
Voltage
48 V
24 V
12 V
Current
2.5 A
5A
10 A
Loss
1.25 W
2.5 W
5.0W
% of Total
1%
2%
4%

.'
Design
Guidelines
Choose a
higher voltage
to reduce the
current.
DC systems:
Match the
voltage of the
largest load
For DC bus systems the DC operating voltage of the load equipment

often determines the bus voltage. This can become a problem with
large 12-volt loads and the designer sometimes needs to consider
the economics of a DC/DC converter to step the voltage down from
say 48 volts at the battery to 12 volts at the equipment. Similarly if
there are mixed load voltages usually the largest load determines
the bus voltage since this reduces the size of the rest of the DC/DC
converters.
For systems with AC loads the battery voltage is determined by the

input DC voltage of the inverter. In general all but the smallest
systems
should use at least 48-volt equipment while larger village
AC systems:
systems (> 25 kWh / day) should use 120-volt or 240-volt batteries.
Match the
The largest PV-diesel hybrids in commercial use today use 480 V
input voltage
of the inverter
batteries (which keep the maximum operating voltage below the 600
VDC limit). The following example shows a comparison between the
use of 12V and 48V batteries for a hybrid system with a 4 kW inverter:

18-42
Hybrid Systems
Example:
Effect of Bus Voltage on Currents

A hybrid system uses a 2.5 kW array of 53-watt modules, a 12 kW
rectifier (90% efficient), and an 8 kW inverter (90% efficient).
At 12 V, the following conditions apply:
Parallel Modules
Array Current
Rectifier Current
Inverter Current
= 48 (48 module array @ 1S x 48P)

= 48 modules X 3.5 A (Isc)
= 168 A
= 12,000 W X 0.90 12 V
= 900 A
= 8000 W 0.90 10 V (min)
= 890 A

Parallel Modules
Array Current
Rectifier Current
Inverter Current

= 12 modules X 3.5 A (Isc)
= 42 A
= 12,000 W X 0.90 48 V
= 225 A
= 8000 W 0.90 40 V (min)
= 222 A

Parallel Modules
Array Current
Rectifier Current
Inverter Current

= 5 X 3.5 A (Isc)
= 18 A
= 12,000 W X 0.90 120 V
= 90 A
= 8000 W 0.90 100 V (min) = 89 A
A 100A disconnect for the 120V system may be less than 10% of the
price of the disconnect for the 12V system. Other components will
have cost savings as well.

18-43
Hybrid Systems
Battery Sizing
Stand-alone photovoltaic systems will often provide 5-7 days or more of battery-back
up for the load. For critical applications or in locations where the solar insolation is
poor this may increase to 10-15 days or more.
The autonomy of the battery bank is the number of days that the battery can supply
the load under no charging conditions. Thus a battery bank that can support the
load for 5 days without charging is said to have 5 days of autonomy.
Design
Guidelines
A larger battery reduces
the number of
engine starts
and battery
cycles
Start with a 23 battery
Check the
number of
cycles for the
battery
For a hybrid system the battery bank is usually much smaller, often
2-3 days of autonomy. The reason that the battery can be smaller in
a hybrid system is because there is an on-demand power source.
When the battery gets low the system can put a generator in action
and replenish the battery as needed. In a stand-alone system the
battery functions as a reserve of energy. This reserve should be
kept as full as possible to guard against periods of poor weather.
The battery in a hybrid system is serving a slightly different purpose.
The battery allows the system to control the use of each power
source. By providing storage the battery allows the generator to run
at optimum times. It also provides a means of utilizing the energy
gathered by the array. In advanced hybrid control strategies there
are even times when the system deliberately seeks to leave the
battery at a low state of charge.
The smaller the battery capacity, the more frequent the cycles. This
results in a shorter battery life. A good hybrid design strikes a
balance between a smaller economic battery and reasonable life. A
common error about hybrid performance is that a small battery
increases the amount of diesel energy contributed to the system.
On a first order this in not true. A smaller battery results in more frequent diesel
operation and more generator starts, but the total amount of energy produced is
about the same. By using high-cycling batteries and a reliable start system a hybrid
can use a smaller battery and potentially save significant money.
Verify the
maximum
charge rate is
less than C/5.
Since the size of the battery does determine the maximum charge current the
battery capacity does have an effect on the size of the battery charger. This in turn
affects the maximum size of the generator because the generator should not be too
much larger than the battery charger. Forcing a very small engine size will increase
the total hours of operation, which might not be desirable.

18-44
Hybrid Systems
Example:
Effect of Battery Size on Cycle Life

A 2-day battery will result in approximately 180 cycles per year. What
is the estimated life of the battery if it is rated for 1200 deep cycle
discharges at 25 C?
A 2-day battery will result in approximately 180 cycles per year. 1200 /
180 = 6.7 years at 25 C. Note if the battery temperature is greater
than 25 C for significant periods of time, the life will drop sharply. A
6.7 year cycle life is reasonable for a hybrid system, since the batteries
are not likely to last much longer.

18-45
Hybrid Systems
Generator and Battery Charger Sizing

The generator and battery charger should usually be sized together. This is
because the two components are very closely linked in a hybrid system. The battery
charger size is first limited by the battery size. The charger should not charge the
battery at an excessive rate. This is typically limited to a C/5 rate.
Design
Guidelines
A larger engine reduces
the number of
run hours
Charge the
battery
between C/8
and C/5.
Match the
generator to
the charger
input.
Derate the
engine for
temperature,
elevation and
humidity
The generator must be sized large enough to accommodate the

battery charger. A 7.5 kW charger can not run with a 5 kW
generator. Take into account both the power requirements as well
as the power factor when computing the necessary size. In addition
if the system is an AC bus hybrid then any AC loads that will be on
the system when the generator is running must also be added.
For battery charging a large engine size affects the number of runhours not the amount of energy contributed to the system. In an
otherwise equivalent system a generator twice as large will require
only half of the run time. Reduced run time results in lower
maintenance costs while the added cost of a larger generator is
often very small. In addition the amount of energy produced per
liter of fuel is general greater with a larger generator. It is usually an
advantage to use as large a generator as possible.
Ideally, the generator should be 75-90% loaded. This will improve
the fuel efficiency and reduce maintenance problems caused by
light loading.
The manufacturers stated ratings for generator output are based on

certain conditions of temperature, elevation and humidity. If the
generator is used under different conditions then the generator
output must be derated. The specific deratings should be checked
for each particular manufacturer. Some general guidelines are
presented below.
Load the
generator at
75% to 90%
of full load.
Since engine generators use internal combustion of fuel-air mixtures to produce their
energy they are subject to a derating factor for altitude. An engine will simply not
produce as much power at higher altitudes. The typical derating is -3.5% for every
300 meters above its rated altitude. The rated altitude could be sea level, 150
meters above sea level, or some other starting altitude depending on manufacturer.
Temperature is also a factor due to the change in the volume of air and in the
capability of cooling and exhaust air to carry heat away from the combustion
chambers. Typical derating is -0.36% for every oC above a rated temperature (often
30 oC). Humidity can also cause derating of up to 6%.

18-46
Hybrid Systems
General Engine Derating Factors:

Temperature
-0.36% of full power for every oC above rated temperature
Elevation
-3.5% of full power for every 300 m above rated altitude
Humidity
Per manufacturers recommendations, up to -6% of full power
Example:
Calculation of Engine Derating due to Elevation

A 10 kW diesel generator has a nominal rating of 10,000 watts up to
700 meters altitude. If the generator is installed at 3000 meters, its
output will be reduced.
Derating
=
=
=
-3.5%
300 m
-0.035
300m
-0.27
X (3000m actual - 700m rated )

X
2300m
or
-27 % derating
Output Power @ 3000 meters
=
=
=
10,000 watts X (1-0.27)

10,000 watts X 0.73
7300 watts
So the 10,000 watt generator could really only produce about 7,300
watts at that altitude. If 10 kW were really required, then a larger
generator would have to be installed.
The example above shows that the rated power must be larger than then required
power to allow for the derating. You would divide the required power by [1 - the
derating factor] to get the sea level rating of the inverter you would need.
Power @ rated conditions =

Power REQUIRED at installed conditions

[1 - Derating Factor]
18-47
Hybrid Systems
Example:
Calculating Rated Engine Size to Give Required Size
Following the above example, in to actually have 10 kW output available at 3000

meters, you would need to install a 14 kW rated model:
10,000 Watts=
(1- 0.27)
13,700 watts, rounded to 14 kW
Generator Energy Contribution

The more energy the generator provides the smaller the PV array can be. This
reduces the initial cost of the system but also increases the fuel and maintenance
costs of the engine.
There are several simplified approaches to determining the amount of energy
contributed by the generator and the total number of run-hours. These are useful for
determining a rough estimate of the generator performance. In order to make
accurate predictions about generator performance more detailed computer-based
simulations are used.
The best approach to energy contribution is to determine the desired number of run
hours and then estimate the amount of energy produced in the system. As above
this method requires careful consideration of the efficiencies and losses. This
calculation must take into account the efficiencies and losses through the system.
A second approach is to just determine a percentage of the total energy. For
example a designer might want the DEG to provide 50% of the annual energy
requirements. This condition then determines how many hours the engine runs.
However this method is arbitrary and can result in systems that are not well
optimized.
Note: It is very important to keep in mind that engine run-hour estimates in these
types of calculations are rough estimates only. Because the total number of hours
can be small any small change in the actual values can result in a very large
percentage change.
Example:
Consequences of Small Changes in Diesel Calculations

Suppose the duty cycle for an engine is estimated to be 10%. If the
actual duty cycle is 12%, the total hours of operation change from 876
per year to 1051 or an additional 175 hours. This may not seem like a
large amount. However, it reflects a 20% additional increase, which
will be reflected in fuel consumption, maintenance requirements, and
so on. If the diesel fuel tank were sized with only an extra 10%
capacity, the engine would run out of fuel!
The above example illustrates the importance to treating engine calculations

carefully and allowing for ample design margins.
18-48
Hybrid Systems
Exercises

/
62

4666

1

,666

"
2

62

%

/:-
1
:-
4666

"
2

0

966

62

,;66"
2

18-49
Hybrid Systems
PV Array and Other Renewables

The photovoltaic array in a hybrid system provides energy without the need for fuel
or frequent maintenance. PV energy therefore helps reduce the operating costs and
required maintenance of the system. The proper tilt angle and the correct amount of
PV required for a hybrid system are slightly different than for a stand-alone PV
system.
Design
Guidelines
A larger array
results in
more energy
from PV and
reduces the
amount of
energy from
other sources
Tilt the array
to get the best
annual output.
Array Tilt Angle

Stand-alone photovoltaic systems need to maximize the insolation
from the short winter sun to minimize battery size and system cost.
This usually means tilting the array surface 10-20 degrees more
than the latitude angle to face the low winter sun.
But in the case of a hybrid system there is no longer the need to
maximize any particular seasonal performance from the solar array.
Array output can be optimized for the greatest output over the
entire year letting the diesel generator fill in during any particular
season as necessary.
Set the Array Tilt Angle Equal to

Latitude for Maximum Annual Output
Size the PV
array
between 25%
and 75%
contribution.
The angle that results in maximum annual output from a solar

module or array is usually the latitude angle of the site. At this angle
the sun travels exactly perpendicular to the surface of the modules
during the spring and fall solstice, and the winter and summer sun paths are equally
spaced away from perpendicular.
Example:
Best Tilt Angle for Annual Output

An example will illustrate the maximum energy production that results
from the array being tilted to the latitude. We can look at a single
SM55 module operating in Denver, Colorado at latitude 40 N:
Array Tilt
Angle
20
30
40
50
60
Annual
Wh / year
93,849
97,031
97,506
95,454
90,959

Lowest Month
Wh /day (Month)
158 (Dec)
186 (Dec)
207 (Dec)
220 (Dec)
228 (Dec)
18-50
Highest Month
Wh /day (Month)
329 (Jun)
314 (Jun)
293 (Sep)
278 (Mar)
272 (Oct)
Hybrid Systems
As the array surface tilt angle increases from 20 to 60 , the total

annual energy output increases and then decreases, with the
maximum occurring somewhere in between. This maximum output is
97,506 Wh/year at 40 tilt, the same angle as the latitude.
Note that at 40, the worst-case output (Dec.) is only 207 Wh/day,
while at 60 the worst-case output rises to 228 Wh/day. For a standalone system, you would choose the 60 tilt angle to minimize the size
of the array and the discharge of the battery bank. However, with a
hybrid system, you do not have to rely on the solar array to recharge
the battery during bad weather--you have the generator.
This illustrates that to maximize the output of a solar array, and thereby
minimize the energy required from a diesel generator as backup, the
best angle to choose is the latitude of the site.
Annual Energy Contribution

The effect of adding more renewables to a system is to reduce the amount of
generator run time. At the same time the renewables also increase the initial cost of
the system. The general range PV contribution is between 25% and 75% of the total
annual load.
A large part of the economic advantage of a hybrid system comes from more
effective capture of solar energy. For this reason it is usually advisable to size the
solar array so that no energy is wasted even in the highest solar month. Since the
battery is smaller than a comparable stand-alone system a hybrid with a very large
renewable energy fraction may not be able to store all the renewable energy
produced.
Because of the variability of solar insolation from day to day the practical limit is
about 90% PV contribution in the best month in order to avoid regulating the array
unnecessarily. This will typically limit the annual effective solar contribution to 7080%. If the energy is wasted then the system efficiency drops and may result in an
economically poor design.
In some design situations there are specific goals for energy contributions during
certain seasons. For this reason it may be acceptable to design the array so that
some energy is regulated during summer months.

18-51
Hybrid Systems
Checklist for Hybrid Sizing

Site and Project Information
System Description:
Site:
Site Latitude (deg):
Site Elevation (m):
Average Temperature:
Avg. Insolation: (tilted):
DC Loads
Description
Quantity
Watts
Hours/day
Daily Demand
(DC Whrs)
Total
AC Loads
Description
Quantity
Watts
P.F.
VoltAmperes
Hours/day
Daily Demand
(AC Whrs)
Total
Total AC Whrs
=
Inverter Effic.

18-52
+
AC Input (Whrs)
=
DC Whrs
Hybrid Systems
Total Load Whrs
Sizing Checklist, Continued.

DC
Volts
DC
Amps
AC
Volts
kW
kVA
Hz
Phase
1 or 3ph
Other Characteristics
Rectifier
Fuel
Type
Total
Derating
Continuous
Rating
Surge
Rating
Estimated
Peak
Estimated
Surge
Capacity
(kWh)
Type
Number in
Parallel
Number in
Series
Quantity
Tilt
Angle
Number in
Parallel
Number in
Series
Generator
Inverter
Total
Load
Battery
PV Array
Other
Renewable

18-53
Hybrid Systems
Simplified Sizing/Design Example

The following example illustrates some of the principles discussed in this section.
Proposed Application
A customer has telecommunications equipment that needs to operate a remote site.
The customer has had poor experience with diesel engines in the past and is
concerned about the amount of maintenance required at this distant location.
The technical load is 30 A continuous at 48 V DC. There is a small air conditioning
unit 600 watts at 220 V AC, 50 Hz that is only used when technicians are on site.
Further discussions with the customer determine that the technicians typically bring
a portable generator with them to the site for test equipment and power tools.
The proposed site is located at an elevation of 1500 m. The average site
temperature is 30 C during the summer season. The climate is dry so humidity is
not an issue. The average solar insolation provides 5.5 kWh/day.
Loads
The telecommunications load is 30A at 48VDC.
30 A x 48V = 1440 W = 34.6 kWh/day = 12,629 kWh /year
Based on the description of the intended use for the air conditioner it is
recommended that the portable generator power this load. A separate power hookup can be made to allow the generator to be connected to the air conditioner and to
allow technicians to power other equipment. This approach greatly simplifies the
design of the system and eliminates the need for an inverter.
Hybrid Configuration
Because of the separation of the technical DC loads from the air conditioner a DC
bus is the most suitable.
Bus Voltage
Since the only loads are 48V DC this is the most appropriate choice for the bus
voltage.
Battery
A battery capacity with 3 days autonomy to a maximum 80% depth-of-discharge
yields:
34.6 kWh/day load = 34.6 X 3 / 0.8 = 129.75 kW
129.75 kWh / 48V = 2703 Ah per string

18-54
System Design Hybrid Systems
Suppose we have the following battery cells to choose from:

Capacity
2000 Ah
2500 Ah
3000 Ah
4000 Ah
The 3000 Ah capacity cells are the most appropriate, since they are the smallest
size that provides at least 3 days of autonomy. The maximum charge rate of C/5 for
this size battery is approximately:
3000 / 5 = 600 Amps.
Battery Charger and Generator
Suppose we have the following rectifiers to choose from at 48V (Efficiency = 0.80):
Amps
Input (kW)
Phase
100
6
1 ph
150
9
1 ph
200
12
3 ph
400
24
3 ph
800
48
3 ph
The 400A rectifier is a little smaller than the maximum allowed charge rate, but will
still charge the batteries at an appropriate rate (about C/8).
We next determine the size of the generator. Suppose that the following generators
are available and the ratings are given for 900 m and 25 C:
Capacity
Fuel
Phase
25 kW
Diesel
3 ph
27.5 kW
Diesel
3 ph
30 kW
Diesel
3 ph
40 kW
Diesel
3 ph
First we calculate the range to provide a 75-90% loading on the engine.

Max
Min
24 / 0.75 = 32 kW
24 / 0.90 = 26.7 kW
We need somewhere between 26.7 and 32 kW installed output from the generator.
We must now allow for any deratings due to the elevation and temperature.
Temperature
Elevation
Total derating
-.36% / C x (30 - 25) = -1.8%

-3.5% / 300 x (1500 - 900 ) = -7.0%
1.8 + 7.0 = 8.8%
The rated range must be:

Max 32 kW / ( 1- 0.088) = 32 / .912 = 35.1 kW
Min 26.7 kW = 26.7 / .912 = 29.3 kW
The 30 kW engine would be a good match.

18-55
Generator Contribution
The customer has stated a preference to reduce the amount of generator run time.
However, given the size of the load it will not be economical to power the loads
exclusively from PV. We might select 500 hrs as the maximum desired generator
run time in one year (typically 1 maintenance interval per year).
The rectifier output is 400A x 48 V = 19.2 kW. Assuming an additional 10% system
losses (0.90 multiplier) and overall battery efficiency of 80% we can calculate the
energy output of the rectifier / generator combination as:
500 x 19.2 x 0.90 x 0.80 = 6,912 kWh/year
PV Array
The PV array needs to make up the remaining portion of the annual load = 12,629 6,912= 5,717kWh /year. The average insolation is 5.5 peak sun hours and we
assume the following multipliers for array output:
High temperature
Dirt
Battery Eff
=
=
=
0.85
0.9
0.8
The size of the array is calculated as:

Array Size (kW)
Annual Energy (kWh/year)

Peak Sun Hrs x 365 days x 0.85 x 0.9 x 0.8
5717
5.5 x 365 x 0.85 x 0.9 x 0.8
4.65 kWp = 4650 Wp
The array voltage is already determined by the voltage of the bus to be 48 V. If we

select SM50 modules with a rated power of 50 Watts, then the array needs to
consist of:
4650 / 50 = 93 modules, rounded up to the next multiple of 4 = 96 modules.

18-56
Rules of Thumb for Sizing

With so much flexibility it is difficult to do any hybrid sizing by hand. It really requires
a computer program to balance the many factors and allow quick modification to see
the effect. However, it is possible to give some very general rules of thumb to get
you started with some hybrid designs. Keep in mind that these guidelines are very
general and rough. Further refinement is always possible.
Start with the average daily load for the system in kWh/day. Then some simple
ratios can be established for the other components of the system such as the size of
the generator, rectifier, battery capacity and inverter. And a rough relationship can
be given between hours of generator run and total generator energy in the year.
From this information array size can be estimated given some weather information.
The factors below are separated for a DC bus system with DC loads only and an AC
bus system with AC loads only.
Units DC Bus Hybrid

Daily Load Demand kWh/da = Y
y
Inverter Power
kW (Not applicable)
AC Bus Hybrid
=Z
Battery Capacity
kWh
= 4.3 Y
=5Z
Rectifier Rating
kW
= 0.48 Y
= 0.56 Z
Generator Rating
kW
= 0.67 Y
= 0.78 Z
Annual Generator
Energy
kWh/yr = Gen Run Hrs X 0.4 Y
Annual Total Energy kWh/yr = 365 Y

Annual PV Energy
= 0.3 Z
= Gen Run Hrs X 0.4 Z
= 365 Z
kWh/yr = Annual Total - Annual = Annual Total - Annual

Generator Energy
Generator Energy
PV Array Pmax
kW = Annual PV Energy
Rating
Peak Hrs X 365 X 0.6

= Annual PV Energy
Peak Hrs X 365 X 0.5
(Note: The derivation of each factor is given at end of the section)

The following example illustrates using these guidelines to do a rough hybrid system
sizing.

18-57
Example:
Hybrid Sizing Using Rules of Thumb Method

A remote village will have a hybrid photovoltaic diesel power system
installed. The typical daily energy demand for the village has been
estimated to be about 24 kWh/day. Since the loads are AC, an AC
bus is most appropriate.
Inverter Power
0.3 x 24
7.2 kW
Perhaps this would be rounded up to the next commercially available

capacity, probably 10 kW. Typically at this size, the input DC voltage
would be 48 or 120 VDC.
Battery Capacity
5 x 24
120 kWh
The Ah of capacity will depend on the input DC voltage of the inverter.

For example, if the input voltage were 48 volts, the capacity would be
2500 Ah. At 120 VDC the capacity would be 1000 Ah.
Rectifier Rating
Generator Power
=
=
0.56 x 24
0.78 x 24
=
=
13.4 kW
18.7 kW
Again, these might be rounded up to the next commonly available size.

Assume that the recommended interval between maintenance is about
500 hours, and that we do not want to have more than one
maintenance trip per year.
Annual Generator Energy
=
=
500 hours x 0.4 x 24

4800 kWh per year
Total Annual Energy Demand
24 kWh/day x 365
days/year
8760 kWh
=
Annual Energy From PV
=
8760 - 4800
=
3960 kWh
Assume that the region receives about 5 peak hours of insolation
daily.
Array Power Required
=
=
3960
5 X 365 X 0.5
4.3 kW
So we find that a 4300 watt array would provide the energy that the
generator did not supply during its 500 hours of annual operation.

18-58
Derivation of Hybrid Sizing Rules of

Thumb
1. Assumed
Parameters:
2. Assume the total

daily load is given
as:
3. The daily amount
of DC energy
drawn from the
battery is:
4. The total energy
available in the
battery is:
5. The total battery
capacity is:
All Systems
Battery bank autonomy
=
3 days
Max battery depth of discharge =
70%
Max battery rate of charge =
C/10
Rectifier efficiency
=
80%
Generator derating
=
90%
General system efficiency =
90% (Due to wiring losses,
etc.)
Battery total energy efficiency
=
80%
PV Array temperature factor
=
90%
PV array dirt factor
=
90%
Inverter efficiency
=
85% (AC Bus systems only)
DC Bus Systems
AC Bus Systems
Y = daily DC load (kWh)
Z = daily AC load from inverter
(kWh)
=Y
= Z / inverter efficiency
= 1.2 * Z
=3*Y
= 3 days * Z / inverter efficiency

= 3.6 * Z
= 3 * Y / Max. DOD
= 4.3 * Y
= 3 * Z / (inverter eff * max.

DOD)
=5*Z
= C/10 of the battery, increased = C/10 of the battery, increased
by system losses.
by system losses.
6. The rectifier
rating is based on
the size needed
= C/10 * Battery Capacity
to deliver the
.9
maximum charge
rate to the battery: = 4.3 * Y / 10
.43 * Y
.9
.9
= .48 * Y

= C/10 * Battery Capacity

.9
= 5 * Z / 10
.9
.5 * Z
.9
= .56 * Z
18-59
6. The generator
rating is based on
the size needed
to power the
rectifier:
7. The annual
amount of energy
produced by the
generator in
kWh/year is given
by:
8. The inverter
rating is assumed
to be:
9. The annual
amount of energy
produced by the
PV array in
kWh/year is given
by:
10. The PV energy
is related to the
array rating as
shown by:
= rectifier rating increased by

the rectifier efficiency and
generator derating
= rectifier rating increased by

the rectifier efficiency and
generator derating
= .48 * Y
.9 * .8
= .56 * Y
.9 * .8
= 0.67 * Y
= Gen Run Hrs * final DC load
output
= 0.78 * Y
= Gen Run Hrs * final AC load
output
= Gen Run Hours * Gen Rating

* rectifier eff * gen derate *
system losses
= Gen Run Hours * Gen Rating

* rectifier eff * gen derate *
system losses * inverter eff
= Gen Run Hours * .66Y * .8 *

.9 * .9
= Gen Run Hours * .78Z * .8 *

.9 * .9 * .85
= Gen Run Hours * 0.4*Y
= Gen Run Hours * 0.4*Z
(Not applicable)
= 0.3 * Z (Typical ratio of load

to continuous rating)
= Total Annual Energy - Energy = Total Annual Energy - Energy

produced by generator
produced by generator
= 365 * Y - Gen Run Hours *
= 365 * Z - Gen Run Hours *
0.4*Y
0.4*Z
Array Energy =
PV Array
kW * peak hours * 365
days/year * deratings
where deratings = heat factor
* dirt * battery eff
= .85 * .9 * .8
So the PV array
rating is given by::
= 0.6
= Annual PV Energy
peak hours * 365 * 0.6
Array Energy =
PV Array
kW * peak hours * 365
days/year * deratings
where deratings = heat factor
* dirt * inverter eff * battery
eff
= .85 * .9 * .85 * .8
0.5
= Annual PV Energy
peak hours * 365 * 0.5


18-60
Hybrid System Maintenance

Since a hybrid system is a cross-breed between two standalone power systems its
maintenance requirements will reflect each of those systems. However, careful
design of the hybrid system will preserve much of the maintenance free
characteristics of a PV system with only a limited amount of diesel maintenance.
PV System Maintenance
The maintenance of a PV array in a hybrid system is no different than standard
maintenance in a regular PV system. A good check of the system will include:
1.
2.
3.
4.
Checks for damaged modules.

Checks for loose electrical connections.
Cleaning the array surface if unusually dirty.
Checks for any loose mechanical connections.
Engine Generator Maintenance

The maintenance of a diesel engine generator (DEG) is proportional
to the number of hours of operation.
This is a very important point in a hybrid system. Even though the energy
contribution of the diesel segment may be 50% or more the duty cycle is typically
much less than this often 10% or less. Because maintenance is related to the
number of hours rather than the total energy delivered this offers a tremendous
advantage to a hybrid system.
Consider the following maintenance schedule which is typical of a small, air-cooled,
prime diesel engine:
Maintenance Task
Interval
Oil Change
250 hrs
Decoke
1,500 hrs
Overhaul
6,000 hrs

!

18-61
An operating manual lists the following Three Levels of Diesel Maintenance

requirements:
Frequency
Task(s) To Perform
Qualifications
250 hours
Can be
Oil and filter change.
performed by a
Inspect air and fuel filters, fuel
trained technician
system, starter battery, and
at the site.
system electrical connections.
1,500
Decarbonization (top end
Can only be
hours
overhaul): In addition to oil
performed by a
change taskstrained engine
mechanic, site
Replace air and fuel filters
work is possible.
Remove the cylinder head and
clean valves, valve seats,
injector nozzles, etc.
Replace all top end gaskets.
6,000
Full overhaul. In addition to tasks Requires a
hours
listed above
competent
mechanic.
Perform full engine overall.
Typically
Clean all combustion and
performed in a
inlet/outlet chambers.
repair shop which
Replace all engine gaskets.
means the engine
Replace camshaft / crankshaft
bearings, valves, valve springs, must be
transported to the
injectors, fuel pumps, pistons,
shop, and a
piston rings, starter battery,
replacement
start solenoid and engine
engine provided
starter, fuel solenoid, fuel
to the site.
pipes, and other parts as
necessary.
"!#
Cost Estimate
$25 to $100
plus travel if
required.
$250-500 plus
travel for a
typical two
cylinder air
cooled engine.
$1,500-3,000,
plus
transportation
and loaner of
replacement
engine.
Note that many standby engines have only a 100-hour oil change interval that, if
running in a prime power application, would necessitate nearly 90 oil changes per
year -- almost twice per week! Engines made for prime power often have options
that can extend the primary maintenance interval to 500, 1000 or even 1500 hours
dramatically reducing the maintenance requirements.
Using the hybrid sizing example in the previous section the generator runs
approximately 500 hours per year vs. 8760 hours for a continuous prime diesel. The
following table combines the maintenance schedule with the operation of the
engines to calculate the total number of maintenance visits required:

18-62
Maintenance
Interval
Hybrid
Prime Power
Task
(Hours)
(500 hours)
(8760 hours)
Oil Change
250
2 per yr
35 per yr
Decoke
1500
1 every 3 yrs
6 per yr
Overhaul
6000
1 every 12 yrs
3 every 2 yrs
$%&!'
(
)*

In addition the higher loading of the hybrid engine will keep the engine running more
efficiently reducing some of the common problems with diesel generators -- clogged
fuel injectors, heavily carbonized valves, slobbering and glazing. This means that
the intervals for extended maintenance items such as top-end overhauls and full
overhauls might be extended even further than for a variable load prime diesel.

18-63
Battery Maintenance
The second most maintenance intensive item in a hybrid system is the battery.
Unlike batteries in PV systems that are treated quite gently hybrid system batteries
are exposed to frequent deep discharges (typically two per week or more) and
variable rate charges (often at high rates). This last point is very important -charging must be coordinated between the various generators.
For example a system may include a diesel engine, a couple of wind turbines and a
substantial size PV array. The engine/battery charger is designed for a high rate
recharge (C/6 or so). If the battery charger is on and the wind suddenly picks up or
the sun comes out the charge rate may exceed the battery manufacturers
recommendations causing damage to the battery and a potentially dangerous
situation with excessive gassing and hydrogen evolution. If VRLA batteries are used
this point is even more critical.
In general the batteries should be checked at regular intervals. Unless there are
serious access constraints to the site the batteries should be examined at least
every 3 months. They should be filled with water, equalized (if the controller does
not have an automatic equalization function) and checked for tight connections,
evidence of corrosion, etc. Once a year a complete record should be made of
individual cell voltages and any suspect cells tested and replaced if necessary.
Since most hybrid systems use flooded cycling batteries they often use increasing
amounts of water as they age. This should be taken into consideration when
planning maintenance visits. In a village type system a local villager can be trained
to check the water level at weekly intervals, keep records and report to the
maintenance crew if there are any thirsty cells. In telecom systems this is not
possible; however, careful monitoring of separate battery strings might show if one is
drawing excessive amounts of current during charge -- a sure sign of a problem cell.
In general the use of a high quality controller and battery charger will help ensure
lower maintenance requirements and longer life of any hybrid system battery.

18-64
Controls Maintenance
The amount of maintenance on the system controls is usually very small and
consists primarily of checking that all the wiring connections are tight and that there
are no problems with the fuses or surge protection.
Some systems with on-site data storage may use this time to download the data or
to replace storage modules for transportation back to a central site.
It is also a good idea to use the maintenance interval as a chance to review the
system operation. Does the system operate as intended? Is there any unusual
behavior that requires attention. Reviewing the performance data can be a way to
identify potential problems before they become serious. The system set points and
control strategies can sometimes be adjusted to accommodate new changes in the
load demand or different conditions than were originally expected.

18-65
Applications and Case Studies

Telecommunications
Telecommunications applications are often candidates for hybrid system for several
reasons. Many telecommunications loads are relatively large and a hybrid system
helps offset some of the initial cost. Companies in the telecommunications industry
often have experience with diesel engines and already have trained technicians who
can perform the necessary maintenance. Also telecommunications equipment often
is located in places where there may be very limited insolation during a portion of the
year or where there is not sufficient area for a full size array.
Telecommunications loads that might be power by a hybrid include:
Microwave repeaters
Fiber optics relays
Rural telephony
Cellular phone sites
Television transposers
Telecommunications power sources demand very high reliability. Some systems will
include a second engine generator as a backup. This is based on the idea that a
system should have duplicates any essential component or system. Sometimes this
is referred to as "1 + 1 redundancy." Another approach used to assure performance
is to over-size some of the components. A telecommunications system may include
a much larger battery or have a very high PV array contribution. These features
allow the system to operate for a longer period of time after a diesel failure, hopefully
giving a repair crew sufficient time to reach the site and make repairs.
A system designed for telecommunications must also give thought to protecting the
load equipment. Unlike a village power system where all the loads are physically
separate from the power system a telecomm hybrid usually needs to include some
space for the load equipment. This should be removed from the generators and
batteries. In addition there may be heating and cooling requirements for the load
equipment. The system designer must include the power necessary to heat or cool
the enclosure in addition to the load equipment draw.

18-66
Village Power
Many rural communities in the developing world do not have electricity. A small
amount of electricity can make a dramatic change in the lifestyle of these people by
providing lighting, refrigeration and access to television and radio programs. Hybrid
systems can be used for rural electrification projects. Of those communities that are
already electrified many are powered by a diesel generator and already have
distribution lines. In communities without power a hybrid offers the ability to supply
power while reducing fuel costs, pollution and noise.
Village load profiles are often well suited for a hybrid system particularly if there is a
significant lighting component. Since the lights are primarily used during the evening
hours the load level may be several times the daytime value. This results in very
inefficient prime power diesel systems, but the same load variation can be handled
by a properly designed hybrid power system. The presence of a generator to supply
power on-demand is also a cost-effective means to power large commercial loads
(e.g. an ice-making facility) that would be very costly for a stand-alone system.
Hypothetical Village Load
4.8 kW Peak / 50 kWHr/Day
5.00
4.50
4.00
Load, kW
3.50
3.00
2.50
2.00
1.50
1.00
0.50
0.00
0:00
2:00
4:00
6:00
8:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00
0:00
Time of Day

+,-

There are many social aspects to bringing power into a village. How will the
lifestyles of the villagers change? What new uses for the power will they find? The
size of the load can often change according to special occasions or holidays.
Sometimes the total load in the village grows dramatically after the system has been
installed.

18-67
Facility Power
Another application of hybrid systems is for facility power. This could be any
isolated site with a collection of diverse loads. Some example of facility applications
include the following:
Hotels
Government offices
Medical clinics
Military facilities
Recreation Sites (Marinas, Campgrounds, etc.)
While the characteristics of each system can vary significantly these applications
require fairly large amounts of power and typically include AC loads. In addition the
load demand is rarely constant and may be composed of a number of variable
loads. It is important to carefully define the load requirements of the system and
ensure that the hybrid will have sufficient capacity to meet peak loads.
Since many facility power applications are existing sites the hybrid power system is
often intended to supplement an existing diesel-based power system. In such
circumstances the primary customer concern is to reduce the amount of diesel
operation. This is in effort to reduce the amount spent on fuel and maintenance.
Pollution and noise may also be significant secondary considerations especially for
ecologically sensitive sites.
One feature of a facility power system is that someone is usually designated to
provide maintenance for the system. This can be an advantage over village power
projects where on-going maintenance can be difficult to support.

18-68
Design Examples
Pulau Pemanggil, Three Islands Project, Malaysia
The Pulau Pemanggil system is located in Malaysia and is designed to provide
power for a small community located on an island near the coast. The original load
requirements were calculated as follows:
ITEM
Avg.
No. of
Wh/Day Units
per unit
276
20
380
3
834
1
464
1
1048
1
3970
1
740
1
Daily
Energy
Wh/Day
5520
1140
834
464
1048
3970
740
Annual
Energy
(kWh)
2014.80
416.10
304.41
169.36
382.52
1449.05
270.10
Domestic House
Domestic House + Shop
Mosque
Community Hall
Police Post
School
Clinic (excluding vaccine storage
system)
Power Plant House
904
1
904
329.96
Total Average Wh/Day
14620
5336.30
Add 10% Safety Factor (to account
1462
533.63
for cable losses, load variations,
etc).
Total Average Energy required.
16082
5869.93
.**,'
Array
Battery
Generator
Inverter
Design Load
7.2 kW peak
124 kWh @ 120 Volts
26 kW diesel generator
10 kW, 240 VAC, 50 Hz, 1 phase
16.1 kWh/day @ 4.5 Sun hours
The system was originally designed as a PV system with the generator primarily
used for backup. The PV array was to supply most of the loads. After the system
was installed the loads grew significantly to the point where the generator supplies a
substantial portion of the loads.
PV Array The photovoltaic array uses high-efficiency, single crystal modules.
There are three ground mounted arrays.
Controls The system uses one set of controls for the PV array and another set of
controls for the generator. In addition there is a data acquisition system that records
performance data for the system.
Battery The system uses a single string of sealed batteries.
18-69
Block Diagram - Pulau Pemanggil

26.4 kW
Generator
11.3 kW
Rectifier
7.1 kW
PV Array
10 kW
Inverter
124 kWh
Battery
240 VAC
Loads

/#
0+*1

For additional information on this system contact Siemens Showa Solar, Singapore.

18-70
SunWize Power Station

The SunWize Power Station is a series of pre-packaged hybrid designs that are
designed to server telecommunications or village power applications. The system
design offers an excellent example of combining all the important factors of a hybrid
system performance, cost, reliability and system integration. In addition the
system is designed as a pre-assembled unit that can be tested in the factory. This
significantly improves the overall quality of the product.
There are several versions of the Power Station. The main components and
features for a PSG1500 system are shown below. It is important to recognize that
ratings and performance depend on the specifics of the installation.
Array
Battery
Generator
Inverter/Rectifier
Design Load
1500 W peak
24 kWh @ 24 Volts
4.8 kW
4 kW, 120 VAC, 60 Hz, 1 phase / 24 VDC
6.5 kWh/day AC loads @ 4.5 Sun hours
PV Array The photovoltaic array uses high-efficiency, single crystal modules.

These modules provide higher amp-hour per watt output.
Array Structure A self-supporting structure provides a framework to mount all the
components. This design uses the weight of the system to ballast the system. The
array structure is adjustable from 30-60 degrees tilt.
Inverter/Battery Charger The system uses a bi-directional inverter/battery
charger. This unit provides both the AC output as well as allowing current from the
generator to charge the battery.
Batteries The battery bank is composed of flooded 2V cells that are contained in a
dedicated enclosure. The battery enclosure is ventilated to prevent the
accumulation of hydrogen.
Generator This system uses an air-cooled propane generator.
Controls The system is completely automatic with the inverter controlling the
generator and DC regulator controlling the solar array functions. The controls and
distribution are located in a weather resistant enclosure.

18-71
Block Diagram - SunWize Power Station

120 VAC
Loads
4.8 kW
Generator
4 kW
Inverter/
Rectifier
1.5 kW
PV Array
24 kWh
Battery

/#
0*#

For additional details refer to the brochure located in the Appendix. The SunWize
Power Station is manufactured by SunWize Technologies, Inc. located in Kingston,
NY, USA.

18-72
McPherson Peak System

The McPherson Peak system* is an example of a large telecommunication hybrid.
This system powers cellular repeater equipment and was designed and installed by
Solar Electric Specialties. The system is located near Santa Barbara, California. As
a telecommunications power system the hybrid system is designed for high reliability
and low maintenance:
Array
Battery
Generator
Inverter
Design Load
11 kW peak
200 kWh @ 24 Volts
4.8 kW 24 VDC Output, Dual Units.
1.5 kW, 120 VAC, 60 Hz, 1 phase / 24 VDC
70 kWh/day @ 4 Sun hours
PV Array The photovoltaic array uses single crystal modules. About 75% of the
array is mounted on the shelter the rest is ground mounted.
Generators The engine generators are propane fueled. There are two DC
generators for the site. One generator is designated as the primary engine and the
other serves as backup (1 + 1 redundancy).
Note that the array size is very large while the generator is relatively small for the
given load. This system was designed to operate at 80-90% PV contribution but
also to allow for significant load expansion. This way the load may increase without
any need to add additional hardware.
System Controller The system controller at this site allows for remote monitoring
of key system values including fuel levels. This allows the customer to access the
site via a telephone modem and determine the status of the system.
Inverter There is a small inverter to provide power to 120 VAC loads within the
shelter. Most of the equipment loads operate directly off the 24-volt power bus.
Batteries The battery bank is composed of sealed batteries. Because the
generator provides a relatively small amount of charge current the risks of
overcharging are well managed.
Shelter/Enclosure The load equipment is contained in a 24' concrete shelter.
The shelter also provides a mounting location for some of the PV array. Cooling
fans ensure that the shelter temperature does not exceed the design limit.
* Note: There are actually two hybrid systems located on McPherson peak both of
serving similar loads. The system in this example is slightly larger than the other
system.

18-73
Block Diagram - McPherson Peak System

4.8 kW DC
Generator
4.8 kW DC
Generator
70 kW
PV Array
200 kWh
Battery
24 VDC
Loads
1.5 kW
Inverter
120 VAC
Loads

/#
0

For additional information on this system contact Solar Electric Specialties in Willits,
California, USA.

18-74
Exercises
*
2.324

!
"


18-75
(End of Chapter)

18-76
CHAPTER EIGHTEEN
HYBRID SYSTEMS
Multiple Energy Sources - Flexible Approaches
18-1
18-1
Fundamentals of Hybrid Power Systems

On-Demand Power Generation
Better Use of Renewable Energy Production
Load Matching
Applications of Hybrid Systems
Advantages of PV/Diesel Hybrid Systems
Disadvantages of PV/Diesel Hybrid Systems
Comparisons of Various System Types
18-2
18-2
18-3
18-5
18-6
18-7
18-9
18-10
Hybrid Configuration and Operation

Power Sources
Bus Configurations
Hybrid Operation
18-11
18-11
18-16
18-19
Component Selection
PV Array
Battery
Engine / Generator
Control Systems
Battery Chargers / Rectifiers
Inverters
Structures and Enclosures
18-23
18-23
18-24
18-25
18-30
18-33
18-36
18-37
PV / Diesel Hybrid Sizing and Design

Bus Configurations
Derivation of Hybrid Sizing Rules of Thumb
18-38
18-39
18-41
18-59
Hybrid System Maintenance

PV System Maintenance
Engine Generator Maintenance
Battery Maintenance
Controls Maintenance
18-61
18-61
18-61
18-64
18-65
Applications and Case Studies

Telecommunications
Village Power
Facility Power
Design Examples
18-66
18-66
18-67
18-68
18-69

18-77
Chapter 18 Answers
Hybrid Systems

Looking at the table of insolation data, the worst case for the system tilted at 55 is 2.84
peak sun hours and occurs in June. In comparison, the average number of sun hours
for the hybrid system tilted at 25 is 4.53. Therefore, the first advantage to using the
hybrid is that we can design around an insolation value of 4.53 sun hours, rather than
2.84.
The highest insolation for the system tilted at 55 is 4.94 sun hours, occurring in
February. A stand-alone system designed for the worst case of 2.84 will lose much of
this energy. The amount of excess energy is
Maximum Insolation
Worst Case
4.94
2.84
1.74
This implies that 74% of the required energy is wasted in February.

a. PV might be preferred to wind or micro-hydro in an area where construction is
difficult. Both wind turbines and micro-hydro turbines can require significant
construction. Very little civil works are required for a small PV array. Other issues that
might make wind or micro-hydro less desirable are: difficulty in obtaining permits, local
wind patterns, water access, undesirability of obstructing water flow, etc.
b. Where there is a steady, significant wind flow, a WTG might be preferable to PV.
Under the proper conditions, wind can provide a very low cost energy source.
Additional factors that might make wind more favorable might be: extreme seasonal
solar variations (e.g. the arctic), limited available ground area, etc.
c. A thermoelectric generator might prove the best solution if there is a cheap source of
energy close by. This is common is gas field applications where the TEG can use
plentiful natural gas. Another common application is in extreme cold climates. Not only
does the TEG operate better under such conditions, it can also provide heat to keep
equipment from freezing.
Note that other answers are possible.
18-1
Hybrid Systems

1. DC Bus (DC Loads Only) This system configuration makes the most sense for
large DC loads. The generator provides power to the rectifier, which then charges the
battery. It is the simplest to design and control, and it does not require a sophisticated
inverter. Good applications for this type of system include telecommunications
systems.
2. DC Bus (DC and AC loads) A DC bus hybrid with mixed loads has some of the
same control simplicity as a basic DC bus system. AC loads are connected to a
dedicated inverter that draws from the battery. This works best if the AC loads are
generally small and do not have much variation. This system again works well for
telecommunications, perhaps for a system including an equipment shelter with lights or
other small AC loads.
3. AC Bus (AC Loads) This type of system has the most flexibility in supplying
variable AC loads. The generator is connected in parallel with the output of the inverter.
This allows generator power to flow directly to the loads and is very efficient. The
inverter must be more sophisticated and the control system needs to include more
features. This type of system works well for a village with many AC loads or a facility
power system, e.g. a small eco-tourist hotel.

An AC bus hybrid can still supply DC loads directly from the battery bank. The bidirectional inverter keeps the battery charged, supplying any DC loads that may be
connected. Note that for larger hybrid system, the battery bank voltage may be too high
for most normal DC loads.

Choose any three of the following factors:
Battery Size
The larger the size of the battery bank, the fewer cycles the generator will make.
A large battery bank will also reduce the number of cycles on the battery, thereby
extending its life.
Generator/Rectifier Size
The larger the generator / rectifier combination, the shorter the time necessary to
recharge the battery bank. This does not change the total amount of energy
provided by the diesel; it just reduces the hours of generator operation.
18-2
Hybrid Systems
Array Size
The larger the PV array, the more renewable energy is produced and the less
energy will be required from the generator.
Inverter Size
The size of the inverter affects the size of the AC loads that the system can
handle. In an AC bus system, it also has the same effect as the rectifier.
Other items, such a load use, weather and control strategies are also possible.

a. System B is the cycle-charged system. System A is a prime-power diesel system.
b. Although the engine sizes can vary by the application, System B probably would
have a larger engine. The larger the engine size in a cycle-charged system, the fewer
the overall hours of operation (same as for a hybrid system). A large engine will require
a large rectifier to properly load the engine. Increasing the size of the generator in
System A does not reduce the run time (the engine is always on). A large engine in a
prime-power system will actually reduce the efficiency of the system.
c. For System A, the issues will be those associated with a high number of run hours.
This includes frequent oil changes and overhauls. The system will also require frequent
re-fueling. System B will have lower maintenance requirements in terms of oil changes
and maintenance. However, the start system will have to be reliable, so the start
system and starting battery will need to be properly maintained.

We use the general engine derating figure given in the text. The derating is calculated
as:
Derating
-3.5% X (4000 m actual - 1000 m rated)

300m
-0.035 X (3000)
300
0.35
or 35% derating
Output Power @ 4000 m =
50,000 Watts X (1 - .35)
32,500 Watts or 32.5 kW
18-3
Hybrid Systems

We now need to take into account deratings for both temperature and altitude. The
altitude derating was calculated in the previous problem to be 35%. The temperature
derating is:
-0.36% X (35 C - 25 C rated)
.0036 X 10
= .036
= 3.6% derating
The final output will be the rated power multiplied by both derating factors:
Output Power
50,000 Watts X (1 - .35) X (1 - .036)
50,000 X .65 X .964
31,330 Watts or 31.3 kW @ 4000 m, 35 C

We first calculate the derating factor for the elevation.
Derating
-3.5% X (1900 m actual - 700 m rated)

300m
-0.035 X (1200)
300
.14 or 14%
The necessary rated power is then calculated by:

Power @ rated conditions =
Power REQUIRED
[ 1 - Derating Factor]
20,000 Watts
[ 1 - .14 ]
23,260 Watts or 23.3 kW @ rated conditions
Therefore, in order to have 20 kW available at 1900m, the engine must have a rating of
at least 23.3 kW at the rated altitude of 700m.
18-4
Hybrid Systems

The sizes of the various installed components (array, battery, rectifier and generator)
may be different than the sizes resulting from the "rules of thumb." There are a number
of reasons for these kinds of differences, including:
Customer requirements, including higher availability

Availability of equipment sizes
Different design goals
Different operational strategies
Cost parameters
It is important to remember that the rules of thumb are a rough guide. The hybrid
system designer still needs to evaluate the sizing and selection of components to meet
his/her specific needs.
18-5
Hybrid Systems
Chapter Nineteen
Installation and
Troubleshooting
As you read this chapter, consider the benefits of planning and the costs of a mistake
once you have begun the installation process. As we have seen throughout this
course, the calculations needed for a typical photovoltaic installation are simple. It is
the proper consideration of all the small details that distinguishes a professional
installation from an amateur one.
Also consider the cost of repair or maintenance visits to a remote site compared to the
cost of initially planning the system for extra reliability. There is never a free lunch,
even with a photovoltaic system. But you can pay a little at the beginning of the life of a
system, or a lot more over time, especially when it is difficult or costly to access the site.
The information presented in this section is intended as a survey of technical topics,
and does not constitute a complete course covering all aspects of installing systems. It
is expected that the student will continue his or her own education, to insure that proper
safety and design concerns are included in any system design, installation,
maintenance and repair.
The technical information and suggestions for installation, operation, use and
maintenance made herein are based on Siemens Solar Industries' knowledge and
experience. They are believed to be reliable, but such information and suggestions do
not constitute a warranty, expressed or implied.
Careful planning, knowledge of electricity, and safe practices will reward the
system installer with a dependable, low maintenance system.

19-1
System Design Installation & Troubleshooting
Safety Issues
The safety record for photovoltaic power systems has been excellent, and we in the
industry want to keep it that way. There is the danger of physical harm from working
with equipment, as well as the danger of electrical shock from working with solar
modules, batteries, inverters, and engine generators.
National Electric Code

In the previous System Wiring chapter we discussed Section 690 of the National
Electric Code in how it addresses photovoltaic power systems. Many issues dealing
with proper design and installation of photovoltaic systems are covered in that section,
and many other issues are referenced from that section to previously written sections
on wiring, grounding, installing, etc. You should become familiar with the Section 690,
or refer any questions regarding proper electrical methods to a trained and
knowledgeable electrician.
Adhering to the NEC helps you avoid the following:
Shock hazards - dry and wet leakage.
Wiring that will deteriorate.
Inadequate or no over-current protection and disconnects.
Non-standardized wiring methods.
Inadequate grounding.
Poor workmanship.

19-2
General Safety Practices

The following are general guidelines for working safely with photovoltaic systems.
Know the system you are working on.
Know the test equipment and tools you are working with.
Never work on a photovoltaic system alone.
Review first aid procedures. Know how to get help.
The best safety system is an alert mind, a skeptical mind, and a slow hand.
Follow National Electrical Code and local codes.

19-3
Safe Practices With Batteries

Batteries deserve their own special attention here because they are potentially
dangerous if improperly handled, installed, or maintained. Dangerous chemicals and
high voltages and currents are potential hazards. Anyone who is working with batteries
must first be familiar with safety procedures and system design. (See also the
ANSI/IEEE standard 937-1987, Recommended Practice for Installation and
Maintenance of Lead-Acid Batteries for Photovoltaic Systems.)
The following suggestions for safety are presented to aid in properly and safely
handling, installing, checking and replacing batteries in photovoltaic power systems. All
batteries should be installed and handled in accordance with the manufacturer's
instructions and local codes and regulations. Observance of proper techniques will help
to insure long battery life, low maintenance and safe system operation.
Personal Safety
Remove any jewelry from around neck or from hands or wrists before working
around batteries.
Use non-metal hard hats to avoid possible electrical shock.
Wear protective clothing while working with batteries. This includes acid-resistant
gloves, apron, and eye protection.
Have fresh water easily accessible in case acid splashes on skin or eyes.
Have baking soda easily accessible in case of acid spills.

19-4
Conditions for Safety in Working with Batteries
Keep open flames and sparks away from battery area.
Discharge body static electricity by touching a grounded conductor before touching

battery terminal posts.
Disconnect battery bank from any sources of charging or discharging before working
on batteries.
Have exposed battery terminals covered by strong non-conducting caps or covers.
Design battery area with adequate ventilation, and adequate protection from the
environment.
Lift batteries only in a manner approved by their manufacturer, and never by the
terminal posts.
Use lifting aides, dollies and other moving aides to transport batteries. Do not
simply pick up batteries and carry over terrain. Plan the transportation process.
Tools for Working with Batteries
Tools should be shorter than the length between terminal posts, to reduce the
possibility of dropping and causing a short circuit.
Have non-working or grip end of metal tools covered with a strong non-conductive
plastic coating or tape.

19-5
Exercise:

!
"#

$%

&'

()
!

*%

!

19-6

!

Planning the Installation

Following is a general discussion of the importance of planning and preparedness for
the installation of PV systems.
The suggested process steps presented here are broken into two phases: the planning
stage followed by the installation. Key to a successful installation is a site visit in the
planning stage to finalize the system design and to determine the layout of the
equipment at the site. By visiting the site first and proper planning, the system designer
will avoid many costly mistakes allowing for a trouble-free installation.
The most important first step in the planning stage is to have proper documentation of
the system design including electrical schematic if available along with information on
the major components: controls, inverter, etc. The designer should be familiar with the
overall system design, the major components and other considerations specific to the
project before visiting the site.
In planning a PV system installation the system designer should account for the
following general and specific considerations:

19-7
General Installation Considerations
Be familiar with all relevant documentation on the system including schematic,

information on the various components, system operation, etc.
Make provisions to meet all permits and codes
Make sure to follow proper installation and operation safety procedures
Plan to consider aesthetics and architectural compatibility
Specific System Considerations
System Sizing and Load Optimization

All pertinent data should be collected and calculations of array and battery size
should be made. The client should be involved in deciding those issues that
involve judgment, such as autonomy period in battery sizing, criticality of the
loads, load sensitivity to voltage ranges and waveform. All efforts should be
made to use efficient loads. The system sizing process may iterate many times
with options examined to arrive with the best design.
Consider shading and aesthetics in choosing the array mounting locations
Plan for lightning protection: grounding, surge arrestors, etc.
Plan to locate batteries, controls, inverters, and electronic components such that
they are protected and access is controlled (locked room).
Plan to protect PV panels from animals, people, and falling objects. Consider
fencing surrounding the array and structure.
Consider the shortest possible wire runs, to minimize voltage drop/wire sizes, and
wire protection, in the placement of equipment.

19-8
Site Visit and Analysis

If at all possible, the designer, installer and client should visit the proposed site for the
system. If no visit is possible, then the designer/installer should have extensive
photographs or maps available to become familiar with the terrain and anticipate any
problems with actually installing the system. A visit will also allow for better planning on
the placement and sizing of equipment as well as the logistics of bringing equipment to
the site. Array mounting or location may be greatly influenced by the solar access at
the site, and this may affect later equipment considerations and sizing calculations.
Equipment for Site Analysis

When visiting the site it is recommended to take along the following tools.
a solar siting device to determine shading for various times of the year
a compass for azimuth
a measuring tape and a notebook
Solar Siting Device

A solar siting device is a device that allows you to view the path of the sun throughout
the year and also any objects that might block the sun and cast shadows on the array.
An example of a solar siting device is shown in the figure.
The dome allows viewing of the entire horizon without having to look directly at the sun.
By looking down on the dome, the installer can see the outline of any trees, structures,
or other objects that might obscure the suns path.
Specially designed cards slip under the dome and indicate the hourly position of the
sun for each month of the year. The installer can move the siting device around at the
site to find a location where the array will not be shaded during peak hours of insolation,
usually between 9:00 a.m. and 3:00 p.m. during the shortest day of the year.

19-9

The device can even be used to estimate quantitatively the amount of shading that
might occur due to objects in the suns path. Values are assigned to each hourly
segment that indicate the percentage of total solar path contained in each hourly
segment. Since the sun is up for fewer hours during the winter, each hour of the suns
path during those months will constitute a greater percentage of the whole than an hour
segment during the summer.
The installer can draw the horizon onto the card, including any objects that intercept the
suns path. Each hourly segment that is obscured during a particular month can be
added up to give the total percentage of the suns path that is blocked. This can then
be subtracted from the expected total solar insolation for a typical day during that month
to give the real insolation that would be available. But all effort should be made at the
outset to site the array away from objects that block the suns path.

19-10
Array Spacing From Objects

One particular aspect to examine on the site visit is the best location for the array.
There may be a conflict between the desire to keep the array as close to the battery
and loads (to minimize voltage losses in the wires) and the best location for receiving
solar radiation. Structures, trees, chimney, fences, and other potential shading objects
should be avoided. Also the prospect of vandalism should be considered, and may
impact the location of the array.
Finding room for placement is usually not an issue for remote sites where ample space
and options for locating may be available. For restricted rooftop mounting, or for large
arrays where land cost is a factor, a detailed analysis of land cost vs. power loss may
be advisable.
Locate Array Away From

Tall Objects
s un
The proper distance

depends on the latitude,
the time, and the height of
the nearest tall object
h e ig h t
a rra y
c ritic a l
a ng le
d is ta n c e

An array should be placed far enough away from an object so that there is no shading
between the hours of best insolation, usually from 9:00 a.m. to 3:00 p.m., on the day
with the longest shadows, December 21 in the Northern Hemisphere and June 21 in the
Southern Hemisphere. It is very helpful to have a solar site locator to determine
this (as discussed earlier). If you do not have access to a siting device, use the
following calculations to insure that the array will be located away from potential
shading.

19-11
A graph is presented to aid in deciding where to place an array for no winter shading.
Read up from the latitude of the site to the curve for the hour when no shadows are to
reach the array (again usually 9:00 a.m.). Then read across to find the Spacing Factor.
Multiply this factor times the height of the object to calculate the distance the array must
be placed away from the object.
Distance from object to array
Object Height x Spacing Factor
Spacing Factor For

No Array Shading
Spacing Factor
24
20
16
12
8
4
0
0
20
40
60
Latitude (degrees)
7 a.m.
8 a.m.
9 a.m.

Example:
An array is to be located near Las Vegas, Nevada (latitude 36 degrees

north) and no shadows are to reach the array after 9 a.m. on
December 21.
Reading up from 36 degrees to the 9 a.m. curve, the spacing factor is 2.
If a nearby housing structure is 15 feet high, the distance from the house
to the array is given by
Distance = 2 x 15 = 30 feet away.

19-12
Array Orientation and Azimuth

Normally, an array should face true south in the Northern Hemisphere, not just towards
the magnetic south indicated by a compass (and true north in the Southern
Hemisphere). The deviation of magnetic north from true north is called the magnetic
declination. Consult local magnetic declination maps to find the azimuth correction that
should be applied to a compass reading for your particular site.
The declination map for the United States is shown below. The map shows east and
west declinations. For example, if your location is in the region of the map where the
east declinations are shown, you should say the north point of a compass points
about XXX degrees east of true north. So true north is actually that many degrees
west of where the compass points to North.
Example:
The northern pan handle of Texas falls between the 10o E and 12 oE
lines of the map. This means that the north point of a compass points
about 11o EAST of true north. So true north would actually be about
11o WEST of where the needle points.


19-13
Array Considerations
When laying out the array, the system designer should consider the following:
Mechanical Considerations
Aesthetics

Must support all loads including wind and snow

Should maintain strength over array lifetime
Provide access for array maintenance
Protect array from animals and vandalism
(e.g. install fencing around array site)
Provide ample space for air cooling
Avoid the use of dissimilar metals
Use appropriate wiring and protection
Minimize electrical circuit losses such as voltage
drop
Select equipment based on expected moisture,
temperature, radiation, and salt environmental
conditions
Consider safety, lightning protection, and
grounding
Blend array with building lines and colors
Tilt array at slope and orientation
Avoid harsh contrasts and patterns
Make mounting hardware inconspicuous
19-14
Equipment Location
The site visit is also an opportunity to plan the location of the control equipment, battery
bank, and other wiring concerns. Ease of access to the equipment and safety should
be considered. Building new cabinets or structures, or modifications of existing ones,
may be planned at this time.
The location of a ground rod or locating of a suitable grounding pipe should be planned.
The quality of the soil may impact the type of grounding scheme used. A great distance
between the array and the control and battery systems may cause the designer to
consider two ground rods, one at the array and another near the battery and controls.
This would impact the grounding wiring plan.
Equipment location is very important. One should consider the following:
Install controls, power conditioning equipment, instruments, and batteries in such a

way that access is controlled
Plan to post a clearly written method for disconnecting power to equipment before
servicing and for reconnecting
Consider that the batteries will need to be vented and protected from sparks. Do
not locate the controller or inverter over the batteries
Do not locate an engine generator in the same room as the batteries if they are
vented lead-acid type
Plan for properly sized electrical disconnect switch or circuit breaker for all source
circuits and to isolate major components such as the controller and the array

19-15
Foundation Design
Purpose of Foundations
The purpose of a foundation is (a) to prevent the array from sinking; (b) to maintain
proper array orientation with respect to the sun; and (c) to prevent the array from
"uprooting" during severe winds. Many factors are involved when choosing an array
foundation: site accessibility, local topography, soil properties, local building codes, and
availability of labor.
Types of Foundations
The slab type foundation requires a large mass of concrete and relatively flat terrain.
The slab can be poured at the site or prefabricated slabs can be transported to the site.
It is thus not well suited for remote applications where the cost of transporting cement
to the site is prohibitive. Nor is it desirable in very rugged terrain because of the site
excavation that would be required before placing the foundation.
The block type foundation is most ideally suited for rugged terrain and remote locations
because it is relatively light and transportable, and can be prefabricated wherever
cement and equipment are available. Little excavation is needed and the blocks can
usually be positioned fairly easily, minimizing alignment problems. Blocks must be
assembled with reinforcing steel, and all cavities fully filled with concrete or mortar.
A variation of the block type foundation is to use sauna tubes (with concrete) or wooden
pressure treated posts to make a beam foundation. In this type the tubes or posts are
set in below the frost line. Beams of wood, steel or aluminum create a frame across
the tops of the posts and tubes. Bolted to this frame is the module mounting hardware,
typically a standoff type. The beam foundation is a compromise between the slab and
block types. It is well suited for "rolling hill" terrain, and provides easy alignment
between adjacent arrays.
There is no foundation that is applicable to all situations all the time, for any such
foundation would have to be grossly over designed and would be uneconomical in most
situations. The foundations discussed represent the most common designs used for
PV power systems. Although variations in these foundations exist, one of them will
generally be suitable for a particular application.
If a pedestal or tracker structure is to be used, consult the manufacturer for the
recommended foundation. Often simply setting the vertical pole in a hole filled with
concrete will be enough.

19-16
Block or Beam Type Foundation

Slab Type Foundation


19-17
Loading Values
The foundation weight must be determined as a part of the foundation design process.
The weight depends upon expected array loading and soil type.
Array loading includes wind force pushing against or lifting the array, snow pack that
might accumulate on the array, and the dead weight of the modules. The Siemens
Solar Standard Support Structure has been designed to withstand 50 lb./ ft2 loading
acting directly on the solar panels. It would take a 140-mph wind (225 km/h) to produce
this force or one meter of snow. A foundation should be designed to support an array
during these "worst load" conditions.
Forces on Foundation from Standard Support Structure
R4
R1
R2
R3

The reaction forces to this loading are shown below, both for horizontal and vertical
forces. The values shown are not the required foundation weights, but are the forces
that will be transmitted to the foundation during worst load conditions. For example, if
an 8-module structure at 45-degree tilt was subject to worst load conditions, the vertical
force transmitted to one front foot would be 75 lbs. and a rear foot would experience
700 lbs. vertical force. The front foot would be pushed horizontally with 325 lbs. and
the rear foot would feel 300 lbs. push horizontally. To remain stationary, the foundation
must react with at least this much force in the opposite direction.
19-18
The reaction forces have both positive and negative values because the forces act in
both directions. The forces will be in one direction when the array is loaded from the
front, but will act in the opposite direction when the loading is on the back of the array.
Vertical forces are resisted by the sheer weight of the foundation. Resistance to
horizontal forces is related to soil density, cohesiveness, and soil aggregate and to
foundation weight and design. In general, an engineering soil analysis may need to be
conducted on site, and the results analyzed to determine required foundation weight
and design.
The values presented below are not weights, but the force that the foundation must
resist. The final weight and shape of the foundation to resist these forces are derived
from knowing the soil characteristics as well as the bulk weight of the foundation
members. All values are for both directions with respect to the R vectors.
LOAD FORCES ON FOUNDATION

Tilt Angle
R1
R2
R3
R4
525
425
350
325
300
300
275
275
300
275
325
250
250
250
250
250
250
250
250
250
275
325
525
125
150
175
225
250
300
325
375
425
400
575
675
700
700
700
700
700
700
725
725
850
4-Module Structure
15
20
25
30
35
40
45
50
55
60
65
175
150
150
125
100
75
50
50
50
50
125
425
300
200
125
50
50
50
50
75
100
125
8-Module Structure
15
20
25
30
35
40
45
50
55
60
65
250
125
100
50
50
50
75
150
200
275
475
300
175
200
250
275
300
325
350
350
350
425
!"
#$%

19-19
Array Mounting Configurations

There are a number of ways to mount photovoltaic panels. The two major types are
ground mounts and roof mounts. Ground mounts are usually done in rack mount or on
a pedestal, as is the case for tracking. Roof mounts are either standoff or directly
integrated onto the roof.
Ground Mount - Rack
Most common and preferred mounting configuration

Works well also on flat roofs
Fixed azimuth orientation that can be adjusted for
optimum tilt
New construction or retrofit
Used for smaller arrays
Allows flexible tilt and azimuth orientation

Can be used on rough terrain
Wind loading must be considered
Roof Mount - Standoff
Mount arrays > 8 cm above roof to allow air flow

Minimize number of roof attachments
Hinged panels provide safe accessibility
Pre-assembled sub arrays can minimize installation
time
Roof Mount - Direct
Attach directly to roofing felt

Operates 15 to 30 degrees C hotter than equivalent
rack mount
May work well with new thin film materials
Ground Mount Pedestal or Tracker
!"
&$

19-20
Battery Location and Compartment

Design
Batteries should be located in a dry location out of the elements. They should be on a
level surface that is structurally sound for the weight required. When batteries are
racked they can weigh several tons over a small surface area (< 10 ft2 or 1 m2).
Batteries should not be in contact with cold surfaces such as concrete, and should also
be located where access is controlled such as a lockable room or battery compartment.
The battery should be enclosed in a container, box or room that allows for: (a) proper
ventilation of evolved gasses; (b) safe and easy access for maintenance and
replacement; (c) reduced exposure to environmental extremes and temperature swings;
(d) restricted access to untrained or unauthorized persons; and (e) isolation from
potential hazards such as sparks from generators.
If the temperature of the site will get very cold, burying the battery compartment should
be considered. An insulated space below the frost line will keep the batteries at a more
reasonable temperature throughout the year. The space should be heavily insulated
and made watertight.
The battery compartment should be vented to meet the relevant code requirements. A
large battery bank may require fans, while a small battery bank may need only a vent
hole or tube. Filling the vent tube with loosely packed fiberglass or other suitable
material is one way to allow escape of hydrogen gas without allowing gusts of wind from
outside to cool the chamber. Another means of venting batteries is to drill a hole in
each battery cap and attach it to a plastic tube that is then vented to the outside through
a central receiving pipe. Recombination caps or catalytic recombiners will minimize
battery gassing and reduce the need for ventilation.
A battery box can be built using plywood and rigid insulation or other suitable materials.
It should allow access for maintenance, and allow safe lifting or sliding for removal of
batteries when necessary. The box should be closed with a lock or other means to
prevent casual contact by unauthorized persons. It is good practice to place batteries
in a plastic vessel to contain any spills of the electrolyte.
In many small systems, batteries are placed outside in weatherproof or rainproof metal
or fiberglass boxes (e.g. NEMA 3r, NEMA 4x). In these cases, it is important to
insulate the boxes to minimize temperature extremes that can reduce battery life.
Venting may be necessary if the batteries are a flooded lead acid type. It is important
that the batteries are isolated from HOT electrical components that might ignite the
hydrogen gas.

19-21
System Component Pre-Wiring and

Testing
For all but the very small and simple systems, it is a good practice to pre-wire as much
of the system as possible. Pre-wiring forces the installer to plan the installation to fit the
location, and to have all pieces and tools required for a safe and correct assembly.
Faulty equipment can be identified and replaced, mistakes can be avoided and design
improvements can be made.
Electrical Connections
All wire connections, switches, receptacles and breakers should be mounted in
electrical boxes. Following are the common ways of connecting wire.
Wire or Conduit to a Box
Wire to Wire
Wire to Terminal

Use strain relief connectors with round wire such

as Type SO if the wire will be pulled.
Use cable connector or Romex connector for nonmetallic sheathed cable (NM).
Use conduit couplings to connect lengths of
conduit.
Three acceptable wire-to-wire connections are
crimp, solder, or wire nuts. Crimping with solder is
preferred.
All wire to wire connections must be mounted in an
electrical box.
When crimping, use heaving duty industrial type
crimpers.
For large wire or when multiple wires are
connected to one terminal, use screw lugs or bolton connectors.
For small wire use ring terminals and spade
terminals (non-insulated are preferred).
Spade terminals can be used in non-vibrating
applications, ring terminals for connections that
can loosen.
Crimping and soldering the ring or spade is
recommended particularly in areas of high
humidity.
When crimping, use heaving duty industrial type
crimpers.
19-22
!" '(
')!*$
Control Center
A primary candidate for pre-wiring is all the control and main safety equipment. This
typically includes the main disconnect switches for the array and battery, as well as the
charge controller(s) and any alarms and metering equipment, and AC and DC
distribution boxes or load centers. One of the great advantages of pre-wired and codeapproved power centers such as the products made by Ananda Power Technologies is
that you do not have to pre-wire and test all the main control and safety equipment.
It is often desirable to pre-wire and actually pre-assemble these components onto a
plywood board for transportation to the site as an assembled unit. Careful
measurements must have been taken at the site to insure the board will fit properly with
the required safe spacing from walls or other equipment.
Even if transporting to the site fully assembled is not possible, pre-assembly is a good
way to test the overall layout and design, and insures that all details are checked. Ends
of wires can be wrapped with a number or letter code, and a legend made, so that in
the field the proper connections can be made quickly.
Locate the control center to avoid temperature extremes. Use insulated enclosures for
cold weather, shaded enclosure for hot weather. Many small system controllers are
mounted in rainproof boxes such as a NEMA 3r. This should be separate from any
battery compartment unless you are using maintenance free batteries that do not emit
gasses.
Array
The array should be pre-assembled and wired, to again insure that all hardware is
included. Module interconnect wires can be pre-cut and have soldered crimp
connections or wire stripped ready for installation into the module junction box. The
path for the main wires from an array can be planned, and means for securing these
wires against wind can be planned. The method for grounding the array and mounting
structure should be worked out.

19-23
Battery
Planning for batteries involve how to safely handle and move them. Due to their large
weight, it may be necessary to use a mechanical lift to move them. In addition, it is
important that there is sufficient width allowed to move batteries in and out without
damaging them.
It is best to have batteries delivered a few days before installation. The battery supplier
should charge them before delivery. Before taking them to the site, the installer should
verify their state of charge and if necessary bring all batteries to an equal full charge.
This is also the best time to screen batteries. Quality control can be a problem with
batteries in many Third World countries. By checking voltages and observing how they
take a charge, batteries can be screened for weak cells and poor performance before
placing them in the field.
Pre-wiring of the battery bank insures that proper lugs or clamps are obtained and presoldered to heavy gauge wire. Also a method can be practiced for attaching any
temperature sensor to the battery cases, for charge controller temperature
compensation.
Other Wiring
The full system of array, battery, controls, loads, and inverters should be planned. Wire
and conduit can be obtained and cut, and interconnects and fittings acquired. The
mounting and wiring of any inverter can be planned. An AC and/or DC load center or
distribution box can be obtained and pre-wired with appropriate circuit breakers or
fuses. All switches for loads can be examined for the best way to interconnect the
wires. Using closed lugs on the switches will assure reliable operation even if the
connecting screws loosen slightly.

19-24
Planning The Logistics Of Installation

Besides the equipment to be installed, there are other elements that must be taken into
account for a successful installation. These primarily concern the logistics on how to
get there, how to get the equipment there, and the planning of what will get done, when,
and by whom.
If the installation is to be done at some remote site where access to comforts and
plentiful tools will not be possible (often the case for PV power systems!), it becomes
even more critical that all the logistics of the work be thought out ahead of time.
Transportation Plan
Appropriate transportation equipment must be obtained. The vehicles must be large
enough to carry all the equipment in a safe and secure manner. All possibilities for
weather should be considered to be prepared for snow, wind storms, rain, and rough
terrain or other problems.
There may be a need for small lifting and transporting equipment such as two-wheel
hand trucks or dollies to move the batteries or inverters.
Human Needs Plan

Planning for human needs can include sleeping and housing arrangements, taking
adequate food and water or insuring that such will be easily available near the site.
First aid equipment should be transported.

19-25
Tool Selection
Some tools for installing and maintaining photovoltaic power systems are listed below.
Of course other equipment for building and construction will be useful. Careful planning
of the installation process will help prevent loss of time due to missing tools and
accessories.
Generally Required or Advised
Optional
Digital Voltmeter, with at least 10-amp

current capability, spare batteries, external
50-100 amp current shunt.
Temperature probe for voltmeter
Clamp-on DC ammeter.
Soldering Iron and solder
Compass and Solar Site Locator
Knife
Tilt Angle Indicator, or plumb line and

protractor
Hammer
Light meter
Socket Set, English and Metric

Tape Measure
Metal Hack Saw, Wood Saw
Wrenches: Specific sizes, for all mounting
bolts; Adjustable, for unexpected on-site
problems; Vise-Grips, for variable and heavy
duty
Chisel
AC Amp meter
Screw Drivers: Flat Blade, in sizes for all

mounting hardware; Phillips, in sizes for all
mounting hardware; Small jewelry size, for
adjusting controls
Wire Crimping, Stripping and Cutting Tool(s)
Electrical Tape
Miscellaneous for connections: Split bolts,
wire nuts, lugs, solder-less connectors
!"


19-26
Installation Processes
As opposed to the pre-installation processes at some comfortable location, the actual
installation process involves working quickly and safely at the installation site to install
and commission the PV power system and loads.
Foundation Alignment
The most critical element of foundation installation is alignment. The points where the
mounting structure is to meet the foundation must be level and any mounting bolts must
be spaced correctly. It is critical that careful measurements be made both for spacing
and for flatness.
The orientation of the foundation must face true (solar) south (unless local climate or
load factors require facing away). Care should be taken to adjust for magnetic
corrections to compass readings.
Array Assembly Procedure

Once the mounting structure is in place, it is time to attach the panels and wire them
together with interconnects. Make sure you have thought through your wiring plan
before starting and, if necessary, refer to a schematic. It is important at this stage to
remind oneself about safety. Large arrays can produce currents sufficient to shock
and/or hurt you. If the array is exposed to sunlight while wiring, cover the array
with blanket or other opaque covering to reduce the hazard of electrical shock.
Be careful to follow proper wiring techniques including using correct colors, fasteners,
connections, etc. Mark wires as positive or negative if a color convention cannot be
followed. Leave proper documentation if necessary.
The following general procedure outlines the steps to install Siemens Solar modules on
a Siemens Solar Standard Support Structure, consisting of two long "C" channels and
two adjustable support legs with angle feet. Some of this procedure will also apply to
other mounting, such as on trackers. Consult the manufacturer of any mounting
structure for recommended procedures. If you are using the Siemens Solar Standard
Support Structure, follow the guide enclosed with each structure.

19-27
Exercise:

+ ,

.

/

01
+ &

,

2/

. '
20

34
5

.

.

6

'

'

"

21

2

2

1

!

4
"

7

/

/

"#

8$
3

.

"

0

00


19-28
Step
Array Installation Procedure:
Check that all tools and mounting hardware are available and in
working order. Check that foundation is properly installed and
locations for mounting feet (if applicable) are spaced correctly.
Remove modules from shipping boxes and inspect. Check electrical
output of each module at Voc, Isc and using a 50 watt 6 ohm resistor to
approximate maximum power.
Place modules face down, side-by-side, on a clean flat non-abrasive
surface. The shipping boxes will work as a "table" if desired. Arrange
the modules with the junction boxes located so that series and parallel
wiring will be efficient.
Place the two support structure "C" channels on top of the modules,
parallel, with the "C" facing inward and the mounting holes facing down
aligning with the mounting holes in the modules.
Attach the modules to the structure channels using a 1/4" stainless
steel bolt, a flat washer, a lock washer and a nut in each of four places.
Fasten only finger tight at this point. Continue for all modules on the
structure.
NOTE: Up to 8 modules can be mounted on a 9-foot long structure.
There should be about 1 foot of empty space left between the lowest
module and the ground, to allow for debris to blow free and not build up
against a module surface. Modules should be mounted flush to the top
of the structure. Empty spaces can be filled with a cross-brace,
available from Siemens Solar.
Attach all module interconnect wires. Use sunlight resistant type UF
wire or equivalent. Remove about 5/8" of insulation (3/4" for AWG #8
wire). Do not use wire lugs. Insert stripped wire through foam seal and
under terminal screw slider. Tighten screw securely using a
screwdriver. Attach large sized strain relief clip enclosed with modules.
For small wires (AWG #14 and #12), use the second smaller strain
relief clip also.
Assemble the two rear support legs to the correct length to give the
proper final tilt angle. Use four 3/8" bolts, washers and nuts to fully
assemble each support leg.
Attach four-foot angles to foundation.
Attach each rear support leg to a rear foot angle using 3/8" bolt, flat
washer, locking washer and nut.
Move module assembly into place with modules face up, and attach
bottom of side channels to front foot angles.
Lift up assembly and bring ends of rear support legs up to meet the
assembly. Attach each rear support leg to a side channel using a 3/8"
bolt, washer, locking washer and nut.
Tighten all 1/4" nuts and bolts to 80 in-lbs., and all 3/8" nuts and bolts
to 230 in-lbs.
8
9
10
11
12
!"


19-29
Battery Bank Assembly

First and foremost, review battery safety procedures discussed at the beginning of this
chapter.
Do not mount the batteries directly on a concrete floor. Mount onto wooden or other
non-conducting rails.
Make sure the batteries are fully charged, and that the electrolyte level is at
manufacturer's recommended level. Check all cell voltages and write down on a status
sheet for later comparison.
Handle batteries with extreme care, and use tools carefully. The greatest danger will
occur if wires are hastily connected, or if tools are dropped onto the bare battery
terminals. All connections should be "walked through" a few times, perhaps with
another installer present to confirm, before actual wiring is done.
Make sure to place a sign at the batteries warning unauthorized personnel about the
dangers.
Control Center
Assemble all controls, disconnect switches, alarms and meters and load centers as preplanned.
Make sure there is the required safe space between any boxes and walls or pipes or
other equipment.
Check that all connections are secure and clean, that all wiring is color coded or
marked for correct polarity and that all wire lugs are fastened tight to their wires.
Before hooking up the inverter, make sure the polarity is correct. Most inverters can
be damaged if hooked up with the wrong polarity.
Leave a schematic of the system wiring (secured to the control center if there is
appropriate room) for later maintenance and troubleshooting. Label all main switches
so that an uninformed person with the operation guide can safely and easily disconnect
the system in an emergency.

19-30
Other Wiring
Connect all equipment with proper wire and conduit if necessary.
Make sure all wiring is secured by gently but firmly pulling on all connections. Check
that insulation is not cut during installation work, or when wrapped around corners or
near edges of conduit.
Make sure that the color coding of the wire used is consistent and conforms to local
codes.
Commissioning
The full system should be tested and put into commission before the installers leave the
site.
All safety equipment and switches and circuit breakers should be cycled and tested.
Some controllers and inverters have a very specific hook-up sequence to bring them
on-line. Be prepared to follow that sequence.
All hard-wired loads should be operated and all outlets should be tested.
Restricted equipment should be locked and the keys secured.
A system description with circuit diagrams and an operation/maintenance plan should
be presented to the client. The designer or installer should read through the
operation/maintenance plan with the client. Perhaps have them sign a copy asserting
that they have read the instructions. A copy of the operation/ maintenance plan should
be left with the client. All major components and switch gear should be labeled and be
easily accessible for service.

19-31
Maintenance and Troubleshooting

A maintenance program should be developed and followed for every photovoltaic power
system. The maintenance program should include consideration for structural,
electrical and mechanical components of the system. It should be clearly and simply
written so that the system owner can easily maintain the system.
The maintenance requirements of a typical photovoltaic power system are quite simple
compared to those for a typical conventional power system such as a diesel generator.
There are very few mechanical components to wear out, and little need for adjustments
after the system has been first commissioned. However, like anything placed in the
environment for years, a system will benefit from occasional attention to corrosion and
proper performance.
Maintenance includes procedures that are expected and normal for the system.
Troubleshooting covers procedures that should be followed if system performance is
out of normal tolerances. All safety precautions previously mentioned should be
followed.
Batteries Are The Main Focus

The most important maintenance item for a solar electric system is taking care of the
batteries to maximize their service life. Care of the battery is primarily a matter of
monitoring and maintaining their state of charge. Each battery type requires certain
maintenance. The most commonly used battery is flooded lead-acid and needs to be
watered often and equalized periodically.
A well-designed system will come with some type of voltage or state-of-charge meter.
The system owner should be taught how to use the meter to find out the battery stateof-charge. Ideally batteries kept between 70% and 100% state-of-charge (SOC)
indicates a healthy system and leads to long battery life. Consistently low SOC is
indicative of excessive use of the PV system. Understanding the usage of the battery
(excessive or light) will be very important in determining maintenance intervals, service
life, and will help in diagnosing system problems in the future.

19-32
Sample Maintenance Schedule

A sample maintenance schedule is presented below to indicate typical frequencies of
maintenance actions.
Weekly Maintenance
Observe battery state charge
Monthly Maintenance
Check and add water to the battery electrolyte

Wipe electrolyte residue from top of batteries
Inspect the charge controller for proper indicator light
sequence
Equalize batteries if specific gravity difference between
any adjacent cells is greater than 15 points
(e.g. 1.250 vs. 1.265)
Observe controller for proper indicator light sequence
Inspect array for broken panels. Wash array.
Annual Maintenance
Check array wiring for physical damage and wind

chafing
Check array mounting hardware for tightness
Inspect inverter. Remove dust or dirt.
Inspect system wiring for poor connections. Look for
signs of excessive heating
Inspect battery terminals for corrosion. Clean and
apply anti-oxidant grease as necessary
Inspect controller for proper operation
Verify output from the array
(Isc and Voc, and if possible Imp and Vmp)
!"
+$, $

19-33
Battery Maintenance
Physical Considerations
Observe water level in every battery cell and fill to lever indicated with distilled water.
Inspect all terminals for corrosion and loosened cable connections. Clean and tighten
as necessary. After cleaning, add anti-oxidant to exposed wire and terminals.
Examine outer surface of batteries for film of liquid. This may indicate excessive
gassing, bubbling electrolyte out of battery, it also may indicate a controller problem.
Confirm that batteries are not sitting directly on floor. Some elevation allows for
cleaning and refilling water to flow completely away from battery cases, and prevents
bottom of battery from operating at lower temperature that top of battery.
Make sure the battery enclosure is well ventilated. Inspect any ventilation pathway or
system for blockage.
Electrical Considerations - Determining Battery State of Charge

(SOC)
There is some controversy as to the best way to determine battery state of charge. It is
generally accepted among designers that the most accurate way is to use a hydrometer
to measure the specific gravity of the battery cells (if batteries are vented). This method
is cumbersome, can expose the installer to acid, and can inadvertently lead to
contamination of the cells. As a result, for ease and general accuracy, a voltage
reading (while batteries are at rest, not being charged or discharged) is preferred by
many in the industry. Fortunately more and more controllers are coming with
reasonably accurate voltmeters that make the appropriate adjustments for temperature,
charge and discharge cycle. This is certainly the trend for the future.
To determine specific gravity, remove battery caps and measure specific gravity of all
cells using an accurate hydrometer. Determine the battery state of charge by referring
to manufacturer's literature for the mix of electrolyte specified. It is common in hot
climates to install batteries with a reduced concentration of acid (lower specific gravity)
to reduce internal battery corrosion due to the high temperatures. The greatest
accuracy occurs when corrections for electrolyte temperature are included. Always
wear gloves and eye protection when working with the acid electrolyte.

19-34
To determine voltage, measure the open circuit voltage of each battery cell, or of each
block of cells if internally connected. The best time of day is late afternoon, after the
array has had the most time to charge the battery. Ideally, this should be done after
disconnecting any array and load connections and waiting at least 15 minutes after any
charging or discharging has occurred. Sometimes it is helpful to first discharge the
batteries for a few minutes to remove any surface charge from the battery plates.
Remember to compensate for temperature.
Use the manufacturer's chart of state of charge vs. Voltage (or the figure presented in
the chapter on Battery Technology) to determine the state of charge from the voltage
reading. If the batteries must remain connected, then measure the charging or
discharging current. Use the manufacturer's charts for voltage at different currents to
estimate the battery state of charge.
Presented below are typical values for state-of-charge, specific gravity and voltage, for
lead acid batteries in temperate climates (25 deg C):
Battery Voltage
SOC
Specific Gravity
12 volt
24 volt
100%
1.265
12.68
25.35
90%
1.250
12.60
25.20
80%
1.235
12.52
25.05
70%
1.225
12.44
24.88
60%
1.210
12.36
24.72
50%
1.190
12.28
24.56
40%
1.175
12.20
24.40
30%
1.160
12.10
24.20
20%
1.145
12.00
24.00
10%
1.130
11.85
23.70
0%
1.120
11.70
23.40
!"
!- .-/ $$&'

19-35
Array Maintenance
Physical Considerations
Replace any broken modules. If full replacement is not possible at time of visit check if
the electrical requirements allow for wiring around the broken module, to allow
continued array output without the broken module.
Wash module surface as needed, using soft cloth and water. Perform cleaning in early
morning or wait until evening, to avoid thermal shock to glass. Perform washing only
when modules are not in direct sunlight, when the sun is positioned below the horizon.
Check module tilt angle using tilt indicator or level and protractor. Tilt angle should be
within five degrees of that specified by sizing calculations.
Check module azimuth angle using compass. Compensate for local deviations of true
or solar North from magnetic North. Normally azimuth is true south in the Northern
Hemisphere and true north in the Southern Hemisphere. However different angles may
be chosen depending on local weather conditions and load profile matching. Azimuth
angle should be within 15 degrees of that specified by sizing calculations.
Confirm that no objects cast shadows on the array surface between 9 a.m. and 3 p.m.
Verify that all bolts are secure, and that the mounting structure is well attached to any
foundation.
Examine all wiring connections for corrosion or looseness. Clean and tighten as
necessary.
Check that all junction boxes are covered and that any seals are in place.
Inspect modules for broken cells, de-lamination, and discoloration. These alone may
not result in any noticeable deterioration in electrical output, but may be useful if other
problems are observed.

19-36
It is very helpful to periodically verify the electrical output of the array.
Measure Short Circuit Current (Isc) using clamp-on meter or passing current directly
through meter. Isc is directly proportional to sunlight intensity. Use a reference cell or
light meter to independently check light level. The ratio of light level to 1000 w/m2
should be the ratio of Isc to literature value.
Measure Open Circuit Voltage (Voc) using meter. Voc decreases with increasing
temperature. Use a reference cell or light meter to independently check light level and
a temperature gauge to measure ambient (air) temperature. Use the graph of cell
temperature above ambient temperature to estimate actual cell temperature. Calculate
drop in Voc for operation above 25 deg.C by using voltage factors presented in Chapter
8. Compare calculated value to measured value.
Actual module current will probably not be at literature values, due to low irradiance and
high temperatures. A rough guideline for module current into a battery is given below.
If a module is connected to a 12-volt battery and operating under common outdoor
conditions (irradiance about 800 w/m2 and ambient air temperature about 30 deg.C),
then the following currents should be observed:
Typical Field Values

36 cell module (75 W)
nominal current 4.4 amps
!"

19-37
3.5 amps
2.5 amps
2.4 amps
1.6 amps
Field Measurement Using Resistors

If possible, measure array voltage and current at some intermediate point between Isc
and Voc (preferably near Vmp) by using some dummy load. Use 6-ohm (approximate)
resistor for a single 40-50 watt 12-volt module. Multiply this resistance by the number
of modules in series and divide by the number in parallel to determine resistance to use
for an array. This resistance will operate the array near maximum power. This
measurement gives more information about the curve shape of the array.
Resistancearray =
6 ohms
# Modules in Series
# Modules in Parallel
R esistive Load Sized To

C heck M odule
Current
(amps)
R = Vmp/Imp = 6 ohms
(Vmp, Imp)
2
1
0 0
8
12
16
Voltage (volts)
20

0
R e s is tiv e L o a d S iz e d T o
C h e c k A rra y
R = 6 ohm s
# S e r ie s
# P a r a lle l

19-38
Electronic Equipment Maintenance

Check that all controls, alarms, meters, etc. are securely attached.
Follow manufacturer's procedures to verify that all controls, etc. are functioning properly
and are calibrated. If possible, include verifying the voltage setting of the charge
controller, to insure that batteries are not over- or under-charged.
Inspect all connections for corrosion and loosened wire connections. Clean and tighten
as necessary.
Inspect relay contact points for pitting and corrosion. Replace if damaged.
Load Maintenance
Lubricate all moving parts of motors and loads as recommended by the manufacturer.
Check refrigerator door seals. Clean refrigerator coils.
Clean or replace any air filters.

19-39
Wiring and Safety Equipment

Maintenance
Check that all safety disconnects are functioning.
Examine all wiring connections for corrosion or looseness. Clean and tighten as
necessary.
If possible measure actual voltage drop along wires during normal operation.
Check that all wiring is secured, and inspect insulation for wear, especially at bends and
where secured.
Confirm that all fuses are conducting.
If possible, safely introduce a short circuit path and check that circuit breakers are
functioning.
Verify that all metal objects involved in the electrical system (boxes, raceways, conduit)
are properly grounded, and that the grounding path is continuous all the way to the
ground rod or grounding object.
Optionally check potential between grounding rod and surrounding earth with proper
equipment.
Verify that properly sized wire has been installed. Determine the maximum current that
might flow for each circuit, and the total length of each circuit. Use charts (for example
in Chapter 8) to determine the proper size wire so that there is not more than 5%
voltage drop from array to battery and from battery to load.

19-40
Troubleshooting
The following guide is presented for troubleshooting of generic photovoltaic power
systems. A more detailed and relevant guide can be created for a specific system with
input from the various component manufacturers.
Symptom Chart
Symptom
Probable Cause
Array
Wiring
Controller
Load
Low battery voltage
A,B,C,D,E
F,G,H
I,J,K,M
High battery voltage
Load inoperative
A,B,C,D,E
System erratic or noisy
I,L
F,G,H
I,J,N
F,G,H
I,M
!"
0
Exercise:

( "#

!

7

.


19-41
Corrective Action Chart

Cause
Corrective Action
Array
A
B
C
D
E
Loose wires
Improper wiring
Broken module
Shading
Improper
orientation
Tighten
Reconfigure
Replace
Remove obstruction, or relocate array
Reorient
Excessive voltage
drop
High resistance
connections
Short circuit
Use larger size wire
Wiring
F
G
H
Controller
I
J
K
L
M
N
Loose
connections at
terminal block
High resistance at
relay contacts
Inoperative
Relay fused shut
Electrical
interference
Load shedding
occurred
Clean and tighten

Repair
Tighten
Replace controller
Replace controller
Replace controller
Install filter, or remove source
Wait for battery to recharge, check actual
load usage against sizing calculations.
Load
O
P
Q
Failed
Excessive current
demand
Generating
electrical noise
Replace. Check for improper voltage.

Check actual load usage against sizing
calculations, check load for failure.
Filter the circuit
!"


19-42
Field Evaluation
An example of useful field evaluation procedures is presented next. Field technicians
or engineers investigating the status or performance of an installed system can use
such worksheets. All parameters may not apply in all systems, but these provide you
with the basis for a thorough assessment of how a system was installed and is working.
The first field evaluation worksheet is for complete system evaluation. This can be
used for commissioning a system or for checking a system after some time of
operation. It is especially useful in the context of checking a system installed by some
other organization than your own, either by government agencies checking the quality
of subsidized systems, or by a company called in to service a system installed by some
other organization.
The second field worksheet is for checking the operation set points of a charge
regulator. This simple procedure is needed in the field because the system may not
cycle through its extremes of operational values during a typical field visit within a short
period of time. The field technician or engineer must force the regulator to operate to
its set point values to check for proper operation.

19-43
Solar Photovoltaic System Field Evaluation Worksheet

PRELIMINARY INFORMATION
Date: ____________________
Inspector Name __________________________
System Location:
___________________________________________________________
System Description:
___________________________________________________________
General Application (End Use):
______________________________________________________________________________
______________________________________________________________________________
PHOTOVOLTAIC MODULES AND ARRAY

Manufacturer: _______________
Model #_______________
Array Configuration:
__________ series X __________ parallel
Total Number of Modules:
__________ (#)
Array Mount Design/Type
_____________________________ (ground/roof, rack, tracking)
Array Tilt Angle
__________ (deg)
Array Azimuth _________ (deg)
Seasonal or Daily Adjustable Tilt or Tracking (describe) _________________________________
2
Module/Array Specifications @ STC (1000 W/m , 25 C)

Module
Open-Circuit Voltage (Voc)
__________
Short-Circuit Current (Isc)
__________
Maximum Power Voltage (Vmp) __________ (V)
Maximum Power Current (Imp)
__________
Maximum Power (Pmp)
__________
Array
(V)
(A)
(A)
(W)
__________
__________
__________ (V)
__________
__________
(V)
(A)
(A)
(W)
BATTERY SUBSYSTEM
Battery Manufacturer:
Battery Chemistry/Type
Battery Bank Configuration
Total Number of Batteries
_______________
Model #: _______________
_______________
(antimony, calcium, VRLA)
_______________ series X _______________ parallel
_______________ (#)
o
Battery Specifications at 25 C
Cell/Battery
Capacity
__________ (Ah)
Nominal Voltage
__________ (V)
Bank
__________ (Ah) Rate __________ (C/#)
__________ (V)
Open-Circuit Voltage
at 100% SOC __________ (V)
at 50% SOC __________ (V)
Specific Gravity at 100% SOC __________ (#)
at 50% SOC __________ (#)
Cycle Life
____________ (cycles) __________ (20% DOD)

____________ (cycles) __________ (50% DOD)
____________ (cycles) __________ (80% DOD)

19-44
ELECTRICAL LOADS
1.
2.
3.
4.
5.
6.
7.
8.
Device
______________
______________
______________
______________
______________
______________
______________
______________
DC/AC
__________
__________
__________
__________
__________
__________
__________
__________
Amps/Watts
__________
__________
__________
__________
__________
__________
__________
__________
Hrs/Day
__________
__________
__________
__________
__________
__________
__________
__________
Ah/Day
__________
__________
__________
__________
__________
__________
__________
__________
Seasonal Variation in Load? ___________________________________________________

Special Load Performance (water volume, illumination levels, etc.) _____________________
___________________________________________________________________________
Total Electrical Load

DC Loads: Peak Amps ________ (A) Avg. Daily ________ (Ah)
AC Loads: Peak Watts ________ (W) Avg. Daily ________ (Wh) Avg. Daily ______ (Ah)
Total Average Daily Load DC & AC
________ (Ah)
CHARGE CONTROLLER/VOLTAGE REGULATOR

Manufacturer
Regulation Type
____________________
Model # _______________
____________________ (series, shunt, PWM, constant-voltage)
Controller Specifications:
Nominal Voltage
Maximum Rated Currents: PV Array
________ (V)
________ (A)
Load
________ (A)
Array Regulation Set Points:

VR:
Array Regulation Switching Element:
________ (V)
________ (SS/EM)
VRH:
Leg:
________ (V)
________ (pos/neg)
Load Disconnect Set Points:

LVD:
Load Disconnect Switching Element:
________ (V)
________ (SS, EM)
LVDH: ________ (V)

Leg:
________ (pos/neg)
Temperature Compensation: __________ (mV/ C/cell) Type: ________ (probe/board)

Other Controller Functions/Characteristics (meters, indicators, alarms, etc.)
_________________________________________________________________________________
_________________________________________________________________________________
_________________________________________________________________________________
_________________________________________________________________________________

19-45
INVERTER/POWER CONDITIONER
Manufacturer: ____________________
Model #: _______________
Waveform
____________________ (sine, quasi-sine, square)
Output Voltage Regulation
__________ (%)
Harmonic Distortion
__________ (%)
Minimum Input/Operating Voltage
__________ (V)
LVD Function On Inverter DC Input
__________ yes
________ no
________ (V)
Inverter Specifications (25 C)

Nominal DC Input Voltage
__________ (V)
Nominal AC Output Voltage
__________ (V)
Frequency _________ (Hz)
Maximum Continuous Output Power
__________ (W)
Standby Power Requirement (zero load) __________ (W)
Surge Capability:
__________(W)
________ (minutes)
__________(W)
________ (minutes)
Nominal Efficiency (resistive load)
__________(%) at 25% Load
__________(%) at 50% Load
__________(%) at 75% Load
__________(%) at 100% Load
Other Inverter Features (battery charger, disconnects, over-current protection, etc.):
____________________________________________________________________________________
____________________________________________________________________________________
OTHER SYSTEM SPECIFICATIONS AND INFORMATION

Over-current Protection and Disconnect Devices (list types, locations and ratings):
__________________________________________________________________________________
__________________________________________________________________________________
Grounding (system grounding, equipment grounding, earth electrode):
_________________________________________________________________________________
_________________________________________________________________________________
Surge Protection Devices (list types, locations and ratings):
_________________________________________________________________________________
_________________________________________________________________________________
Blocking and Bypass Diodes (list types, locations and ratings):
__________________________________________________________________________________
__________________________________________________________________________________
Conductors and Wiring (list diameters {sizes}, types, lengths and color codes):
PV Array Interconnects: _____________________________________________________________
PV Array J-Box to Controller: ________________________________________________________
Controller to DC Loads: _____________________________________________________________
Battery to Controller: _______________________________________________________________
Battery to Inverter: ________________________________________________________________
Inverter to AC Loads: _______________________________________________________________
Other Information:
__________________________________________________________________________________
__________________________________________________________________________________

19-46
SYSTEM EVALUATION AND DATA MEASUREMENT RECORD

Photovoltaic Array
Open-Circuit Voltage (Voc)
short-circuit Current (Sic)
Irradiance (sunlight)
Temperatures:
Array
Sub Array
Array
________ (V)
________ (V)
________ (A)
________ (A)
2
________ (W/m )
O
O
________ ( CR) Ambient
________ ( CR)
2
Normalized Array Short-Circuit Current (1000 W/m )

________ (A)
Within Specification?
________ yes ________ no
Array to Battery Charging Current
Battery Voltage
2
Normalized Charging Current (1000 W/m )
________ (A)
________ (V)
________ (A)
Irradiance ________ (W/m )
Battery Subsystem
o
Battery Temperature ________ ( C)

Open-Circuit Cell Voltages:
1. ______ (V) 2. ______ (V)

4. ______ (V) 5. ______ (V)
7. ______ (V) 8. ______ (V)
10.______ (V) 11.______ (V)
Avg. ________ (V)
Minimum ________ (V)
3. ______ (V)
6. ______ (V)
9. ______ (V)
12.______ (V)
Maximum ________ (V)
Specific Gravities:
3. ______ (#)
6. ______ (#)
9. ______ (#)
12.______ (#)
Maximum ________ (V)
1. ______ (#) 2. ______ (#)

4. ______ (#) 5. ______ (#)
7. ______ (#) 8. ______ (#)
10.______ (#) 11.______ (#)
Avg. ________ (V)
Minimum ________ (V)
Need for Equalization (difference of more than 0.15 SG): ________________________

Battery State of Charge Estimate ________ (%)
Battery Load Test (C/1-C/0.3, 5% of capacity):
________ (Voc)
________ (A)
________ (Vend)
________ (min)
________ (Vrest, min)
Electrical Loads
DC Load Currents 1. ________ (A) 2. ________ (A) 3. ________ (A) Voltage ______ (V)
4. ________ (A) 5. ________ (A) 6. ________ (A)
AC Load Power
1. ________ (W) 2. ________ (W) 3. ________ (W)
Total DC Load Current ________ (A) Total AC Load Power ________ (W)
All Operational? ________ yes ________ no
Special Load Operational Performance (water volume/head, illumination levels, etc.)
_________________________________________________________________________________
19-47
Charge Controller Operation

Voltage Drops Through Controller:
PV to Battery
________ (V) at Current ________ (A)
Battery to Load
________ (V) at Current ________ (A)
Set Point Tests:
Controller Regulation
Load Disconnect
Within Specifications
Quiescent Current
________ (VR)
________ (LVD)
________ yes
________ (A)
________ (VRH)
________ (LVDH)
________ no
Other Functions - Operational Performance: _______________________________________________

____________________________________________________________________________________
____________________________________________________________________________________
Inverter Operation
DC Voltage (V)
__________
__________
__________
DC Current (A)
__________
__________
__________
Standby Power at Zero Load
DC Power (W)
__________
__________
__________
AC Power (W)
__________
__________
__________
Efficiency (%)
__________
__________
__________
__________ (W)
System Operation
Steady-State Operation (no controller regulation, no changes in load or irradiance)
PV Array Current
________ (A)
DC Load Current
________ (A)
Inverter DC Current
________ (A)
Controller Quiescent Current
________ (A)
Sum of all currents should add to zero _________ (check)
Other Noteworthy Comments and Data

____________________________________________________________________________________
____________________________________________________________________________________
____________________________________________________________________________________
____________________________________________________________________________________
____________________________________________________________________________________
____________________________________________________________________________________
____________________________________________________________________________________
___________________________________________________________________________________

19-48
(This page is blank)

19-49
In-Field Measurement of Charge Controller Set Points

Why a Procedure Is Needed to Measure Charge Controller Set Points In the Field
In general, it is not possible to determine charge controller set points in the field during system operation
because:
Often the system will not be in regulation.

More often, the system will not be in a load disconnect condition.
The time need to observe these conditions in actual system operation may be a very long time.
For these reasons, a procedure is needed to force the regulator to operate to its set point values, to allow
the field personnel to witness first-hand whether the regulator is operating within acceptable limits.
Charge Controller Set Points - Review
Regulation Voltage (VR): the maximum voltage the controller allows the battery to reach before the
array is disconnected.
Regulation Voltage Hysteresis (VRH): the voltage span between the VR and the voltage at which
the array is reconnected to the battery.
Low Voltage Disconnect (LVD): the battery voltage at which the controller disconnects the load,
defining the maximum DOD.
Low Voltage Disconnect Hysteresis (LVDH): the voltage between the LVD and the voltage at which
the load is reconnected.
Battery Voltage
Voltage Regulation (VR)

Array Reconnect Voltage (ARV)
Load Reconnect Voltage (LRV)

Low Voltage Disconnect Hysteresis (LVDH)
Low Voltage Load Disconnect (LVD)
Charging
Discharging
Time
19-50
Procedure and Methodology

The charge controller must be isolated from the system to perform set point tests.
Disconnect the PV array, battery and DC loads via switches or circuit breakers provided in the system.
Remove all conductors from the terminals of the charge controller.
Connect a small battery or electrolytic capacitor to the charge controller battery terminals.
The battery voltage must be the same as the nominal charge controller voltage.
If a battery is used, the capacity should be low, approximately 5-10 amp-hours (ensures that the battery can be cycled in a
short period of time)
If a capacitor is used, the working voltage must be above the array open-circuit voltage. The capacitance should be between
5,000 and 50,000 micro-farads.
Observe correct polarity!

Connect a PV module (or series string of modules) to the charge controller array terminals.
The array nominal voltage must be the same as the nominal charge controller voltage.
The module (or series string) should have a low output current, but large enough to quickly charge the small battery if used.
Approximately one to three amp modules should be sufficient.
Observe correct polarity!

Connect a resistive load to the charge controller load terminals.
The resistive load should be large enough to discharge the battery at a high rate, at least C/1 but not greater than C/0.25.
Furthermore, the load must consume a lower current than the test PV array provides.
The charge controller should now be cycling between the voltage regulation and array reconnect set
points.
Connect a voltmeter to the charge controller battery terminals.
The voltmeter must have the capability to record minimum and maximum voltages.
Wait for the controller to complete a full regulation cycle, and note the minimum and maximum voltages
recorded on the meter.
These voltages will be the array regulation (VR) and array reconnect set points.
Next, reset the minimum and maximum voltage recording for the voltmeter.
Disconnect the test array from the charge controller array terminals (or shade the array).
Leave the load connected and adjust the load for a current higher than the PV array current.
Wait until a minimum voltage is recorded.
This will be the load disconnect voltage (LVD).
Reconnect (or uncover) the array and wait until a maximum voltage is recorded.
This will be the load reconnect voltage.

19-51
Charge Controller In-Field Test

Configuration
Charge Controller
Under Test
Test PV Module
or Array
Variable
Resistance
Load
Voltmeter
Small Battery or
Electrolytic Capacitor

19-52
Special Considerations
The relative sizes of the test battery (or capacitor), the PV array and test load must be correct to
ensure the operation of the controller between all set points.
Disconnects and overcurrent devices should be used on the battery connection to the charge
controller during the tests.
Special controller algorithms and timing circuits may require special considerations for these tests.
A variable resistor may facilitate the performance of these tests.
Repeat the test a few times to ensure that accurate data is measured. Average the final results from
all of the measurements.

19-53
[End of Chapter]

19-54
CHAPTER NINETEEN
INSTALLATION AND TROUBLESHOOTING
19-1
Safety Issues
National Electric Code
General Safety Practices
Safe Practices With Batteries
19-2
19-2
19-3
19-4
Planning the Installation

Site Visit and Analysis
Foundation Design
Battery Location and Compartment Design
System Component Pre-Wiring and Testing
Planning The Logistics Of Installation
19-7
19-9
19-16
19-21
19-22
19-25
Installation Processes
Foundation Alignment
Array Assembly Procedure
Battery Bank Assembly
Control Center
Other Wiring
Commissioning
19-27
19-27
19-27
19-30
19-30
19-31
19-31
Maintenance and Troubleshooting

Batteries Are The Main Focus
Sample Maintenance Schedule
Battery Maintenance
Array Maintenance
Electronic Equipment Maintenance
Load Maintenance
Wiring and Safety Equipment Maintenance
Troubleshooting
19-32
19-32
19-33
19-34
19-36
19-39
19-39
19-40
19-41
Field Evaluation
19-43

19-55
Chapter Twenty
Economic Analysis
It is not uncommon to have a request to compare renewable generators (wind, PV
hydro) with non-renewable generators (diesel engines, grid-connected systems). It
is very difficult to compare these technologies fairly since their cost structure is
entirely different. For example, a diesel generator has a low initial cost while a PV
system to supply the same energy requirement is significantly more expensive. On
the other hand, the PV system, once installed, uses no fuel and has very low
maintenance costs, while the diesel generator requires constant purchases of fuel
and maintenance on a regular basis. An approach to economic analysis that looks
at the total life costs of these different systems is required. This is discussed next.
Why Life Cycle Analysis Is Important

Life cycle cost (LCC) analysis is a tool used to compare the ultimate delivered costs
of technologies with different cost structures. Rather than comparing only the initial
(capital) costs or operating costs, life cycle cost analysis seeks to calculate the cost
of delivering a service over the life of the project. This means adding the capital
costs and the operating costs over all the years of operation of the project, then
dividing by the total kilowatt-hours (kWh) of electricity provided. This gives a value in
cost per kWh, no matter what technology is used to deliver the electricity.
The following sections give an overview of the factors and methods used to provide
a life cycle costs analysis for comparing renewable and nonrenewable energy
systems, along with a more detailed explanation of "net present value analysis."

20-1
System Design Economic Analysis
System Costs
To properly account for all the expenses in a power system that operates over many
years, three different types of costs should be analyzed. These are: (1) initial capital
costs of purchasing equipment and installation; (2) recurring costs that occur every
year of operation such as fuel and maintenance; and (3) nonrecurring costs that may
occur on an irregular basis, such as equipment replacement or repairs. Each of
these costs is discussed next.
Capital Costs
Generators
The cost of a generator which uses fossil fuel (gasoline / petrol, diesel, or LPG) as
its fuel source varies depending on the construction of the engine. Typically,
gasoline is used for portable generators because of the high weight / power ratio of
these engines. However, these engines are typically only designed for a lifetime of
2,000 hours or less, and require near continuous maintenance. In addition, the
storage problems (flammability and "shelf life") associated with gasoline make
diesel or LPG the fuel of choice for prime power generation.
Diesel generators are typically specified either as "standby" or "prime" duty cycle
engines. The prime rating is usually less than 90% of the standby rating. In
addition, prime diesels usually have longer maintenance intervals than "standby"
engines, savings costs later in the life cycle process. Typical life of a diesel
generator is 20,000 to 40,000 hours. Single generators are usually sized to meet
annual peak loads, which means that there are typically sized at 6-8 times the
average load.
The cost of generators depends on the fuel, "duty cycle" and size. Gasoline
generators are typically less expensive than diesel or LPG generators. Units
designed for "prime" duty are usually significantly more expensive than those rated
for "contractor" or "standby" duty. Finally, larger generators typically cost much less
than small generators. AA typical example is that a generator which is four times as
large will only cost two as much as a base case generator.
For specific examples, a recent "Grainger" catalog lists a gasoline powered 4 kVA
"contractor" generator at $1,089 (approx. $270 per kVA), while a "standby" diesel
generator rated for 8 kW continuous sells for $7,500 (approx. $940 per kW of
generator capacity) and a 35 kW standby generator sells for approximately $15,000
(approx. $430 per kW).

20-2
BOS costs such as control systems, fuel tanks, fuel containment structures,
generator buildings, redundancy (dual engine systems) and storage / power
electronics (batteries and rectifiers for telecom systems) add additional costs to a
generator system. Translated into terms of $/kW of load, the capital cost of diesel
systems is usually $10K-$30K per kW of load.
PV Modules
PV modules are typically sold in terms of $/W(p) (or "dollars per peak watt") for a
specific module size. Although there are some discounts for bulk purchases, the
cost of a PV system over a wide range of power outputs is much more linear than
the cost of a diesel generator.
A typical PV array will cost $6,000 to $8,000 per rated kilowatt ($36K-$48K per kW
of load), plus array structure and BOS. The array that is required to meet a specific
load is typically sized for worst case solar insolation, and is thus 6-8 times the rating
of the load. Thus a 6 kW(p) PV array might be required to meet a 1 kW load in a
typical environment.
Batteries
PV systems usually have batteries equivalent to 5-10 days of storage, with the
batteries rated to 80% of design capacity. Thus, a 1 kW PV system would have
120-240 kWh of useful capacity, equivalent to 150 to 300 kWh of rated capacity.
Batteries usually cost $125-$250 per kWh of capacity, so the net cost of a battery for
a 1 kW load PV system is $19K-75K.
Diesel based telecom systems usually have batteries rated to 3-8 hours of load, and
so are significantly less expensive than PV batteries.
BOS
A PV system must include a support structure for the PV array, charge controller to
regulate battery charging, and housing for the system battery. A diesel based
telecom system usually includes housing for the telecom equipment, batteries, and
battery chargers and often for the diesel generators as well.
Line Extensions
Capital costs for line extensions are usually based on the distance that the load is
away from an established distribution transformer. Typical line extension costs for a
small system range from $1K to $5K per km, plus the actual connection charges ($12K). Stringent environmental requirements for remote sites can increase the cost of
extension by up to a factor of 10 times.

20-3
Recurring Costs
Recurring costs are those costs that are incurred every year of operation. For a
fossil-fuel generator, this includes fuel and engine maintenance costs, while for a PV
system, it is typically limited to checking the batteries one to four times per year.
Fuel costs are usually segregated from maintenance costs because of a difference
in inflation rates between the two categories. If the system is financed, the recurring
costs include loan payments.
Fuel
Annual fuel costs are dependent on the fuel consumption and the cost of fuel at the
site. Fuel consumption in a gas or diesel generator is a function of how highly the
engine is loaded. For example, a diesel engine that is loaded at 25% of its rated
capacity will still use 40% of its full load fuel consumption. Another way to express
these numbers is as specific fuel consumption or kWh per liter of fuel. A fully loaded
diesel has a specific fuel consumption of approximately 3.1 kWh/L, meaning that it
will deliver 3.1 kWh for every liter of fuel consumed. At 25% load, specific fuel
consumption goes down to approximately 1.9 kWh per liter, while at 12% load the
engine will only deliver approximately 1.2 kWh for every liter of fuel used. This is a
very important factor in comparing renewable and nonrenewable technologies, since
most prime diesel plants operate at an average load of between 12% and 25% of
rated capacity. Well-designed hybrid power systems, however, operate the engine
at 75-95% of its rated load, making them much more fuel-efficient.
The cost of fuel at the site is the sum of the purchase price of fuel plus the delivery
costs. While delivery costs for some sites are negligible, the delivery costs at many
remote sites (mountaintop telecom installations, remote villages, and island
communities) can easily double or even triple the purchase price of fuel.
Maintenance
System maintenance is the other recurring (annual) cost. This category includes
equipment maintenance, site maintenance, system supervision, etc.
For PV and wind systems, the generators typically require very little maintenance.
However, if there are batteries in the systems, these will require inspection and
topping up approximately every 3-6 month.

20-4
Generators have much more complex maintenance requirements. A typical

example is an air-cooled diesel generator used for prime power applications. An
operating manual lists the following Three Levels of Diesel Maintenance
requirements:
Frequency
Task To Perform
250 hours
Oil and filter change.
Qualifications
Can be
performed by a
Also inspect air and fuel filters, trained technician
fuel system, starter battery, and at the site.
system electrical connections.
1500 hours Decarbonization (top end
Can only be
overhaul). In addition to oil
performed by a
change tasks, replace air and
trained engine
fuel filters. Remove the
mechanic, site
cylinder head and clean valves, work is possible.
valve seats, injector nozzles,
etc. Replace all top end
gaskets.
6000 hours Full overhaul. In addition to
Requires a
tasks listed above, perform full competent
engine overall. Clean all
mechanic.
combustion and inlet/outlet
Typically
chambers. Replace all engine
performed in a
gaskets. Replace camshaft /
repair shop which
crankshaft bearings, valves,
means the engine
valve springs, injectors, fuel
must be
pumps, pistons, piston rings,
transported to the
starter battery, start solenoid
shop, and a
and engine starter, fuel
replacement
solenoid, fuel pipes, and other
engine provided
parts as necessary.
to the site.
Cost Estimate
$25 to $100 plus
travel if required.
$250-500 plus
travel for a typical
two-cylinder aircooled engine.
$1,500-3,000,
plus
transportation and
loaner of
replacement
engine.
For a prime power engine which is running continuously (8,760 hours per year), the
sample maintenance schedule listed above would require 35 oil changes, 6
decarbonizations, and 1.5 overhauls per year. Engines made for prime power often
have options that can extend the primary maintenance interval to 500, 1,000 or even
1,500 hours, dramatically reducing the maintenance requirements. Note that many
standby engines have only a 100-hour oil change interval, which would necessitate
nearly 90 oil changes per year almost twice per week!
Loan Payments
If the system is financed through a bank or loan agency, the amortized loan
payments can be treated as a recurring cost rather than a capital cost.

20-5
Non-recurring Costs
The last major category of costs are those which only happen every once in a while.
These include replacement of major components, major system maintenance tasks,
etc. For remote power systems, typically non-recurring costs include replacements
of engines/generators and replacement of system batteries.
Engine Replacement
Generator lifetimes are typically dependent on the construction and fuel type. A
portable gasoline / petrol generator will probably last only 500-2,000 hours. A diesel
generator has an effective lifetime of 20,000 to 50,000 hours, and LPG engines that
are based on diesel construction can last 100,000 hours or more.
Battery Replacement
While many telecom-style batteries are rated for 20 year lifetimes in float
applications, the harsh environments and cycling duty of remote PV or hybrid power
system typically limit useful battery life to 7-10 years at best. Life cycle cost
techniques can be used to help evaluate battery tradeoffs -- e.g. should you use a
high quality battery which will last 14 years, or use a lower cost battery and simply
plan on replacing it every 5-7 years.
Exercise:


20-6
Economic Factors
Life Cycle Term (term)
In performing a life cycle cost analysis, the primary economic factor is the "life cycle" of the
system. The length of "term" of the analysis is chosen to be the service life of the longestlived component. For example, a system which compared the costs of a diesel generator
and PV system over the course of a only a single year would ignore the fact that the PV
system will continue to operate for many years without additional expenses.
In the case of PV system comparisons, the useful life of a PV module is on the order of 2030 years, so this should be the "term" chosen for the analysis. Similarly, a comparison
between standard "prime" diesel generators and "cycle charged" systems would probably be
based on the life of the battery in the cycle charged system, or about 10 years.
Discount Rate (DR)

The next most important factor in NPV analysis is the "discount rate." This is the factor that
describes the changing value of money over time. It is basically equivalent to the amount of
money you could make with your capital if you chose to invest it in a bank or other
investment program rather than in a power system. The discount rate is applied to the price
of something in the future to bring it back to an equivalent value today.
Typical values for the discount rate at 7-15% when used in this type of analysis. In general
higher discount rates make future expenses less important, and favor putting off
expenditures to the future. And in general, lower discount rates keep the impact of future
expenses high, and favor capital intensive projects such as renewable energy systems.
Fuel & General Escalation (FE, GE)

Cost Escalation accounts for the fact that components and services traditionally get more
expensive over time. For remote power systems, this applies to fuel, maintenance costs
and replacement parts. Traditionally, fuel costs are escalated at a separate (and slightly
higher) rate than other costs. This represents the volatility of the international energy
market. While it is very difficult to predict long term trends in fuel costs, it is important to
pick a value that has some real world representation for the specific situation being
analyzed.
Typical annual escalation rates for remote power systems are 5-10% for fuel and 3-8% for
non-fuel expenses. Of course, local economic factors may cause large variations in these
numbers.

20-7
Net Present Value

A mathematical method called "Net Present Value" (NPV) analysis takes into
account the fact that the value of money changes over time. This method seeks
to calculate the costs in terms of "constant" currency. This is often expressed in
"constant 1995 dollars" or some similar expression. The basis of NPV analysis is to
express a series of future costs in terms of present currency, taking into account the
changing value of money.
An item or event that has a cost in the future will probably cost more than the same
item or event today due to general price escalation. However, money invested
today to pay for that future cost would also increase in value, depending on how that
money today is invested. The net present value of an event or item occurring in the
future is in a sense the equivalent amount of money you would need to invest today
to be equal to the cost of the event or item at that future date. The idea of NPV is
best expressed in an example:
Example:
A full tank of fuel for a generator costs $100 today. It will need to be replaced
next year. If the discount rate is 10%, then the amount of money needed
today to pay for $100 of fuel next year will be given by
PV
1
1+ 0.10
$100
$90.91
So $90.91 is the present value of the future $100 needed to pay for the fuel in
one year. If the $90.91 was put aside this year and invested at the discount
rate of 5%, then in one year we would have $100.
If there is price escalation happening as well, then the future cost of the tank
of fuel will be slightly higher than the price today. If the price escalation of
diesel fuel is 5%, then the price in one year will be given by
FV
(1 + 0.05)
$100
$105.00
So you would need to have $105.00 in one year to pay for the fuel.
The NPV of next years fuel will be the escalated price discounted back to the
present. This involves combining the two formulas above.
NPV
(1 + 0.05)
(1+ 0.10)
$100
$95.45
So the net present value of the tank of fuel to be purchased next year is
$95.45 in todays money. Putting this much aside today will increase in value
at the discount rate of 10% and be enough to cover the escalated price of
$105.00 that will be needed one year from now.

20-8
Exercise:

!

!

"

#

$
%
&

'
(
!

)

*

'

!

+

,
*

-.
)

&
/01+!222

/032!222
)
-
&
/0+2!222

/04!+22
* /

)
03222

5$
%$% !

3+6
76

76
16


20-9
LCC Analysis Methodology

The basic life cycle cost analysis method is as follows:
1. Calculate the initial (capital) cost of the system;
2. Calculate the annual recurring fuel cost, then multiply by a factor which accounts
for the discount rate (DR), fuel inflation rate (FE) and life cycle term (term);
3. Calculate the annual recurring maintenance cost, then multiply by a factor which
accounts for the discount rate, non-fuel general inflation rate (GE) and life cycle
term;
4. Determine the schedule for each non-recurring cost and multiply the cost by a
factor to account for the discount rate and inflation rate in the year of occurrence;
5. Add these four costs together and divide by the total number of kWh produced to
determine the life cycle cost of energy.
Capital Costs
Capital costs are input into the formula with a factor of 1.0 since they are assumed
to occur in "year 0," that is, before the start of system operation.

20-10
Recurring Fuel Costs

The formula for life cycle fuel costs is:
1 + FE 1 + FE Term

LCCFuelCost = AnnFuelCost *
1

DR FE 1 + DR
(Eqn 20.1)
where
AnnFuelCost
is the annual fuel expenditure
FE
represents Fuel Escalation
DR
represents Discount Rate, and
Term
is the life cycle term
(Note: If the discount rate is equal to the fuel escalation rate, the first term has a
"division by zero" which will cause it to go to infinity. This can be remedied by using
a number with a very small difference, i.e. 0.06999 instead of 0.07 for a 7% discount
rate. In any event, if the escalation factor is the same as the discount rate, then the
Fuel Factor (the term in the brackets) is simply equal to the life cycle term. So the
total life cycle fuel cost would simply be the annual fuel expenditure times the
number of years in the term. The same concern applies to the Maintenance Factor
described in the next section)
Recurring Maintenance
The formula for life cycle maintenance costs is:
1 + GE 1 + GE Term

LCCMaintCost = AnnMaintCost *
1

DR GE 1 + DR
(Eqn 20.2)
where
AnnFuelCost
is the annual non-fuel expenditure
GE
represents General Escalation
DR
Term
is the life cycle term

20-11
Non-Recurring Costs
The formula for nonrecurring costs is:
LCCReplCost =
1 + GE RY
ItemCost
*
1 + DR
(Eqn 20.3)
where
ItemCost
is the nonrecurring expenditure in

present day costs
GE
represents General Escalation
DR
RY
is the "Replacement Year"
Note that the costs for each non-recurring cost need to be calculated separately,
then summed together to get the total nonrecurring costs.
Final Result - Life Cycle Energy Cost

The formula for the life cycle energy cost is:
LCC$ / kWh =
CapitalCosts + LCCFuelCost + LCCMaintCost + LCCReplCost

Term * 365 * kWh / d
(Eqn 20.4)
where
LCC$/kWh
is the life cycle cost per kWh energy
LCCFuelCost
is calculated above
LCCMainCost
is calculated above
LCCReplCost
is calculated above
Term
is the life cycle cost term, and
kWh/d
is the daily kWh output of the system

20-12
Exercise:

8

!

,

5595
.
0
,

55: 5
.
0
,

55"5
.
0
,

*,
550;*,.
0;*,

5
' <

9<

"
*,
& 5<@
9
<@
:
5

$ A 5

" B
-


'<
9<
"
?,;
55

95

42
=6
>6
346
42
0+2!222
04!222
: 5

0+22
&5
3
032!222
"B
B32
20-13
Sample Analyses
This section presents analysis of four different power system options for a sample
application, a small village power supply. The costs associated with these
calculations are estimated only -- they are not intended to be a formal analysis of the
options presented. Rather, they are intended to show the method used and how a
comparison of various power solution options might be presented. A thorough
evaluation would require more detailed designs, actual system and operating costs,
and a careful selection of economic factors.
The analyses will compare four typical energy solutions:
Pure Standalone PV System

PV Diesel Hybrid System
Prime Diesel System
Utility Grid Extension

20-14
Sample Application Parameters

The sample application is a small rural village in Madras, India. The initial load is
estimated to be 42 kWh AC per day, with a 15 minute peak load of approximately 8
kW, with occasional surges up to 16 kW (recall the continuous and peak power
graph in the chapter on Load Estimation). The site is moderately remote (15 km
from the nearest power line), but accessible for the entire year by dirt road.
Site Factors
Location
Madras,
India
13 deg.N
Load Factors
Daily
Economic Factors
Life Cycle
20 years
Term
Latitude
Annual
15,330
Discount
12%
kWh
Rate
Longitude
80 deg.E Total in
306 MWh Fuel
10%
Life Cycle
Escalation
Distance
15 km
Surge
20-40 kW General
7%
from grid
Estimate
Escalation
Fuel Cost
$0.40 /liter
(delivered)


42 kWh
20-15
PV System
A PV array to meet this load in this location would be approximately 16.5 kWp (4
series X 55 parallel 75 Wp modules). A battery bank with 5 days of autonomy is
recommended with a total rated capacity of 196 kWh DC @ 10 hour rate (4080 Ah
@ 10 hr @ 48 volts).
The major cost during the life of the system would be the replacement of the battery
bank. It is estimated that the battery will need replacement after year 7 and again
after year 14. A small cost for battery maintenance trips during each year is
included.
The resulting LCC analysis shows that the total life cycle (20 year) system cost,
including initial capital and maintenance and battery replacement, comes to
$168,486 NPV (discounted to constant dollars). The system produces 306,000 kWh
in 20 years, for a LCC Cost/Energy ratio of $0.55/kWh.

20-16
L ife C ycle C o st A n alysis

S ys tem D e sc rip tio n
P ure P V
L o c atio n
M ad ra s, Ind ia
In itial C a p ital C o sts

Item
S ola r A rray
B attery B an k ( kW h @ $/W h )
S ystem C on tro ls
In ve rte r ( kW c on tin uo u s @ $ /w att)
B uilding
C ivil W o rks (in sta llation )
R ec u rrin g C o sts
Fuel
--------------
A nn ua l H ou rs
--------------
M ainte na nc e
G en era l
Q ua ntity
2 20
1 96
1
1 0.0
0
0
C os t
$ 42 0.0 0
$ 0.1 4
$ 1,0 00
$ 0.7 0
$0
$0
E xte nd ed C os t
$ 92 ,40 0
$ 27 ,44 0
$ 1,0 00
$ 7,0 00
$0
$0
$0
$0
$0
$0
T o ta l Initia l C ap ita l
$ 12 7,8 40
L ite rs/h ou r
--------------
C os t/u nit
--------------
F re qu en cy/yea r
1 .0
C os t/E ve nt
$ 35 0
A nn ua l C os t
--------------
E xte nd ed C os t
$ 35 0
$0
$0
$0
$0
T o ta l R e currin g C o st
$ 35 0
N o n -R e c u rrin g Item s
Item
C os t (tod ay)
B attery R ep l + 10 % la bo r
$ 30 ,18 4
B attery R ep l + 10 % la bo r
$ 30 ,18 4
Economic Factors
Item
LCC Period (years)
Discount Rate
Fuel Escalation (FE)
General Escalation (FE)
A ve ra g e E n e rg y R eq u ire m e n t
4 2 kW h/da y
Value
20
12%
10%
7%
Y ea r
N P V F ac tor
1
0 .95 5
2
0 .91 3
3
0 .87 2
4
0 .83 3
5
0 .79 6
6
0 .76 0
7
0 .72 6
8
0 .69 4
9
0 .66 3
10
0 .63 3
11
0 .60 5
12
0 .57 8
13
0 .55 2
14
0 .52 8
15
0 .50 4
16
0 .48 2
17
0 .46 0
18
0 .44 0
19
0 .42 0
20
0 .40 1
T o ta l N o n-R ec urrin g C o st
N P V C os t
$0
$0
$0
$0
$0
$0
$0
$ 20 ,94 6
$0
$0
$0
$0
$0
$0
$ 15 ,21 5
$0
$0
$0
$0
$0
$ 36 ,16 1
NPV Factor
16.64
12.82
Cost Summary
Initial Capital
All Non-Recurring Costs
Annual Fuel
Annual Maintenance
$127,840
$36,161
Annual Cost
-------------$350
NPV Factor
16.64
12.82
Total LCC Cost (NPV)

Total Energy (kWh)
Cost/kWh (NPV)

Life Cycle Cost

$127,840
$36,161
$0
$4,485
$168,486
306,600
$0.55
20-17
The graph below illustrates that the array is sized large enough to operate all the
loads even during the lowest insolation month (July), and that there is excess array
energy produced in every other month of the year. The array is deliberately
oversized so that it can adequately operate the load in every month. The oversizing
is evident by the fact that the array potential is larger than the load demand
throughout the year (except for July). The largest oversizing is evident in March,
where the array could have produced almost 40% more energy than was needed by
the load.
No generator is included in the standalone system, so there are no generator starts
in any year.
S ys t e m P e rf o rm a n c e
1
2000
1800
1600
1400
1200
1000
800
600
400
200
0
JAN
0
FEB
MAR
APR
G e n S ta rts
MAY
JUN
JUL
M o nthly Lo a d
AU G
SEP
O CT
A rra y (us e d )
N OV
DEC
A rra y (p o te ntia l)

!

20-18
A cash flow analysis shows that there is a large initial outlay for all the equipment,
and a battery replacement in year eleven. All other costs are small. The total
accumulated cash flow (not discounted) comes to $278,333 over the life of the
system.
C as h F lo w Ana lys is - P u re P V S ys tem

$300,000
$140,000
A c c u m u la te d C o s ts
$ 2 7 8 ,3 3 3
$120,000
$250,000
$100,000
An n u a l C o s t
$200,000
$80,000
$150,000
$60,000
$100,000
$40,000
$50,000
$20,000
$0
$0
1
10 11 12 13 14 15 16 17 18 19 20 21
Ye a r
"#$% &

20-19
Prime Generator
A prime diesel generator power system would include two (redundant) generators,
to extend life and insure reliable power availability. They will be sized at 20 kW
each, but could range from 20-40 kW, to handle the possible surge requirements of
a variable village load profile (recall the discussion of surge requirements for
complex large systems in the chapter on Loads). The lowest of the possible sizes is
chosen here, to be most conservative in the cost comparison.
The generators would be alternated, so they would both age approximately the
same, each operating for 4380 hours per year. The manufacturer would
recommend replacement after 30,000 hours of this relatively continuous operation,
which would occur after 7 years of system operation (30,000 hours / 4380 hours
each generator = 6.8 years), and then again after year 14 of the system life.
Fuel and maintenance costs would add up to a considerable expense during the
system life. The fuel consumption rate is estimated to be 3.0 liters/hour as the
generator is expected to operate on the average at only about 25% load. Operating
for 8760 hours/year at this rate consumes 26,400 liters/year at a cost of $10,775.
Maintenance is expected for oil changes every 250 hours, decoking every 1,500
hours and major overhaul every 6,000 hours (with full replacement needed after
30,000 hours).
including initial capital and fuel, maintenance and generator replacement, comes to
$354,816 (discounted to constant dollars). The system produces 175,200 kWh in 20
years, for a LCC Cost/Energy ratio of $1.16/kWh.

20-20
Life Cycle Cost Analysis

System Description
Prime Diesel
Location
Madras, India
Initial Capital Costs

Item
Diesel Gen: Power (kW) =20
Fuel System
Building
System Controls
Civil Works (installation)
Recurring Costs
Fuel
Diesel fuel (liters)
Maintenance
Oil Change
Decoke
Overhaul
General
Non-Recurring Items
Item
Quantity
2
1
0
1
0
(Sug g e sted g e ne ra tor size ra ng e a nd fue l ra tes

to meet surge requirements
for va ria b le villa g e loa d )
21
42
kW
3.12
5.96
liter/hour @25%
Cost
$12,000
$1,000
$10,000
$1,000
$10,000
Extended Cost
$24,000
$1,000
$0
$1,000
$0
$0
$0
$0
$0
$0
Total Initial Capital
$26,000
Liters/hour
Annual Hours @ 25% Loa d
8760
3.00
3.0
Period (hours) Frequency/year
250
35.0
1500
5.8
6000
1.5
0.0
Cost/unit
$0.41
Annual Cost
$10,775
Extended Cost
$5,256
$1,752
$2,190
$0
$0
Total Recurring Cost
$9,198
Cost (today)
DEG Replace +10% labor
$26,400
DEG Replace + 10%labor
$26,400
Economic Factors
Item
LCC Period (years)
Discount Rate
Average Energy Requirement

42 kWh/day
Value
20
12%
10%
7%
Cost/Event
$150
$300
$1,500
$300
Year
NPV Factor
1
0.955
2
0.913
3
0.872
4
0.833
5
0.796
6
0.760
7
0.726
8
0.694
9
0.663
10
0.633
11
0.605
12
0.578
13
0.552
14
0.528
15
0.504
16
0.482
17
0.460
18
0.440
19
0.420
20
0.401
Total Non-Recurring Cost
NPV Cost
$0
$0
$0
$0
$0
$0
$0
$18,320
$0
$0
$0
$0
$0
$0
$13,307
$0
$0
$0
$0
$0
$31,627
NPV Factor
16.64
12.82
Cost Summary
Initial Capital
Annual Fuel
Annual Maintenance
$26,000
$31,627
Annual Cost
$10,775
$9,198
NPV Factor
16.64
12.82

Total Energy (kWh)
Cost/kWh (NPV)

Life Cycle Cost (NPV)

$26,000
$31,627
$179,314
$117,874
$354,816
306,600
$1.16
20-21
A cash flow analysis shows that there is an initial outlay for all the equipment, and
then generator replacement in year 8 and 15 of system life. The escalated cost of
generators after 7 and 14 years is shown. The fuel and maintenance costs
continually add up during system life, and also increase in cost as their values
escalate over time. The total accumulated cash flow (not discounted) comes to
$420,312 over the life of the system.
This analysis shows the sensitivity of long-term prime diesel system operation cost
to the discount rate value used. There are substantial costs occurring in far future
years. The higher the discount rate, the more discounted the far future costs are
reduced in the LCC type analysis (with all future costs being brought back to initial
constant dollars). So different discount rates can end up drastically affecting the life
cycle cost of such a system with much of its cost occurring in the far future.
C a s h F lo w A n a ly s is - P rim e D ie s e l
$1,400,000
$160,000
$1,226,510
$140,000
$1,200,000
$120,000
$100,000
$800,000
$80,000
$600,000
$60,000
Accumulate d C ost
An nu al C ost
$1,000,000
$400,000
$40,000
$200,000
$20,000
$0
$0
1
10 11 12 13 14 15 16 17 18 19 20 21
Y e ar
' ( "#$% &

20-22
PV-Diesel Hybrid
The hybrid system designed for this example included enough solar array to supply
40% of the annual energy load, with the generator operating to provide the
remaining energy throughout the year. The battery needed for such a system can
be quite small, as there is no need for many days of autonomy because the
generator is available on demand. A battery bank of 2 days of autonomy was
used, with the generator turning on and recharging the battery whenever it dropped
to 75% depth of discharge (deep cycling batteries). This resulted in a total battery
capacity of only 101 kWh (C/10 rate), a reduction of 50% of the capacity needed for
the Pure PV system.
One generator is installed with a capacity of 13 kW, sized based on charging the
battery bank at its maximum allowable charge rate. The fuel consumption rate is
estimated to be 4.8 liters/hour, as the generator is expected to operate at full load
when it is needed to recharge the battery. This is of course larger than the rate used
for the prime diesel example, because the generator was only operating on the
average at about 1/4 of full load. The hours of operation at this full rate are now
estimated to be only about 1296 hours/year (instead of 8760 hours), only about 15%
of the amount of hours for the prime diesel system example.
Operating for 1296 hours/year at this full power rate consumes 6220 liters/year at a
cost of $2,550. This is less than 25% of the cost for the prime diesel system.
Maintenance is estimated at the same periods as for the prime diesel system, with
oil changes every 250 hours, decoking every 1,500 hours and major overhaul every
6,000 hours. Maintenance costs are estimated to come to $1253/year, less than
15% of the prime diesel costs.
At the end of the full system life term of 20 years, the diesel is expected to have
operated on 25,920 hours total (20 years X 1296 hours/year), less than the
estimated life of 30,000 hours. So the generator will not have to be replaced during
the system life, compared to two replacements during the prime diesel systems life.
The battery life is expected to be similar to the pure standalone system, due to the
more intensive deep cycling of the batteries but less sulfation resulting from less
time at partial state of charge. The generator is expected to start a total of about
108 times per year, so the batteries will cycle down to 75% depth of discharge at
least that many times. This would mean about 756 such deep cycles in 7 years.
Although some battery manufacturer literature states longer expected cycle life to
this depth of discharge, the accumulated loss of life over real time was estimated to
result in the need to replace the batteries after year 7 and again after year 14 of the
system life.
including initial capital and fuel, maintenance and battery replacement, comes to
$141,300 (discounted to constant dollars). The system produces 175,200 kWh in 20
years, for a LCC Cost/Energy ratio of $0.46/kWh.
20-23
PV-Genera tor Hybrid System Sizing Report

Loc a tion Info rm a tion Whid b e e Isla nd , WA.
Arra y Inform a tio n:
Mo d ule Qua ntity

Mo d ule Typ e
Arra y Po w er
76
SP75
5.7 kW
Ge ne ra tor Info:
Pow e r:
Fue l Co nsum p tion
Outp ut De ra ting Fa c to r
13 kW
4.80 lite rs/ ho ur
100 %
Ba tte ry Inform a tio n: Da ys of Auto no m y:

Ba tte ry Vo lta g e Effic ienc y
Ba tte ry Cha rg e Effic ienc y
Ba tte ry Energ y Ca p a c ity
(a d just b y 1.3 to C/ 10 ra te )
Ba tte ry Cha rg e Ca p a c ity
Ba tte ry No m ina l Volta g e
Ba tte ry Ma x. DOD
2
88
90
132
101
2745
48
75
Inverter Inform a tio n: Effic ie nc y:
85 %
Re c tifier Inform a tion: Effic ie nc y:

Outp ut Ra ting :
80 %
275 a m p s
Da ily
Mo nthly Arra y kWh Ge n kWh
Loa d kWh Loa d kWh Sup p ly
Sup p ly
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NO V
DEC
ANNUAL To ta ls:
PV/ Diesel Ra tio
42.0
42.0
42.0
42.0
42.0
42.0
42.0
42.0
42.0
42.0
42.0
42.0
d a ys
%
%
kWh @ C/ 108
kWh @ C/ 10
Ah
Vo lts
%
# G en
Sta rts
Ho urs of
Fue l
Op era tionUse d (liters)
1302.0
1176.0
1302.0
1260.0
1302.0
1260.0
1302.0
1302.0
1260.0
1302.0
1260.0
1302.0
562.3
553.5
627.2
579.1
545.5
453.5
422.9
461.4
486.1
509.5
481.4
509.5
672.0
672.0
672.0
672.0
756.0
756.0
924.0
840.0
756.0
840.0
756.0
756.0
8
8
8
8
9
9
11
10
9
10
9
9
96.0
96.0
96.0
96.0
108.0
108.0
132.0
120.0
108.0
120.0
108.0
108.0
461
461
461
461
518
518
633
576
518
576
518
518
15330
6192
9072
108
1296
6220
PV 40%

Ge ne ra tor 59%
20-24

System Description
Hybrid System
Location
Madras, India

42 kWh/day

Item
Diesel Gen Power (kW) =13
Fuel System
Building
Rectifier
Solar Array
Battery Bank ( kWh @ $/Wh , 10 hr rate )
System Controls
Inverter ( kW continuous @ $/W )
Civil Works (installation)
Quantity
1
1
0
1
76
101
1
10.0
0
Recurring Costs
Fuel
Diesel fuel
Annual Hours
1296
Liters/hr
4.80
Maintenance
Oil Change
Decoke
Overhaul
General
Period (hours) Frequency / year

250
5.2
1500
0.9
6000
0.2
0.0
Non-Recurring Items
Item
Cost
$6,000
$1,000
$10,000
$0
$420.00
$0.14
$1,000
$1.00
$10,000
Extended Cost
$6,000
$1,000
$0
$0
$31,920
$14,190
$1,000
$10,000
$0
$0
$64,110
Cost/liter
$0.41
Annual Cost
$2,550
Extended Cost
$777
$259
$216
$0
$0
$1,253
Cost (today)
Battery Repl + 10% labor
$15,609
Battery Repl + 10% labor
$15,609
Economic Factors
Item
LCC Period (years)
Discount Rate
Value
20
12%
10%
7%
Cost/Event
$150
$300
$1,000
$300
Year
NPV Factor
1
0.955
2
0.913
3
0.872
4
0.833
5
0.796
6
0.760
7
0.726
8
0.694
9
0.663
10
0.633
11
0.605
12
0.578
13
0.552
14
0.528
15
0.504
16
0.482
17
0.460
18
0.440
19
0.420
20
0.401
NPV Cost
$0
$0
$0
$0
$0
$0
$0
$10,832
$0
$0
$0
$0
$0
$0
$7,868
$0
$0
$0
$0
$0
$18,700
NPV Factor
16.64
12.82
Cost Summary
Initial Capital
Annual Fuel
Annual Maintenance
$64,110
$18,700
Annual Cost
$2,550
$1,253
NPV Factor
16.64
12.82

Total Energy (kwh)
Cost / kWh (NPV)

20-25
Life Cycle Cost

$64,110
$18,700
$42,438
$16,052
$141,300
306,600
$0.46
The graph below shows the array and generator performance over a typical year.
The array is sized to meet only 40% of the load, so the generator is operated every
month, varying from 8 starts in January-April (about once every 3-4 days) to 11
starts in July (about once every 2-3 days). All of the array potential output is utilized,
in contrast to the pure standalone system where the array met the load in July and
was oversized all the rest of the year, peaking at nearly 40% oversize in March.
System Performance
1400
1200
1000
800
600
400
200
0
JAN
FEB
MAR
Gen Starts
APR
MAY
JUN
JUL
AUG
Monthly Load
SEP
OCT
Array (used)
NOV
12
11
10
9
8
7
6
5
4
3
2
1
0
DEC
Array (potential)
)* ! +

20-26
A cash flow analysis shows that there is an initial outlay for all the equipment, and
then battery replacement in year 8 and 15 of system life. The escalated cost of the
battery bank in those future years is shown. The fuel and maintenance costs are
quite small compared to the prime diesel system. The total accumulated cash flow
(not discounted) comes to $349,600 over the life of the system.
C a s h F lo w A n a ly s is - H y b r id S y s t e m
$ 7 0 ,0 0 0
$ 4 0 0 ,0 0 0
$ 3 4 9 ,6 0 0
$ 3 5 0 ,0 0 0
$ 6 0 ,0 0 0
Annual C ost
$ 2 5 0 ,0 0 0
$ 4 0 ,0 0 0
$ 2 0 0 ,0 0 0
$ 3 0 ,0 0 0
$ 1 5 0 ,0 0 0
A c c u m u la te d C o s
$ 3 0 0 ,0 0 0
$ 5 0 ,0 0 0
$ 2 0 ,0 0 0
$ 1 0 0 ,0 0 0
$ 1 0 ,0 0 0
$ 5 0 ,0 0 0
$0
$0
1
10 11 12 13 14 15 16 17 18 19 20 21
Year
,(* "#$% &

20-27
Grid Connect
A grid connection would entail the costs of extending the line, estimated for this
example at $6,000 / km, and the cost of a step-down transformer (estimated at
about $7,000). There is no diesel fuel, but there is an electricity charge estimated to
be $0.10/kWh, and an annual maintenance cost for the line of about $30/km.
including initial capital and electricity and maintenance, comes to $128,279
(discounted to constant dollars). The system produces 175,200 kWh in 20 years, for
a LCC Cost/Energy ratio of $0.42/kWh.
Of course, this analysis does not account for any of the issues associated with grid
power in remote areas, such as power quality or outages. There can be real
economic costs attached to low voltages and poor power factor, in that customers
might have to pay for power conditioning equipment at their end of the wire, or may
have unacceptable operation from motors or electronic equipment. And power
brown outs or outages will have very measurable impact on productivity. This
aspect of the grid extension alternative do not show up from this simple type of
analysis, and should also be considered when comparing true costs of various
power solutions.
The cash flow analysis shows the large initial cost of extending the grid and then the
small-accumulated cost of the electricity and line maintenance over the life of the
system.
C as h F lo w An a lys is - U tility E x te n sio n
$120,000
$250,000
$213,322
$100,000
An n u a l C o st
$80,000
$150,000
$60,000
$100,000
$40,000
Acc u m u la te d C o s t
$200,000
$50,000
$20,000
$0
$0
1
9 10 11 12 13 14 15 16 17 18 19 20 21
Ye a r
-"#$.+ % &

20-28

System Description
Utility Extension
Location
Madras, India

Item
Utility Line Extension (km @ $/km)
Transformer
Recurring Costs
Fuel
Electricity
Maintenance
Line Maintenance
Non-Recurring Items
Item
Economic Factors
Item
LCC Period (years)
Discount Rate

42 kWh/d
Quantity
15
1
Cost
$6,000
$7,000
Extended Cost
$90,000
$7,000
$0
$0
$0
$0
$0
$0
$0
$0
$97,000
Annual kWh
15,330
Cost / kWh
$0.10
Extension Length Frequency/year

15
1
Annual Cost
$1,533
Extended Cost
$450
$0
$0
$0
$0
$450
Cost (today)
Value
20
12%
10%
7%
Cost / km
$30
Year
NPV Factor
1
0.955
2
0.913
3
0.872
4
0.833
5
0.796
6
0.760
7
0.726
8
0.694
9
0.663
10
0.633
11
0.605
12
0.578
13
0.552
14
0.528
15
0.504
16
0.482
17
0.460
18
0.440
19
0.420
20
0.401
NPV Cost
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
NPV Factor
16.64
12.82
Cost Summary
Initial Capital
Annual Fuel
Annual Maintenance
$97,000
$0
Annual Cost
$1,533
$450
NPV Factor
16.64
12.82

Total Energy (kWh)
Cost/kWh (NPV)

Life Cycle Cost

$97,000
$0
$25,512
$5,767
$128,279
306,600
$0.42
20-29
Comparison of Results
The real value of a LCC analysis is not in the absolute number but in a comparison
of values from various alternative choices. Now we can compare the four energy
solutions side-by-side, to see how their values and components relate.
Economic Factors
Hybrid
System
$64,110
$42,438
$16,052
$18,700
Prime Diesel
Pure PV
Initial Capital
$26,000
$127,840
Annual Fuel (NPV)
$179,314
$0
Annual Maintenance (NPV)
$117,874
$4,485
$31,627
$36,161
(NPV)
Total LCC Cost
$141,300
$354,816
$168,486
Cost/kWh (NPV)
$0.46
$1.16
$0.55
Accumulated Cash Flow
$349,600 $1,226,510 $278,333
/"" ! "#$"
Utility
Extension
$97,000
$25,512
$5,767
$0
$128,279
$0.42
$213,322
One useful method of comparison is to graph the component costs. Each solution is
graphed below, with the initial capital, annual fuel and maintenance costs, and any
major non-recurring expenses shown. This makes it easier to see which alternative
has the largest initial capital costs, and which has greater costs over time. For
example, the diesel has the lowest initial costs of all, but accumulates the greatest
operating costs of all.
Life Cycle Cost Comparison (NPV)

$400,000
All NonRecurring
Costs (NPV)
$350,000
$300,000
Annual
Maintenance
(NPV)
$250,000
$200,000
$150,000
Annual Fuel
(NPV)
$100,000
$50,000
Initial Capital
$0
Hybrid
System
Prime Diesel
Pure PV
Utility
Extension

20-30
Notice that the benefits of the hybrid solution over the pure PV solution are not just
in a lower LCC, but also in a dramatically lower initial cost.
Another interesting method of comparison is to look at the cash flows over time.
Each alternative is graphed below against the same scale, to more easily see when
their accumulated (non-discounted) expenditures cross one another. Each system
is assumed to be producing the same amount of total energy annually, so this
comparison shows when each system becomes cost effective compared to the
others.
For example, in the first years of the systems, the prime diesel system is producing
energy for the lowest accumulated cost. At about 5 years, the accumulated costs of
fuel and maintenance push the prime diesel system above the total accumulated
cost of the hybrid system. Therefore after about 5 years, the hybrid system is the
most cost effective (total energy / total expenditure). And after about 7 years, even
the pure standalone PV system, with its very high initial costs, is more cost effective
than the prime diesel system.
Cash Flow Analysis
Accumulated Cash Flow
$1,400,000
$1,200,000
$1,000,000
$800,000
$600,000
$400,000
$200,000
$0
0
10
12
14
16
18
20
Year
Hybrid System
Prime Diesel
Pure PV
Utility Extension


20-31
Parametric Analysis
In this analysis, various factors, or parameters, are varied while holding all other
factors constant to determine the sensitivity of the final result to the initial
assumptions. Typical factors to compare through parametric analysis include:
Fuel Cost
Discount Rate
Inflation Rate
Maintenance Rate
Grid Extension Distance
Grid Electricity Cost
Battery Cost
Diesel Cost
PV / Diesel Ratio
Sensitivity to Fuel Cost

As an example, the initial analysis presented above used a fuel cost of $0.40 per
liter. By varying this value only, the following parametric analysis table can be
developed:
Alternative
Fuel Cost ($/liter)
Sensitivity
0.20
0.40
1.00
Standalone
PV
Prime Diesel
0.55
0.55
0.55
none
0.86
1.14
2.00
high
Hybrid
0.39
0.46
0.66
moderate
Grid
0.42
0.42
Extension
'/"""
0.42
none
This analysis shows that PV and grid options are not at all sensitive to fuel cost
(since they use no fuel directly), while hybrids are moderately sensitive and diesel
systems are highly sensitive. Therefore any assumptions about diesel fuel cost
have a tremendous effect on the life cycle costing of the Prime Diesel alternative,
and should be validated as much as possible for this type of analysis to have real
meaning.

20-32
Sensitivity to Discount Rate

Varying the discount rate shows another example of parameter sensitivity. This
factor should be of major concern when comparing various projects that have
significant costs in the future.
Alternative
Discount Rate
Sensitivity
6%
12%
18%
Standalone
PV
Prime Diesel
0.66
0.55
0.49
low
2.00
1.16
0.76
high
Hybrid
0.66
0.46
0.36
moderate
0.38
low
Grid
0.50
0.42
Extension
)/""(
This analysis shows that the capital intensive projects (pure standalone PV and grid
extension), where most of the expenses occur at the beginning of the project, are
not very sensitive to discount rate, while the project with the highest proportion of
recurring expenses in the future (prime diesel) is highly sensitive. Once again, this
points out that assumptions about discount rate must be carefully exposed and
validated for the conclusions of this type of analysis to have meaning. The
assumptions of risk and real financial conditions must be carefully considered,
especially since we are considering alternatives that have very different future costs.
A high discount rate for example will make a Prime Diesel alternative look
inexpensive compared to a PV solution, whereas a low discount rate will show a
Prime Diesel system to be very expensive over time.

20-33
LCC Sensitivity
Another interesting analysis is to evaluate the total Life Cycle Cost of the hybrid
system for different PV/Diesel ratios. It is interesting to find the ratio of PV energy to
generator energy that gives the lowest life cycle cost. Other factors such as initial
cost may take precedence, but this type of sensitivity analysis helps to show how
much of a hybrid the system should be to have the lowest cost over time.
Below is shown the analysis for the system discussed earlier, with PV/Diesel energy
ratio changing from 0/100 (no PV) to 100/0 (no diesel). It is interesting to note that
the total cost over time actually goes down as the size of the solar array increases.
This indicates that at least in this particular example that the costs of fuel and
maintenance in the future outweighs the higher initial cost of solar modules. It is
only at very high PV/Diesel ratios (95% and higher) that the cost actually goes up
again.
Sensitivity of LCC to PV/Diesel Ratio

0.5
LCC ($/kWh)
0.49
0.48
0.47
0.46
0.45
0.44
0.43
0.42
0
20
40
60
80
100
Percent PV
0/""12(
In summary, life cycle cost analysis is a powerful tool that can be used to compare
technology options for remote power requirements. The techniques described can
easily be used to compare water pumping alternatives, grid connection options and
other choices.

20-34
(End of Chapter)

20-35
CHAPTER TWENTY
ECONOMIC ANALYSIS
20-1
Why Life Cycle Analysis Is Important
20-1
System Costs
Capital Costs
Recurring Costs
Non-recurring Costs
20-2
20-2
20-4
20-6
Economic Factors
Life Cycle Term (term)
Discount Rate (DR)
Fuel & General Escalation (FE, GE)
Net Present Value
20-7
20-7
20-7
20-7
20-8
LCC Analysis Methodology

Capital Costs
Recurring Fuel Costs
Recurring Maintenance
Non-Recurring Costs
Final Result - Life Cycle Energy Cost
20-10
20-10
20-11
20-11
20-12
20-12
Sample Analyses
Sample Application Parameters
PV System
Prime Generator
PV-Diesel Hybrid
Grid Connect
Comparison of Results
20-14
20-15
20-16
20-20
20-23
20-28
20-30
Parametric Analysis
Sensitivity to Fuel Cost
Sensitivity to Discount Rate
LCC Sensitivity
20-32
20-32
20-33
20-34

20-36

Reference For Chapter 3

Cargado por

Información del documento

Título original

Derechos de autor

Formatos disponibles

Compartir este documento

Compartir o incrustar documentos

Opciones para compartir

¿Le pareció útil este documento?

¿Este contenido es inapropiado?

Copyright:

Formatos disponibles

Reference For Chapter 3

Cargado por

Copyright:

Formatos disponibles

Chapter One

Purpose of This Manual

Siemens Solar Basic PV Technology Course

How to Proceed Through These

Siemens Solar Basic PV Technology Course

Siemens Solar Basic PV Technology Course

Siemens Solar Basic PV Technology Course

Welcome to the World of Photovoltaics

Purpose of This Manual

How to Proceed Through These Materials

Siemens Solar Basic PV Technology Course

Siemens Solar Basic PV Technology Course

Fundamentals - Basics of Photovoltaics

What Are Solar Cells

Siemens Solar Basic PV Technology Course

Fundamentals - Basics of Photovoltaics

What Are Solar Cells?

Silicon is abundant (sand)

Light carries energy into cell

Siemens Solar Basic PV Technology Course

Fundamentals - Basics of Photovoltaics

Key Benefits of Solar Electricity

Fuel Is Already Delivered

Siemens Solar Basic PV Technology Course

Fundamentals - Basics of Photovoltaics

Benefits of Solar Electricity

Siemens Solar Basic PV Technology Course

Fundamentals - Basics of Photovoltaics

Generate Where Needed

Siemens Solar Basic PV Technology Course

Fundamentals - Basics of Photovoltaics

Siemens Solar Basic PV Technology Course

Fundamentals - Basics of Photovoltaics

How Solar Cells Work

Internal Field and Electron Flow

Siemens Solar Basic PV Technology Course

Fundamentals - Basics of Photovoltaics

How Solar Cells Change

Siemens Solar Basic PV Technology Course

Fundamentals - Basics of Photovoltaics

Siemens Solar Basic PV Technology Course

Fundamentals - Basics of Photovoltaics

Cells Into Modules

Connect Cells To Make

Siemens Solar Basic PV Technology Course

Fundamentals - Basics of Photovoltaics

Terms and Definitions

The CELL is the basic building block of a manufacturer of solar modules.

It is a collection of cells interconnected by usually flat wire, and includes

A PANEL is a collection of modules physically and electrically grouped together

Siemens Solar Basic PV Technology Course

Fundamentals - Basics of Photovoltaics

An ARRAY is the full collection of all solar photovoltaic generators.

Sometimes an array is so large that it is grouped into SUB-ARRAYS, for easier

-- basic building block in factory

MODULE -- smallest unit that can do

PANEL -- physically connected modules on a

ARRAY -- all solar generators in one

Siemens Solar Basic PV Technology Course

Fundamentals - Basics of Photovoltaics

Basic Concepts of Energy and

Siemens Solar Basic PV Technology Course

Fundamentals - Basics of Photovoltaics