SUBMETIDO PARA 2a REVISO

JOURNAL OF OCEANIC ENGINEERING

Wave-to-wire model and energy storage analysis of

an ocean wave energy hyperbaric converter
Paula B. Garcia-Rosa, Jose Paulo V. S. Cunha, Fernando Lizarralde,
Segen F. Estefen, Isaac R. Machado and Edson H. Watanabe

AbstractThis paper addresses the dynamic modeling and the energy storage analysis of a wave energy
hyperbaric converter, which consists of a set of oscillating bodies (named as pumping modules) linked to
hydropneumatic accumulators and an electric generating unit. A mathematical model of the accumulator
is discussed and a model for the generating unit is
proposed, including a nonlinear model of a Pelton
turbine. Then, the hydrodynamic, mechanical and electrical characteristics of the subsystems that compose
the converter are discussed and the dynamics of the
entire process can be evaluated. Simulation results from
the incident wave to the generated power delivered to
an islanded system or to the grid are presented.
Index TermsEnergy conversion, Ocean wave energy, Wave-to-wire, Gas accumulators, Pelton turbine.

Nomenclature
m, v
Mass and velocity of the floating body.
,
Wave frequency and water surface elevation.
mr , Rr Added mass and radiation damping coefficient.
, w Freshwater and seawater density.
Aw , g Waterplane area and gravitational constant.
Ap , pp Area and pressure of the pump.
Hs , Tp Significant height and peak period.
Qi , Qo Input and output flow.
Va , pa Water volume and pressure in the accumulator.
Aa , xa Area and piston position of the accumulator.
V0 , p0 Initial volume of gas and initial pressure.
VT
Total volume of water and gas in the system.
xN , kN Needle position and proportional constant.
p
pressure at the turbine conduit.
vc
Water velocity in the conduit.
Ac , Lc Length and pipe area of the conduit.
pf , m Field poles and angular speed of rotor.
Jm
Moment of inertia (generator and turbine).
P, T
Power and torque.
R, L
Resistance and inductance.
e, i, Voltage, current and magnetic flux.
Subscripts
n
Number of a floating body.
P. B. Garcia-Rosa and S. F. Estefen are with the Dept. of Ocean
Eng., COPPE, Federal University of Rio de Janeiro, Brazil.
J. P. V. S. Cunha is with Dept.of Electronics and Telecommunication Eng., State University of Rio de Janeiro, Brazil.
F. Lizarralde, I. R. Machado and E. H. Watanabe are with the
Dept. of Electrical Eng., COPPE, Federal University of Rio de
Janeiro, Brazil.

e, m
d, q
r, s
l, g
f, k

Electrical and mechanical quantities.

d and q reference frames.
Rotor and stator.
Leakage and magnetizing indutances.
Field and damping quantities.

I. Introduction
CEAN wave energy is a renewable and non-polluting
source, with a world-wide potential estimated at the
order of 2000 TWh/year [1], that can contribute significantly to the demand for electricity. Although it has not
yet been commercially exploited, research and development on wave energy technologies have been conducted
for more than three decades.
Several wave energy converters (WECs) with different
processes of energy conversion are available from laboratory experiments and prototype tests. Recent reviews
indicates that there are about one hundred projects at
various stages of development [2]. According to the working principle, WECs can be classified as oscillating water
column (OWC) systems, overtopping devices or oscillating body systems. In the oscillating bodies, a floating
body may be connected to a hydraulic system before the
electrical energy conversion, for example [3][4][5], or may
be connected directly to a linear electric generator, for
example [6][7][8][9].
The wave power varies in several time-scales: from wave
to wave (seconds), according to the sea state (hours) and
according to seasonal variations (months). Thus, regardless of the energy conversion process, without any method
of energy storage, the variation of the power output from
a WEC can be quite large, as may occur in other intermittent renewable sources, such as wind or solar energy.
A method of energy storage allows the smoothing of the
output power that is generated from a highly variable
input power [3], as in the conversion system analyzed here.
Some examples of energy storage technologies that
have been investigated and implemented in renewable energy systems are: compressed-air energy storage (CAES),
hydrogen-storage systems, flywheels, batteries, supercapacitors and superconducting magnetic energy storage
(SMES) [10].
In wave energy technologies, there are basically three
methods used as a short-term energy storage (range of
ten seconds to a minute)[4]: potential energy storage in
water reservoirs, that are applied to some overtopping
systems [11], kinetic energy (flywheel) in air turbines at

JOURNAL OF OCEANIC ENGINEERING

OWC systems and gas accumulators, that are applied to

some oscillating bodies [3].
The design of a wave energy conversion system is usually
based on numerical simulations and tests with smallscale models at ocean basin laboratories. The obtained
results with these methods are fundamental to perform
necessary changes in the design and to optimize selected
configurations [12]. Different versions of a WEC may be
evaluated with numerical models at a relatively low cost
and then, a small-scale model is usually tested to validate
the numerical results and to investigate phenomena which
are not evidenced by the simulations [12].
Moreover, the numerical models are useful to evaluate
and estimate the generated power according to typical sea
states of an installation site and to verify the performance
and feasibility of control strategies, which are included to
improve global performance and reliability of the system.
These strategies may be designed to improve the absorption of wave energy and to deal with emergency procedures
(interlocking mechanisms) and the quality of the electricity
supplied.
Designing efficient WECs and their control systems
depend on mathematical modelings of hydrodynamic, mechanical and electrical characteristics of the converter
subsystems. In this framework, some published works, for
example [5], [13] and [14] describe the dynamic model of
different WECs from the incident wave to the electrical
power delivered to the grid. In [15], a generic model of
a system with hydraulic power take-off is developed. A
wave-to-wire model allows prior knowledge of the dynamic
behavior of the plant considering interaction between each
subsystem, which enables the identification of dynamics
that are not evidenced if the models are considered separately.
This paper addresses the wave-to-wire model and the
energy storage analysis of the Brazilian wave energy hyperbaric converter [16]. This energy converter consists of a set
of oscillating bodies (named as pumping modules), a twostage accumulator linked to a hyperbaric chamber, which
is previously pressurized, and an electric generating unit.
The water flow obtained from the oscillating body motion
is displaced to the accumulator, and then, a pressurized
water flow operates a Pelton turbine coupled to an electrical generator. Such arrangement is different from other
oscillating body systems which include hydraulic power
take-off (PTO) mechanisms with high pressure (HP) and
low pressure (LP) gas accumulators. In these systems, a
machine (motor or turbine) is driven by the flow resulting
from the pressure difference between accumulators [4].
Additionally, the applied pressure is in the range 250500
meters of water column (mwc), which is substantially
higher than some proposed concepts.
In previous works ([17], [18], [19], [20]) related to
hyperbaric converters, the system dynamics, considering
both pumping modules and the generating unit, are not
considered. In [17] numerical simulations and experimental
results (in small scale 1:10) of pumping modules are
presented. The problem of optimizing the efficiency of

the primary conversion is discussed in [18] together with

a latching control strategy. In [19], a strategy based on
phase control concepts is also proposed to improve the
absorption of wave energy. In this case, a sliding mass
fixed on the arm is used to modify the dynamics of the
floating body. In [20], a hydraulic turbine speed governor
is proposed based on a linear model of the generating unit
considering constant pressure in the accumulators. The
control strategy is based on cascade control combined with
feedforward of load disturbances.
In this work, the dynamics of the accumulator is discussed and a mathematical model integrating it to the
generating unit is proposed, considering a nonlinear model
of the hydraulic turbine. Then, the hydrodynamic, mechanical and electrical characteristics of the main subsystems that compose the converter are discussed. The
dynamics of the entire process and its energy storage
capacity are evaluated. Simulation results considering the
proposed wave-to-wire model under the action of regular
and irregular incident waves illustrate the performance of
the system.
II. Wave energy hyperbaric converter
The hyperbaric converter is composed of pumping modules, a hydropneumatic accumulator, a hyperbaric chamber and a generating unit [19], as illustrated in Figure 1.
In this system, each pumping module has a floating
body linked to a hydraulic pump through a lever arm.
The vertical motion of the body due to the wave action
induces the pump actuator to displace water obtained from
a closed circuit to a hydropneumatic accumulator, which
has an internal piston separating water from gas. Then,
the pressurized water flow from the accumulator operates
a Pelton turbine coupled to an electrical generator. The
water flow can be controlled through a needle actuator.
A full-scale prototype of this WEC with two pumping
modules and a synchronous generator with 125 kVA is
presently installed at Port of Pecem, Cear
a State, Brazil.
In what follows, an overview of the wave climate in the
installation site is described and a block diagram of the
wave-to-wire model is depicted.
A. Overview of the wave climate in the Brazilian Northern
Coast
Considering a northwest-southeast (NW-SE) coastline
orientation, Brazilian northern coast is exposed to wave
fields generated in both hemispheres and a similar mode
of variability is observed in the wave climate. In the
winter season, locally generated southeastern waves are
predominant while northern swells prevail over local sea
conditions during the summer [21].
Figure 2 illustrates the scatter diagram representing the
occurrence of sea states at the coast of Port of Pecem.
Wave data were obtained through some in situ measurements using a Datawell Directional Waverider buoy
during a 5-year program led by the National Institute of

JOURNAL OF OCEANIC ENGINEERING

electric
generator

hyperbaric
chamber
needle
actuator

lever arm

hydropneumatic
accumulator

arm
movement

WAVE
ENERGY

pelton
turbine

support
platform

sea wave

hydraulic
pump

floating body

ELECTRIC
ENERGY

water jet

water recirculation

Schematic of the wave energy converter.

Fig. 1.

% Occurrence
8
9

2.5

III. Pumping modules model

The dynamics of the pumping module is described by
the motion of the floating bodies. Here, we assume heave
oscillatory motion, identical bodies, and neglect wave
interaction effects between bodies. Then, the governing
equation for a body n is given by [23],

1.5

H (m)

is more appropriate for large-signal performance analysis

of the turbine [22] and takes into account the dynamics of
the accumulator and the oscillations both in pressure and
flow output.

0.5
4

10

12

14

16

Tp(s)

Fig. 2.

Sea state occurrence at the Port of Pecem, Brazil.

Waterways Research (INPH) and then, an analysis of the

wave climate was done in [21].
Initial analysis have shown a predominance of small
waves, between 1 1.75 m of significant height, despite
some recordings of waves higher than 2 m. Peak period
values are predominantly short, between the range 5 7 s,
but the occurrence of waves with long periods (12
20 s) coming from the North Hemisphere have also been
recorded. The occurrence of these swells coincides with the
cold seasons (and some associated storms) in the North
Atlantic [16].

m v n (t) = fhn (t) + fsn (t) + fpn (t) ,

(1)

where the body oscillates under the action of hydrodynamic forces (fh ), hydrostatic force (fs ) and the vertical
force due to the PTO mechanism (fp ). Since linear hydrodynamic theory [23] is considered, the hydrodynamic
forces are given by the superposition of the excitation force
(fe ), which is the force on the body held fixed in incident
waves, and the radiation force (fr ), which is the force due
to the body oscillation in the absence of waves.
Thus, the vertical motion of the body satisfies the
integro-differential equation [24]
Zt
[ m + mrn () ] v(t)
+ K(t ) vn ( ) d +

(2)

gw Aw

vn ( ) d = fen (t) + fpn (t) ,

B. Description of the wave-to-wire model

A block diagram of the wave-to-wire model describing
the hyperbaric converter is presented in Figure 3. It can
be noted that the pressure oscillations may affect both
pumping module and generating unit dynamics. Since the
accumulator connects these subsystems, from the development of its dynamic model we can evaluate the oscillations
in the system variables (e.g., pressure, flow, turbine speed
and electric generated power), which are induced by the
oscillatory behavior of the waves.
In [20] a linear model representing the small-signal
performance of the turbine was considered. Such model
does not consider large variations in the input power that
can occur because of the intermittent characteristic of the
waves. In this work, a nonlinear model is described, which

(3)

where the kernel of convolution term is known as fluid

memory term and is given by
Z
2
K(t ) =
Rr ()cos [ (t ) ] d .
(4)

In this case, the power take-off (PTO) force is the force

applied by the pump to the body, given by
(
0, vn (t) > 0 ,
fpn (t) = Ap ppn (t) , ppn (t)
(5)
pa , vn (t) < 0 ,
Equation (5) indicates that the body motion injects water
into the accumulator exclusively during the descending
motion of the body, when the pressure inside the pump

JOURNAL OF OCEANIC ENGINEERING

incident wave

input
flow

pumping unit
dynamics

pressure

needle
position

Fig. 3.

acummulator
dynamics

conduit
dynamics

output flow

turbine
dynamics

mechanical
torque

electrical load

rotor
dynamics

speed

generator
dynamics

electrical
torque

Basic block diagram of the wave-to-wire model.

becomes equal to the chamber pressure pa (high-pressure

stage). During the ascendant motion of the body, the
pressure inside the pump becomes negligible (low-pressure
stage) [17].
The wave excitation force is given by [23]
r
2g 3 Rr ()
fe (t) = Fe () cos t , Fe () = A
, (6)
3
when considering regular waves of frequency and amplitude A. For irregular waves,
fe (t) =

N
X

fe,i (t) ,

A correlation between numerical simulations and experimental results of the absorbed power per floating
body (in kW) were done in [17], validating the model.
The experimental results were obtained with a small-scale
model (1 : 10) with four pumping modules, tested in
the ocean basin from the LabOceano at COPPE, Federal
University of Rio de Janeiro (Brazil), as illustrated in
Figure 4. The model was submitted to different regular
waves, with heights and periods representatives of the
Brazilian South coast.

(7)

i=1

which is obtained by a linear superposition of N regular

waves. This is a numerical approximation of real waves,
that are characterized by defining a spectrum with a
significant wave height Hs , a peak period Tp and by using
a random phase generator [23].
Finally, the water flow input into the accumulator Qin (t)
can be calculated by
(
0,
vn (t) > 0 ,
Qin (t) =
(8)
Ap vn (t), vn (t) < 0 ,
where the velocity v for each body n is determined by the
solution of (2). Then, the total input flow is
Qi (t) =

k
X

Qin (t) .

(9)

n=1

The power of the incident wave train is expressed in kW

per meter of a wavefront as
Pw = kw Hw2 Tw .

(10)

where Hw and Tw are, respectively, the height and the

g2
103 kg/m/s4 ,
period of the incident wave and kw = 32
for sinusoidal regime or kw 0.49 kg/m/s4 for irregular
regime [23]. Note that for irregular waves Hw = Hs is the
significant wave height and Tw = Tp is the peak period.
The mean power extracted by each floating body during
a time interval T is given by
Z
1 T
Pa =
fpn (t) vn (t) dt .
(11)
T 0

Fig. 4.

Small-scale pumping modules at LabOceano.

IV. Accumulator Model

The internal section of the hydropneumatic accumulator
is divided into air and water separated by a piston, as
illustrated in Figure 5. Thus, when the inflow of water
Qi is greater than the flow output Qo (water released
to the hydraulic turbine), the piston rises up, storing the
captured energy in the form of compressed air. In a dual
form, when the flow output is greater than the inflow,
the piston moves down, releasing the energy that was
previously stored. Based on this fact, the position of the
internal piston xa (t) is governed by the law of conservation
of mass,
dxa (t)
Aa
= Qi (t) Qo (t) .
(12)
dt
In [4], the gas compression/expansion process inside
the accumulators is considered as an isentropic process

JOURNAL OF OCEANIC ENGINEERING

[22]

piston

vc (t) = kN xN
Ha

Qi

dvc (t)
= Ac (p pa ) .
(17)
dt
The output flow from the accumulator Qo is proportional
to the velocity of the water in the conduit, that is, Qo (t) =
Ac vc (t). Thus, from (17), the water flow rate is given by
Lc Ac

Qo

Schematic representation of the hydropneumatic accumulator.

Fig. 5.

(negligible heat transfer) during a specific sea state. In

this study, we consider a simplified approach where the
gas compression/expansion is an isothermal process. This
approach makes simpler the integration between the models of the accumulator and the generating unit, without
significative error.
Then, the pressure pa (from Fig. (5)) is obtained by the
expression of an ideal gas:
(13)

where the volume of gas (V ) can be calculated by the

difference between the total volume of water and gas (VT )
and the volume of water in the accumulator,
V (t) = VT Va (t) .

(16)

From assumption (H2), the acceleration of water column

in the conduit is characterized by the Newtons second law
of motion and can be expressed as

xa

pa (t)V (t) = p0 V0 ,

p
2gp .

(14)

dQo
Ac
=
(p pa ) .
dt
Lc

According to assumption (H4) and considering fixed

power losses (Pl ), the mechanical output power is given
by
Pm = Qo p Pl .
(19)
B. Electric Generator model
A synchronous machine is used as the electric generator
in the hyperbaric converter. The model of this machine
takes into account the dynamics of the stator, field circuit
and damper windings. Figure 6 shows the equivalent circuits represented in the rotor reference frame (dq frame).
The equivalent circuits provide a visual description of the
machine model, which is described in [26].

After deriving (13) and (14) as function of time, we

have the following non-linear equation for the accumulator
pressure,


pa
dVa
dpa
=
.
(15)
dt
VT Va
dt
V. Generating unit model

(a)

Rkd
Llkd

Rs

rq

ekd
ikd

Ll

Llf d
id
ed

Lgd

Rf d
if d

This section presents the nonlinear model of the hydraulic turbine, which includes the water dynamics in the
conduit, and the model of the electric generator.

ef d

(b)

Rkq2

A. Hydraulic Turbine Dynamics

The turbine and conduit characteristics are determined
by equations relating the acceleration of water column,
the velocity of water in the conduit and the turbine mechanical power [25]. Additionally, we consider the following
assumptions:
(A1) The pressure loss in the conduit and needle are
negligible;
(A2) The water pipe is inelastic and the water is incompressible;
(A3) The velocity of the water varies directly with the needle opening and with the square root of the net pressure;
(A4) The mechanical power of the turbine is proportional
to the product of pressure and water flow through the
turbine.
Thus, from Bernoullis equation and from assumption
(H3), the velocity of the water in the conduit is given by

(18)

Llkq2
Rs

rd

Ll

ekq2
ikq2

Llkq1
iq
eq

Lgq

Rkq1
ikq1
ekq1

Fig. 6. Equivalent electric circuit of the synchronous generator:

(a) d-axis and (b) q-axis.

The instantaneous power at the machine terminal is

calculated as
3
Pe = (ed id + eq iq ) ,
(20)
2

JOURNAL OF OCEANIC ENGINEERING

and the electromagnetic torque is [26]

3 pf
(d iq q id ) .
(21)
4
Moreover, according to Newtons second law and neglecting friction, the rotor motion is governed by the
following differential equation

TABLE II
Fluctuations for different values of the chamber
volume Vc .

Te =

dm
= Tm Te .
(22)
dt
When there is an unbalance between the torques, the net
torque causes rotor acceleration (or deceleration). Since
the generator is a synchronous machine, the generated
frequency is proportional to the rotor speed.
Jm

VI. Simulation results

The simulations concerning the mathematic model of
the hyperbaric converter were performed considering respectively one floating body (section VI-A) and two bodies
(section VI-B) in the pumping unit. The floating body is
a cylindric buoy, which has predominantly vertical motion
and mass m = 17 103 kg. The added mass mr and
radiation resistance Rr coefficients were computed by the
software WAMIT [27].
The accumulator area is Aa = 0.57 m2 , VT = 6.3 m3 ,
V0 = 5.5 m3 and p0 = 2.5 MPa (255 mwc). The parameters of the synchronous generator (WEG GT A202CM V J,
Brazil) are given in Table I.
TABLE I
Parameters of the synchronous generator.

Parameter
Rated power
Rated voltage
Rated frequency
Field poles
Stator resistance

Symbol
SN
e
f
pf
Rs

Value
125
380
60
4
0.016

Unit
kVA
V
Hz

ohms

A. Energy storage analysis

The first simulation results are related to an analysis
of the energy storage of the hyperbaric converter. As the
wave power has an oscillatory nature, the instantaneous
absorbed power by the floating bodies can suddenly fluctuate from zero to maximum values. The aim of such analysis
is to numerically evaluate the smoothing characteristic of
the system in order to obtain acceptable power quality on
the output for fast oscillations of the incident wave (range
of seconds). The quality of the electricity supplied must
meet requirements with respect to limits of variations in
frequency and voltage, as well as the level of reliability [22].
In this framework, voltage and frequency controllers have
to be included in the system. Here, as our goal is an energy
storage analysis of the system, we consider the open-loop
performance of the plant and then, the terminal voltage
and frequency fluctuations due to the variable power input
are evaluated. In general, an acceptable limit of frequency

Qo %
p%
Pm %
Pe %
f %
e%

0.1 Vc
3.321
6.544
9.643
0.043
3.001
0.050

Vc
0.796
1.614
2.372
0.010
0.740
0.012

10 Vc
0.090
0.180
0.274
0.001
0.081
0.010

TABLE III
Fluctuations for different values of the accumulator
volume Va .

Qo %
p%
Pm %
Pe %
f %
e%

0.1 Va
0.802
1.620
2.429
0.010
0.750
0.004

Va
0.796
1.614
2.372
0.010
0.739
0.004

10 Va
0.793
1.555
2.364
0.010
0.738
0.004

fluctuation is 2% [28] and an acceptable limit of voltage

fluctuation is 5%.
The generator is connected to an islanded system and
the converter is submitted to incident regular waves of
period T = 6 s and amplitude A = 2.8 m, with a wave-power
level of approximately 49 kW/m. In these simulations, the
oscillations on the variables are calculated according to the
following equation


xmax xmin
100% ,
(23)
x =
xmin
where x is the variable, xmax is the maximum value
observed and xmin is the minimum value of the variable
during a simulation.
Tables II and III illustrate the obtained oscillations for
the pressure head, the output flow, the mechanical and
electrical powers, the generated frequency and the voltage,
when the volume of the hyperbaric chamber Vc and the
volume of the accumulator Va are respectively multiplied
and divided by a factor equal to 10.
It can be observed that in both cases (changes in Vc and
Va ), the steady-state values of the variables have larger
oscillations as the volume of the equipments are smaller.
However, the variables have the lowest oscillations for the
highest value of the chamber volume (10Vc ), when compared to the highest volume of the accumulator (10Va ).
Indeed, the volume of the accumulator is divided into
water and gas, and the chamber has the greatest amount
of gas. The chamber is responsible for smoothing the
oscillations observed, but note that the discharge duration
of the storage is enhanced by increasing the accumulator
volume. From (12) and considering constant flow output
at rated value, the full power duration of the storage is
estimated at the order of 30 s.

JOURNAL OF OCEANIC ENGINEERING

(a) incident wave

2

(m)

1
0
1
2

20

40

60

20

40

60

20

40

60

80

100
t (s)
(b) input flow

120

140

160

180

80
100
120
t (s)
(c) output flow to the turbine

140

160

180

120

140

160

180

100
t (s)
(e) pressure

120

140

160

180

80
100
t (s)
(f) mechanical power

120

140

160

180

120

140

160

180

0.04
0.03

0.02

TABLE IV
Electrical power, frequency and voltage fluctuations for
different values of the combined moment of inertia Jm .

Simulations were performed in open loop, where the

needle position was chosen in order to maintain the
mechanical average power matched to the load power
(islanded case). In a real situation, the needle must be
controlled to keep the synchronous speed of the generator
[20][22].
The flow output Qo (Fig. 7c) is about the average value
of the flow input Qi (Fig. 7b), which shows the smoothing
characteristic obtained by the hyperbaric chamber. From
Fig. 7d and Fig. 7e, the oscillations in the piston position
and the pressure can be observed.

Q (m /s)

The frequency fluctuations are about the same oscillations of the output flow and the oscillations of the
electrical power are greatly reduced when compared to the
mechanical power. It can be observed that for the smallest
volume chamber (0.1Vc ), the frequency fluctuations exceed
2%.
A new trend for energy storage in renewable energy
systems is to combine several storing technologies [10]. In
the hyperbaric converter, a storage system that integrates
gas accumulators and flywheel energy storage can be
applied. Flywheels are very commonly used due to the
simplicity of storing kinetic energy in a spinning mass
[10]. Table IV illustrates the obtained oscillations for the
electrical power, frequency and terminal voltage when a
flywheel with different moments of inertia (combined with
the moment of inertia of the generator and the turbine) is
applied in the hyperbaric converter.

0.01
0
3

6.65

A Pierson-Moskowitz spectrum [29] is used to generate

a incident wave train with significant wave height of Hs =
1.4 m, a peak period of Tp = 6.2 s and random phase. This
sea state is representative of the wave climate in the Port
of Pecem, as illustrated by Fig. 2. The wave-power level is
5.96 kW/m. Two different methods are commonly used to
simulate time-series of ocean waves given a target power
spectrum: the random phase scheme and the random
coefficient scheme [30]. This paper considers the first one.
Figures 7, 8 and 9 show the simulation results, when
the hyperbaric converter generator is an islanded system
and when it is connected to the grid. Each simulation was
performed for the same time-series, thus the variables of
the pumping unit and the mechanical power exhibit the
same results.

80

100
t (s)

(d) piston position

1.2

x (m)

1
0.8
0.6
0.4

20

40

60

20

40

60

20

40

60

80

p (MPa)

2.6
2.55
2.5
2.45
2.4

14.8

B. Islanded system and grid connection

6.55
6.5

P (kW)

It can be noted that the frequency fluctuations are

smaller with greater values of the inertia, as expected.
The flywheel affects directly the fluctuations on generator
speed and consequently, the frequency fluctuations. Besides, the buffering capacity of the chamber affects directly
the fluctuations on output flow, pressure and mechanical
power and consequently, this affects the fluctuations on
electrical power, frequency and terminal voltage. This
analysis can be used to correctly design the energy storage
system in such way as to have acceptable power oscillations
without oversizing the main equipment. For an islanded
system, a solution with a small chamber and a flywheel
could be a solution for storage with reduced structural
costs.

x 10

6.6

10 Jm
0.001
0.070
0.001

5 Jm
0.002
0.141
0.001

Q (m /s)

Pe %
f %
e%

Jm
0.010
0.740
0.004

14.4
14
13.6
13.2

80

100
t (s)

Simulation results - Pumping unit variables and

mechanical power for islanded/connected system.
Fig. 7.

In the islanded system (Fig. 8), the torque oscillations

imposed to the machine are filtered by the inertia of the
rotor and the flywheel. Thus, the average value of the
generated power is about 14 kW) and the power oscillations are small, around 9.8 103 %. Note that, when the
instantaneous mechanical power is greater or smaller than
the load power, the speed fluctuations reaches up to 0.3%
of the rated value (Fig.8b). The obtained terminal voltage
is about 380 V with oscillations of 103 %.
When the hyperbaric converter generator is connected
to the grid (Fig. 9), the initial conditions are chosen such

JOURNAL OF OCEANIC ENGINEERING

that the generated voltage has the same magnitude, phase

angle and frequency of the grid, to provide synchronism
between generator and grid. In a real situation, a synchronism relay should be adopted.
From Fig. 9b, it can be noted that the machine keeps
the synchronism with the grid. This is due to the strong
characteristic of the simulated grid, which imposes the
synchronous speed of the machine. However, it can be
inferred that when the system is connected to the grid,
as the rotor speed is imposed by the grid, the rotor inertia
does not act as a filter and then the oscillations observed
in the mechanical power are transformed directly into
electrical power oscillations (Fig. 9a). The average value
of the generated power is about 14 kW but the power
oscillations are approximately 3%. The terminal voltage
is 380 V.
(a) electrical power
14.8
Pe (kW)

14.4
14
13.6
13.2

20

40

60

20

40

60

20

40

60

80

100
t (s)
(b) rotor speed

120

140

160

180

80
100
t (s)
(c) terminal voltage

120

140

160

180

120

140

160

180

m (rpm)

1810
1805
1800
1795
1790

e (volts)

385

380

375

80

100
t (s)

Simulation results - Generating unit variables for

islanded system.
Fig. 8.

(a) electrical power

14.8
Pe (kW)

14.4
14
13.6
13.2

20

40

60

20

40

60

80

100
t (s)
(b) rotor speed

120

140

160

180

80
100
t (s)
(c) terminal voltage

120

140

160

180

m (rpm)

1810
1805
1800
1795
1790

e (volts)

385

380

375

20

40

60

80

100

120

140

160

180

t (s)

Simulation results - Generating unit variables for

connected system.
Fig. 9.

VII. Conclusions and Future works

In this study, a wave-to-wire model of the wave energy
hyperbaric converter was presented. The main subsystems of the converter were described and the dynamic
models were integrated in order to evaluate the energy
storage of the system. The dynamics of the accumulator
was discussed and a model for the generating unit was
proposed, including a nonlinear model of the hydraulic turbine. Then, the hydrodynamic, mechanical and electrical
characteristics of the main subsystems were discussed and
the dynamics of the entire process could be evaluated.
Simulation results considering the proposed dynamic
model under the action of regular and irregular incident
waves illustrate the performance of the system. The largest
values of the chamber volume result in the smallest oscillations in the variables, such as pressure, flow, mechanical
power and turbine speed (frequency), as expected, but
resulting in higher costs. The power oscillations can be
further reduced by including a flywheel in the generating
unit. The combination of two storing technologies (gas
accumulators and flywheel) can make possible the size
reduction of the chamber, which can reduce the associated
costs.
The proposed model is essential to correctly design the
complete system in such a way as to have acceptable power
oscillations without oversizing the main equipment. The
acceptable limits of power oscillation depend on whether
the converter is connected to an islanded system or a weak
or strong power grid.
A full-scale ocean wave energy hyperbaric converter
with 125 kVA nominal electrical power is currently being
built at Port of Pecem, Ceara State, Brazil. In future
works, experimental results obtained from this full-scale
converter will be compared with the simulation results of
the wave-to-wire model presented in this article. Furthermore, this model can be used to estimate and evaluate
the performance of control strategies, for example, speed
governors and voltage controllers and strategies to improve
the power absorption. The integration between control
strategies and interlocking mechanisms under different
wave climate is an ongoing research work.
Acknowledgment
The research work presented in this paper was supported by CAPES, CNPq, FAPERJ and Tractebel Energia
(ANEEL R&D Program).
References
[1] A. Cl
ement, P. McCullen, A. Falc
ao, A. Fiorentino, F. Gardner,
K. Hammarlund, G. Lemonis, T. Lewis, K. Nielsen, S. Petrocini,
M. T. Pontes, P. Schild, B. O. Sj
ostr
on, H. C. Sorensen, and
T. Thorpe, Wave energy in Europe: Current status and perspectives, Renewable and Sustainable Energy Reviews, vol. 6,
no. 5, pp. 405431, 2002.
[2] A. F. O. Falc
ao, Wave energy utilization: A review of the technologies, Renewable and Sustainable Energy Reviews, vol. 14,
no. 3, pp. 899918, 2010.
[3] R. Henderson, Design, simulation, and testing of a novel hydraulic power take-off system for the pelamis wave energy
converter, Renewable Energy, vol. 31, no. 2, pp. 271283, 2006.

JOURNAL OF OCEANIC ENGINEERING

[4] A. F. O. Falc
ao, Modelling and control of oscillating-body
wave energy converters with hydraulic power take-off and gas
accumulator, Ocean Engineering, vol. 34, pp. 20212032, 2007.
[5] C. Josset, A. Babarit, and A. H. Cl
ement, A wave-to-wire
model of the SEAREV wave energy converter, Proceedings of
the Institution of Mechanical Engineers - Part M: Journal of
Engineering for the Maritime Environment, vol. 221, no. 2, pp.
8193, 2007.

[6] I. A. Ivanova, O. Agren,

H. Bernhoff, and M. Leijon, Simulation
of wave-energy converter with octagonal linear generator, IEEE
Journal of Oceanic Engineering, vol. 30, no. 3, pp. 619629,
2005.
[7] C. Bostrom, R. Waters, E. Lejerskog, O. Svensson, M. Stalberg,
E. Stromstedt, and M. Leijon, Study of a wave energy converter connected to a nonlinear load, IEEE Journal of Oceanic
Engineering, vol. 34, no. 2, pp. 123127, 2009.
[8] S. Tyrberg, R. Waters, and M. Leijon, Wave power absorption
as a function of water level and wave height: Theory and
experiment, IEEE Journal of Oceanic Engineering, vol. 35,
no. 3, pp. 558564, 2010.
[9] K. Rhinefrank, A. Schacher, J. Prudell, T. K. A. Brekken,
C. Stillinger, J. Z. Yen, S. G. Ernst, A. von Jouanne, E. Amon,
R. Paasch, A. Brown, and A. Yokochi, Comparison of directdrive power takeoff systems for ocean wave energy applications,
IEEE Journal of Oceanic Engineering, vol. 37, no. 1, pp. 3544,
2012.
[10] J. M. Carrasco, L. G. Franquelo, J. T. Bialasiewicz, E. Galvan,
R. C. P. Guisado, M. A. M. Prats, J. I. Leon, and N. MorenoAlfonso, Power-electronic systems for the grid integration of
renewable energy sources: A survey, IEEE Transactions on
Industrial Electronics, vol. 53, no. 4, pp. 10021016, 2006.
[11] L. Margheritini, D. Vicinanza, and P. Frigaard, SSG wave
energy converter: Design, reliability and hydraulic performance
of an innovative overtopping device, Renewable Energy, vol. 34,
no. 5, pp. 13711380, 2009.
[12] J. Cruz, Ocean Wave Energy: Current Status and Future Prepectives. Springer Series in Green Energy and Technology, 2008.
[13] F. Wu, X.-P. Zhang, P. Ju, and M. J. H. Sterling, Modeling and
control of AWS-based wave energy conversion system integrated
into power grid, IEEE Transactions on Power Systems, vol. 23,
no. 3, pp. 11961204, 2008.
[14] M. Amundarain, M. Alberdi, A. J. Garrido, and I. Garrido,
Modeling and simulation of wave energy generation plants:
Output power control, IEEE Transactions on Industrial Electronics, vol. 58, no. 1, pp. 105117, 2011.
[15] K. Ruehl, T. K. A. Brekken, B. Bosma, and R. Paasch, Largescale ocean wave energy plant modeling, in Proceedings of the
2010 IEEE Conference on Innovative Technologies for an Efficient and Reliable Electricity Supply (CITRES, Massachusetts,
USA, 2010, pp. 379386.
[16] S. F. Estefen, P. R. Costa, E. Ricarte, and M. M. Pinheiro,
Wave energy hyperbaric device for electricity production,
in Proceedings of the 26th International Conference on Offshore Mechanics and Arctic Engineering, vol. OMAE-2007, San
Diego, USA, 2007.
[17] S. F. Estefen, P. T. T. Esperanca, E. Ricarte, P. R. Costa, M. M.
Pinheiro, C. H. P. Clemente, D. Franco, E. Melo, and J. A.
Souza, Experimental and numerical studies of the wave energy
hyperbaric device for electricity production, in Proceedings of
the 27th International Conference on Offshore Mechanics and
Arctic Engineering, vol. OMAE-2008, Estoril, Portugal, 2008.
[18] P. B. Garcia-Rosa, J. P. V. S. Cunha, F. Lizarralde, S. F.
Estefen, and P. R. Costa, Efficiency optimization in a wave
energy hyperbaric converter, in Proc. Int. Conf. on Clean
Electrical Power, Capri, Italy, 2009, pp. 6875.
[19] P. R. Costa, P. B. Garcia-Rosa, and S. F. Estefen, Phase
control strategy for a wave energy hyperbaric converter, Ocean
Engineering, vol. 37, no. 17, pp. 14831490, 2010.
[20] P. B. Garcia-Rosa, J. P. V. S. Cunha, and F. Lizarralde, Turbine speed control for an ocean wave energy conversion system,
in Proc. American Control Conf., Saint Louis, Missouri, 2009,
pp. 27492754.
[21] E. Ricarte, S. F. Estefen, A. L. T. Mendes, and C. E. Parente,
Wave climate analysis for a wave energy conversion application
in brazil, in Proceedings of the 26th International Conference on

[22]
[23]
[24]
[25]

[26]
[27]
[28]
[29]
[30]

Offshore Mechanics and Arctic Engineering, vol. OMAE-2007,

San Diego, USA, 2007.
P. Kundur, Power System Control and Stability. USA: EPRIPower System Engineering Series McGraw-Hill Inc., 1994.
J. Falnes, Ocean Waves and Oscillating Systems: Linear Interaction including Wave-Energy Extraction. USA: Cambridge
University Press, 2002.
W. E. Cummins, The impulse response function and ship
motions, in Schiffstechnik, vol. 47, no. 9, 1962, pp. 101109.
E. D. Jaeger, N. Janssens, B. Malfliet, and F. V. D. Meulebroeke,
Hydro turbine model for system dynamic studies, IEEE Transactions on Power Systems, vol. 9, no. 4, pp. 17091715, Nov.
1994.
P. C. Krause, O. Wasynczuk, and S. D. Sudhoff, Analysis of
Electric Machinery. IEEE Press, 1995.
C. H. Lee and J. N. Newman, WAMIT User Manual Version
6.3, 6.3PC and 6.3S, 6.3S-PC, USA, 1998.
IEC 60034-1, Rotating electrical machines Part 1: Rating and
performance, 2004.
L. Moskowitz, Estimates of the power spectrums for fully
developed seas for wind speeds of 20 to 40 knots, Journal of
Geophysical Research, vol. 69, no. 24, pp. 51615179, 1964.
S. Elgar, R. T. Guza, and R. J. Seymour, Wave group statistics
from numerical simulations of a random sea, Applied Ocean
Research, vol. 7, no. 2, pp. 9396, 1985.