Documentos de Académico
Documentos de Profesional
Documentos de Cultura
BOTTLENECKS
IN PROCESS PLANTS
Peter L. King
Lean Dynamics, LLC
October 2011
AGENDA
Definition of a bottleneck
Bottlenecks in process plants
Capacity Constraint Resources (CCRs)
Finding the bottleneck ~ Value Stream Mapping
Root causes of bottlenecks
Bottleneck management ~ Theory of Constraints
Bottlenecks can move!
Successive bottlenecks
Additional examples
-->
Thus bottlenecks must be identified
and opened up as much as practical
And bottlenecks must be very well managed
© 2010 - LEAN DYNAMICS LLC - ALLRIGHTS RESERVED 4
BOTTLENECKS IN
PROCESS PLANTS
5
© 2010 - LEAN DYNAMICS LLC - ALLRIGHTS RESERVED
BOTTLENECKS IN PROCESS PLANTS
SHAPE MANUFACTURING
STORAGE
SILOS
PACKAGING
FLAKE MANUFACTURING
9
© 2010 - LEAN DYNAMICS LLC - ALLRIGHTS RESERVED
CAPACITY CONSTRAINT IN CEREAL PLANT
SHAPE MANUFACTURING
because of poor
scheduling – is
called a Capacity
Constraint
PACKAGING
Resource FLAKE MANUFACTURING
10
© 2010 - LEAN DYNAMICS LLC - ALLRIGHTS RESERVED
DIAGNOSING THE BOTTLENECK
One of the best tools for finding the bottlenecks in your process is a
11
© 2010 - LEAN DYNAMICS LLC - ALLRIGHTS RESERVED
VSM COMPONENTS
AGGREGATED
DEMAND
DEMAND
(MONTHLY)
SALES & MANAGEMENT CUSTOMER FORECASTS
CUSTOMER
CAPACITY FORECAST SERVICE
(MONTHLY) REPS ORDERS
MASTER PRODUCTION
RAW MATERIAL SCHEDULE (MONTHLY)
ORDERS (MONTHLY)
Frequency 12 Trucks/Day
SHEET CALENDARING WRAPPING Lot Size 648K Sq Ft
SLITTING CHOPPING
FORMING BONDING PACKAGING
6
(3) 5
(3) 5
Transp time 3 Days
(4) (4) 6 LABELING 4
Invtry 6.3M lbs Effective Invtry 2500 R Effective Invtry 3650 R Invtry 4000 Invtry 322M sq Ft
Capacity 11.8 8.9 Effective Effective Effective
SUPPLIER 2 Capacity 10.3 29
Material Flow
Days 16 Days 12.6 Days 21 Capacity
Days 7 Capacity Capacity 200 Days 42
TAKT 9.5 TAKT 8.3
# # # TAKT 7.2 # TAKT 24 TAKT 120
Order Lead Frequency 1 Car/Day 6 (Master Rolls/Hr) 50 200 1000 # SKUs 2,000
1 Week SKUs SKUs (Master Rolls/Hr) SKUs (Bonded Rolls/Hr) SKUs (Slit Rolls/Hr)
Time (Cut Rolls/Hr)
Lot Size 140K lbs Utilization 80%
# SKUs 4 Utilization 93% Utilization 70% Utilization 83% Utilization 60%
Transp time 7 Days Lead time 15 Min Lead time 17 Min Lead time 10 Min Lead time 10 Min Lead time 8 Min
Yield 87% Yield 87% Yield 98% Yield 100% Yield 100%
Reliability 90% Reliability 98% Reliability 95% Reliability 98% Reliability 98%
UPtime 73.6% UPtime 61% UPtime 69% UPtime 98% UPtime 98%
# SKUs 50 # SKUs 200 # SKUs 1000 # SKUs 1800 # SKUs 2000
Batch size 1 roll Batch size 1 roll Batch size 1 roll Batch size 1 Slit Roll Batch size 1 Cut Roll
EPEI
C/O time
C/O loss
9 days
1 hr
2 Rolls
EPEI
C/O time
13 days
45 Min
EPEI
C/O time
7 days
5 Min
EPEI
C/O time
7 days
0
Data boxes
EPEI
C/O time
7 days
0
KEY
K = 1,000
M = 1,000,000
C/O loss ~0 C/O loss ~0 C/O loss 0 C/O loss 0 B = 1,000,000,000
Avail time 168 hr/wk Avail time 168 hr/wk Avail time 168 hr/wk Avail time 168 hr/wk Avail time 168 hr/wk
Shift schd 3x8x7 Shift schd 3x8x7 Shift schd 3x8x7 Shift schd 3x8x7 Shift schd 3x8x7
3 M 42 Days
16 Days 12.6 Days 21 Days 7 Days NVA Time = 99 Days
15 Min 17 Min 10 Min 10 Min 8 Min VA Time = 73 Min
Timeline
Example
A salad dressing line has weekly demand of 45,000 cases for the flavors
made on that line
Three eight-hour shifts per day, 5 days per week
Production continues through lunch, breaks, shift changes
Available time = 3 x 8 x 5 = 120 hours/week
Takt = 45,000 cases /120 hours = 375 cases/hour
So the plant must produce 375 cases per hour to satisfy customer
demand
© 2010 - LEAN DYNAMICS LLC - ALLRIGHTS RESERVED 13
VALUE STREAM MAP DATA
Effective Capacity
It represents realistic expectations, not perfection
It should represent what we can realistically do today
Utilization
A measure of how fully occupied a process step is, how busy it is
FIBER CUTTER/BALER
Effective
Cycle TimeCapacity
(Capacity) 30 BALES/SHFT
Root causes of bottlenecks can TAKT 35 BALES/SHFT
be diagnosed from the data boxes Utilization 117%
Lead Time 9.6 MIN
This is a B/N! Yield 97%
Machine failures are the primary Reliability 70%
Uptime 60%
culprit
# SKUs 14
Changeover times also Batch Size 1 BALE
contribute to the constraint (11% EPEI 4 DAYS
of capacity) C/O Time 45 MIN
C/O Losses 1%
Both must be improved to Available Time 168 HRS/WK
resolve the B/N Shift Sched 8X3X7
No of Operators 1
17
© 2010 - LEAN DYNAMICS LLC - ALLRIGHTS RESERVED
DIAGNOSING THE BOTTLENECK
RESIN REACTOR
Effective
Cycle TimeCapacity
(Capacity) 50 GAL/MIN
TAKT 55 GAL/MIN
Utilization 110%
This is also a B/N! Lead Time 2 HR
Yield and reliability are not Yield 97%
the problem Reliability 98%
Uptime 74%
Changeover times are the # SKUs 12
issue here – 22% of the Batch Size 8000 GAL
available capacity is EPEI 36 HR
consumed in C/Os C/O Time 40 MIN
C/O Losses 1%
Available Time 168 HR/WK
Shift Sched 12 hr X 2 sh X 7 d
No of Operators 1
18
© 2010 - LEAN DYNAMICS LLC - ALLRIGHTS RESERVED
ROOT CAUSES OF BOTTLENECKS
19
© 2010 - LEAN DYNAMICS LLC - ALLRIGHTS RESERVED
TYPICAL BOTTLENECK ROOT CAUSES
21
© 2010 - LEAN DYNAMICS LLC - ALLRIGHTS RESERVED
MANAGING THE BOTTLENECK
Once the bottleneck has been identified and its root cause(s)
determined, it’s time to open it up, to increase flow.
22
© 2010 - LEAN DYNAMICS LLC - ALLRIGHTS RESERVED
CONSTRAINT MANAGEMENT PROCESS
23
© 2010 - LEAN DYNAMICS LLC - ALLRIGHTS RESERVED
CONSTRAINT MANAGEMENT PROCESS
IDENTIFY THE
BOTTLENECK
Theory of Constraints
Eliyahu M. Goldratt, 1990 24
© 2010 - LEAN DYNAMICS LLC - ALLRIGHTS RESERVED
CONSTRAINT MANAGEMENT PROCESS
SUBORDINATE
Don’t let the B/N be constrained by
EVERYTHING TO
upstream or downstream limitations
THE BOTTLENECK
A B C
A B C
BUFFER 48 GPM
EMPTY 100 GPM
CAPACITY CAPACITY
CAPACITY 100 GPM
TAKT 50 GPM INVENTORY TAKT 50 GPM BUFFER TAKT 50 GPM
27
© 2010 - LEAN DYNAMICS LLC - ALLRIGHTS RESERVED
B/N MANAGEMENT ~ THEORY OF CONSTRAINTS
A B C
BUFFER 48 GPM
EMPTY 100 GPM
CAPACITY CAPACITY
CAPACITY 100 GPM
TAKT 50 GPM INVENTORY TAKT 50 GPM BUFFER TAKT 50 GPM
28
© 2010 - LEAN DYNAMICS LLC - ALLRIGHTS RESERVED
CONSTRAINT MANAGEMENT PROCESS
30
© 2010 - LEAN DYNAMICS LLC - ALLRIGHTS RESERVED
BOTTLENECKS CAN MOVE WITH PRODUCT MIX
SHEET
BONDING
FORMING
1
1
6
1.5 1.5
The bottleneck may move as the process cycles through the various
products being made.
In his case, both forming and bonding have enough capacity to meet
customer demand
31
© 2010 - LEAN DYNAMICS LLC - ALLRIGHTS RESERVED
PRODUCT-SPECIFIC BOTTLENECK
SHEET
BONDING
FORMING
1
1
6
1.5 1.5
PRODUCT 432A
When forming products with very high basis weight, forming must run
at slower lineal speeds - mass flow is the limiting factor
Thus forming becomes the bottleneck when making these products
32
© 2010 - LEAN DYNAMICS LLC - ALLRIGHTS RESERVED
PRODUCT-SPECIFIC BOTTLENECK
SHEET
BONDING
FORMING
1
1
6
1.5 1.5
PRODUCT 4516F
33
© 2010 - LEAN DYNAMICS LLC - ALLRIGHTS RESERVED
PRODUCT-SPECIFIC BOTTLENECK
Other examples:
34
© 2010 - LEAN DYNAMICS LLC - ALLRIGHTS RESERVED
WHY DOES THIS MATTER?
SHEET
BONDING
FORMING
1
1
6
1.5 1.5
PRODUCT 432A
SHEET
BONDING
FORMING
1
1
6
1.5 1.5
PRODUCT 4516F
SHEET
BONDING
FORMING
1
1
6
1.5 1.5
1.5 1.5
PRODUCT 4516F
There may be other steps which are bottlenecks – this may have
been masked by the most obvious bottleneck
An accurate VSM would have brought this to light
However….. The primary bottleneck can restrict flow in a way that
masks restrictions at other steps
Or…… managers may be making assumptions about bottleneck
locations without benefit of a VSM
37
© 2010 - LEAN DYNAMICS LLC - ALLRIGHTS RESERVED
EXAMPLE – SALAD DRESSING
38
© 2010 - LEAN DYNAMICS LLC - ALLRIGHTS RESERVED
EXAMPLE – SALAD DRESSING
BN #2 BN #1 BN #3 BN #4
BOTTLE
BOTTLE BOTTLE CASE
HOMOGENIZER FILLING
CAPPER LABELER PACKER
MACHINE
SURGE 300 Bot/Min 500 BPM ACCUMULATING 360 BPM
EFF CAPACITY 60 GPM EFF CAPACITY EFF CAPACITY EFF CAPACITY EFF CAPACITY 33 cases.min
TANK 400 BPM 400 BPM BOTTLE 400 BPM
TAKT 75 GPM TAKT TAKT TAKT TAKT 34 cases/min
CONVEYOR
UTILIZATION 125% UTILIZATION 133% UTILIZATION 80% UTILIZATION 111% UTILIZATION 101%
OEE 90% OEE 85% OEE 94% OEE 82%
OEE 73%
Salad Dressing NOTE: All capacities and Takt are relative to 24 oz bottles,
Bottle Filling Line which is the primary size run on his line
WAREHOUSE
PALLET PALLET
SHRINK CASE
STRETCH LABEL
WRAP PATTETIZER
WRAPPER PRINTER
TUNNEL
APPLICATOR
EFF CAPACITY 122 cases/min EFF CAPACITY 100 cases/min EFF CAPACITY 224 pallet/hr EFF CAPACITY 170 pallet/hr
TAKT 34 cases/min TAKT 34 cases/min TAKT 26 pallet/hr TAKT 26 pallet/hr
UTILIZATION 28% UTILIZATION 34% 12%
UTILIZATION UTILIZATION 15%
OEE 90% OEE 94%
OEE 92% OEE 76%
40
© 2010 - LEAN DYNAMICS LLC - ALLRIGHTS RESERVED
EXAMPLE – SALAD DRESSING
The cost of fitting the bottle filler with new nozzles was easily
justifiable
However, the cost of eliminating all 4 potential bottlenecks
was not
The decision was made to build a new line to handle the
increased demand
This also provided additional capacity for future demand
growth
They ultimately would have gone that way
The VSM and bottleneck analysis got them to the right
decision much sooner
And avoided the cost of a set of stainless steel nozzles.
41
© 2010 - LEAN DYNAMICS LLC - ALLRIGHTS RESERVED
BOTTLENECKS IN PROCESS PLANTS
bottleneck?
Solution
Give slitters an equal priority for maintenance
Implement TPM
Result
Slitter Utilization > 85%
Slitters no longer a bottleneck
Flow matches demand
UPPER
LAYER
PREPARATION
SUBSTRATE
PREPARATION
UPPER
LAYER
PREPARATION
That won’t relieve the bottleneck
– it’s still the bottleneck
But at least it won’t suffer from Laminator or
Trimmer downtime
Bottlenecks matter!
They constrain material flow
They prevent you from meeting customer demand
They require inventory
Bottlenecks in process plants are usually the equipment, not staffing
Adding people won’t help, nor will overtime, extra shifts
So you must confront the root causes
Opening the bottleneck requires performance improvement
Yield improvement
Reliability improvement (TPM)
Changeover reduction (SMED)
Whenever there are bottlenecks, Theory of Constraints should be applied
Productivity Press
May 2009
(302) 239-1667
(302) 528-2700