cfsccomparison of single cell and parallel cell li-ion testing
TRANSCRIPT
C f S CComparison of Single Cell and Parallel Cell Li-ion Testing
John HalpineThe Aerospace Corporationp p
NASA Aerospace Battery WorkshopNov. 15th -17th, 2011
© The Aerospace Corporation 2011
Agenda
• Test Set-up• Comparison of well-matched cells in modules with single cellsComparison of well-matched cells in modules with single cells
– Voltage– Temperature
Current sharing– Current sharing• Performance of degraded cells in modules
– Individual cell performanceV lt– Voltage
– Current sharing• Comparison of well-matched modules with degraded modules• Summary
[email protected] Technology Department
Single Cell vs Parallel Modules
3.6
3.4
3.5
(V)
Average Results
From:
3.2
3.3
Cel
l Vol
tage
Single Cell Module
6 single cells2 3-cell modules1 6-cell module
3
3.1
0 200 400 600 800 1000Cycle Numbery
[email protected] Technology Department
Single Cell vs Parallel Module: selected results
3.6
3.4
3.5
(V)
3.2
3.3
Cel
l Vol
tage
(
3
3.1Cell 4 Cell 5 3Cell Mod2
0 100 200 300 400 500 600 700 800 900 1000Cycle Number
[email protected] Technology Department
Temperatures during cycling
34
30
32
s (C)
24
26
28
Cell T
empe
ratur
es
20
22
24
Single Cells 3 Cell Modules 6 Cell Module
Temperature drops due to a cell stopping cycling
0 100 200 300 400 500 600 700 800 900 1000
Cycle Number
[email protected] Technology Department
Current Sharing Between Cells in a Module
2.04
2.06
2.08
rren
t (A)
1.98
2
2.02
End
of D
isch
arge
Cur
Cell1_CurrCell2_CurrCell3_Curr
3-Cell Modules
1.94
1.96
0 100 200 300 400 500 600 700 800 900 1000
Cycle Number
2.08
2.02
2.04
2.06
ge Current (A
)
Cell1_CurrCell2_CurrCell3_Curr
1.96
1.98
2
End of Discharg
[email protected] Technology Department
1.94
0 100 200 300 400 500 600 700 800 900 1000
Cycle Number
Current Sharing during Discharge
3 8
4
‐1.94
‐1.92
‐1.9
V)A)
Cell1_CurrCell2_CurrCell3_CurrV l (V) 3 8
4
‐1.94
‐1.92
‐1.9
V)A)
Cell1_CurrCell2_CurrCell3_CurrVoltage(V)
3.2
3.4
3.6
3.8
‐2.06
‐2.04
‐2.02
‐2
‐1.98
‐1.96
Mod
ule Voltage
(V
Discharge
Current (A Voltage(V)
Cycle 100 3.2
3.4
3.6
3.8
‐2.06
‐2.04
‐2.02
‐2
‐1.98
‐1.96
Mod
ule Voltage
(V
Discharge
Current (A Voltage(V)
Cycle 500
3‐2.1
‐2.08
0 5 10 15 20 25 30
Discharge Time (minutes)
Cycle 100
3‐2.1
‐2.08
0 5 10 15 20 25 30
Discharge Time (minutes)
Cycle 500
3.6
3.8
4
‐2
‐1.98
‐1.96
‐1.94
‐1.92
‐1.9
Voltage
(V)
Curren
t (A)
Cell1_CurrCell2_CurrCell3_CurrVoltage(V)
3
3.2
3.4
‐2.1
‐2.08
‐2.06
‐2.04
‐2.02
2
0 5 10 15 20 25 30
Mod
ule V
Discharge
Cycle 900
[email protected] Technology Department
0 5 10 15 20 25 30
Discharge Time (minutes)
Summary of Single Cell vs Module
• No noticeable difference in performance• Significant impact of temperature on performanceSignificant impact of temperature on performance
– particularly towards the end of life• Current sharing may assist in identifying weak cells
[email protected] Technology Department
Degraded Cells Performance
3.9
4Cell 1 Cell 2 Cell 3cell 1 cell 2 cell 3
3 6
3.7
3.8
(V)
3.4
3.5
3.6
Cel
l Vol
tage
Capacity loss ~10%
3.1
3.2
3.3 Energy Loss ~ 20%
3
0.1 0.3 0.5 0.7 0.9 1.1 1.3 1.5 1.7 1.9 2.1Discharge Capacity(Ah)
[email protected] Technology Department
Degraded Modules Stress Cycling
3.4
3.5
3.6
tage
(V)
6Cell
3Cell Mod1
3Cell Mod2
3.2
3.3
End of Discharge
Volt
3
3.1
0 100 200 300 400 500 600 700
Cycle Number
[email protected] Technology Department
Current Sharing within Modules
‐0.5
0
Cell 1 Cell 2 Cell 3
‐2
‐1.5
‐1
Cell Cu
rren
t (A)
‐3
‐2.5
0 1 2 3 4 5 6
Discharge Capacity (Ah) ‐0.5
0
Cell 1 Cell 2 Cell 3 Cell 4 Cell 5 Cell 6
6-Cell Module
‐2
‐1.5
‐1
Cell Cu
rren
t (A)
3-Cell Module
‐3
‐2.5
0 2 4 6 8 10 12
Discharge Capacity (Ah)
[email protected] Technology Department
Discharge Capacity (Ah)
Current Sharing in Degraded Modules with Cycling
‐1.3
‐1.7
‐1.5
‐1.3Cu
rren
t (A)
cell1
‐1.7
‐1.5
ell C
urrent (A
)
‐2.3
‐2.1
‐1.9
nd of D
ischarge
Cell
cell2
cell3
cell4
cell5
2 3
‐2.1
‐1.9
of Discharge
Ce
cell1
cell2
cell3
‐2.5
0 100 200 300 400 500 600 700
En
Cycle Number
cell6
‐2.5
‐2.3
0 100 200 300 400 500 600 700
End o
Cycle Number
[email protected] Technology Department
Comparison with well-matched module results
Off-set due to known degradation: 1/3 of cycle life 1 out of 3 cells could not support cycling conditions
3.5
3.6expected decrease in cycling due to degradation
3.3
3.4
Vol
tage
(V)
3.1
3.2Cel
l
Degraded Module New Module
30 200 400 600 800 1000
Cycle Number
[email protected] Technology Department
Comparison with well-matched module results
Off-set based on end-of-discharge voltages
[email protected] Technology Department
Summary of Module Testing with Degraded Cells
• The degraded modules demonstrate a reduced life that is in-line with known degradation of the individual cellwith known degradation of the individual cell
• This degradation was not observed through either capacity measurement or module voltage trends
• The testing highlighted that module end of discharge voltage is notThe testing highlighted that module end of discharge voltage is not a good indicator of state of health of the module by itself– Current sharing between cells may be required in order to provide a
more complete indication of the state of health of the modulemore complete indication of the state of health of the module• Extrapolation of life estimates using the end-of-discharge voltage
demonstrate 20% difference in cycle life capability based on the test set-up used. p
[email protected] Technology Department
Acknowledgements
• Justin Lee and Jasen Ross provided significant technical support• The Aerospace Corporation is gratefully acknowledged forThe Aerospace Corporation is gratefully acknowledged for
supporting this work as part of its Mission-Oriented Investigation and Experimentation program
[email protected] Technology Department
Abstract
• Stress cycle testing was set up and run on Li-ion 18650 cells to compare cell and module performance. The cells were either cycled individually, in 3-cell parallel modules or 6-cell parallel modules The cells and well matched modules demonstrated no noticeable difference inparallel modules. The cells and well-matched modules demonstrated no noticeable difference in performance during the stress cycling. During the testing it was noticed that slight differences in temperature of 2-3°C had a large impact on performance. These differences were caused by chiller outages or test stoppages that resulted in reduced heat loads on neighboring cells, the changes in voltage were particularly pronounced at end-of-life. Modules were observed to have less variation in performance than single cells, as might be expected, due to the modules averaging the individual cell performance differences. From the testing on well matched-modules it appears that single cell life test data can be used for module performance as well. Stress cycle testing on modules with degraded cells in the module was also completed. Comparison with modules with no degraded cells has been made The degraded modulesComparison with modules with no degraded cells has been made. The degraded modules demonstrated a slight reduction in capacity, with a much larger reduction in cycle life. This cycle life degradation was not evident in either initial capacity measurement or module end-of-discharge voltage trends. The testing highlighted that module end of discharge voltage is not a good indicator of state of health of the module. Extrapolation of life estimates using the end-of-di h lt d t t 20% diff i l lif bilit b d th t t tdischarge voltage demonstrate 20% difference in cycle life capability based on the test set-up used. Current sharing between cells may provide a much better indication of state of health of the module.
[email protected] Technology Department