1. Effective Modeling of Temperature Effects on Lithium Polymer Cells
F. Baronti, G. Fantechi, E. Leonardi, R. Roncella, R. Saletti
University of Pisa, Italy Dept. of Information Engineering: Electronics, Computer Science and Telecommunications

2. Outline Introduction Cell model Model characterization
Experimental set-up
Test description
Parameter extraction
Thermal model
Model validation
Conclusions

3. Introduction – Lithium Battery
Lithium batteries very promising for next generation hybrid and electric vehicles
Model
Type
Capacity
[Ah]
Max continuous current [A]
Energy density
[Wh/kg]
Power density
[W/kg]
Cycle Life
Charge
Discharge
Saft VH AA 1500
Ni-MH
1.5
4.2 69
194 NA
Saft VL34P
Li-ion 33
500
120
128
7378
Kokam SLPB H
LiPo 31
155 62
133
1334
>800
ThunderSky-LYP40AHA
LiFePO4 40 85
1707
>3000
Altairnano 50Ah
Nano Lithium Titanate 50
300 72
719
>4000

4. Introduction - Lithium Battery (cont’d)
Battery management systems (BMSs) needed for safe and reliable operation of the vehicle battery
Monitor cell voltage and temperature
Evaluate state-of-charge (SOC) and state-of-health (SOH)
Accurate cell model
to be embedded in the BMS for SOC and SOH estimate
to be used for BMS design and simulation
In this work, a scaled model of LiPo cell is considered
Kokam SLPB723870H4 1.5 Ah, 1.5 A max charge current, 30 A max continuous discharge current

5. Electrical Cell model Model Parameter Symbol Dependence Cell capacity
Ccapacity -
Self discharge
Rself_discharge
Ohmic resistence
Rseries
Icell, SOC, T
Long transient RC
Rt_long / Ct_long
Short transient RC
Rt_short / Ct_short
M. Chen et al. “Accurate Electrical Battery Model Capable of predicting Runtime and I-V Performance,” IEEE Trans. Energy Convers., vol. 21, no. 2, June 2006

6. Experimental set-up
Standard set-up combined with a custom-designed temperature-controlled chamber
NI USB
TTi LD300 80V-80A
TTi QL355TP 35V-5A

7. Temperature-controlled chamber
Cell encased in two symmetrical halves. Each of them contains:
Array of digital temperature sensors
82 W Peltier TEC

8. Temperature effect on cell behavior
1C rate discharge at different cell temperatures
As known, temperature significantly affects cell behavior decreasing the usable cell capacity
Actual capacity is about 1.4 Ah while the nominal one is 1.5 Ah
end-of-discharge voltage

9. Cell model parameter extraction
Test procedure:
Test conditions:
Charge current (Itest): 0.5C, 1C rates
Discharge current (Itest): 0.5C, 1C, 5C, 10C, 20C rates
Cell temperature: 10, 25, 35 °C
C - Used to signify a charge or discharge rate equal to the capacity of a battery divided by 1 hour.
Init phase:
1C charge
1 h pause
1C discharge
Pause phase:
1 h pause
Test phase:
charge/discharge cycle
current and temperature of interest
5 min pause after 1% SOC variation and every 9% SOC variation

10. Cell model parameter extraction (cont’d)
Model parameters derived from the cell voltage transient during the 5 min pauses
charge cycle
pause
discharge cycle

11. Cell model parameter extraction (cont’d)
Icell = Itest
pause (Icell = 0)
Icell = Itest

12. Cell model parameters Open Circuit Voltage (OCV)
Extracted from 1C charge/discharge
Small dependence on cell temperature

13. Cell model parameters (cont’d)
Let’s have a look at Rtot=Rseries+Rt_long+Rt_short
As expected, Rtot increases at lower temperatures

14. Thermal Model
The electrical model has been improved with a first order thermal model
Cth and Rth estimated by observing the thermal evolution of the cell
Param.
Value
Cth
50 W/°C
Rth
14.7 °C/W

15. Model validation
Model implemented in Matlab/Simulink® using multidimensional LUT for model parameters
Model 25 °C
Model
Model T
Max
RMS
Abs error
398 mV
49 mV
181 mV
14 mV
% error
9.61 %
1.45 %
5.6 %
0.39 %

16. Model validation (cont’d)
Model 25 °C
Model
Model T
Max
RMS
Abs error
521 mV
26 mV
478 mV
17 mV
486 mV
% error
12.6 %
0.69 %
11.5 %
0.44 %
11.7 %

17. Conclusions Experimental characterization of a 1.5 Ah Kokam LiPo cell
Accurate modeling including thermal effects
Model implemented in Matlab/Simulink®
Very good matching between experimental and simulated dynamic cell behaviors
Model accuracy significantly improves if temperature-dependent parameters are used
Starting point for the design and simulation of the battery pack and BMS using 31 Ah Kokam LiPo cells targeting a Fuel Cell HEV