Today: Battery pack modeling and BMS

Conventional Battery System

n + +
cells Battery
(+protection) DC bus Electric drive
Management DC-DC VDC propulsion
in Vbat
series System (BMS) converter components

_ _

Control bus Vehicle

controller

• Battery pack: Stack battery cells in parallel and series
• BMS: Monitor battery state (voltage, SOC, SOH), perform cell 
balancing, communicate with vehicle controller

1
Packaging: Single Cylindrical Cells

2
Packaging: Prismatic Cells

3
 Battery Packs

Nissan Leaf
Ford C‐Max Energi

4
Temperature Effects: Cell Characteristics

Cnom = 20 Ah cell tested at 0.5C charge/discharge rate

Reference: S. Chackoa, Y. M. Chunga, Thermal modeling of Li‐ion polymer battery 
for electric vehicle drive cycles, Journal of Power Sources, Volume 213, 1 
September 2012, Pages 296–303

5
Temperature effects: aging and cycle life
• Crystal formation around the electrodes, reduces the 
effective surface area
• Passivation: growth of a resistive  layer that impedes the 
chemical reactions
• Corrosion consumes some of the active chemicals

Aging leads to:
• Increased series resistance, reduced power rating
• Reduced capacity

End of life: drop to 80% of the original power or capacity rating

All of aging processes are accelerated with increased 
temperature

6
Battery system electrical circuit model
assumption: identical cells

7
 Battery Simulink Model
ECEN5017 2
Battery Model (D)
SOC
Mp = number of cells in parallel
Open-circuit voltage
Ns = number of cells in series
as a function of SOC
SOC
1-D T(u)
1
Ibat 1
Ibat >= 0 -1/(Mp*Cnom*3600) 100
s Vz SOC VOC_cell
Integrator Percent
etaC
Cnom Limited Conversion Voltage-vs-SOC
Coulomb Switch (capacity in Ah) Characteristic
efficiency Vsp

Series resistances

Rp/Mp
Vsn
positive Ibat Series resistance Ns 1
Vcell Vbat
for Ibat>0 Vbat
Number of
cells in series
Rn/Mp
Vr
negative Ibat Series resistance
for Ibat<0

Diffusion
1
R1
R1/Mp Vh
tau1.s+1
R1 (diffusion) Diffusion
low-pass
output
voltage
loop
Ibat Vh

Hysteresis

8
 Battery Simulink Model
Vbat
Vbat

Ibat

Repeating SOC
SOC
Sequence
Stair Li-ion battery
Ibat

Ebat

1 Scope
-1 s
Product Integrator
Gain

Example:  Mp = 2, Ns = 100

Ro+ = Ro‐ = 10 m, SOC(0) = 50%, Cnom = 5 Ah, 

VM = 20 mV, tH = 50 s, R1 = 10 m, 1 = 100 s

9
Battery waveforms (an example)
400
350
300

100
50
0
0 10 20 30 40 50 60

10
0
-10
6
x 10
4
2
0

92.3%
10
Charging

11
 CCCV Charging example
390

380

370

360

100

50

0
0 10 20 30 40 50 60
10

-10

12
Effects of cell mismatches and the need for 
battery management system (BMS)
Mismatches in cell characteristics
• Capacity
• Leakage (self‐discharge) current
• Other parameters: series resistance, …
Mismatches in cell SOC’s during operation

Reference: Davide Andrea, Battery Management Systems for Large Lithium‐Ion Battery 

Packs, Artech House 2011, http://book.liionbms.com/

Animations at: http://liionbms.com/balance/index.html

13
Cell versus system capacity [Ah]
Discharge, assuming equal 
starting SOC=100%

Q  Ci 1 SOCmin,i 

Cmin  minCi 

Q  Cmin 1  0 

Csystem  Cmin

14
Charging and discharging a system with 
mismatched cells

System must sense individual cell 
voltages or (better) employ 
individual cell SOC estimators

15
System capacity with unbalanced SOC’s
Discharge, assuming unequal 
starting SOC’s

Q  Ci SOCo ,i  SOCmin,i 

Q  Csystem  minCi SOCo ,i  SOCmin,i 

SOCmin,i  0

Csystem  Cmin

16
 Passive “top” balancing

Animated example: http://liionbms.com/balance/index.html
17
Example: TI bq76PL536A

18
19
20
21
 Active balancing

Example: switched‐capacitor charge shuttling

22
Active balancing architectures

http://liionbms.com/php/wp_passive_active_balancing.php
23
Cell versus system capacity [Ah] with 
charge re‐distribution
Discharge, assuming equal 
starting SOC=100%

C system  Ci

24