Battery management and energy management
Simon Schwunk, Matthias Vetter, Nils Armbruster
Fraunhofer Institute for
Solar Energy Systems ISE
FVEE-Workshop
Ulm, 20. Januar 2010
© Fraunhofer ISE
Agenda
Stationary applications
Battery management
Architecture
State determination for lead-acid batteries
Results
Energy management
Architecture
Results
Automotive applications
Architecture
State determination for lithium–ion batteries
Photo: INES
© Fraunhofer ISE
Stationary Battery Management
DC/DC-converter
intensive charge bus
Q Increased flexibility by =
splitting the battery into
several strings =
Q DC/DC-converter for DC-bus
making intensive full
charges +
Q Appears as a single
battery for the connected
system
String String String String
1 2 3 4 -
© Fraunhofer ISE
Charging strategies: Example lead-acid batteries
CCCV CCCVCC
Q Standard procedure Q Besonders geeignet für
verschlossene Bleibatterien
Q Often two constant voltage
phases with different voltage Q Ia-Phase: Abstände von 1-2
levels Monaten
© Fraunhofer ISE
Charging strategies for lead-acid batteries
13.2
Discharge characteristics
for a capacity test after 2.) intensive full charge
12.6
1. CCCV charge similar to a
Einzellockspannung [V]
full charge in solar
12.0
applications
1.) constant voltage charge
2. CCCVCC charge with 11.4
intensive full charge
(112% of capacity) Block 1
10.8 Block 2
Æ Very good experiences
with maintenance free 0 20 40 60 80 100
lead-acid batteries in Kapazität [% C10]
field tests Bild: Fraunhofer ISE, Freiburg, Germany; Solarpraxis AG, Berlin, Germany
© Fraunhofer ISE
State of charge determination
© Fraunhofer ISE
State of health
Q Determination of battery
capacity
Q Comlpete discharge of the
battery
Q Calculation of U10=f(I10)
Q Attenuation of U10
Q Correction of Offset
Q Correction by temperature
© Fraunhofer ISE
Battery management system
Q Four battery states
Q Charging and
discharging priorities
according to state
Q Within a state priority
given according to state
of charge
© Fraunhofer ISE
Results – Frequeny of SOC for battery system
2006: without BMS 2007: with BMS
© Fraunhofer ISE
Results – Frequeny of SOC for string 4
2006: without BMS 2007: with BMS
© Fraunhofer ISE
Energy management system
Objectives of the operation strategy
Q Supplying the load with sufficient energy and power
Q Optimized usage of photovoltaic energy
Q Long battery life
Q Minimized usage of auxiliary power generator (e.g.
diesel) for low fuel consumption and long life times
© Fraunhofer ISE
Energy management system
Characteristics of good operation
Q Produced energy is preferably consumed directly and
not stored within the battery
Q For lead acid batteries the state of charge should not fall
below 40 to 50 %
Q Regular full charges in the case of lead-acid batteries
Q No usage of auxiliary power if there is sufficient
photovoltaic energy available
Q Little start / stop cycles and minimum run times for
auxiliary power generators
© Fraunhofer ISE
Energy management system – scheduling fuel cell
Q Optimized use of the fuel
cell because of limited
lifetime
Q Direct supply to avoid
storage losses
Q The schedule is determined
for the next day
© Fraunhofer ISE
Fraunhofer System Research for Electromobility FSEM
Energy Interface: Energy storage
Power
transport and power network Recycling
generation
distribution ↔ vehicle New vehicle concepts
¾ Funding: German Ministry for Education and Research BMBF
(economic stimulus package II)
¾ Budget: 30 Mio. Euro, plus 14 Mio. Euro for Invests in economic stimulus package I
¾ Project head: Prof. Dr. Ulrich Buller (Senior Vice President Research Planning
of the Fraunhofer-Gesellschaft)
¾ Main project coordinator: Prof. Dr.-Ing. Holger Hanselka (Director of Fraunhofer LBF)
¾ Duration: June 2011
© Fraunhofer ISE
Associated Fraunhofer Institutes
In sum: 33 Institutes at 22 locations
© Fraunhofer ISE
Demonstrator for Passenger Cars and Public Transport
Wheel-hub motor
(including power electronics) Battery
Fraunhofer E-Concept Car Typ 0 AutoTram®
FrECC0 (Photo Fraunhofer IFAM) (Photo Fraunhofer IVI)
© Fraunhofer ISE
Hybrid Storage System
360 V
Hybrid Storage System Vehicle
Controller
Model Based Energy Management
(Electric and Thermal)
DC
Battery Electric Thermal DC
Management Safety Safety
(SOC, SOH) Concept Concept
Auxiliary
Electric
Hybrid Storage Loads
Power Storage Energy
Thermal
(Li Cells, Storage Cooling System
Loads
Supercaps) (Li Cells)
© Fraunhofer ISE
Pumping System Vehicle Battery
Cooling Circuit Control Device Charger
Energy and Battery
Management – Architecture GND_BAT
String 3
String 2
String 1
Current Measurement
Energy management system as central
Shutdown
Communication Bus
Safety
control unit One Cell Temperature
Monitoring
Decentralized battery management
Battery Module
BMS
(12 Cells in series)
BMS BMS (Microcontroller, 12
+ +
...
Kalman filter,
system for each single battery module Battery Module Battery Module
LTC6802 b Voltage
Measurement and Cell
Balancing)
Æ Determination of state of charge and
state of health of each single cell BMS BMS BMS
possible +
Battery Module
+
Battery Module
+
Battery Module
Æ Slightly higher costs for processors, but EMS
(Energy Management
some kind of frontend for acquring BMS BMS BMS
System)
Embedded System
measuring data is necessary anyway + + +
Battery Module Battery Module Battery Module
Redundancy:
Æ Additional energy demand for quick
Communication Bus
2x EMS
2x Communication
.....
.....
.....
processor is negligible (decrease of
efficiency of battery system by 0.05 %)
BMS BMS BMS
+ + +
Battery Module Battery Module Battery Module
© Fraunhofer ISE GND_BAT
Pumping System Vehicle Battery
Cooling Circuit Control Device Charger
Main Tasks of Energy GND_BAT
String 3
String 2
String 1
Management System Current Measurement
Shutdown
Communication Bus
Safety
One Cell Temperature
Monitoring
Supervision of BMS modules
Battery Module
BMS
(12 Cells in series)
BMS BMS (Microcontroller, 12
+ +
...
Kalman filter,
LTC6802 b Voltage
Battery Module Battery Module Measurement and Cell
Management of communication with BMS modules
Balancing)
BMS BMS BMS
Data acquisition, logging and fault tracking +
Battery Module
+
Battery Module
+
Battery Module
EMS
Main tasks of model based electric and thermal BMS BMS BMS
(Energy Management
System)
Embedded System
management: +
Battery Module
+
Battery Module
+
Battery Module
Redundancy:
Communication Bus
2x EMS
¾ Reduction of motor power (via vehicle controller),
2x Communication
.....
.....
.....
e.g. in case of low SOCs or critical temperatures BMS BMS BMS
+ + +
¾ Distribution of energy fluxes of power and energy Battery Module Battery Module Battery Module
storage (in case of a hybrid storage)
¾ Set values for on- and off-board chargers
GND_BAT
¾ Set values for cooling system
User information: Cruising range estimation
Diagnosis for pack or module exchange
Safety of pack (Switches in +/- connections)
© Fraunhofer ISE
Pumping System Vehicle Battery
Cooling Circuit Control Device Charger
Tasks Battery Management System GND_BAT
String 3
String 2
String 1
Current Measurement
Shutdown
Communication Bus
Safety
One Cell Temperature
Monitoring
Module battery management system:
Battery Module
BMS
(12 Cells in series)
BMS BMS (Microcontroller, 12
+ +
...
Kalman filter,
LTC6802 b Voltage
Battery Module Battery Module Measurement and Cell
¾ Battery module front end
Balancing)
¾ Battery module controller
BMS BMS BMS
+ + +
Battery Module Battery Module Battery Module
Main tasks battery module front end:
EMS
(Energy Management
System)
Embedded System
BMS BMS BMS
¾ Measurement of voltage and temperature +
Battery Module
+
Battery Module
+
Battery Module
Redundancy:
Communication Bus
2x EMS
2x Communication
¾ Switches for cell balancing
.....
.....
.....
Main tasks of battery module controller: +
BMS BMS
+
BMS
+
Battery Module Battery Module Battery Module
¾ Preprocessing of measurement values
¾ State of charge determination of single cells GND_BAT
¾ State of health determination of single cells
¾ Control of cell balancing
Safety functions
¾ Identification of critical values (e.g. voltage levels,
temperatures) and communication to EMS
© Fraunhofer ISE
State of charge determination
Ah counter: Integration of measurement errors
Most conventional approaches:
¾ Use of some kind of OCV correction in combination with Ah counting
Æ Recalibration of the SOC value via OCV consideration needs resting phases
Flat OCV characteristic with hysteresis for LiFePO4
© Fraunhofer ISE
State of charge determination
Æ Approach: Kalman-Filter
More insensitive against measurement errors
No resting phases necessary for recalibration of SOC
wk vk+1
Fast identification of starting values
Improved performance for Measurement:
System Input x k +1
Measurement:
System Output
B’ + H’ +
aged batteries uk z k +1
xk
Process A’ Z -1
Recursive state estimator
Optimal estimator for processes H
with gaussian noises zˆ k +1
~
z k +1
Suitable only for linear systems K -
Output:
−
xˆ k +1 xˆ k +1
For non-linear systems: Extended
Estimation
B + +
or Unscented Kalman-Filter M odel A
x̂ k
Z -1
(Kalman Filter) M easurement
Process M odel M odel
© Fraunhofer ISE
State of charge determination
Approach: Extended Kalman-Filter (EKF)
Extension of Kalman Filter approach for non-linear systems:
Æ Linearization within the operating point
LiCoO2 LiFePO4
1.0 1.0
0.8 0.8
0.6 0.6
SOC
SOC
0.4 0.4
0.2 0.2
EKF RCModell EKF RCModell
AhZähler AhZähler
0.0 0.0
0 50 100 150 0 50 100 150
t h t h
© Fraunhofer ISE
State of health determination
Principle of Dual Extended Kalman Filter
¾ Two decoupled parallel Kalman Filters
¾ Exchange of computed states of state filter (state of charge) and of
weight filter (state of health)
Initialization
State Filter
Compute a priori Compute a
estimation posteriori value
(Prediction) (Correction)
−
x̂ New measured
− value
Θ
ˆ
Weight Filter
Initialization
Compute a priori Compute a
estimation posteriori value
(Prediction) (Correction)
© Fraunhofer ISE
State of health determination
Aged battery:
80 % SOH
¾ Cathode: NMC
¾ Anode: Carbon
2.45 Ah, 3.6 V
© Fraunhofer ISE
Conclusions
Battery systems for electric vehicles
Energy and battery management systems are similar both in stationary and automotive
applications in principal
But specific boundary conditions must be taken into account:
Very specific properties of batteries in automotive applications make e.g.
approaches for state determination based on rules difficult to implement Æ model
based approaches are more straight forward to adapt to different battery types
Centralized BMS in stationary applications easier to handle since there are less
dynamics and safety can be more easily assured Æ module BMS in automotive
applications ensures safety, reliable state determination and long life times
Thermal management is more important in automotive applications than in
stationary applications
In automotive applications there are not multiple power generators available at the
same time
© Fraunhofer ISE
Thank you for your attention
Contact:
Simon Schwunk
[email protected]
© Fraunhofer ISE
