Cell Balance and End of Life

In any battery pack design it is only as strong as the weakest link [4], one bad cell or group of cells in the series string will control the total power and energy available from the pack. This means it is important to match the cells and to keep them balanced throughout the lifetime use of the pack. However, at some point the cell balance or unbalance will be too big to be corrected by the balancing circuit in a reasonable time window thus affecting the end of life. Once this unbalance gets to a point where the cells have less energy than needed to meet the basic requirements this will then be end of life for the pack. Or at the very least a repair process will be required where the failing cells are replaced.

At the beginning of the life of a battery pack we assemble cells with all of the cells in series matched to within ~20mV. 

During use the BMS will further balance the cells to within 1% and as we see a roughly 1V swing in open circuit voltage (OCV) from fully charged to discharge, then 1% is around 10mV. 

That is the biggest OCV difference between the highest and lowest cells in series of 10mV. If the cells drift apart for any reason such as differences in leakage current the BMS will re-balance the string when most appropriate. This balancing can take a long time as balancing currents tend to be very small (~1 to 3mA/Ah).

Normally balancing occurs during a low power charge event.

However, at some point the cell to cell differences will increase, but when are these differences too big? What level of unbalance is too much? When would this signal end of life for the pack?

One Cell Failing

Let us consider the case where one cell is not balancing in a 6s pack, for ease we will consider this to be a 1p configuration. Using a passive resistive based balancing system we would be switching the resistive loads across the good cells to dissipate energy and to keep the cell voltage below the maximum allowed during a charge event.. Thus allowing the cell at lower capacity to be charged.

However, these balancing currents will be very small, let us assume 3mA/Ah. Assuming these cells are 10Ah and the capacity different is 8Ah that means we have a balancing current of 30mA and it would take 8/0.03 = 267 hours to balance. I would suggest that at this point we have a failed pack. A 12 hour charging event would allow the pack to increase in capacity by ~3.6%.

Turning this calculation around we can see that if the unbalance is increasing by >3.6% per use and you are charging it for 12 hours each time it will be losing capacity. You can expand this very simple calculation in a number of ways to allow you to generate requirements and limit cases for the BMS with respect to cell balance and end of life.

This post has been built based on the support and sponsorship from: About:EnergyAVANT Future MobilityQuarto Technical ServicesTAE Power Solutions and The Limiting Factor. 

You could improve the balancing circuit and move to an active balancing approach [1,2]. However, this comes at the cost of extra components and will still have a current limit.

Further Questions

  • Do you log the fact that it was not possible to re-balance the cells?
  • How many further attempts do you have at balancing?
  • When does reduced performance equate to end of life?


  1. Yevgen Barsukov, Battery Cell Balancing: What to Balance and How, Texas Instruments
  2. Sihua Wen, Cell balancing buys extra run time and battery life, Texas Instruments
  3. Thorsten Baumhöfer, Manuel Brühl, Susanne Rothgang, Dirk Uwe Sauer, Production caused variation in capacity aging trend and correlation to initial cell performance, Journal of Power Sources, Volume 247, 2014
  4. S. Rothgang, T. Baumhofer and D. U. Sauer, Diversion of Aging of Battery Cells in Automotive Systems2014 IEEE Vehicle Power and Propulsion Conference (VPPC), Coimbra, Portugal, 2014

1 thought on “Cell Balance and End of Life”

  1. I think there can be an in depth approach to balancing based on the intricacies of minimizing balancing current.

    You need to be able to predict the Self-Discharge based on time, temp and capacity. Heavy testing needs to be performed in order to validate Balancing Approach.

Leave a Comment