Life-long availability of the battery system

BMS algorithm for state of health determination in hybrid and electric vehicles

5. May 2017 | Engineering Service

With the increasing application of lithium-ion battery technology in the automotive industry, lifetime battery-aging behavior is an important topic for today’s xEV vehicle applications. Aging influences available battery capacity which is a crucial value for EV/PHEV applications, as well as battery internal resistance a crucial value for PHEV/HEV applications. Knowing the aging status (SOH, State of Health) of battery systems is important, since OEMs, workshops and even customers are interested in the SOH of the battery and some Battery Management System (BMS) functions need to be adjusted during operation to ensure the availability of the battery system throughout its lifetime.

Technical Challenges of Aging Prediction

Battery cell aging mechanisms are susceptible to two kinds of effects:

  • Calendric aging (depending on SOC and temperature)
  • Cyclic aging (depending on depth of discharge and C-rate)

Both cases result in capacity fade and internal resistance increase, which is a direct drawback for system performance.
To get sufficient information for the onboard aging model such as capacity fading prediction and internal resistance prediction, exhaustive battery cell testing must be performed. These tests include storage and cycling under defined conditions for a long period of time.

Alternatively, accelerated aging can be applied – for example, using very high operating temperatures during tests for cyclic aging, resulting in faster aging and, therefore, time saved. However, in this case, the extrapolated fuzziness of the results for estimating cell behavior over a lifetime under normal operating conditions might harm the accuracy of aging prediction over the full product life cycle, compared to the non-accelerated approach. In both cases (accelerated and non-accelerated aging analysis), unadapted static models might not be accurate enough to perform a reasonable aging prediction over a lifetime; they should be adapted online during operation when comparing measured and estimated battery parameters using real-time vehicle data.

FEV’s Approach

FEV developed an online adaption concept without using complex aging prediction models. This offers a potential reduction in effort during battery cell testing, as well as in complexity of the BMS software and hardware. The concept is to determine SOH by monitoring fundamental BMS functions, calibrated at the beginning of life (SOC calculation and Power Prediction) of battery operation, since these BMS functionalities will be affected by aging: The system will not be capable of delivering the power given by the power prediction due to the increase in internal resistance, and SOC will then no longer be calculated correctly, assuming beginning-of-life battery capacity, which will be reduced. Monitoring is performed by way of a smart comparison of the operating data and related expected values under specific operating conditions. Furthermore, the BMS functions are adapted according to this comparison to ensure availability of the battery system over the course of its lifetime.


This approach was designed to be a cost- and effort-optimized method of determination for SOH during a battery life cycle and to adapt relevant BMS functions accordingly. The goal is not to replace established methods, but may improve system behavior in development projects that are very cost and time sensitive.


Graphics - battery aging

State-of-the-art approach to determine SOH during battery operation










Graphics - battery aging

Signal and functional dependencies for SOH determination and function adaption