How Does a BMS Estimate State of Health (SOH)?
A BMS estimates state of health (SOH) by tracking capacity fade and rising internal resistance over time, comparing a battery’s current performance against its original, beginning-of-life values. Like SOC, SOH can’t be measured directly — it’s calculated from patterns in how the battery behaves as it ages.
SOC vs SOH: A Quick Distinction
SOC answers “how full is the battery right now?” SOH answers a different question entirely: “how much total capacity can this battery hold compared to when it was new?” A battery at 50% SOC could have 100% SOH or 80% SOH — the two numbers are independent and serve different purposes.
The Two Main Signs of Aging
Capacity Fade
Capacity fade is the gradual loss of how much total energy a battery can store. A pack that held 100 kWh when new might only deliver 85 kWh after several years of use. This is the most intuitive way to think about SOH, since it directly represents how much range or runtime has been lost to aging.
Rising Internal Resistance
As a battery ages, chemical changes inside the cells, including growth of the SEI layer and electrode degradation, cause internal resistance to climb. Higher resistance means more voltage drop under load, reduced peak power delivery, and more heat generated during charging and discharging. This is why an older battery might feel less punchy under acceleration even before its total capacity has dropped much.
How the BMS Actually Estimates SOH
Capacity-Based Estimation
By comparing how much energy a full charge-to-discharge cycle actually delivers against the battery’s original rated capacity, the BMS can calculate a straightforward capacity-based SOH percentage.
Internal Resistance Measurement
The BMS can apply a known current pulse and measure the immediate voltage response, or use more advanced electrochemical impedance analysis, to measure resistance directly. Comparing this to the beginning-of-life resistance gives a complementary health indicator, especially useful for catching power fade before capacity loss becomes significant.
Cycle Counting
Since capacity loss and cycle count are correlated for most lithium-ion chemistries, tracking the number of completed charge-discharge cycles offers a simpler, though less precise, way to estimate remaining battery life.
Model-Based and Machine Learning Methods
More advanced BMS platforms combine multiple inputs, voltage, current, temperature, and historical usage patterns, into predictive models or neural networks trained to estimate SOH more accurately than any single method alone.
Why Combining Capacity and Resistance Matters
Capacity fade and resistance increase don’t always move together at the same rate. A battery could show relatively strong capacity retention while its internal resistance has climbed significantly, meaning it delivers a similar range but noticeably less peak power. Tracking both gives a fuller picture than either measurement alone.
What Counts as “End of Life”
Most manufacturers define end of life as roughly 70-80% of original capacity, or a defined increase in internal resistance, whichever threshold is reached first. Reaching that point doesn’t mean the battery stops working — it means it no longer meets the original performance specification.
| Method | What It Measures |
|---|---|
| Capacity-based estimation | Total energy storage compared to original rating |
| Internal resistance measurement | Power delivery capability and heat generation |
| Cycle counting | Simple correlation-based estimate from usage history |
| Model-based/ML methods | Combines multiple signals for higher accuracy |
Conclusion
A BMS estimates state of health by tracking two core aging signals, capacity fade and rising internal resistance, and combining them with cycle counts and increasingly sophisticated predictive models. Since neither measurement alone tells the complete story, modern BMS platforms blend multiple approaches to give the most accurate possible picture of how much life a battery has left.
FAQs
Can SOH improve over time?
No, SOH generally only declines, though its rate of decline can slow with good charging habits and moderate temperatures.
Is SOH the same across every cell in a pack?
No. Individual cells can age at different rates, which is one reason cell balancing and monitoring remain important even as a pack ages.
How often is SOH recalculated?
It’s typically updated gradually over time as new charge cycles and usage data accumulate, rather than instantly like SOC.
