Direct answer: Battery storage SOH accuracy describes how close a reported state-of-health estimate is to the true remaining capacity. The reading can drift because SOH is estimated from voltage, current and temperature rather than measured directly, and because temperature, rest and operating history all move the estimate. A controlled capacity test is the reference check, and asset decisions should account for the uncertainty in the reported value.
State of health is usually defined as the ratio of current capacity to the capacity when the battery was new. It matters for warranty, replacement planning and reserve margin, but a battery management system cannot measure that ratio directly during normal operation. It estimates the value, and the estimate carries error.
This guide explains where that error comes from and how to interpret a reported SOH reading. It is a planning reference; the exact estimation method and any test procedure must come from the documentation for the installed system.
What SOH is really estimating
A useful SOH value is a capacity ratio, but capacity is not an instantaneous measurement. The system must infer it from how much charge the battery accepted, how the voltage responded, and how those responses compare with a reference model. The result is a model output, not a direct reading like a temperature.
Because the definition can vary, confirm what the product reports. Some systems report capacity relative to nameplate, others relative to a field test, and the two can differ. The same percentage can mean different things on different products.
Confirm also whether the value is a fleet or a module-level number. A module-level estimate can be more useful for maintenance, while a system-level value smooths out individual differences.
Why capacity-based SOH is hard to measure
Measuring true capacity requires a controlled charge and discharge under defined current and temperature. Normal operation rarely provides that. Partial cycles, varying current and changing temperature all mean the system never sees the clean reference conditions that a capacity test provides.
As a result, the estimator works with incomplete information and must fill the gaps with a model. The model is usually good, but its confidence depends on how close recent operation has come to the reference conditions.
A field estimate therefore inherits the uncertainty of the operating history. Two identical batteries on different duty cycles can show different SOH estimates even when their true capacities are similar, because the estimator has seen different evidence.
Estimation methods and their errors
Common approaches include tracking cumulative charge passed, matching voltage behavior to a model, and counting cycles against a degradation curve. Each has a characteristic error. Coulomb counting drifts with sensor offset, model matching is sensitive to temperature and age, and cycle counting ignores calendar aging unless it is added separately.
A system often blends several methods, which reduces but does not eliminate error. The reported SOH should therefore be treated as an estimate with a range, not a precise number.
The error also depends on how the estimator was tuned for the specific cell chemistry. A model tuned for one operating window may be less accurate when the site runs a very different duty cycle, so the same product can report different accuracy on different projects.
Temperature and rest change the reading
Temperature moves both the voltage and the available capacity, so an estimate made when cold can differ from one made when warm even on the same battery. A long rest can also change the reading because voltage relaxes and the estimator sees a different state than during cycling.
This is why SOH can appear to jump after a rest or after a temperature change without any real change in the battery. Comparing readings across conditions requires care.
Because of this, a single reading should not be compared across seasons without normalizing for temperature. A trend built from readings at a similar temperature and state of charge is far more reliable than a trend that mixes summer and winter values.
A capacity test is the reference
A controlled capacity test, run at the manufacturer’s specified current and temperature, provides the closest thing to a true SOH reference. The test measures delivered capacity directly and can be compared with the reported estimate to quantify the error.
| Source | How it works | Typical limitation |
|---|---|---|
| Model estimate | Infers SOH from voltage and current | Drifts with conditions |
| Coulomb counting | Accumulates charge passed | Sensor offset over time |
| Cycle counter | Counts cycles against a curve | Ignores calendar aging |
| Capacity test | Measures delivered capacity | Needs controlled conditions |
The capacity test is the audit that keeps the other methods honest. Running it periodically, under the same conditions each time, shows the true trend even when the day-to-day estimate wanders.
Degradation is not linear
Capacity fades faster early in life for some chemistries and then slows, while calendar aging continues in the background. A linear assumption built from a short operating history can overstate or understate the remaining life. The SOH trend, combined with the product’s expected degradation profile, gives a better forecast than a straight line drawn through two points.
Using SOH in asset decisions
When SOH feeds a warranty, replacement or reserve decision, use the range and the test history rather than a single displayed percentage. A value that is drifting upward after a rest may be a condition effect, while a steady downward trend confirmed by capacity tests is the signal that matters.
Keep the test records with the temperature and current so each measurement is comparable. A trend line of controlled tests is more useful than any single estimate from the field.
If the estimate and the test disagree by more than the product’s stated tolerance, investigate before acting. The gap may indicate a sensor issue, a model configuration error, or a test that was run outside the defined conditions.
Frequently asked questions
What is battery storage SOH accuracy?
It is how close a reported state-of-health estimate is to the true remaining capacity. SOH is estimated from voltage, current and temperature, so the reading carries error that changes with conditions.
Why does SOH change after a rest period?
Rest changes temperature and voltage relaxation, which moves the estimator even when the battery itself has not changed. Compare readings made under similar conditions to see the real trend.
How do I get an accurate SOH value?
Run a controlled capacity test at the manufacturer’s specified current and temperature, then compare the measured capacity with the reported estimate. Repeat under the same conditions to build a trend.
Can I use SOH for a warranty claim?
Use the controlled capacity test results and the test history rather than a single displayed percentage, and follow the warranty’s defined test method. The reported estimate alone may not be the agreed measure.
What to send for a technical review
Share the reported SOH, the capacity test history, the test conditions and the installed system model with the AllGrand engineering team. The team can review the estimate against the product’s reference method.
Review the lithium energy storage product family or read more battery storage technology guides. AllGrand Group has manufactured batteries since 1987 and supplies energy storage systems, lead-acid batteries and cells to customers in more than 100 countries. SOH interpretation must be confirmed for the actual product.
Sources and further reading
These sources were used to verify general state-of-health and degradation concepts. The comparison table and interpretation guidance are original AllGrand editorial work.