How to Calculate Battery State of Health (SOH)

Definition and measurement context

SOH is defined as the ratio between a battery’s present capability and its rated capability at beginning of life (BOL). Capability may be specified as usable capacity at a reference C-rate, internal resistance at a defined state of charge (SOC), or a composite indicator combining several degradations. The resulting percentage is unitless but traces directly to laboratory measurements anchored in standards such as IEC 62660 for automotive cells or IEEE 1491 for stationary storage.

Field programs typically pair periodic capacity checks with electrochemical impedance spectroscopy (EIS) or direct current internal resistance (DCIR) tests. Aligning the discharge protocol with the original factory test is essential; otherwise the comparison mixes methodological noise with actual ageing.

Variables, units, and reference conditions

To quantify SOH, collect the following variables while holding ambient temperature, SOC window, and rest times consistent with the commissioning test:

• C_rated (Ah): the manufacturer’s rated capacity at the reference discharge rate and temperature.
• C_meas (Ah): measured dischargeable capacity using the same current profile and cut-off limits.
• R_rated (mΩ): BOL internal resistance, typically measured at 50% SOC via a short pulse or EIS sweep.
• R_meas (mΩ): current internal resistance at the identical SOC and test method.
• n (cycles): total equivalent full cycles accumulated, used for degradation modelling and validation, though not always needed in the base formula.

Record environmental metadata—temperature, humidity, rest durations—because they explain outliers and inform correction factors. The recommended logbook mirrors the cycle history tables maintained for assets evaluated with the EV TCO Battery Degradation Calculator, ensuring comparability across planning tools.

Core formulas and composite SOH

The simplest capacity-based state of health is the ratio SOH_capacity = (C_meas / C_rated) × 100% which captures lost lithium inventory and electrode degradation manifesting as reduced ampere-hours. However, ageing also elevates internal resistance, throttling peak power and accelerating heat generation even when usable capacity remains acceptable. A composite metric therefore averages the retention of capacity with the inverse of resistance growth:

SOH_composite = 0.5 × [ (C_meas / C_rated) + (R_rated / R_meas) ] × 100%

The weighting coefficient (0.5) assumes equal importance between energy throughput and power capability—an assumption appropriate for grid firming batteries governed by revenue stacking models such as the ones optimised when applying the Levelized Cost of Storage walkthrough. Operators with power- or energy-dominant duty cycles may adopt alternative weights, but the averaging framework remains consistent.

Step-by-step calculation workflow

1. Stabilise the cell or pack. Charge or discharge to the reference SOC (often 100% for capacity tests and 50% for resistance), allow thermal equilibrium, and log the rest period.
2. Measure discharge capacity. Apply the defined current profile, integrate current over time to produce ampere-hours, and record C_meas. Flag interruptions such as protective cut-offs.
3. Measure internal resistance. Execute a pulse or EIS sequence at the same SOC, ensuring instrumentation bandwidth matches the specified test. Convert the voltage response to resistance, yielding R_meas.
4. Compute ratios. Divide capacity and resistance readings by their rated values. If resistance testing is not available, note the omission explicitly—the composite formula defaults to unity for the resistance term.
5. Average and convert to percentage. Apply the composite formula or the capacity-only variant, multiply by 100, and round to two decimals for reporting consistency.
6. Contextualise with cycle count. Compare the obtained SOH to the expected degradation trajectory at the recorded cycle number. Deviations beyond tolerance warrant re-running the test or investigating operational stressors.

Automating these steps with a reproducible script or a calculator reduces transcription errors. When the workflow feeds asset models—such as revenue projections tied to the LCOS calculator—consistent rounding and units are non-negotiable.

Validation techniques

Treat the computed SOH as a hypothesis about electrochemical health. Validate it through multiple lenses:

• Mass balance check: Verify that measured coulombic efficiency across the test cycle aligns with historical averages; large deviations often indicate instrumentation drift.
• Model fit: Feed SOH, temperature, and cycle count into degradation models (Arrhenius-derived or data-driven) to confirm the estimate lies within predicted confidence intervals.
• Power capability correlation: Use high-frequency load tests to ensure that the observed voltage sag matches the resistance-derived component of SOH.
• Redundancy across instruments: Re-test with an alternate cycler or impedance analyser when the SOH trend breaks monotonicity.

Document every validation artefact in the asset log. Doing so facilitates audits and supports cross-team collaboration when reconciling SOH with warranty triggers or energy market obligations.

Limits, assumptions, and interpretation

The composite formula assumes linear weighting between energy and power indicators. Batteries operating in strictly energy-arbitrage applications may down-weight the resistance term, while high-power systems (frequency regulation) could elevate it. Furthermore, the ratio framework ignores recovery phenomena such as lithium plating re-dissolution after rest. Treat the output as a conservative snapshot rather than an immutable truth.

Be cautious with results above 100%. They typically arise from calibration drift, elevated temperatures increasing apparent capacity, or rounding interactions when R_meas is slightly below R_rated. Repeating the test under tighter control or applying temperature correction factors usually resolves the anomaly.

Finally, communicate uncertainty. Reporting SOH with a ± tolerance derived from repeated trials or Monte Carlo propagation builds trust with stakeholders who must align maintenance plans, warranty coverage, and financial projections.

Worked example

Consider a 100 Ah battery module originally rated at 4.0 mΩ DCIR. After 1,200 equivalent full cycles, laboratory testing produces C_meas = 88 Ah and R_meas = 5.0 mΩ. The ratios are C_meas/C_rated = 0.88 and R_rated/R_meas = 0.80. Averaging yields 0.84, and multiplying by 100 results in 84%. The module therefore retains 84% of its composite capability. If resistance data were unavailable, the capacity-only SOH would be 88%, signalling why transparency about missing measurements is critical.

Feeding this SOH into augmentation planning models affects both capital forecasts and dispatch strategy. For example, lowering the state-of-charge operating window to slow further resistance growth might shift revenue expectations, a sensitivity already captured when you compare LCOS outputs before and after the degradation update.

Automating the computation

Manual calculations invite transcription errors and inconsistent rounding. Embedding a deterministic calculator inside your reporting workflow ensures that capacity tests, resistance checks, and documentation share the same formula and formatting conventions. The embedded tool below enforces unit consistency, handles optional resistance inputs, and reports the result with two-decimal precision. For day-to-day runtime planning, you can complement this with the Battery Life Calculator so operational teams interpret SOH alongside expected discharge durations.