Energy storage battery state of health (SOH) assessment is a critical evaluation process that quantifies the degradation level of battery systems used in grid-scale storage, renewable integration, and backup power applications. Through standardized capacity testing, internal resistance measurement, impedance analysis, and data-driven diagnostics, SOH assessment enables operators to maximize asset life, ensure operational safety, and support residual value estimation. This article provides a comprehensive technical overview of SOH evaluation methodologies, applicable standards, testing parameters, and practical applications for energy storage battery cells, modules, and packs.
Technical Overview
State of Health (SOH) is a quantitative indicator that describes the degree of degradation of an energy storage battery relative to its fresh, factory-calibrated condition. It is most commonly defined as the ratio of the present usable capacity to the rated capacity, expressed as a percentage: SOH = (Q_present / Q_rated) × 100%. Alternatively, SOH can be defined through the ratio of present internal resistance or impedance to the initial value, since resistance growth and capacity fade are the two dominant degradation signatures in lithium-ion and other rechargeable chemistries.
Battery degradation in energy storage systems (ESS) originates from complex electrochemical mechanisms, including loss of lithium inventory (LLI), loss of active material (LAM) at both electrodes, electrolyte decomposition, thickening of the solid electrolyte interphase (SEI), lithium plating under high-rate or low-temperature charging, and current collector corrosion. These mechanisms manifest externally as capacity fade, power fade (increased impedance), accelerated self-discharge, and reduced coulombic efficiency. A rigorous SOH assessment program captures these indicators under controlled laboratory or field conditions and translates them into actionable health metrics.
The industry typically considers a battery to have reached its "end of first life" when SOH falls to approximately 80%, while end of second life is often defined at 70% or below. Between these thresholds, batteries may still deliver reliable service in less demanding applications, which makes accurate SOH evaluation essential for echelon utilization (second-life) decision-making, warranty enforcement, insurance underwriting, and asset valuation. International and national standards organizations, including IEC, ISO, UL, IEEE, and GB (Chinese national standards), have published test procedures that underpin consistent, comparable SOH reporting across laboratories and manufacturers.
Testing Samples
Lithium Iron Phosphate (LFP) Cells and Modules
Lithium iron phosphate (LiFePO4) is the dominant chemistry in modern grid-scale stationary storage due to its long cycle life, high thermal stability, and favorable cost. SOH assessment samples typically include prismatic or cylindrical LFP cells (ranging from tens to several hundred ampere-hours), modules assembled from multiple cells, and complete battery packs integrated with a battery management system (BMS). LFP exhibits a very flat open-circuit voltage curve across a wide state-of-charge (SOC) range, which complicates SOC estimation and places greater emphasis on capacity verification and impedance-based diagnostics.
Ternary Lithium Cells (NMC/NCA)
Nickel manganese cobalt (NMC) and nickel cobalt aluminum (NCA) chemistries, with their higher energy density, are found in commercial and industrial storage systems, hybrid installations, and repurposed electric vehicle batteries. These samples require careful attention to lithium plating indicators, gas generation, and impedance evolution, as degradation modes can progress more rapidly at elevated temperatures or high states of charge.
Other Chemistries
SOH assessment also applies to lead-acid battery banks in legacy stationary installations, vanadium redox flow batteries (where electrolyte imbalance and stack resistance are key health indicators), sodium-ion cells, and hybrid supercapacitor modules. Each chemistry requires adapted test protocols reflecting its specific voltage windows, rate limits, and degradation characteristics.
Field-Deployed Battery Systems and BMS Data
In addition to laboratory cell and module samples, SOH evaluation increasingly incorporates field data harvested from operating containerized BESS (battery energy storage system) installations. Historical charge-discharge curves, BMS-logged voltages, temperatures, and coulomb counters provide the raw material for non-intrusive, continuous health estimation without taking the asset offline.
Testing Items
A comprehensive SOH assessment evaluates a structured set of electrical, electrochemical, and thermal parameters. The core testing items include:
- Actual capacity and capacity retention — measured through full charge-discharge cycles at a specified rate (commonly 0.5C or 1C) and reference temperature (25 °C ± 2 °C), compared against rated capacity to calculate capacity fade percentage.
- State of Health (SOH) calculation — derived from capacity ratio, resistance ratio, or a weighted combination of both indicators.
- DC internal resistance (DCIR) — determined from the voltage response to controlled current pulses at specified SOC points, per hybrid pulse power characterization (HPPC) style procedures.
- AC impedance / electrochemical impedance spectroscopy (EIS) — frequency sweep analysis (typically 0.01 Hz to 10 kHz) revealing ohmic resistance, charge transfer resistance, and diffusion behavior correlated with aging mechanisms.
- Open-circuit voltage (OCV) characteristics — OCV-SOC curve mapping after controlled rest periods to detect voltage curve shifts indicative of lithium inventory loss or electrode imbalance.
- Self-discharge rate — capacity or voltage loss measured over extended storage (e.g., 28 days) at controlled temperature, indicating internal micro-shorts and SEI instability.
- Coulombic efficiency (CE) — ratio of discharged charge to charged charge over defined cycles; declining CE signals parasitic side reactions.
- Power capability and rate performance — maximum sustained charge/discharge power at various SOC levels, relevant for grid services requiring high-rate response.
- Cycle life verification — accelerated aging under defined duty cycles (charge rate, discharge rate, depth of discharge, temperature) with periodic capacity checkpoints to construct capacity-versus-cycle-number curves.
- Calendar aging — storage testing at elevated temperatures and defined SOC levels to quantify time-dependent degradation independent of cycling.
- Cell-to-cell consistency analysis — statistical evaluation of capacity, resistance, and voltage distribution across cells in a module or rack to identify weak cells and imbalance risk.
- Thermal and safety screening — surface temperature rise during cycling, heat generation rate, and non-destructive safety checks as supporting health indicators.
Testing Methods
Standard Capacity Verification Test
The reference method for SOH determination is the full-cell capacity test. The sample is first conditioned and fully charged using the manufacturer's specified protocol, then discharged at a constant current (typically 0.5C) to the cut-off voltage in a temperature-controlled chamber. The discharged capacity, corrected for temperature, defines the present capacity from which SOH is calculated. Testing is generally aligned with procedures described in IEC 62660-1 (secondary lithium cells for EV applications), IEC 62619 (safety and performance for industrial/ESS applications), UL 1973 (batteries for stationary applications), and GB/T 36276 (lithium cells and batteries for electrical energy storage in China).
Pulse Resistance and Power Characterization
DC internal resistance is measured by applying short, controlled current pulses (for example, 10-second discharge and regenerative charge pulses) at discrete SOC points while recording the instantaneous voltage response. The resulting DCIR values, tracked over the battery's life, reveal power fade. A resistance-based SOH is computed as SOH_R = (R_initial / R_present) × 100% or, in some frameworks, tracked as a growth ratio. Elevated resistance growth above 50–100% of initial values is a common criterion for power-limited end of life.
Electrochemical Impedance Spectroscopy (EIS)
EIS applies small-amplitude sinusoidal perturbations across a wide frequency range to deconvolute the battery's internal processes. The high-frequency intercept captures ohmic resistance (electrolyte, contacts, current collectors), the mid-frequency semicircle reflects charge transfer and interfacial resistance growth linked to SEI thickening, and the low-frequency tail relates to solid-state diffusion. Fitting spectra to an equivalent circuit model (e.g., Randles-type circuits with additional RC elements and Warburg diffusion) allows degradation mode separation and supports early fault detection.
Incremental Capacity and Differential Voltage Analysis
Incremental capacity (IC, dQ/dV) and differential voltage (DV, dV/dQ) analysis transform slow-rate charge-discharge curves into characteristic peak signatures. Peak position shifts and intensity changes correlate with specific degradation modes such as LLI and LAM, enabling mechanism-level diagnosis that simple capacity tests cannot provide. This method is particularly valuable for LFP systems despite their flat voltage plateau, as the technique amplifies subtle voltage features.
Model-Based Estimation Methods
Equivalent circuit models (ECM) — typically consisting of an ohmic resistor in series with RC networks — are parameterized from pulse test data and combined with recursive estimation algorithms such as extended Kalman filters (EKF), unscented Kalman filters (UKF), or particle filters to jointly estimate SOC and SOH in real time. Electrochemical pseudo-two-dimensional (P2D) models and single-particle models offer deeper physical fidelity, mapping aging to electrode-level parameters at the cost of higher computational burden.
Data-Driven and Machine Learning Approaches
Data-driven methods, including Gaussian process regression, support vector machines, random forests, and deep neural networks (such as LSTM and transformer architectures), learn the mapping between easily measurable features (partial charge curves, temperature profiles, impedance features, cycling statistics) and health labels obtained from reference tests. These approaches enable rapid SOH estimation from partial charging data without full discharge cycles, making them well suited to online BMS implementation and fleet-level health monitoring.
Non-Destructive Characterization Techniques
Supporting laboratory techniques include ultrasound scanning for gas detection and delamination, X-ray computed tomography for structural inspection, and thermal imaging during cycling. These methods complement electrical testing by identifying mechanical and morphological degradation that precedes measurable electrical failure.
Standards and Reference Frameworks
SOH assessment programs are commonly structured around guidance from IEC 62619 and IEC 62660 series, IEC 62933-5-2 (safety for electrical energy storage systems), UL 1973, ISO 12405 series, IEEE 1188 (stationary battery maintenance, historically lead-acid focused), and GB/T 36276, GB/T 34131 (BMS for energy storage), and related GB/T standards for capacity and cycle test methods. Accredited third-party laboratories conduct these evaluations under ISO/IEC 17025-accredited quality systems to ensure traceable, defensible results.
Applications
SOH assessment delivers measurable value across the entire lifecycle of energy storage assets:
- Commissioning and acceptance testing — verifying that newly delivered BESS containers, racks, and modules meet contractual capacity and resistance specifications before commercial operation, protecting buyers against underperforming assets.
- Periodic health audits and performance guarantee verification — annual or semi-annual capacity verification supporting capacity availability guarantees, availability payments, and performance-based service contracts in grid services projects.
- Warranty claims and dispute resolution — providing independent, standards-based evidence of degradation beyond manufacturer warranty thresholds, often required by insurers, financiers, and asset owners.
- Second-life (echelon utilization) grading — retired EV and ESS batteries are screened, graded, and re-sorted by measured SOH, internal resistance, and self-discharge rate before redeployment in less demanding stationary storage applications.
- Asset valuation and financial due diligence — residual capacity data underpins fair market valuation for battery leasing, resale, refinancing, and insurance underwriting of storage portfolios.
- Predictive maintenance and safety management — continuous SOH and impedance trending enables early detection of weak cells, developing internal shorts, and thermal runaway precursors, allowing scheduled replacement before failures occur.
- BMS algorithm development and benchmarking — laboratory reference SOH data calibrates and validates onboard estimation algorithms, improving accuracy across large deployed fleets.
- R&D and supplier benchmarking — comparative cycle life and calendar aging studies guide cell selection, chemistry qualification, and design margin decisions for new storage products.
FAQ
What SOH value indicates that a battery needs replacement?
Most grid storage warranties define end of first life at 80% SOH, at which point the battery can no longer deliver its guaranteed usable energy. In practice, replacement decisions consider SOH alongside power capability, cell consistency, and economic factors; many systems continue operating profitably between 70% and 80% SOH with derated performance.
How is SOH different from SOC?
SOC (state of charge) describes how much energy is currently stored in the battery relative to its present maximum capacity — it changes minute by minute during operation. SOH describes how much of the battery's original capability remains after long-term degradation — it changes slowly over months and years of use. SOC answers "how full is the tank now," while SOH answers "how large is the tank today."
Can SOH be measured without a full discharge?
Yes. Partial charging segment analysis, impedance measurements, OCV curve matching, and model-based or machine learning estimators can infer SOH from routine operating data without removing the battery from service. Full charge-discharge reference tests remain the most accurate method and are typically reserved for periodic calibration and acceptance testing.
How often should SOH assessment be performed?
Field estimation should be continuous via the BMS, with laboratory-grade reference capacity tests recommended annually or at major contractual milestones. Systems operating under aggressive cycling, high ambient temperatures, or frequent high-rate grid services benefit from more frequent verification, for example every six months.
Which factors most strongly accelerate battery degradation in storage applications?
The dominant stress factors are elevated temperature, prolonged residence at high SOC, high charge rates (particularly at low temperature, which promotes lithium plating), deep discharge cycling, and high cumulative throughput. Controlling the thermal environment and moderating SOC operating windows can substantially extend service life.
How long does a laboratory SOH assessment take?
A single capacity verification test on a cell or module typically takes one to several days including conditioning and temperature stabilization. A full SOH assessment package combining capacity, DCIR, EIS, self-discharge, and consistency analysis generally requires one to three weeks, while cycle life verification studies extend over months by design.
Which standards govern SOH and performance testing for storage batteries?
Key references include IEC 62619, IEC 62660-1/-2, IEC 62933 series, UL 1973, ISO 12405 series, and GB/T 36276 together with GB/T 34131 for battery management systems. Selecting the applicable framework depends on the target market, chemistry, and whether the assessment supports safety certification, performance verification, or warranty purposes. Accredited third-party laboratories can advise on the most appropriate combination of standards for a specific project.