Battery State of Health (SOH)

I. Product Definition

Battery State of Health (SOH) is a key metric reflecting the degradation level and overall health condition of a traction battery. It compares the battery's current performance against its factory-fresh state, typically expressed as a percentage—where 100% indicates that the battery's condition matches its original factory specifications perfectly. As one of the core parameters within the Battery Management System (BMS), an accurate SOH assessment is essential for ensuring operating safety and maximising service life.

II. Definition and Calculation Method

The most direct definition of SOH is the ratio of current maximum usable capacity to factory rated capacity:

$SOH=frac{C}_{\text{Current}}{C}_{\text{Rated}}times100$SOH=fracCCurrent​CRated​times100

Where C Current represents the maximum dischargeable capacity in the battery's current state, and C Rated denotes its original nominal capacity.

However, battery degradation is a multi-dimensional process, meaning capacity alone cannot provide a comprehensive evaluation of health status. Beyond capacity fade (which reflects energy capability), internal resistance increase (reflecting power capability) serves as another critical SOH indicator. Furthermore, parameters such as cycle count, voltage drop, and power fade are widely used to define SOH.

The Chinese standard "State of Health Evaluation Indicators and Estimation Error Test Methods for Electric Vehicle Traction Batteries" (T/CSAE 184-2021), published in 2021, systematically introduced five core characterisation parameters for evaluating battery SOH for the first time:

• Discharge capacity retention rate

• Discharge energy retention rate

• Energy efficiency retention rate

• Peak power retention rate

• DC internal resistance change rate

Additionally, the standard incorporates the rate of change in voltage range and variance as secondary diagnostic metrics to evaluate battery health more rigorously and comprehensively.

III. Key Factors Influencing SOH

Battery degradation is an irreversible chemical process influenced by several key factors:

• Charging Rate: The most significant contributor to SOH deterioration.

• Depth of Discharge (DoD): Deep discharging (running the battery completely flat) accelerates cell ageing.

• Temperature: Extreme heat or cold impairs battery life. Crucially, while higher ambient temperatures can enhance cell efficiency, the heightened air conditioning load under extreme heat can paradoxically accelerate capacity loss.

• Driving Habits: Aggressive driving styles, such as harsh acceleration and heavy braking, place higher electrical loads on the pack.

• Calendar Ageing: Even when left unused, batteries naturally degrade over time.

IV. SOH Estimation Methods

SOH cannot be directly measured like voltage; its estimation relies primarily on indirect BMS algorithms. Currently, mainstream approaches fall into three main categories:

1. Direct Measurement (Offline Method)

SOH is evaluated by directly taking measurements of capacity, internal resistance, and electrochemical impedance spectroscopy (EIS). While straightforward and versatile in principle, this method requires specialised test equipment, making it impractical for everyday on-road vehicle use.

2. Model-Based Method

This approach utilises equivalent circuit models (ECM), electrochemical models, or empirical degradation models to predict SOH. For instance, a second-order RC equivalent circuit model estimates the battery's ohmic internal resistance in real time, mapping this value to derive SOH. Such methods require minimal computing resources but offer limited model adaptability.

3. Data-Driven Method

Machine learning algorithms (such as support vector machines and neural networks) establish non-linear mappings between external operational features and SOH. This method does not depend on internal electrochemical mechanisms and delivers high accuracy when backed by reliable training data, though it demands substantial computing power.

Currently, hybrid approaches combining data-driven algorithms with empirical models are gaining traction, delivering high prediction accuracy whilst reducing processing demands on the BMS.

V. SOH Thresholds and Replacement Criteria

Industry standards define clear SOH thresholds regarding when a traction battery should be retired or replaced:

80%—Retirement Threshold

Under widely accepted industry standards, when a battery's SOH drops to 80%, it reaches the benchmark retirement threshold for automotive lithium-ion batteries. This baseline originates from GB/T 31484-2015 ("Cycle Life Requirements and Test Methods for Traction Battery of Electric Vehicle"). At a capacity retention rate of ≥80%, battery performance remains stable without requiring replacement; once retention falls to ≤80% alongside a noticeable drop in driving range, a formal SOH diagnostic test is recommended.

70%—Mandatory Failure Line

Under national regulatory standards, battery SOH must remain above 70%. A reading below 70% constitutes an outright inspection failure, meaning the vehicle cannot clear mandatory roadworthiness inspections and must have its battery pack repaired or replaced before returning to the road.

However, critics argue that using an arbitrary 80% cutoff as a blanket retirement standard has obvious limitations. Some packs with an SOH below 80% exhibit a shallow degradation gradient and remain perfectly usable; conversely, batteries with SOH above 80% that display steep degradation curves pose greater operational risks. Consequently, the degradation gradient (the rate of capacity loss relative to cycle count) is increasingly viewed as a more objective health indicator than a standalone SOH figure.

VI. Conclusion

Battery State of Health (SOH) serves as the primary benchmark for assessing battery degradation. Precise SOH estimation is vital for ensuring electric vehicle operating safety, refining energy management strategies, and pinpointing the ideal time for pack replacement. Driven by advancements in BMS technology, SOH tracking is progressing from basic capacity ratios to multi-parameter holistic evaluation frameworks—and from offline bench testing to real-time, data-driven onboard estimation—delivering more accurate and dependable battery health management solutions for EV owners.