Battery SOH describes how much capability remains, while RUL estimates how long that capability will remain usable. SOH is evaluated primarily through capacity retention, internal-resistance growth, impedance, temperature behavior, and safety-related indicators. RUL is estimated by modeling how those measurements will evolve until a defined end-of-life threshold is reached. During second-life grading, precision battery testing systems provide the repeatable measurements needed to distinguish usable cells from cells that are likely to mismatch, underperform, or fail early.
Second-life grading is only as reliable as the degradation data behind it. Precision testing converts inconsistent end-of-life cells into comparable evidence about capacity, resistance, operating behavior, and future lifetime.
What SOH Measures
Capacity Retention
SOH commonly expresses the battery's available capacity relative to its initial capacity:
[ SOH_{capacity} = \frac{C_{current}}{C_{initial}} \times 100% ]
A cell with 80% of its original capacity has experienced greater degradation than a cell retaining 95%, assuming the measurements use comparable test conditions.
Capacity is usually determined through controlled charge-discharge cycling, often using a constant-current, constant-voltage charging protocol followed by a defined discharge procedure.
Internal Resistance and Impedance
Internal resistance generally increases as a battery ages. A resistance-based SOH indicator can be expressed as:
[ SOH_{resistance} = \frac{R_{initial}}{R_{current}} \times 100% ]
This indicator complements capacity retention because two cells may have similar capacity but very different power capability, heat generation, or voltage drop under load.
Impedance spectroscopy and related electrochemical measurements can provide additional information about degradation mechanisms, although they require controlled instrumentation and careful interpretation.
Operating and Safety Behavior
SOH evaluation should also consider cell voltage, temperature response, charge acceptance, self-discharge, and behavior under dynamic loads. These measurements help identify cells that appear healthy under a simple capacity test but behave abnormally during real operating conditions.
State of Charge (SOC) and State of Power (SOP) are related but different quantities. SOC describes available charge at a particular moment, while SOP describes the power the battery can safely deliver or accept.
How RUL Is Estimated
Defining the End-of-Life Threshold
RUL is not meaningful until the system defines what “end of life” means. A common convention is an available capacity of 80% of the initial value, but the applicable threshold depends on the cell chemistry, application, safety requirements, and operating conditions.
Resistance growth, loss of power capability, excessive temperature rise, or safety limits may also define EOL. Therefore, the threshold used for an electric vehicle may differ from the threshold used for a stationary energy-storage system.
Tracking Degradation Over Time
RUL estimation uses historical measurements to project future degradation. Capacity fade, resistance growth, impedance changes, temperature behavior, and load history can all contribute to the estimate.
Controlled aging tests are valuable because they produce performance records across many charge-discharge cycles. These records establish the degradation trajectory needed to identify acceleration points, including the aging “knee” where capacity or performance begins to decline more rapidly.
Model-Based and Data-Driven Methods
Several prognostic approaches are used:
- Extended Kalman Filters (EKF): Estimate hidden battery states from measurements such as voltage and current while updating the estimate as new data arrives.
- Support Vector Machines (SVM): Classify health conditions or map measured features to SOH and RUL estimates.
- Time-series models such as ARIMA: Model statistical trends in degradation measurements over time.
- Particle filters: Represent uncertainty in nonlinear or variable-load battery behavior and can estimate events such as end of discharge.
- Other machine-learning models: Learn degradation patterns from large aging datasets, provided the training data reflects the cells and conditions being evaluated.
No model eliminates the need for high-quality measurements. An algorithm can process noisy data, but it cannot reliably recover degradation information that was never measured or was measured inconsistently.
Validating the Prediction
RUL models should be evaluated against known aging results rather than accepted solely because they fit historical data. Common metrics include Mean Absolute Error (MAE), Root Mean Square Error (RMSE), relative prediction error, and variation across repeated tests.
Validation should also examine whether errors increase under different temperatures, load profiles, cell batches, and aging conditions. A model that performs well on one laboratory dataset may not generalize to returned battery packs with unknown usage histories.
Why Precision Testing Matters in Second-Life Grading
Returned Cells Are Not Uniform
End-of-life packs can contain cells with different capacities, resistance values, temperatures, degradation mechanisms, and usage histories. Cells that came from the same pack are not automatically equivalent after years of operation.
This variation makes nameplate ratings and pack-level measurements insufficient for detailed second-life selection. Individual cells or modules require comparable diagnostic tests before they are reassembled.
Capacity Determines Usable Energy
A precision cycler measures how much charge a cell can actually store and deliver under defined conditions. This provides the evidence needed to sort cells by capacity retention and to select modules that meet the energy requirements of the intended application.
For example, an operator may use an 80% capacity threshold as a screening rule. That threshold must be applied consistently, with the same temperature, current, voltage limits, rest periods, and measurement procedure across the evaluated cells.
Resistance Determines Power and Heating
Capacity alone does not determine suitability. A cell with acceptable capacity but unusually high resistance may experience larger voltage sag, reduced power capability, and greater heat generation during operation.
Precision resistance and impedance measurements help identify these cells before assembly. This is particularly important when modules will operate under high current or frequent cycling.
Consistent Data Enables Matching
Cell matching is a central second-life challenge. Cells with substantially different capacity or resistance can develop unequal SOC levels and uneven stress during charging and discharging.
Accurate grading supports better decisions about:
- Cell and module sorting.
- Series and parallel grouping.
- Reassembly strategies.
- Inventory classification.
- Expected system performance.
- Monitoring and protection requirements.
The objective is not merely to find cells that still work. It is to assemble groups whose behavior remains sufficiently consistent throughout the intended operating window.
Controlled Testing Creates a Defensible Baseline
Laboratory testing systems control current, voltage, temperature, cycling sequence, and data capture. This makes results repeatable and allows engineers to compare returned cells against common acceptance criteria.
The same baseline also supports later RUL monitoring. Without an accurate initial measurement, a future decrease in capacity or increase in resistance cannot be quantified reliably.
Understanding the Trade-offs
Testing Time Versus Throughput
A detailed capacity and aging assessment can require long charge-discharge cycles, rest periods, and repeated measurements. Testing every returned cell extensively may therefore reduce processing throughput.
A practical grading process often uses staged testing: rapid screening first, followed by more detailed characterization for cells near acceptance boundaries or intended for demanding applications.
Measurement Precision Versus Cost
Higher-precision cyclers, thermal control, impedance instruments, and safety systems increase equipment cost. However, inadequate precision can create more expensive consequences through incorrect grading, premature failure, warranty exposure, or unsafe module assembly.
The required accuracy should be tied to the application risk and the separation needed between acceptance categories.
Model Accuracy Versus Generalization
A sophisticated prognostic model may achieve low error on a controlled dataset but perform poorly on cells with different chemistries, ages, temperatures, or operating histories. Reported error values are meaningful only when the test conditions and validation population are clearly understood.
RUL should therefore be reported with uncertainty or confidence information where possible, rather than as an apparently exact date or cycle count.
Capacity Thresholds Versus Actual Application Needs
An 80% capacity threshold is a widely used convention, not a universal rule for every second-life application. A stationary system with modest power demands may tolerate a different degradation profile than a system requiring high peak power, rapid charging, or strict thermal limits.
Grading criteria should reflect the complete application requirement, including energy, power, safety, temperature, cycling frequency, and expected service life.
Making the Right Choice for Your Goal
The testing strategy should reflect what the second-life system must reliably deliver.
- If your primary focus is energy capacity: Prioritize controlled capacity testing, consistent SOC and temperature conditions, and capacity-retention thresholds that match the storage application.
- If your primary focus is power performance: Emphasize internal resistance, impedance, voltage sag, thermal response, and dynamic-load testing in addition to capacity.
- If your primary focus is accurate RUL prediction: Build a repeatable aging dataset, define EOL explicitly, track degradation across cycles, and validate models with MAE, RMSE, relative error, and repeated-condition testing.
- If your primary focus is safe module reassembly: Use cell-level grading and matching based on capacity, resistance, self-discharge, temperature behavior, and abnormal operating responses.
- If your primary focus is high processing throughput: Combine rapid screening with targeted precision characterization for borderline or high-value cells.
Reliable second-life decisions come from combining precise measurements, application-specific thresholds, and validated predictions of future degradation.
Summary Table:
| Indicator | Measurement | Why It Matters |
|---|---|---|
| SOH (Capacity) | Capacity retention (%) | Determines usable energy and sorting by remaining capacity |
| SOH (Resistance) | Internal resistance/impedance | Affects power capability, voltage sag, and heat generation |
| RUL | Model-based/data-driven prediction | Estimates remaining lifetime before EOL threshold |
| Precision Testing | Controlled cycling, impedance, etc. | Provides repeatable, accurate data for grading and matching |
Ensure accurate SOH/RUL evaluation and reliable cell grading with KINTEK's precision battery testing systems. Our portfolio covers the entire cell fabrication and testing workflow, from slurry mixing to testing systems, designed for second-life applications. Contact us today to optimize your grading process and maximize ROI — talk to a specialist now!