Second Life Battery Capacity Grading — PatSnap Eureka
Second Life Battery Capacity Grading Technology 2026
Retired EV packs with 70–80% remaining state of health are a rapidly growing resource for stationary storage and grid services. Accurate, fast capacity grading is now the critical industrial bottleneck.
From Retired EV Pack to Second-Life Asset: The Grading Imperative
Second-life lithium-ion batteries — retired EV packs with 70–80% remaining state of health — represent a rapidly growing resource for stationary energy storage, grid services, and mobility applications. As EV fleet retirements accelerate through the late 2020s, capacity grading has become a critical industrial bottleneck separating viable repurposing from premature recycling.
The core technical domains in this dataset span electrochemical characterization methods such as Incremental Capacity Analysis (ICA), differential voltage analysis, and electrochemical impedance spectroscopy (EIS); data-driven SoH estimation using LSTM and neural networks; fast non-destructive grading protocols using rebound voltage and AC/DC resistance; and sorting logic based on capacity, chemistry, aging history, and predicted remaining useful life.
The patent record within this dataset spans 2001 to 2026, while literature concentrates between 2018 and 2024, indicating evolution from early-stage capacity estimation into a rich applied science domain. Full discharge characterization can take 4–10 hours per module; technologies offering sub-hour grading — rebound voltage methods, AC/DC resistance proxies, partial charge ICA — represent the primary near-term IP battleground.
Among the 10 patent records retrieved in this dataset, China accounts for 6 filings, with assignees including state-grid utilities, automotive research institutions, and battery manufacturers. Hefei Gotion High-Tech Power Energy Co., Ltd. is the only assignee in retrieved records with both EP and US filings on the same core grading technology, indicating active international IP protection.
Patent Activity by Technology Cluster and Filing Period
Within this dataset, patent activity clusters around four core technology approaches, with a clear acceleration in AI-augmented and digital-twin-enabled methods in the 2023–2026 window.
Patent Filings by Technology Cluster (Dataset Snapshot)
Voltage-signal-based fast estimation and ML/data-driven SoH methods each account for the largest share of identified patents in this dataset, reflecting commercial pressure to minimize test time.
↗ Click bars to explorePatent Filings by Period — Second-Life Battery Grading (Dataset Snapshot)
Recent filings (2023–2026) account for the largest identifiable cluster in this dataset, driven by LSTM-based sorting and IoT digital twin patents from Chinese and international assignees.
↗ Click bars to exploreKey Second-Life Deployment Contexts for Graded Battery Modules
Capacity grading determines which retired battery modules are routed to stationary storage, EV charging infrastructure, residential energy management, photovoltaic integration, or recycling. Each deployment context places distinct demands on grading accuracy and speed.
Stationary Grid Energy Storage
Studies confirm retired EV packs at approximately 75–80% SoH are viable for peak shaving, ancillary services, and solar-storage integration. Capacity grading determines whether modules are routed to residential or commercial storage tiers. Referenced in a 2023 critical comparison of Li-Ion aging models for second-life applications and a 2022 test method development study for renewable energy applications.
Stationary StorageResidential Building Energy Management
Retired LFP packs from EVs have been studied for smart-building storage across multiple LCA scenarios. A 2022 study on SoH estimation of second-life lithium-ion batteries under real profile operation focused specifically on Nissan Leaf modules deployed in residential self-consumption and fast-charging scenarios. Capacity grading determines per-cell and per-module usability for building integration.
Building StorageEV Charging Infrastructure Stations
Mobile and fixed fast-charge stations for EVs have been identified as second-life use cases requiring moderate SoH and precise capacity knowledge. A 2023 study on second-life batteries modeling for performance tracking in a mobile charging station presents characterization and modeling specifically for this deployment context. Both moderate capacity accuracy and cycle-life predictability are required.
EV ChargingPhotovoltaic-Integrated Storage Systems
A 2022 mathematical modelling and simulation study of a second-life battery pack with heterogeneous state of health demonstrates that uneven SoH across modules — which grading must account for — significantly affects PV-integrated system performance. Bridge Green Upcycle Corp.’s 2026 WO patent explicitly generates recycling recommendations for batteries failing second-life thresholds, including PV storage requirements.
PV IntegrationKey Patent Assignees in Second-Life Battery Grading (Retrieved Records)
In this dataset, Hefei Gotion High-Tech Power Energy Co., Ltd. is the only assignee with filings in both EP and US jurisdictions on the same core rebound voltage grading technology. Changsha Automobile Innovation Research Institute accounts for 2 of the 10 retrieved patent records, both filed in 2025, focusing on LSTM-based capacity decay path prediction for retired battery sorting.
Top Assignees by Filing Count — Second-Life Battery Grading (Dataset Snapshot)
↗ Click bars to exploreHefei Gotion High-Tech Power Energy
Hefei Gotion filed 2 patents in this dataset — one EP (2025) and one US (2026) — on the same core rebound voltage scatter-plot grading methodology, making them the only assignee in retrieved records with active cross-jurisdictional protection on a single grading technique. Their method uses discharge capacity C1, endpoint voltage V1, rebound voltage V2, and remaining capacity C2 to build prediction models enabling rapid grading without lengthy test cycles. The US filing is the PCT national-phase counterpart refining the methodology for manufacturing-scale grading.
China — CN (EP & US filings)Changsha Automobile Innovation Research Institute
Changsha Automobile Innovation Research Institute filed 2 patents in this dataset, both in 2025 (CN), covering a sorting method, system, device, and medium for retired lithium-ion batteries using LSTM models trained on capacity increment curves and non-linear capacity loss zones. These filings represent a shift toward predicting future capacity decay paths and identifying abrupt degradation breakpoints rather than measuring current SoH alone. Both patents are directed at second-life suitability assessment for industrial-scale sorting pipelines.
China — CNFour Innovation Signals from 2025–2026 Filings
The most recent filings in this dataset (2025–2026) signal a transition from laboratory-scale SoH measurement toward industrial AI-augmented sorting pipelines, continuous digital twin monitoring, and chemistry-aware routing systems.
LSTM-Driven Degradation Trajectory Prediction
The dual 2025 CN filings from Changsha Automobile Innovation Research Institute represent a methodological shift: rather than measuring current SoH alone, these systems predict the future capacity decay path, including identification of abrupt breakpoints in degradation rate. This enables proactive routing to second-life tiers based on predicted stability. The LSTM model is trained on capacity increment curves and non-linear capacity loss zones from historical cycling data.
ML Plus Chemistry-Specific Simulation Hybrid
Bridge Green Upcycle Corp.’s 2026 WO filing explicitly combines ML-based state estimation with chemistry-aware aging simulation models, enabling differentiated second-life recommendations based on whether the battery is NMC, LFP, or another chemistry. This is a critical capability as mixed-chemistry retired EV fleets scale up. The patent also generates explicit recycling recommendations for batteries that fail second-life thresholds, covering the recycling-vs-reuse decision boundary.
Fast Grading Methods vs. Full Electrochemical Characterization
Click any row to explore further.
| Dimension | Fast / Non-Destructive Grading | Full Electrochemical Characterization |
|---|---|---|
| Test Duration | Sub-hour (rebound voltage, AC/DC resistance, short discharge) | 4–10 hours per module (full discharge cycle) |
| Primary Techniques | Rebound voltage analysis, OCV characteristic points, CC-CV mode transition timing, AC/DC resistance measurement | Incremental Capacity Analysis (ICA), EIS, constant-current/constant-voltage full cycles, differential voltage analysis |
| Degradation Insight | Capacity estimate only; limited mechanistic information | Mechanistic degradation mode identification (loss of lithium inventory, loss of active material) via dQ/dV peak analysis |
| Representative Patent | Hefei Gotion High-Tech (2025, EP): rebound voltage V2 and remaining capacity C2 scatter-plot model | Audi Aktiengesellschaft (2022, DE): distinctive points in OCV aging curves for non-invasive capacity determination |
| Validation Scale | 506 cells, 203 modules, 3 battery packs (Nissan Leaf) per 2023 fast estimation literature study | 24 modules from 6 commercial EVs per 2019 ICA characterization study; 50 Ah pouch cells (50 cells) per 2023 health indicators study |
| AI Integration | LSTM trained on capacity increment curves (Changsha, 2025); feedforward neural network on batch data (Guangdong, 2025) | Physics-informed and semi-empirical degradation models; Gaussian process and neural network SoH estimation |
| Chemistry Handling | Dongfeng dual-chemistry (NCM/LFP) lifespan mapping (2024, CN); Bridge Green chemistry-specific simulation (2026, WO) | EIS shown superior for detecting subtle degradation differences between modules at module level (2021 study) |
| Industrial Suitability | High — designed for throughput-sensitive repurposing workflows; primary near-term IP battleground per strategic analysis | Lower — standard in laboratory and R&D contexts; used for benchmarking and model training datasets |
Frequently Asked Questions: Second-Life Battery Capacity Grading
According to the content of this dataset, retired EV packs entering second-life workflows typically retain 70–80% remaining state of health, making them viable for stationary energy storage, grid services, and mobility applications without full replacement.
Full discharge characterization can take 4–10 hours per module according to patents and literature in this dataset. Speed is commercially critical because test time is a primary bottleneck in industrial repurposing workflows; sub-hour methods such as rebound voltage analysis, AC/DC resistance proxies, and partial-charge ICA are described as the primary near-term IP battleground.
ICA converts voltage-capacity curves into dQ/dV spectra whose peak positions and amplitudes correlate with degradation modes such as loss of lithium inventory and loss of active material. In this dataset, a 2019 study applied ICA and CV-phase analysis to 24 modules from 6 commercial EVs, and a 2023 study evaluated 31 health indicators on 50 Ah pouch cells, identifying 13 highly aging-sensitive indicators including Coulombic efficiency and constant current charge ratio.
China (CN) accounts for 6 of the 10 patent records retrieved in this dataset. Active Chinese assignees include Changsha Automobile Innovation Research Institute, State Grid Zhejiang Power Company Huzhou Supply Branch, Yunnan Power Grid Energy Investment Co., Ltd., Guangdong Energy Storage Detection Technology Co., Ltd., Anhui Deyi Energy Technology Co., Ltd., and Dongfeng Motor Group Co., Ltd.
Traditional SoH measurement determines a battery’s current state; LSTM-based methods as filed by Changsha Automobile Innovation Research Institute (2025, CN) predict the future capacity decay path, including identifying abrupt breakpoints in degradation rate. This enables proactive routing to second-life tiers based on predicted future stability rather than current state alone.
As mixed-chemistry EV fleets retire simultaneously, grading systems that cannot differentiate NMC from LFP degradation profiles will produce unreliable outputs. Two filings in this dataset address this: Dongfeng Motor Group’s 2024 CN patent maps capacity intervals between chemistries to normalize remaining usable capacity, and Bridge Green Upcycle Corp.’s 2026 WO patent uses chemistry-specific aging simulation to generate differentiated second-life recommendations.
Data and insights on this page are based on a limited patent and literature dataset and are for reference only. Figures may not represent the complete technology landscape.