Source-linked AI summary
Battery Second Life : A Review of Experimental Studies
Marwan Hassini, Eduardo Redondo-Iglesias, Pascal Venet
TL;DR
Experimental studies on retired batteries have not been synthesized, despite the relevance of reuse to reducing battery-related environmental impacts. This review examines experimental work and the assumptions used to assess retired batteries. It reports that proposals for differing end-of-life health thresholds remain insufficiently supported by experiments, which could give them greater leverage.
Problem
Experimental findings on retired batteries had not been synthesized, although reuse can reduce virgin-material extraction and improve resource use.
Source: §1 · pp. 2–3Method
The review critically evaluates experimental studies and compares common practices for choosing retired batteries, assessing dispersion, and characterizing battery health.
Source: §3 · p. 9, §6 · p. 33Results
The 20–30% capacity-loss threshold remains common despite proposals for end-of-life thresholds that vary with battery health state, based on observations and simulations.
Source: §3.1 · p. 10Takeaways & Limitations
The authors argue that experimental results supporting alternatives to the common threshold could give those proposals greater leverage.
Source: §3.1 · p. 10The review may have missed relevant studies because “second life battery” is ambiguous and battery terminology lacks global consistency.
Source: §2.3 · p. 7
Abstract
from arXiv · showhide
As electric vehicle volume increase, the interest for reusing their retired batteries grow. Giving a second life to batteries reduce their environmental impact and provide access to more affordable energy storage systems. This lifespan extension is of interest in a wide range of fields as the reuse offer a solution for end-of-life batteries management and the development of renewable energies. Although several review articles have described opportunities and barriers related to reuse, to our knowledge none have synthesized insights from studies that conducted experiments on retired batteries. This article aims to address this gap by providing a detailed analysis of more than 150 articles. This article's main findings are the proposition of a new vocabulary for batteries end-of-life, an evaluation of state of health and dispersion in retired batteries as well as the synthesis of aging and safety studies on these batteries. In a nutshell, this review show that the performance of retired batteries can be high, their dispersion should be managed and their degradation can be slow and safe. Limitations of current battery research ideas for future experiments and ideas for future studies are also presented.
1. Introduction
Battery reuse can reduce the environmental burden of producing new batteries, but reviews have not synthesized experimental findings on retired batteries. This review aims to address that gap and make research assumptions and findings more visible and comparable.
1. Introduction
- Reusing retired batteries can reduce virgin-material extraction and improve resource use as battery demand grows.The passage links growing battery demand to concerns about raw-material extraction and processing.Source: §1 · p. 2
- Earlier reviews covered reuse opportunities and barriers, but had not synthesized findings from experiments on retired batteries.The authors identify this as a gap despite the growing experimental literature.Source: §1 · p. 3
- Without synthesized experimental evidence, studies commonly assume batteries need replacement after performance degradation and neglect cell-to-cell variation.The authors say these assumptions affect energy-system design and management, and economic and environmental cost studies.Source: §1 · p. 3
- Experimental studies on retired batteries often remain one-off efforts, limiting visibility and comparison across works.The authors argue synthesis is needed to organize results and clarify the state of knowledge, controversies, and information gaps.Source: §1 · p. 3
- The review aims to critically synthesize experimental findings to assess research progress, identify opportunities, and increase the visibility of existing work.The authors present this as support for more realistic and incremental research.Source: §1 · p. 3
2. Method
The review applies PRISMA-guided literature selection and synthesizes experimental studies of retired batteries, while clarifying terminology and identifying limits in the available data and sample.
2.1. Literature review
- The review followed PRISMA principles, searched six literature databases, and screened studies for relevance, scope, duplication, and peer-review status.References and citing papers were also checked for relevant publications.Source: §2.1 · pp. 4–5
2.2. General review of existing studies about second-life batteries
- 157 articles covering more than 15,000 retired batteries were investigated, and first-life duration was unknown for 90% of batteries.The review covered studies published from 2012 to 2026.Source: §2.2 · pp. 5–6
- Study volume rose after 2021 as reviewed technologies shifted from cobalt-rich LCO and LMO toward LFP, with cell-scale studies predominating.The review notes that cell assembly studies are needed to understand cell-to-cell variation.Source: §2.2 · p. 6
2.3. Study limitations
- The review proposes distinct battery-life-cycle terms and notes that ambiguous use of “second-life battery” may cause relevant studies to be missed.It distinguishes terms including retired, reused, remanufactured, repurposed, and spent batteries.Source: §2.3 · pp. 7–8
2.4. How to compare results ?
- Only 7% of reviewed articles shared raw data, and digitizing plotted data may lose points where large, homogeneous datasets overlap.The authors recommend publishing such results in tables or separate files.Source: §2.4 · p. 8
2.5. A representative sample ?
- The review is a snapshot with uncertain representativeness for future retirements; 72% of reviewed studies used cells, despite complex battery disassembly.The authors note that disassembly costs may limit the economic interest of battery reuse.Source: §2.5 · p. 8
3. Choosing retired batteries
The section assesses retired-battery suitability through state-of-health evidence, cell dispersion, and rapid characterization, questioning reliance on a fixed capacity-loss threshold.
3.1. State of health evaluation of retired batteries
- The review tests the common 20–30% capacity-loss end-of-life threshold against experimental data, noting that it persists despite varied battery uses and applications.The threshold was defined in automotive standards in the 1990s and remains widely used, although aging studies often stop before reaching it.Source: §3 · p. 9, §3.1 · p. 9
- Retired-battery performance spans 0% to over 110%, with a median of 84.5%, a mean of 79.9%, and a Weibull fit.The synthesis collected characterization results from 104 articles covering more than 10,000 retired cells.Source: §3.1 · p. 10, Table 4 · p. 10
3.2. Evaluation of performance dispersion
- Most tested retired batteries have capacity spread below 2% without outliers, but 20% of studies report spread above 4% and up to 12% outlier cells.Spread is standard deviation divided by mean capacity; an outlier is more than 1.5 standard deviations from the mean. Production, assembly, and use can contribute to cell heterogeneity.Source: §3.2 · pp. 11–12
3.3. Fast performance assessment
- The proposed suitability procedure has four steps: defect detection, degradation-mechanism evaluation, performance assessment, and dispatch to reuse or recycling.The first two steps need pack-level field methods, which the review says have received limited research attention.Source: §3.3 · p. 12
- Charge–discharge testing benchmarked below 2.5% error in under 10 minutes, while impedance spectroscopy reached below 1.9% in under 3 minutes and 30 seconds.The impedance approach requires setting state of charge to 100%; both approaches need further assessment of duration, cost, and practical deployment.Source: §3.3 · pp. 15–16
- Incremental capacity analysis is widely studied, but inconsistent health-indicator definitions and accuracy metrics make results difficult to compare across studies.The review renames curve features as peaks and valleys according to their voltage.Source: §3.3 · pp. 17–18
- Ultrasonic measurement is a novel non-destructive assessment option, but its time and energy consumption remain insufficiently studied for industrial suitability.Further work is needed to evaluate this technique and develop other non-destructive techniques.Source: §3.3 · p. 18
- No single assessment technique appears suitable for every battery model; the review calls for application-matched indicators, common metrics, and open datasets.It also identifies pack-level defect detection and characterization time and cost as research gaps, and proposes matching battery-passport information to reuse needs.Source: §3.3 · pp. 18–19
4. Aging of reused batteries
Experimental studies show that reused batteries can have widely varying degradation paths and lifespans, while conditions, cell dispersion, and aging history shape durability. The review also synthesizes evidence on deep degradation and safety, where findings identify both promising behavior and areas needing further study.
4.2. General observations
- No common degradation pattern or post-threshold acceleration emerged, and second-life duration showed no general correlation with first-life state of health.Many batteries reached thousands of cycles and state of health below 60%.Source: §4.2 · p. 21, Figure 6 · p. 21
4.3. Focus on short lifespan studies
- Short lifespans generally followed harsh aging conditions, while some campaigns ended early by design rather than because cells degraded rapidly.In one study, cells charged at 0.9C had shorter lives than cells charged at 0.45C; another campaign stopped after 300 cycles despite homogeneous, slow degradation.Source: §4.3 · pp. 22–23, Figure 7 · p. 22, Figure 8 · p. 23
4.4. Focus on long lifespan studies
- Second-life campaigns exceeded 9,000 cycles, with room temperature, limited current, and restricted state-of-charge ranges among conditions associated with long durability.Long lives also occurred below 80% state of health; in one study, cells cycled between 45–55% SoC under frequency-regulation-like use reached nearly 10,000 cycles.Source: §4.4 · pp. 24–25, Figure 9 · p. 24, Figure 10 · p. 24
4.5. Focus on dispersion evolution
- Performance dispersion increased significantly in all four reviewed studies, making spread and outliers important considerations for reused batteries.Dispersion was calculated as capacity standard deviation divided by mean capacity.Source: §4.5 · pp. 25–26, Figure 11 · p. 25
4.6. Focus on laboratory studies
- Deeply degraded cells often retained familiar degradation-curve shapes, while usage changes altered aging dynamics in ways influenced by prior operating history.Five studies aged cells to 35% state of health; reported curves were mostly linear or square-root-shaped. Other experiments found reduced degradation speed after usage changes, while nonlinear aging was more affected by new operating conditions.Source: §4.6 · pp. 26–27
4.7. Focus on influence of aging on safety
- Aged-battery thermal stability depends mainly on degradation mechanisms, and aged cells dissipate less energy during runaway than new cells.Avoiding abusive conditions that encourage lithium plating is advised; results on SEI growth and abusive overcharging are sometimes contradictory, motivating further research.Source: §4.7 · pp. 28–29
5. Synthesis of review insights
The review finds that retired batteries can retain substantial performance and support long second lives, while dispersion, inconsistent research practices, and limited evaluation methods complicate reuse decisions. It proposes clearer terminology and research directions spanning battery selection, aging management, and broader study coverage.
5.1. Structuring battery research
- Cumulative battery research needs shared terminology, experimental data, and comparable methods for evaluating characterization techniques.The review reports that only 7% of reviewed articles shared experimental data; it also argues that state of health is useful for tracking, but not for defining end-of-life thresholds on its own.Source: §5.1 · p. 30
5.2. Choosing retired batteries
- Retired-battery state of health averaged 79.9% with a median of 84.5%, and dispersion exceeded that of new batteries by several orders of magnitude.These findings suggest that common second-life definitions based on a 20% or 30% capacity threshold do not align with experimental results.Source: §5.2 · p. 30
- Benchmark studies estimated state of health with error below 2% in less than 4 minutes, while cell-position and capacity-resistance correlations were unreliable shortcuts.The review favors health indicators highly correlated with state of health and a more holistic assessment that includes defects and degradation mechanisms.Source: §5.2 · p. 31
5.3. Aging of reused batteries
- 36 articles covering more than 890 cells showed that reused batteries can last several thousand cycles and beyond 80% state of health.The review also reports that second-life batteries do not follow one specific degradation pattern.Source: §5.3 · p. 31
- Experimental conditions primarily explain short versus long second-life spans, while first-life history may also affect degradation patterns.Adapting current rate and state of charge to measured capacity can support longer life; safety risks can be minimized by avoiding aging-accelerating conditions.Source: §5.3 · p. 31
5.4. Directions for future research
- The review distinguishes second-life batteries, changed in use or structure, from deeply degraded batteries beyond the popular 80% state-of-health threshold.A battery can belong to both categories, but not necessarily; the proposed distinction is intended to reduce confusion in research.Source: §5.4 · pp. 31–32
- Future work should assess defects, degradation, and dispersion, develop technology-specific study registries, and examine reuse through logistics, economics, and safety perspectives.The review also calls for broader performance measures, including dispersion and outliers, alongside state-of-health tracking.Source: §5.4 · p. 32
6. Conclusion and outlooks
The conclusion describes a review that clarifies current knowledge about retired batteries and surveys experimental studies of their second-life performance and degradation. The dataset produced for the review is shared under an open license.
6. Conclusion and outlooks
- The study clarifies the current state of knowledge regarding retired batteries.It provides a detailed overview of experimental studies on retired batteries’ second-life performance and degradation.Source: §6 · p. 33
- The review focuses on experimental studies of retired batteries’ performance and degradation in a second life.Source: §6 · p. 33
- The dataset produced to conduct the review is shared under an open license.Source: §6 · p. 33
Appendix A. Equations
The appendix describes fitting a generalized extreme value distribution to experimental data and gives parameter definitions. It identifies k as the shape parameter and µ as the location parameter.
Appendix A. Equations
- The MATLAB function fitdist was used to fit a generalized extreme value distribution to the experimental data.The distribution’s probability density function is based on k, β, and µ.Source: Appendix A · p. 34
- k is the shape parameter of the distribution.Source: Appendix A · p. 34
- µ is the location parameter of the distribution.Source: Appendix A · p. 34
Appendix B. Experimental conditions for short lifespan studies
The appendix table describes experimental conditions in aging campaigns associated with short lifespans. It indicates how cell counts are reported and marks known aging-accelerating conditions in bold.
Appendix B. Experimental conditions for short lifespan studies
- The table covers aging campaigns that led to short lifespans, defined as fewer than 500 cycles.Source: Appendix B · p. 34
- The reference column reports the number of short-lifespan cells over the total tested, in the form X/Y.Source: Appendix B · p. 34
- Bold experimental conditions mark factors known to accelerate aging.Source: Appendix B · p. 34
Appendix C. Experimental conditions for long lifespan studies
The table summarizes experimental conditions from aging campaigns that achieved long lifespans, defined here as more than 1,500 cycles.
Appendix C. Experimental conditions for long lifespan studies
- More than 1,500 cycles defines the long-lifespan campaigns covered by this table.The table presents their experimental conditions.Source: Appendix C · p. 37
- The reference column gives the number of long-lived cells over the total cells tested as X/Y.Source: Appendix C · p. 37
- Bold experimental conditions identify factors known to accelerate aging.Source: Appendix C · p. 37
Appendix D. Effect of accelerating factors
The appendix reports no direct correlation between accelerating-factor count and reused-battery lifespan, but indicates faster degradation with more factors and knees mostly at three or more.
Appendix D. Effect of accelerating factors
- Table D.14 presents the study's definitions of accelerating factors.Source: Appendix D · p. 40, Table D.14 · p. 40
- No direct correlation links accelerating-factor count to reused-battery lifespan, though zero or one factor suggests linear degradation.Source: Figure D.15 · p. 40, Appendix D · p. 40
- Faster degradation occurs with more accelerating factors, and aging knees occur mostly with three or more used simultaneously.Source: Figure D.15 · p. 40, Appendix D · p. 41
- Testing should limit accelerating factors and update current rate and state-of-charge range using measured capacity.Nominal values may lead to faster degradation.Source: Appendix D · p. 41
- Nominal-capacity settings can make a 1 C test rate 1.25 C for a retired 8 Ah battery rated at 10 Ah.The passage also notes that aging-related resistance increases can shift the perceived state-of-charge range, potentially accelerating aging and producing knees.Source: Appendix D · p. 41