In 2026, more than 120 GWh of EV batteries worldwide reach end-of-vehicle life, and over half of those packs retain 70-80% of original capacity, creating a second-life supply pool that converges with stationary storage rack demand at roughly $116 per kWh [S3][S4].
Four independently sourced 2026 estimates frame the same market: Roots Analysis values it at USD 1.70 billion in 2026 [S1], GMI puts the broader reuse market at USD 10.96 billion [S5], Fortune Business Insights projects USD 1,267.8 million growing to USD 7,611.3 million by 2034 [S2], and MarketsandMarkets tracks USD 1,200 million in 2024 to USD 4,700 million by 2030 at 25.5% CAGR [S8]. The wide spread reflects scope (lithium-ion only vs. all chemistries) rather than disagreement on direction.
Feedstock: 70-80% SoH packs and the State of Health threshold
Retired traction packs leave the vehicle once usable capacity drops to 70-80% of nameplate, leaving 8-10 years of vehicle service behind them and 1-15 years of viable second life ahead, depending on chemistry and duty cycle [S1]. GMI puts 2025 deployed second-life volume at approximately 11,150 MWh, confirming the move past proof-of-concept into aggregated operational capacity [S5]. Spiers New Technologies led the 2025 reuse market with over 8.5% share, followed by 4R Energy, Connected Energy, B2U Storage, and Renault Group, with the top five holding 33% collectively [S5].
Capacity is scaling fast: roughly 25-30 GWh in 2025 expands toward 330-350 GWh by 2030, a CAGR near 65%, per the 2026 investment-recovery playbook [S4]. McKinsey's earlier estimate that second-life supply for stationary applications could exceed 200 GWh per year by 2030 [S6] now sits inside that wider 330-350 GWh envelope, which also captures telecom, UPS, and forklift flows beyond EV traction packs.
Chemistry split: LFP vs NMC for stationary duty
LFP and NMC dominate the feedstock, with NMC packs carrying higher residual energy density but shorter cycle life once redeployed; LFP packs tolerate deeper DoD and higher cycle counts, which is why most grid-tied second-life projects spec LFP first [S1][S5]. Hydrometallurgical recycling of retired cells now recovers 95% of lithium and cobalt and 97% of nickel, so any pack that fails repurposing grading still carries a measurable material-recovery value rather than hazardous-waste cost [S4].
For stationary storage handling integration, the practical comparison looks like this:
- LFP second-life: typical residual 80-90% nameplate capacity, 3,000-5,000 additional cycles at 80% DoD, lower energy density (~120-160 Wh/kg cell-level), best fit for 2-8 hour duration storage cage builds and behind-the-meter industrial storage rack arrays.
- NMC second-life: typical residual 70-80% nameplate capacity, 1,000-2,000 additional cycles at 80% DoD, higher energy density (~180-240 Wh/kg), best fit for short-duration frequency regulation and commercial UPS strings where footprint matters more than cycle count.
- NiMH and lead-acid second-life: covered in the Roots scope [S1] but rarely economic at 2026 lithium prices; relegated to niche telecom and off-grid replacement duty.
Unit economics: $116/kWh and where the margin lives
The 2026 benchmark cost of repurposed kWh sits at roughly $116/kWh, against new LFP stationary cells landing in the $130-150/kWh range, giving a 15-25% capex discount before integration [S3][S4]. That spread is wide enough to justify second-life procurement for behind-the-meter commercial and industrial users with 4-8 hour duration needs, but it is not wide enough to survive high integration cost, which is why the refurbishment step (testing, grading, re-encoding the BMS, re-warranting) defines whether a project is bankable [S7].
Connected Energy, Nissan, Fortum, Renault, RWE, and Enel X concentrate second-life procurement in Europe, where European Battery Regulation pulls compliance-grade packs out of vehicles earlier than in markets without disposal mandates [S1]. Redwood Materials launched a dedicated second-life business unit in June 2025 and signed a strategic partnership with Crusoe to build microgrids pairing second-life EV batteries with large-scale solar, a useful template for the hybrid solar-plus-storage buildout now repeating across the U.S. Southwest [S1].
Standards, regulation, and the EU Battery Passport clock
Three regulatory threads shape 2026 procurement. First, the EU Battery Passport becomes effective on February 18, 2027, requiring digital chain-of-custody for industrial and EV batteries above 2 kWh; any second-life pack redeployed after that date without passport data is unsellable into the EU stationary market [S4]. Second, the U.S. clean-vehicle critical-minerals threshold sits at 70% in 2026 with FEOC restrictions live, so domestically sourced second-life packs avoid FEOC surcharge that imported packs carry [S4]. Third, IEEE 1547 and UL 9540 govern the inverter and BESS safety certification that any second-life rack must clear before grid interconnection, independent of cell provenance.
For procurement teams, this means the data backbone (SoH history, cycle count, chemistry, BMS firmware version) is no longer a nice-to-have: it is the price of admission to the EU market in 2027 and to U.S. utility RFPs from 2026 forward [S4][S7].
Failure modes and where second-life loses
Second-life is not a universal substitute. Three failure modes bite in 2026. First, residual-capacity variance: packs arrive at the refurbisher in a 70-90% SoH spread, and module-level grading to within 2-3% is required to build a balanced rack; without grading, cell divergence kills the string within 200-400 cycles. Second, BMS obsolescence: older packs ship with firmware and connectors that do not talk to modern inverters, forcing a controller swap that adds $15-30/kWh to integration cost. Third, cycle-life warranty gap: second-life vendors typically warrant 5-7 years or 2,000-3,000 cycles, versus 10-15 years on new LFP stationary cells, which excludes second-life from long-duration utility procurement where 15-20 year asset life is baseline [S4][S5].
For applications outside those constraints (commercial peak shaving, EV charging buffer, microgrid with solar, behind-the-meter demand-charge reduction, and 2-6 hour duration industrial UPS), second-life is competitive on capex and increasingly competitive on LCOS once integration is done well. For long-duration utility storage, front-of-meter frequency regulation with tight SoX control, or any project that needs a 15-year single-vendor warranty, new LFP still wins.
Specifications engineers should pin on a 2026 second-life datasheet
For the inverter and switchgear side, projects still follow the same industrial valve and pressure transmitter discipline used in any BESS: thermal-runaway gas detection, smoke sensors per NFPA 855 spacing, and HVAC sized to keep cells in the 15-35 C operating window where second-life LFP holds its 3,000-5,000 cycle promise [S4].
Trackable signals for the next 6-12 months: EU Battery Passport go-live on February 18, 2027, with first enforcement actions expected in Q3 2027; second-life pack pricing relative to new LFP stationary cells, where any spread compression below 10% kills the refurbishment margin; and the 2026 wave of Fortune 500 fleet retirements, which is the first hard test of whether the 25-30 GWh 2025 baseline actually scales toward the 330-350 GWh 2030 projection.