An EV battery is generally considered end-of-life for automotive use at around 70 to 80% of original capacity. At that point it can no longer deliver the range and power the vehicle was sold on. It is also, by any reasonable standard, still a substantial battery.
The mismatch matters because stationary storage has entirely different requirements. A pack sitting in a charging hub does not need to be light, does not need automotive power density, and cycles on a far gentler profile. Capacity that fails a vehicle specification can be perfectly adequate where weight and peak discharge are not constraints.
Why this matters more here than elsewhere
Storage is frequently the component that decides whether a distributed energy project is viable. In solar-plus-storage charging infrastructure, batteries can dominate capital cost. Reduce that line item meaningfully and marginal projects become buildable.
This is why second-life storage appears in the architecture of the charging work I have designed, both in solar-microgrid hub concepts and in distributed fast-charging for electric motorcycles. It is not primarily an environmental position. It is the intervention that moves the cost structure.
Second-life storage is not a discount on new batteries. It is a different asset with a different risk profile, and it has to be engineered as one.
Grading is the entire discipline
The failure mode in second-life projects is treating retired packs as interchangeable. They are not. Two packs with identical nameplate capacity and identical age can have very different remaining useful life depending on thermal history, charge behaviour and cell-level degradation spread.
Serious deployment therefore requires characterisation before installation: capacity testing, internal resistance measurement, and assessment of cell balance within the pack. Modules that appear healthy in aggregate can hide a weak cell group that will define the pack's real behaviour. Mixing poorly matched modules in one string means the worst performer governs the system.
- Test and grade every pack before it enters a system, without exception
- Group modules by measured state of health, not by nameplate or age
- Specify a battery management system that reports at module granularity
- Design thermal management for degraded cells, which tolerate less
- Plan replacement access from the start; these packs will be swapped sooner
- Define the end-of-second-life route before deployment, not after
Safety deserves more caution than new cells
A degraded lithium pack is less tolerant than a new one. Reduced thermal headroom, uneven internal resistance and unknown mechanical history all narrow the safe operating envelope. Second-life installations should be more conservative than equivalent new-battery systems, not less: better ventilation, tighter monitoring thresholds, physical separation between strings, and clear isolation for maintenance.
This is also where regulatory clarity is thinnest. Standards for repurposed batteries are less developed than for new equipment, which places more responsibility on the engineer. Documenting grading methodology and monitoring thresholds is not bureaucratic overhead; it is how the installation is defended when something eventually needs explaining.
Circularity is a system, not a step
Second life only counts as circular if the third step exists. A pack that reaches genuine end-of-life in a stationary system and then has nowhere to go has deferred a waste problem rather than solved it.
That means recovery pathways have to be planned at deployment, and the operator has to remain accountable for the pack across its whole life. Working on household biogas systems, converting a waste stream into an energy input while returning bio-slurry as fertiliser, taught me that circular systems fail at the least glamorous handoff. Batteries are the same. Collection and recovery are logistics problems, and logistics problems are solved by ownership, not intention.
What makes it work
Second-life storage is not free capacity. It is cheaper capacity with tighter engineering requirements, shorter replacement intervals and a harder monitoring burden. Treated that way, it is one of the most useful tools available for making distributed energy infrastructure affordable across the continent.
Treated as a bargain, it produces installations that underperform, fail early, and set back the case for everyone else. The technology is sound. The discipline around it is what decides the outcome.
Working on something in this space? Start a conversation.