What the evidence shows
A peer-reviewed Nature Energy study published in January 2025 estimated BEV service life at 18.4 years and found later first-use cohorts achieved comparable longevity to combustion vehicles. The MOT data run through 2022; this historical model does not measure battery lifespan.
Older vehicles or current generations?
These sources measure survival, roadworthiness, or breakdowns, not battery life. None provides battery chemistry or model-year splits.
We separate vehicles from before 2023 and 2023 onward. A report’s publication date does not establish a vehicle’s model year or a battery’s manufacturing date.
Vehicle years not establishedGreat Britain MOT survival model
- Vehicles / sample
- 29.8 million GB cars and light-goods vehicles, including 41,600 BEVs; leading BEV makes were Nissan, Tesla, and Renault.
- Vehicle years
- Manufacturer model years are not reported; first-use cohorts span 2005–2017, and mean BEV first-registration year is 2015.1.
- Battery chemistry
- Not captured by MOT data.
- Battery capacity
- Not captured by MOT data.
- Battery manufacture
- Battery production dates not available.
- Observation period
- Nearly 300 million MOT tests, 2005–2022; survival is estimated statistically, not battery lifespan.
Vehicle years not establishedGerman roadside-assistance comparison, 2024 event year
- Vehicles / sample
- German two- to four-year-old BEVs and combustion cars in roadside-assistance records.
- Vehicle years
- Not reported; ADAC gives first-registration years 2020–2022.
- Battery chemistry
- Not reported.
- Battery capacity
- Not reported.
- Battery manufacture
- Pack production dates not reported.
- Observation period
- 2024 attended breakdowns per 1,000 registered vehicles; not battery failures or lifetime data.
Vehicle years not establishedGreat Britain MOT durability analysis, 2024–2026 tests
- Vehicles / sample
- UK class 4 vehicles: cars, taxis, and small vans; nine sibling-model families compared.
- Vehicle years
- Not reported. The '2016–2023' historical-cohort caption is not identified as build/model years.
- Battery chemistry
- Not measured by MOT tests.
- Battery capacity
- Not reported; named early models include Leaf, iOn, and i-MiEV.
- Battery manufacture
- Pack manufacture dates not reported.
- Observation period
- 47.4 million MOT tests, July 2024–July 2026; measures roadworthiness, not battery health.
“Not reported” means the source does not disclose it. Model year, first registration, vehicle assembly and battery manufacture are different dates. Unmatched studies cannot establish how much newer batteries improved.
Which brands do these results describe?
The Great Britain MOT study publishes make-specific BEV survival and mileage estimates, identifies Tesla as the leading BEV make on lifetime mileage, and names Nissan, Tesla and Renault as the largest BEV makes in its sample. This allows make-level study comparisons, not universal brand rankings.
EV-focused brands
Tesla
Second-largest BEV make in the sample (19%); the study names Tesla as the leading BEV make on modeled lifetime mileage.
Established multi-powertrain brands
Nissan · Renault
Largest BEV make in the British MOT sample (49%); the study publishes make-specific estimates.
Third-largest BEV make in the British MOT sample (17%); the study publishes make-specific estimates.
All manufacturers are OEMs. These groups describe brand focus, not a quality ranking or country of origin. EV-focused brands can also sell plug-in hybrids. Results apply to the identified models, batteries and conditions.
Manufacturer and model sources (2)
Includes historical inputs from before 2023. Vehicle cohorts, observation periods and publication dates are listed separately.
This item is clearly marked historical context
The Nature Energy paper was published in January 2025, but its MOT observation data cover 2005–2022. It therefore fits a recent-publication reading list while falling outside a strict 2023–2026 observation window. It should not be placed on a chart labeled recent annual measurements. An article can still use it to explain what long-term survival analysis does and where stronger historical evidence is available. [1][2]
What the researchers found
The study modeled whole-vehicle longevity using nearly 300 million inspection records and estimated average BEV service life at 18.4 years. It found faster improvement across later first-use cohorts (the analyzed cohorts span 2005–2017), but these are not manufacturer model-year labels. This is a modeled population result, not evidence that every contemporary EV has already survived for that duration or that its original battery lasted the whole period. [1][2]
Longevity and roadside reliability differ
A car can suffer a repairable breakdown and remain in service for years. It can also leave one country’s records through export rather than physical failure. ADAC’s April 2025 release offers a different measure using German 2024 roadside incidents: among two-to-four-year-old vehicles, it reports 3.8 breakdowns per 1,000 EVs versus 9.4 for combustion cars. That is useful recent reliability evidence, but it cannot be converted into a lifetime estimate. [3]
A September 2026 BVRLA-commissioned analysis by New AutoMotive reports on Great Britain MOT records, with roadworthiness methods specifying July 2024–July 2026. Among 3–8-year-old Class 4 light vehicles (including cars, taxis and small vans) in the 90,000–120,000-mile band, its MOT test failure rate was 16.5% for BEVs versus 22.1% for petrol; above 120,000 miles, it was 16.0% versus 23.5%. In a separate age-15 check, a table labeled ‘2016–2023 cohorts’ reports that 94.6% of BEVs and 88.1% of petrol vehicles returned for an MOT within 30 months; the report does not identify those labels as model or build years. Non-return can reflect scrappage, export, SORN or missing follow-up; it does not establish why a vehicle left the test record. These commissioned secondary analyses measure MOT roadworthiness and test return, not roadside breakdowns, battery state of health or degradation. [4]
How to use this evidence without overselling it
Keep geography, vehicle age and outcome visible. Do not merge a UK survival model, a German roadside rate and a battery-degradation sample into one universal reliability score. Each measures something different. For an individual purchase, inspect the model’s known repair issues, condition and maintenance history. The study challenges an assumption of inherently short EV lives while leaving uncertainty about present vehicles, future aging and outcomes outside the observed market.
Data period: Historical Nature Energy survival model: MOT observations 2005–2022, published January 2025. ADAC roadside events: 2024. New AutoMotive September 2026 report: roadworthiness methods specify July 2024–July 2026, while a separate age-15 return table is labeled 2016–2023 cohorts without identifying these as model/build years.
What this does—and doesn’t—tell us
- Historical underlying observations are outside the requested recent-data window; no recent-observation chart is included.
- Retirement is inferred from inspection attendance and can reflect scrappage or export; model estimates are not guaranteed lifespans.
- The BVRLA-commissioned report includes Class 4 light vehicles, not a verified passenger-car-only sample. Its methods specify July 2024–July 2026, while other parts describe a 12-month sample; this internal reporting inconsistency limits precision about that test window. Its separate 2016–2023 return-table labels are not established as model/build years and do not identify why vehicles leave MOT records.
The UK survival model uses historical MOT data; a 2026 commissioned MOT analysis adds age- and mileage-matched roadworthiness outcomes. Neither measures battery lifespan.
Found something that needs checking?
Identify the statement, explain your concern, and share original sources for our review.
Keep the conversation curious.
Questions and corrections are welcome. Link your sources, challenge ideas, and be kind to people.
Discussion actions are unavailable until sign-in is configured.
You can still read published comments and their sources.
Loading discussion…