Which Used EVs and Plug-In Hybrids Were Told to Charge Only to 80%? A Census of Every US Battery Recall With a Charge-Limit Remedy

Key facts

• Census size: 48 owner-facing charge-restriction EV/PHEV recall campaigns in NHTSA's data, model years spanning 2012-2026; reports received 2012-05-30 to 2026-09-30.
• The news peg: NHTSA campaign 26V630000 (Volkswagen #93EX), MY2023-2026 Volkswagen ID.4, 22,524 units; defect is that the high-voltage battery may fail and lead to a vehicle fire; interim remedy is charge to a maximum of 80% until repair; final remedy was "under development" at the data pull; interim owner letters expected 2026-11-27.
• The de-facto standard ceiling is 80%: 18 of 28 explicit "charge-to-X%" campaigns use 80%. Chevrolet's five Bolt campaigns use 90%, as does one Jaguar I-Pace campaign (26V067000); two Volvo EX30 campaigns use 70%; two transit-fleet campaigns use 75%.
• It is not a make-specific quirk: charge-restriction campaigns span 18 brands. The three joint-largest brands (Audi, Chevrolet and Ford, five each) hold 15 of the 48 (31%).
• Half also say park outside: 26 of 48 charge-restriction campaigns (54%) also carry a park-outside instruction; only one (2%) adds a do-not-drive order.

By the numbers

• 15,303 distinct post-2010 recall campaigns in the source file.
• 318 of them are EV/PHEV campaigns (gated on a traction-battery / propulsion / rechargeable-energy component).
• 50 keyword-match a charging restriction; 48 are owner-facing after excluding two automatic-only remedies (a Tesla Model S onboard-charger current reduction and a Kia Soul EV software auto-limit).
• 28 of the 48 state an explicit charge-to-X% ceiling.
• Largest charge-restriction campaign by units: Jeep Grand Cherokee/Wrangler 4xe, 25V741000, 320,068 units (do-not-charge). Largest explicit ceiling campaign: Chevrolet Bolt, 21V560000, 57,414 units (90%).
• Earliest in the census: Fisker Karma, 12V241000 (2012-05-30, "instructed not to charge or drive," 19 units).

Based on 15,303 distinct post-2010 NHTSA recall campaigns in FLAT_RCL_POST_2010.txt (zip Last-Modified 2026-10-12 07:02 GMT; max report-received date 2026-10-06), retrieved 2026-10-12 07:22 GMT. Per-campaign verdicts are frozen in the published CSV linked in Sources.

Which used EVs and plug-in hybrids have been told to charge only to 80%?

A charging restriction is an interim safety measure some EV and PHEV recalls impose while the real fix is prepared: cap the charge level, stop charging entirely, or stop fast charging. The table below is every owner-facing charge-restriction campaign in NHTSA's post-2010 file, deduped to the campaign level. "Remedy" is the interim instruction, not the final repair.

The Kia EV6/EV9 and Hyundai Ioniq 5 entries at the bottom of the list are the same campaigns covered in our used electric car battery recall check before buying explainer; another recent EV battery action, the battery-ECU recall covered in our used Toyota bZ4X problems to check before buying piece, is a software fix with no charge instruction and is correctly absent here.

Is the "charge to 80%" rule an EV-wide pattern or concentrated in a few makes?

It is a broad cross-industry interim playbook, not a make-specific quirk. Charge-restriction campaigns span 18 brands, and the three joint-largest brands by campaign count, Audi, Chevrolet and Ford with five each, hold 15 of 48 campaigns (31%). That debunks the common assumption that this is a GM-Bolt-plus-Hyundai story.

Concentration appears only when you group by corporate parent and look at the sharper "charge-to-X%" ceiling subset of 28 campaigns. There, the Volkswagen Group (Audi, Porsche and VW) accounts for 10 of 28 ceiling campaigns (36%), and VW Group plus the Chevrolet Bolt together reach 15 of 28 (54%). Across all 48 charge-restriction campaigns, VW Group (12) and Stellantis (7) together hold 40%.

Two famous EV fire recalls are not in this census, correctly. The Hyundai Kona Electric battery-fire recalls of 2020-2021 used a park-outside remedy, not a charge limit. The original 2019 Audi e-tron fire recall was a coolant-seal fix with no charging instruction.

How to check if a used EV has a battery charging-limit recall

Start with the VIN and the free federal tool. Enter the 17-character VIN at NHTSA.gov/recalls (or the recalls.gov lookup); it returns open recalls on that exact vehicle, including battery campaigns such as 26V630000 for an affected ID.4. The charge-limit instruction itself is not in the one-line recall summary, so open the campaign's Part 573 Safety Recall Report to read the interim remedy and any ceiling, fast-charging, or park-outside guidance.

A VIN recall report, such as a Zilocar VIN check, screens the same recall-presence data NHTSA uses and pairs it with the car's accident and airbag-deployment records, salvage-auction history, odometer check, ownership and listing-price history in one place, which is useful for sizing up a used EV beyond the recall. Whichever tool you use, the last step is the same: take the campaign number to a franchised dealer and confirm whether the final battery repair was completed on that VIN. Neither the free NHTSA tool nor any VIN report shows remedy completion.

What a VIN check can and can't confirm here

A VIN check tells you a recall is present. It does not prove it was remedied. This is the same blind spot we documented for parts not visible in a recall screen in our note on how to check if a used car has counterfeit parts before buying.

Is it safe to buy a used VW ID.4 under the 2026 fire recall (26V630)?

Campaign 26V630000 covers 22,524 MY2023-2026 Volkswagen ID.4s whose high-voltage battery may fail and cause a fire; the interim remedy is to charge only to a maximum of 80%. At the time of this data pull, VW's final remedy was "currently under development," with interim owner letters expected 2026-11-27. By definition, no affected used ID.4 can show a completed battery fix yet, so every copy in this recall is under the 80% charge limit until VW ships the repair.

This is distinct from campaign 26V030000, the separately reported ~44,000-unit (43,881) MY2023-2025 ID.4 recall for battery overheating, whose remedy is a software update plus battery replacement and which carries no charge-limit instruction. Do not conflate the two. 26V630000 is also at least the third ID.4 charge-limit campaign, after the smaller 25V836000 (629 units) and 26V028000 (670 units); it is the largest ID.4 charge-limit action but not the first.

How a "charge to 80%" recall differs from a "park outside" recall

A charge limit tells you how to use the car; a park-outside order tells you where to keep it. They often travel together: 26 of the 48 charge-restriction campaigns (54%) also tell owners to park away from structures, because a battery that can ignite is a risk both while charging and while parked. A do-not-drive order is far rarer here, appearing in just one campaign (2%), the 2022-2023 BMW i4/iX. For the full park-outside picture, see our census of how to check if a used car has a park outside fire recall; for the strictest tier, see how to check if a used car has a do not drive recall.

A used EV or plug-in hybrid can carry both at once. Confirm which instructions apply by reading the specific campaign's remedy text, then verify repair status with a dealer.

Shopping a used EV or PHEV? A Zilocar VIN check surfaces open recall presence alongside the car's accident and airbag-deployment records, salvage and junk-auction history, odometer/rollback check, ownership trail, and past and current listing prices and days-on-market. It screens for recalls like the free NHTSA tool does; it does not confirm a recall was remedied, show firmware status, or track NHTSA investigations, so confirm the final battery fix with a franchised dealer.

Related checks

• How to check if a used car has a park outside fire recall
• Used electric car battery recall check before buying
• How to check if a used car has a power loss recall

Method

Source. NHTSA ODI recall flat file FLAT_RCL_POST_2010.txt, taken from FLAT_RCL_POST_2010.zip (zip Last-Modified 2026-10-12 07:04 GMT; newest Part 573 report-received date inside the file 2026-10-06), retrieved 2026-10-12 07:50 GMT. The archive's CRC was verified before parsing. The file is tab-delimited, 29 fields, no header row. This census reads CAMPNO (field 2), MAKETXT (3), MODELTXT (4), YEARTXT (5), COMPNAME (7), RCLTYPECD (11), POTAFF (12), ODATE (13), RCDATE (16), DESC_DEFECT (20), CONEQUENCE_DEFECT (21), CORRECTIVE_ACTION (22) and NOTES (23), per NHTSA's published RCL.txt field layout.

Universe, N, and the filter chain. 245,855 component rows resolve to 15,303 distinct campaigns on CAMPNO, of which 13,673 are vehicle campaigns (RCLTYPECD = V; the rest are equipment, tyre and child-restraint filings). Gating those to campaigns whose COMPNAME identifies a traction battery, propulsion system, high-voltage system or rechargeable-energy system leaves 318 EV and plug-in-hybrid campaigns. Within those, a charge-restriction keyword classifier (charge-to-N%, limit or maximum charge, reduce or limit state of charge, do-not-charge-until-repair, no DC-fast or Level-3 charging) matched 50 campaigns, after the false-positive phrase "free of charge" was stripped from the text. Each of the 50 was then read and adjudicated for whether the restriction is actually owner-facing: two were excluded as automatic-only, a Tesla Model S onboard-charger current reduction (14V006000) and a Kia Soul EV software auto-limit (23V218000), leaving the 48 campaigns in the table. Everything is deduplicated to campaign level, so a campaign spanning several makes, models or model years counts once.

The classifier is the load-bearing judgement, and it was tested from both sides. Where the keyword search is pointed changes the answer, so the census was recomputed two further ways and both results are published here rather than the flattering one. Pointed at the remedy text only (CORRECTIVE_ACTION plus NOTES), an independently written automated pass returns 44 campaigns and recovers 42 of the 48 above; the only two campaigns it adds are precisely the two automatic-only remedies this census deliberately excludes, which is a check on the exclusion rather than a contradiction of it. Pointed at all four text fields pooled, the same pass returns 49, and the extra matches are false positives: a defect summary saying a vehicle "may stop charging" describes the fault, not an instruction to the owner. The six campaigns the strict pass misses carry their owner instruction outside CORRECTIVE_ACTION — 25V655000's "no Level-3 quick charging" is the clearest case. So 48 is an adjudicated count, not the output of one reproducible filter, and an independent implementation should expect 44 to 49 depending on which fields it reads. The frozen CSV carries both automated verdicts for every campaign, so the adjudication can be checked row by row.

Corroboration. The news-peg campaign reproduces exactly against a second, independent path: api.nhtsa.gov and the Part 573 report both give 26V630000 as 22,524 MY2023-2026 Volkswagen ID.4 units with a charge-to-80% interim remedy, matching the flat file to the unit. 26V030000, the earlier and separate ID.4 battery-overheat recall, independently reproduces at 43,881 units, confirming the two are distinct campaigns rather than one restated twice. All 48 published campaigns were re-joined to the flat file and all 48 resolved, so no row in the table is an artefact of the write-up.

What this cannot show. Unit counts are POTAFF, the manufacturer's potentially affected population, not confirmed defective cars, and they are not additive: supersessions (20V701000 to 21V560000, 24V720000 to 25V741000) and overlapping fleets mean the same vehicle appears under more than one campaign, so no grand total of affected vehicles is asserted anywhere on this page. The census counts campaigns, not vehicles and not owners. A charge ceiling is an interim instruction and several of these campaigns have since received a final remedy, so presence in this table does not mean a car is still restricted today. Most importantly, this file records what manufacturers filed, not what owners did: it cannot show compliance, and it cannot show whether any individual car has had its battery repair completed. Only a franchised dealer can confirm remedy completion by VIN.

Sources

• NHTSA Part 573 Safety Recall Report 26V630 (VW ID.4, charge-to-80% interim remedy) — regulator
• NHTSA Recalls API, recallsByVehicle (corroborates 26V630000: 22,524 units, MY2023-2026, charge-to-80% remedy, VW #93EX) — regulator
• NHTSA ODI recall flat file FLAT_RCL_POST_2010.zip (the primary dataset for this census; Last-Modified 2026-10-12, max report-received date 2026-10-06) — regulator
• NHTSA Datasets and APIs index — regulator
• Frozen per-campaign classification CSV (50 keyword-matched rows, owner-facing verdict, matched keyword classes, units, retrieval provenance) — Zilocar vehicle-data desk
• Scripps News Life / WTKR, "NHTSA says faulty battery modules in VW ID.4s could pose fire risk" (news-peg corroboration) — major news

Per-campaign verdicts, including both automated classifier results, are frozen in the CSV linked above. Full method is in the Method section.

Last verified: 2026-10-12