EV Battery Lifespan: Real Data From 22,700 Vehicles (2026)

The warranty card says eight years. Geotab's dataset of 22,700 vehicles suggests the real number is closer to eighteen, the lithium-ion type used in most modern EVs can last at least a decade before replacement, and the fleet data pushes the ceiling well past that. It is the single most misunderstood number in EV ownership, and it is the one that decides whether the used-EV market makes sense for you.

For any EV built in 2022 or later, sold new in Canada, and operated under typical Canadian conditions: 12 to 18 years of meaningful capacity. That is a wide band because the variance is real, and a false single number is worse than an honest range.

Quick Answer: What to Actually Expect

Modern EV batteries retain 80 to 90 percent of their original capacity after eight years of normal use. That is what Geotab's telematics data says, 22,700 vehicles, the largest independent fleet dataset published to date.

Geotab, analysing 22,700 electric vehicles across its telematics fleet, found average retention of 81.6 percent at the eight-year mark, and only 0.3 percent of batteries in vehicles built after 2022 have required replacement. A separate Nature Energy analysis using Great Britain vehicle testing data estimated that the lithium-ion type used in most modern EVs can last at least a decade before replacement, with the vehicles in that dataset averaging roughly 18.4 years of service life overall.

The industry warranty floor across virtually every brand sold in Canada is eight years or 160,000 kilometres, whichever comes first. It is, to be fair, a number, and it is calibrated to be the number the manufacturer is confident it will never have to pay out. The actual engineering target sits well beyond that.

The skeptical read is that fleet telematics data skews toward newer vehicles operated by professional drivers with routine maintenance, and that consumer ownership will look worse. That objection has merit at the margins, but the Nature Energy dataset draws from consumer vehicles in the UK MOT testing regime, not fleets, and it lands in the same neighbourhood, the convergence between two very different data-collection methods is what should move you.

For a deeper picture of how degradation plays out year by year, the year-by-year breakdown of Canadian battery degradation walks through the 5, 10, and 15-year checkpoints with real data.

What Actually Kills a Battery (and What Doesn't)

Three vectors do the damage: heat, charge cycles, and calendar age. The hierarchy matters, because most owners worry about the wrong one.

Heat is the dominant vector. Sustained high temperatures, pack temperature, not ambient, accelerate the chemical reactions that permanently reduce capacity. This is why Arizona fleets show measurably worse degradation than Ontario ones, and why the "battery in a hot parking lot at 100 percent" scenario is the single most damaging thing an owner routinely does without realising it. The engineers who designed modern EVs knew this from the start: they built active thermal management and continuous battery-health monitoring into the pack precisely because they could not trust owners to avoid the hot-parking-lot problem on their own.

Charge cycles matter less than most people think, and chemistry matters more. Li-NMC batteries using lithium nickel manganese cobalt oxides are the most common in EV, and the lithium iron phosphate battery (LFP) is on the rise, the two chemistries age very differently. LFP tolerates 3,000 to 5,000 full cycles before meaningful capacity loss; NMC typically fewer, but with higher energy density as the trade-off.

Cold is the temperature story owners get wrong. Cold weather temporarily reduces range, sometimes dramatically, 20 to 40 percent in a Canadian February, but it does not accelerate permanent degradation the way sustained heat does. A battery that spends its winters at minus twenty and its summers at plus twenty-five will out-live one that spends its whole year at plus thirty-five. Geotab was explicit about a gap in its own data here: the dataset lacked sufficient cold-only samples to isolate the impact of extreme cold climates, which is an honest limitation worth naming rather than papering over.

Calendar age is the quiet third factor. Even a battery that is barely cycled loses capacity slowly to time itself, roughly one to two percent per year at moderate temperatures. This is the number that puts the ceiling on lifespan regardless of how gently you drive. It also explains why low-mileage used EVs from 2017 sometimes benchmark worse than higher-mileage 2020s, the clock runs on a parked car too.

The Charging Habits That Move the Needle

The single most effective thing an owner can do is keep the state-of-charge between roughly 30 and 70 percent for daily use. A 2023 study from Dalhousie University estimated this alone could extend battery life by up to 25 percent compared with routine full-cycle charging. That is not a rounding-error improvement. That is the difference between selling the car with 82 percent capacity or 68.

Daily DC fast charging accelerates degradation measurably. Occasional DC fast charging does not. The distinction matters because the internet has spent a decade conflating "fast charging is bad" with "fast charging on a road trip once a month is bad", the second claim is not supported by the data. The fast-charging degradation data breakdown covers the actual numbers.

Sequence matters more than people expect. Parking a hot battery at 100 percent in a hot climate is worse than fast-charging in mild weather. If you must charge to full, do it right before you leave, not the night before. If you fast-charge in summer, drive for at least twenty minutes afterwards to let the pack shed heat before parking.

Most battery management systems built after 2019 enforce charge limits automatically and moderate fast-charge speeds when the pack is hot. Gen-1 EVs, meaning most vehicles from 2016 or earlier, lacked this protection, which is why Recurrent's dataset shows the highest replacement rates concentrated in that cohort, with a notable secondary uptick in 2019 and 2021 Teslas. If you are buying used, the model year is the single most important variable.

How Canadian Conditions Change the Calculus

Canadian winters get more airtime than they deserve when it comes to permanent battery health. Cold reduces range temporarily, yes, meaningfully, sometimes brutally, but it does not accelerate the permanent capacity loss that sustained heat causes. A pack in Yellowknife will out-live an identical pack in Phoenix.

Preconditioning is where Canadian owners either protect their battery or waste it. The habits that matter come down to a short list:

• Precondition while plugged into a Level 2 charger, the pack warms on grid power and your usable range stays intact.
• Precondition on battery power in a driveway at minus fifteen and you will spend 5 to 8 percent of usable charge to warm a pack for a commute that will spend more.
• Cap daily charging at 70 to 80 percent unless you need the extra range for a specific trip.
• Never leave the car parked at 100 percent in summer heat, the worst-case scenario for calendar-age degradation.

The federal warranty floor sits at eight years or 160,000 kilometres, and every vehicle that qualified under iZEV, the retired predecessor to today's EVAP rebate, carried at least that. Some brands do measurably worse: the Canadian Hyundai EV battery warranty is eight years or 160,000 kilometres, compared with ten years or 100,000 miles in the United States. Same car, different paperwork. The complete Canadian EV warranty comparison has the brand-by-brand breakdown.

For used buyers, the model-year cutoff is the number to internalise. Recurrent's data shows gen-1 EVs, 2016 and earlier, carrying the highest rate of battery replacements due to deterioration or age. Post-2022 vehicles show a 0.3 percent replacement rate. The gap between those two cohorts is enormous, and it is not linear across the years in between. If you are shopping used, 2020 is a reasonable functional floor, and 2022 is a safer one.

When Replacement Makes Sense, and What It Costs

The functional replacement threshold for most drivers is around 70 percent of original capacity. Below that, the range hit and the charging inconvenience start to compound. The warranty trigger is typically 70 percent, sometimes 70, sometimes 66, always in the paperwork, which is the point at which the manufacturer replaces the pack for free.

Replacement costs have fallen and continue to fall, but the range is still wide: $5,000 to $20,000 Canadian depending on pack size, chemistry, and model. LFP packs sit at the lower end; large NMC packs in performance vehicles at the higher end. Chinese-market replacement pricing is meaningfully lower than North American pricing, and that gap is one of several arguments for why the Chinese-EV entry into Canada matters beyond the sticker price. NIO's battery-swap model sidesteps replacement entirely by treating the pack as a serviceable component that gets swapped in three minutes at a dedicated station rather than a fixed one, you are not buying a battery, you are subscribing to one.

Second-life applications extend the economic value of a battery well past the point at which it stops being useful in a vehicle. A pack at 70 percent capacity is still an excellent grid-storage or home-backup asset, and the resale market for retired EV packs has grown into a real industry rather than a rounding error. The battery is not a consumable that gets thrown away at end-of-vehicle-life. It is an asset that continues to produce value in a different job.

For most buyers financing a 2022-or-newer EV over five to seven years, battery replacement before loan payoff is a statistical outlier, well under one percent of vehicles, on current data. The nightmare scenario that dominates comment sections is a real risk for gen-1 used buyers and a vanishing one for anyone shopping new.

Mark the calendar for Geotab's 2028 update on the post-2022 cohort. If the 0.3 percent rate holds through 2028, the battery-range anxiety that shaped the first decade of EV marketing will be a 2010s problem. The Canadian used-EV market on the other side of that data point is the more interesting trade. If the number climbs past two percent, the used-market math needs redoing and this whole calculation shifts with it.