The used EV market can’t thrive without accurate battery-health data

Figuring out what a used electric vehicle is worth means figuring out the health of its most expensive component — its battery. But today, the battery inside a used EV is something of a black box.

Manufacturing defects, driving behaviors, charging patterns and weather conditions can all significantly impact the health of an EV battery over time. 

But without some way to tell the healthy batteries from the degraded ones, buyers are forced to presume the worst, a factor that drives down prices across the board. That could undermine the growth of a used EV market that will be vital for expanding access to include people who can’t afford the upfront cost of new electric cars.

It’s also bad for the rental car companies, commercial vehicle operators, auction houses and various financial parties in the ​“secondary” markets for used vehicles, where knowledge of a vehicle’s residual value plays a major role in fleet-vehicle economics. That’s why a healthy and lucrative market for used EVs depends on being able to determine what’s happening inside those black boxes.

“There’s a real concern here for secondary buyers,” said Scott Case, CEO of Recurrent, a data-science firm specializing in used EVs. ​“Everyone’s had an iPhone long enough to know that the battery wears down over time. They’re bringing that experience to their auto buying.”

To help solve this information problem, Recurrent pulls data from 45 different models and makes of EVs, with ​“more than 10,000 cars connected and sharing their data every day” via onboard data collection technology, he said.

That growing data set — covering about 100 million miles of driving at present — informs an analytics engine that can take key details, such as the age of the vehicle, odometer readings and the climate where the car has been driven, and churn out an estimate of how much better or worse than average the vehicle’s battery can be expected to be compared to manufacturers’ estimates, he said.

As EVs continue to take a larger and larger share of the overall vehicle market — and as newly instituted federal used EV tax credits and other incentives drive more buyers into the used-EV market — ​“this is information that has to be communicated because it is such an important part of the value proposition,” he said.

Getting used EV battery health data into used EV markets 

Recurrent is one of many companies taking on the challenge of diagnosing EV battery health. Right now, the company sells its reports to auto dealers, makes data available to the AAA Washington drivers’ club, and has also inked a partnership with Black Book, one of the country’s major compilers and publishers of used-vehicle data.

Cox Automotive, owner of automotive businesses including the Kelley Blue Book repository of used-vehicle values and wholesale-vehicle market operator Manheim Auction, is also investing in battery-health diagnostics. Last year, it bought Spiers New Technologies, a provider of ​“EV battery lifecycle management” services including repair, reuse and recycling of end-of-life batteries, as well as the battery analytics used to assess battery health.

Battery packs account for between 30 and 40 percent of an EV’s cost, Cox notes, making them a ​“high-value asset that needs inspection, valuation and servicing.” But the company’s research shows that eight out of 10 people who say they aren’t considering a used EV for their next car are skeptical about battery value and remaining useful life.

Edmunds, another well-known automotive analysis and advisory firm that’s putting a lot of work into assessing EVs, noted in a recently published used EV buyer’s guide that the ​“possibility of diminished battery performance is a major con for used EVs.” This uncertainty can outweigh advantages such as lower maintenance and operating costs of EVs and the fact that they tend to be in better condition than used internal combustion engine vehicles because they tend to be driven less and have fewer parts that are subject to degradation over time, not to mention the availability of federal and state tax credits and incentives.

Just how well can data ease this uncertainty? That depends on how much data is available and how relevant it is to a specific EV, Case said. Recurrent’s approach is ​“really good on common EVs in common locations,” such as a Tesla 3 that’s spent its whole life in the San Francisco Bay Area, where ​“we’re probably within plus or minus five miles of range,” he said. The data yields less certainty for less common models of EVs in regions with fewer EVs.

But variables like EV make and model year, mileage ratings and climate conditions don’t tell the whole story. Charging behaviors can be another major factor impacting battery health. Just like cellphone or laptop batteries, EV batteries last longer when they’re kept within a certain range of charge, between roughly 30 percent on the low end and 80 percent on the high end.

Running them ​“down to zero and [charging] back to 100 percent” can lead to faster degradation, Case said. So can frequently subjecting them to the higher voltages of fast chargers, as opposed to charging them more slowly at the lower voltages typical of home chargers. But that data can be harder to get unless EV drivers agree to allow it to be tracked.

Recurrent does offer access to this more granular data to the EV drivers who subscribe their cars to its data-collection fleet, he noted. ​“For those, we can really eliminate the uncertainty.”

Can battery data prevent used EVs from becoming a market for lemons? 

Lennart Hinrichs, vice president of marketing and strategic partnerships for Twaice, a Munich, Germany–based battery analytics startup, agreed that collecting more data from individual vehicles results in more accurate and valuable findings.

“Statistics-only” data can be helpful, he said, but it ​“comes with a number of drawbacks.” Statistics can’t capture which batteries may have manufacturing flaws, for example, or pinpoint those that are performing poorly due to an improperly calibrated interaction with the rest of a vehicle’s systems.

More accurate findings can emerge from an EV’s onboard diagnostics and those embedded in batteries themselves, he said. Twaice is working with German testing laboratory TÜV Rheinland to conduct a ​“quick test” using both sources of data to derive a ​“specific and precise state of health for that battery,” he said.

Even more precise assessments of battery health, as well as of battery safety, can be done by pulling the real-time data that modern EVs collect and sending to the cloud, he noted. Twaice works with German automakers to predict the health of EV batteries to help them structure warranties and plan ahead for how many batteries they’ll need to manufacture to replace those going out of service years into the future, for example.

More accurate and verifiable data for used EVs is important for automakers and fleet operators to avoid creating a ​“market for ​‘lemons,’” Hinrichs said. That term, coined by 2001 Nobel Prize–winning economist George A. Akerlof in a 1970 paper, describes markets in which expectations of poor quality and a lack of information that could prove otherwise drive down prices and undermine the incentive for improving the products being traded, as with defective, or ​“lemon,” used cars.

In a recently published report with TÜV Rheinland and Autovista Group, Twaice analyzed how verifiable data on battery health could add about 2.4 percent additional value, or about €450 ($468) per vehicle, to the lower bound of values for used EVs that otherwise lack data. That’s €4.5 million for every 10,000 used EV transactions, to put it in terms that apply at scales that matter to the automakers that care about residual vehicle values for leasing structures, warranties and other issues.

Automakers, owners and leasers of vehicle fleets, used car resellers and other auto industry players ​“are looking into finding greater transparency into their value,” he said. ​“The beauty of that is, from a fleet-management perspective, you get risk reduction — technical and safety risks, but also financial risk and warranties, which are all important” for automakers, he said.

Automakers don’t necessarily make this kind of in-depth battery-health analysis easy for consumers to access, however, Case said. That’s particularly true of earlier models of EVs, such as the Nissan Leaf, that experienced their own round of negative publicity for batteries that wore down more quickly than the automakers had predicted.

Early Leaf owners came up with their own ​“hacks” for this, such as the Leaf Spy smartphone application that pulled data from the vehicle’s on-board diagnostics port, he said. ​“But you have to be a Ph.D. chemical engineer to understand what the thing is telling you.”

There are sound reasons for being careful with sharing battery data that’s still subject to quite a bit of uncertainty, he noted. At the same time, automakers and the broader automotive industry will need to find ways to share as much data as possible in an easily understandable way to provide the foundation for a healthy and growing used EV market.

In fact, California is already proposing rules for how battery health must be tracked and made available for used EVs, both for battery recycling purposes and for informing used EV markets, he said. ​“Eventually, every part of the auto ecosystem is going to change based on the transition from combustion engines to EVs.”

Recurrent’s work with AAA is aimed at this type of consumer education, he noted. Beyond offering EV owners the company’s data collection and analysis, Recurrent also provides personalized tips for drivers on how to keep their battery healthy and ​“eventually help them sell that vehicle at a premium,” he said. After all, ​“no one grew up learning from their parents or grandparents how to take care of an EV.”