According to what may be the most comprehensive real-world electric vehicle battery capacity study to date, the Tesla Roadster’s battery has been projected to retain an average 80-85 percent of its charge holding capacity after 100,000 miles.

This data on the 2008-2012 electric car model that’s been in production the longest in this current era of EVs was presented by Plug In America on Saturday at the Teslive event in Milpitas, Calif., and it beats expectations originally set by Tesla.

Projections Tesla made in 2006 were that the car would retain just 70 percent after 5 years of driving 10,000 miles per year, or an average of 50,000 miles.

While the average was shown to be positive, Plug In America’s projections also indicate a considerable degree of variability between vehicles and cautioned against surmising that every Roadster will meet the higher benchmark.

The results were derived from responses by verified Roadster owners from around the world. Another way of parsing the “80-85 percent” per 100,000 mile number is to say the Roadster lost about 3.7 ideal miles of range per 10,000 miles driven.

“Our study also found no discernible effect of climate on battery-pack longevity,” said Tom Saxton, chief science officer for Plug In America, who led the research. “Roadster owners in hot climates are not seeing noticeably different battery capacity profiles than owners in moderate climates.”

Qualified Results

Information from Tesla indicates 2,500 Roadsters in all were built, stats are provided by Tesla for more than 2,100 sold in 31 countries – with a good deal more than this believed to have actually been sold – and production ended January 2012.

The study compared two sets of data – one being 4-5 percent of all Roadster drivers accounting for around 10 percent of all miles driven – 122 actual owners – who responded to a several-minute voluntary survey. The other data set was from 106 Open Vehicle Monitoring System (OVMS) installed by Tesla owners. The latter is an anonymous data collection system. The two sets of data are not from the same people necessarily.

“To protect owner privacy, the OVMS data was contributed anonymously, so I can’t say exactly how much it overlaps the survey data,” said Saxton, “but I expect there’s some overlap. Still, the OVMS data is less self-selected so it adds to the quality of the distribution of the data for the study.”

The two sets of data for the most part agree suggesting reasonable expectation of accuracy and results are not far off of a smaller 2011 survey Saxton did in the Pacific Northwest.

What’s more, the study went in with eyes wide open, and stated since the 122 respondents were voluntary, they may have been biased either in favor or against the Roadster or Tesla, but that now seems less likely.

The respondents were reached online via social media avenues, and not via a comprehensive mailing to every single Roadster owner across the globe. Therefore, the more connected or socially engaged owners were believed more likely to have noticed the survey, and taken the time to answer it.

“I met a number of owners at Teslive who didn’t know about it,” said Saxton when asked about the quality of the data sampling, “and they are presumably far more socially active than the average Tesla owner.”


This said, the 4 percent is a relatively large sample size in qualified terms, and far larger, say, than some nationwide Gallup polls which may only survey a small fraction of 1 percent and deem that statistically significant to draw conclusions from.

The second-most comprehensive study Saxton is aware of is one also done by Plug In America on the Nissan Leaf. There just 1 percent of owners (240 people out of a far larger production volume) answered.

Between the two cars, Tesla owners indicated more favorable feedback in the hot climate areas, but here too, the survey allows for adjustment in time, assuming more data came forth.

“The data does thin out about 50,000 miles driven, so it’s possible a pattern may become visible as we collect more data, but with the current data set no climate pattern is visible,” said Saxton.

Other significant conclusions drawn by the Tesla survey include:

As there is considerable variation among vehicles with similar mileage, an individual
owner’s experience may vary significantly from the average.

The survey found no significant correlation between climate and battery pack longevity.

Individual experience may vary. The survey data for high-mileage vehicles is sparse with little variation in climate among those vehicles, so it’s possible an effect from climate will emerge as more data is collected.

The survey found no significant correlation between vehicle age and battery pack longevity, although the study has no data on the first year of use, nor use beyond 4.5 years.

The calculated amp-hour capacity is the most reliable measure of battery pack capacity. It would be a benefit if this value were readily visible to Roadster owners.

Lastly The survey’s written conclusions observe:

It’s curious that Tesla does not offer any sort of warranty on battery pack capacity, not previously as part of a new Roadster purchase, not as part of the extended warranty they are now offering Roadster owners as their warranties expire, and not even to Model S owners despite the purported improvement in battery chemistry and corresponding increase in both time and miles on the Model S battery warranty.

Plug In America was partly responsible in convincing Nissan to offer such a warranty after relatively small numbers of Leaf owners reported excessive capacity loss, particularly in hot first-wave roll-out states, including Arizona, Texas, and California.


We asked Saxton therefore to clarify his thoughts on Tesla’s “curious” lack of battery capacity warranty.

“Having a warranty that covers capacity loss, like the Nissan Leaf does, gives owners more assurance about the long term utility of their vehicle. That seems like a selling point to me, although it hasn’t been a big issue with current Roadster owners,” he said.

We then followed up with Tesla but were unable to receive a reply before deadline. If we do later, we can update this.

The full survey can be found here.