Long-term Tesla Motors Battery Study from Plug in America

One of the things to consider when making the switch to an Electric Vehicle (EV) is the combination of the battery and electricity IS the fuel that is consumed to power an EV. As such, batteries and range degrade and may eventually need to be replaced. Therefore, one of the questions that these new Model 3 reservation holders ask is “how resilient is the Tesla battery?” or phrased another way, “how long will the battery last?”

The Model 3 announcement did not really cover how different or similar the battery technology in the Model 3 will be from predecessor vehicles from them. The Roadster has a different pack than the Model S and Model X. So, how does one get the comfort of knowing that “Tesla knows what they’re doing with batteries.” I suppose we can just trust them.

Fortunately, that is not our only option. Over the past few years, long-time Electric Vehicle advocate, Plug in America Chief Science Officer, and Tesla Motors Roadster owner Tom Saxton has been conducting several long-term battery studies hosted on the Plug in America site.

For those unfamiliar with Plug in America, they’re the folks that formed out of the advocates that tried to stop the “murder” of the GM EV1 and other Electric Vehicles of that era that was documented in the movie “Who Killed the Electric Car?”  Or, as they describe themselves in their webpage:

Our Mission

Plug In America drives change to accelerate the shift to plug-in vehicles powered by clean, affordable, domestic electricity to reduce our nation’s dependence on petroleum, improve air quality and reduce greenhouse gas emissions.

Our History

Plug In America is a coalition of early adopters. We’re the EV trailblazers – RAV4‐EV drivers, former lessees of Honda EV+, GM EV1, Ford Ranger and Ford Th!nk City electric cars – that passionately advocate for energy independence and clean air. Before 2008, we functioned as a loose network of individuals organized around various websites like dontcrush.com and saveEV1.com. We then coalesced into a chapter of the Electric Auto Association. On January 2, 2008, Plug In America became a separate California non‐profit corporation. On August 18, 2008, we became an official 501(c)(3) public charity!

The battery studies that Tom Saxton have been running for years rely on nearly semi-annual updates from respondents that drive Tesla Roadsters and Tesla Model S as well as the Nissan Leaf and the first generation Toyota RAV4 EV.  In email correspondence with Tom, he has indicated that he is looking to expand the study  in the near future to include the Model X as well.

Tom’s long-term battery study has been invaluable not only to the greater EV community but specifically to Tesla fans as well.  The take rate for participants for the Tesla Roadster study is close to a 7% sample, from what I gather and the Tesla Model S one had a healthy start, but could use more participants.

With the new range numbers from the redesigned front fascia of the vehicle, I am sure I’m not the only one to wonder what the long-term differences would be between a 90D classic fascia vs a 90D new fascia.

Providing a third party study of the effects of long-term battery health enables all concerned with a greater understanding and comfort to know “that Tesla knows what they’re doing.” Furthermore, it gives current non EV drivers a sense of comfort when making the switch to electrically fueled car ownership.

So, if you own a Tesla Roadster and haven’t participated in the study. Or perhaps you’re one of the lucky few to have upgraded to the new 3.0 battery from Tesla, please fill out the Tesla Roadster battery survey.

Perhaps you’re a Model S owner and you’d like to help add to the number of respondents to this study, fill out for the Tesla Model S battery survey.

What has Tom been able to share with the public so far.

Well, for the Roadster, he’s published an entire study three years ago including a paper entitled “Plug in America’s Tesla Roadster Battery Study.” The advent of the 3.0 battery upgrade may require a new study and the addition of almost another three years since the publication of that study might give more information to the study, but that’s entirely up to Tom and his cohorts at PiA.

The Model S Results page is more dynamic than the Roadster results publication.

I have taken screenshots as of April 27, 2016 of a few of the dynamic charts that are provided on the results charts page.

The first chart that caught my eye is the chart on the battery capacity vs. the miles that particular Model S iterations. With new EPA numbers with the launch of the new fascia should further complicate this chart.

Battery Survey - Model S Battery Capacity-Miles

This same chart can be used to also track how a particular respondent’s vehicle matches with the universe of respondents. The Vehicle in black on the chart below shows the performance of my vehicle in relation to other respondents’ cars.

Battery Survey - Model S Battery Capacity-Miles - Specific Vehicle

The third chart that was of interest is the reliability of certain components, namely the Drive Unit, battery, and chargers on the Model S. I wonder if the increased reported failures on chargers for 2014 vehicles resulted in the movement from the old chargers to the new 48A charger.

Battery Survey - Model S Major Maintenance - Model Year

Lastly, the inspiration to my exhorting fellow owners to participate in this survey was the chart of participant vehicles.

Battery Survey - Model S Survey Vehicles

For as many Model S are on the road now, I wonder as to the ability of this study, in its current count, to fully report on the vehicle with a small sample size. The Model S battery survey form is fairly straight forward and serves our common purpose. Tesla has been great, but it’s also good to have interested third parties run a check against what they claim and provide.

Quick update… after firmware patch.

So, the first 90% charge after the firmware of the Model S at 29,000 miles was updated from V6.0 (1.67.125) to V6.0 (2.0.81) and I was able to remotely verify that it was still holding at 226 miles for that charge. This is only a few days after the 29,000 Mile post (also, mileage is now 29,460).

29,000 Miles in the Tesla Model S


We’re getting to the point where we’re close to “doubling up” on the miles on the Model S than our Roadster. It’s got a lot more to do with the utility of a Model S vs. a Roadster as well as the number of miles that I drive for work.

The Model S is a 2013 that we picked up in November 2013 and the Roadster is a CPO that we picked up in September 2013. The Model S had around 20 miles when we picked it up and the Roadster had around 2200 miles when we picked that up.

They’re both great cars in their own rights and we’ve put in close to 15,000 miles on the Roadster at this point.

Ideal Miles – 291


Rated Miles – 251


The battery degradation on the car has shown its effect in the last few hundred miles. My 90% daily charge has gone down to around 226 Rated miles/261 Ideal Miles. Range charges in my most recent charge is 251 Rated miles/291 Ideal Miles. Most parts of the world, winter has a very noticeable effect on range. This is not normally the case in Southern California. With 291 Ideal miles or 9 lost miles. Doesn’t look bad vs 300 miles. But I live in a rated miles world and my 265 is now down to 252, so that sorta hurts, after all 13/265 is almost 5% of usable range gone vs. the Ideal loss around 3% (291/300) though this is also 4.6% (291/306).

I often charge the car to 90% on most days.  The Roadster has a feature called “Top Off” that I mimic on the Model S.  What that means is before I roll off, I extend the charge and run the pre-conditioning to cool the car and battery down.  I wonder if that contributed to the degradation.  Either way, there it is.  Still waiting on Tesla to step it up and provide some sort of warranty and guidelines on what to expect.

If battery loss is linear, then should I expect 20% loss by 116,000 miles?  Or is loss on a curve and 20% is closer to 150,000 or even 200,000 miles?  Or is it accelerated and 5% loss is actually good for 29,000 miles.  Not to mention the fact that I am measuring this based on Rated miles, which is what really works, but should I continue to normalize and look at Ideal miles, which may be ideal, but unrealistic.  Because, as I’ve written, that’s currently closer to a 3-4.5% loss of Ideal miles.

Do I need to change my charging behavior?  Because 9,000 miles ago, I was still range charged to 260 Rated miles.  So, if we plot my two data points.  I lost 5 miles in the first 20,000 miles and 8 more miles in the next 9,000 miles (Rated miles again).  One thing to note is the car has been constantly updating with new firmware.  So, one needs to wonder if these losses were already in the car and that changes in the rated mils is because of firmware changes and “fine-tuning of the algorithm.  If it’s based on our driving, I’ve been rather conservative over the past 29,000 miles.  My average since we picked up the car has been 308 Wh/mile (see the Trip A meter on any of the three pictures.)

Perhaps I should charge the car less and let it dip below 100 miles more often.  Some guidelines from Tesla really would help here.  In the meantime, we’re saving for replacement battery packs for the Roadster and Model S.  Even without some sort of guidelines on what that will cost, the recent 400 mile Roadster upgrade announcement makes me hopeful that the same, or at least a replacement pack price will be announced for the Model S when our pack drops down to a level where either something kicks in to replace it or it gets unbearable and we decide to replace it ourselves.  Tesla Gigafactory anyone?

As I’ve mentioned before, c’mon Tesla be AWESOME again! Publish a Battery Degradation Warranty that we can count on or at least start to provide pricing on replacement Model S packs for folks to start saving up for.

[UPDATE 2015-01-17]
I’ve been asked several times on social media as to the firmware at the time of the publication of this post.

It is V6.0 (1.67.125). The car is actually going in for a patch to fix an unrelated bug with regard to the clock switching from 12h AM/PM to 24h seemingly randombly.)

Now updated to v6.0 (2.0.81) at the Service Center.