Guidance on Battery degredation? 2016 90D 46k miles only 72kwh left?

[Greg Blythe]

very close to making a deal on a 2016 Model S 90D. It has 46k miles. It checks all of the boxes for me, but seems to have a rather low battery capacity. Ive spent hours watching videos and reading forums about how to estimate the range/battery capacity. It all seems to point to a 12% battery degredation. Drove the car today at highway speeds and <40 degrees. Car could only muster ~200 miles. Should I be concerned about the battery degredation? Can this be remedied through "rebalancing?" What is the best way to accurately determine the state of the battery pack?

thanks in advance!

 

[Greg Blythe]

Note the vehicle is a Nov 2016 build date. 100% rated range is 255 miles. I was expecting the rated range to be closer to 275.

 

[WallysRide]

The "good" news is that this battery likely is flattening out in terms of ongoing degradation, since degradation is typically steepest in the first few years. But of course that's only a statistical prediction.

Based upon this prediction and your real-world test of the range, I would suggest focusing on the question of whether you can live with an effective range of ~200 miles during colder months. In cold temps, expect to lose ~15% of range just due to the colder temps (w/o cabin heat) and 25-40% of range if you use cabin heat. So yes, if you're estimating 12% degradation and then adding on further losses due to cold temps, it's perfectly normal to have an effective range of ~200 in the winter.

The other question I would suggest focusing on is your risk tolerance - how comfortable are you with a pack that's turning 7 years old this year and has less than 2 years left of warranty.

Basically, what I'm suggesting is that you not think so much about further degradation, but rather the usability of what's remaining given your driving needs, and your comfort level with an older HV battery. Doing rebalancing tricks might get you a couple more miles displayed on the computer but won't materially improve the actual range you get.

 

[gaswalla]

You got 200 miles in 40° weather? That’s pretty solid. In fact that’s probably as good as you’re gonna see from any 2016 Model S

 

[Greg Blythe]

thanks! both great comments. It does seem to come down to other factors. 6 year old car vs newer? ride quality and reliability issues of an S vs cheaper new verhicle. Ultimately the battery is what the battery is. There is no "perfect" used car.

 

[PCMc]

There are a wide range of factors that impact battery degradation. One of these is time, so the fact the car is 6.5 years old is as significant than the mileage you see on the odometer. Also, it's not just odometer miles, but as explained to me by some I know that develop battery packs, it's more what they call energy throughput, which is the total energy that has gone both in and out of the battery. So 40k miles driving a 200 Wh/mi versus someone driving 40k miles at 400 Wh/mi are also going to see different results.

I think you're also mixing battery degradation with what your energy consumption is per mile. Based upon my June 2016 built MS90D, with 19" wheels, I need to average about 274 Wh/mi to get 1 actual odometer mile per 1 rated miles as displayed by the car. Not sure what you averaged on your 200 mile drive that you referenced. If it's higher than about 270-275, then you're going to see less usable range that the displayed number of rated miles.

I've posted this information in several other threads over the past couple months, but below is plot for my MS90D relative to apparent battery degradation over time. It would suggest that my car over past couple years has started to level out around about 268-270 rated miles, so about 8.5-9% apparent degradation.

How much of what you're seeing is function of cell balance or even the BMS calibration having drifted is difficult to assess. I try to operate my car in a way that I allow the BMS to naturally see a range of 50-80% versus topping up every day. I also tend to periodically take longer range trips which tend to cycle the battery from normal operation over a range of say 20-80% or 10-90% also helping to keep my BMS calibration from drifting too far out of whack. As a result I've never had my range "recover" any practical amount for intentionally trying to balance the pack.

Does 255 rated miles feel a little on the lower side? Yeah, maybe a little based upon my 1-off datapoint. Only you can judge if overall this car meets your needs versus you continue to look for another car.

Whichever way you decide, hope you find something that you enjoy and works well for you.

 

[Greg Blythe]

There are a wide range of factors that impact battery degradation. One of these is time, so the fact the car is 6.5 years old is as significant than the mileage you see on the odometer. Also, it's not just odometer miles, but as explained to me by some I know that develop battery packs, it's more what they call energy throughput, which is the total energy that has gone both in and out of the battery. So 40k miles driving a 200 Wh/mi versus someone driving 40k miles at 400 Wh/mi are also going to see different results.

I think you're also mixing battery degradation with what your energy consumption is per mile. Based upon my June 2016 built MS90D, with 19" wheels, I need to average about 274 Wh/mi to get 1 actual odometer mile per 1 rated miles as displayed by the car. Not sure what you averaged on your 200 mile drive that you referenced. If it's higher than about 270-275, then you're going to see less usable range that the displayed number of rated miles.

I've posted this information in several other threads over the past couple months, but below is plot for my MS90D relative to apparent battery degradation over time. It would suggest that my car over past couple years has started to level out around about 268-270 rated miles, so about 8.5-9% apparent degradation.

How much of what you're seeing is function of cell balance or even the BMS calibration having drifted is difficult to assess. I try to operate my car in a way that I allow the BMS to naturally see a range of 50-80% versus topping up every day. I also tend to periodically take longer range trips which tend to cycle the battery from normal operation over a range of say 20-80% or 10-90% also helping to keep my BMS calibration from drifting too far out of whack. As a result I've never had my range "recover" any practical amount for intentionally trying to balance the pack.

Does 255 rated miles feel a little on the lower side? Yeah, maybe a little based upon my 1-off datapoint. Only you can judge if overall this car meets your needs versus you continue to look for another car.

Whichever way you decide, hope you find something that you enjoy and works well for you.View attachment 907787

Thanks for the thoughtful reply! One of the suggested methods for determining total battery capacity: multiply the average use rate (kwh/mile) with range remaining (mile) and divide by % battery remaining. I calculated this a couple times over the course of the trip yesterday. The result was approximately the same ~72kWh. I believe this approach should allow for eliminating the efficiency variables (tire size, speed, hvac, etc) from the calculation. Though given your efficiency assumption, 270-275 wh/mile would suggest a range of 263. Better than the 255 rated miles as reported at 100%. Based upon 263 (degradation is closer to 10.5%).

The car spent its life in Florida. I don’t have a history of how the car was charged or driven.

Maybe, an appropriate question: how does battery degradation affect value? (-$/%)

 

[Greg Blythe]

I think the last question is perhaps the best question… what’s the value of battery capacity in a used model S?

One way to compare is to look at the difference in values of identically optioned 2016 75d and 90d (kbb estimates this to be ~$3k). Assuming that both batteries saw a 7% degradation. That $3k is buying you about 8.6kWh. Or approximately $350/kWh, and $240/(% degradation)

 

[Greg Blythe]

I think the last question is perhaps the best question… what’s the value of battery capacity in a used model S?

One way to compare is to look at the difference in values of identically optioned 2016 75d and 90d (kbb estimates this to be ~$3k). Assuming that both batteries saw a 7% degradation. That $3k is buying you about 8.6kWh. Or approximately $350/kWh, and $240/(% degradation)

You’d be miles ahead (pun intended) buying a high degradation 90D than a low degradation 75D

 

[Ciaopec]

I have a June build 2016 90D. After more than 155,000 miles my degradation is about as listed by @PCMc. Over the course of that 155,000 miles driven from points as diverse as Chicago and Albany,NY to Biloxi, MS and Melbourne, FL and points in between in weather ranging from -10F to 90F
my Wh/mi has averaged 288. I'm happy with the 273 miles rated range and don't worry too much about it. I don't charge to 100%; don't drive more that about 150-200 miles before stopping anyway. There are so many charging options now compared to when I started driving this car in 2016 that I don't worry about range. Enjoy the car! I think it is one of their best "build years".

 

[2101Guy]

 

[Boza]

That does not work on MS.

 

[PCMc]

I think the last question is perhaps the best question… what’s the value of battery capacity in a used model S?

One way to compare is to look at the difference in values of identically optioned 2016 75d and 90d (kbb estimates this to be ~$3k). Assuming that both batteries saw a 7% degradation. That $3k is buying you about 8.6kWh. Or approximately $350/kWh, and $240/(% degradation)

The answer to the question of what's the value of added battery capacity that has any meaning in this situation is the value that you choose to assign to it. What the extra range is worth to you may be totally different than what it's worth to me or any other person. Just as the value of a car that is multi-coat red may be worth more to you than to me who is quite content with my "dull" midnight silver metallic.

Regarding the right, best, or correct way to determine battery capacity and degradation, you'll find as many "experts" on that topic as you'll find people posting videos on youtube and other platforms professing theirs to be the "right" way, which in my more cynical view at times, is simply them desperately trying to generate clicks and ad revenue. Fundamentally all of these methods are based upon the same fundamental estimates of the BMS system and trying to 'peer inside the black box' to try and infer from them.

Regarding the method you mentioned of using displayed percent remaining and displayed Wh/mi values, I've found that method to not always be that great. Couple reasons for that are the Wh/mi values you'll find displayed and the implied rate constant are not necessarily consistent with the same values you'll see posted in other screens. For example, what I find I must average (my 274 Wh/mi number) so 1 odometer mile = 1 rated mile is not the same constant that Tesla uses in the Energy screen on the main display (appears to be based upon a 288 Wh/mi rate constant). Also, you start to run into numerical errors with roundig and that fact that the percent value is only displayed to integer value, and a displayed value of 50% could be anywhere from 49.51 to 50.49%. That may seem insignificant, but if you're referencing off a SOC of say 20%, that error is getting amplified by a 5x factor.

That's where I came to the view that what's more valuable to me using a consistent method over time and only focusing on the trend. Does it really matter whether my current estimate of ~ 8.5% degradation is the absolute correct value? No, what matters is how is it trending and is it stable, especially if I were to see a sudden drop of say 5-8% which might indicate I have a module about to fail.

There are a few tricks I use to minimize numerical noise in the data I shared. It's based on values that I can pull directly over the API via a python script I wrote. That allows me to get the rated range to two decimal places. However, the SOC percent value is still only available as integer. To combat that I always take values at the end of a charging session where I've let the car shut off at my desired SOC, whether that's 60%, 70%, 80%, or 100%. This way I have a consistent percent reference relative to rounding. And for those that argue that taking ratio of rated range to SOC to get the full 100% value is inaccurate, I'd challenge you to pick out the 25 or so data points from the roughly 1200 points in my graph above. Trust me, I've isolated those and there is no influence of whether I charged to 70%, 80% (my default), or let it fully charge to 100%, sitting to trickle for 1-2 hours before it finally stopped.

Good luck on your decision and hopefully you'll find a car you enjoy.

 

[Greg Blythe]

Thanks all for the feedback and thoughtful analysis! I did decide to buy the car. Decided that the FUSC, Red Pearl, Enhanced AP, 41k miles made up for the battery condition.

First supercharging experience was very troubling! After about 5 minutes, the car (now my car) kicked out several errors: BMS_u011, BMS_f035, BMS_w035, BMS_u016, BMS_f086, BMS_w086, GTW_w018, DI_u014, BMS_f071

At this point, I nearly sh!t myself.

I was able to disconnect from supercharger. Did a system shutdown. Upon restart, the car acted normal. I was able to get home.

needless to say, I’ll talk to service today. Welcome to Tesla!

 

[siraj]

Note the vehicle is a Nov 2016 build date. 100% rated range is 255 miles. I was expecting the rated range to be closer to 275.

Mine is June 2016 build. I used to get 285 when I first bought the car. Now I get 258 on full charge.