-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Model S Battery Degradation data
- Thread starter Maarten ST
- Start date
Electric700
Active Member
Thanks for that.
So, the worst case scenario is approximately a 6% loss of range after about 85,000 km (52,816 miles), which is about a 27 km/17 mile loss. That's also probably with the earlier battery packs (Revision A or B). As new battery pack updates come along with new battery technologies, the range lost over time should be reduced.
So, the worst case scenario is approximately a 6% loss of range after about 85,000 km (52,816 miles), which is about a 27 km/17 mile loss. That's also probably with the earlier battery packs (Revision A or B). As new battery pack updates come along with new battery technologies, the range lost over time should be reduced.
I've got 6% degradation after 30,000 miles, so it seems unlikely I'll still be at 6% in 20,000 more miles...
A brief look at my own data shows that I'm somewhere in the range that the graph shows. Although I'm not sure what 100% is referring to. Not all cars show the same range when new. So the 100% seems to be slightly different for each car.
I'm not sure the charge to 100% and look at the 'rated range' number is a very good way to estimate battery degradation. We know from Tesla that the rated range numbers are based on a computer model and it has changed with each firmware version. So there is a variable that we don't know about. Based on just looking at the 100% charge range, my car has lost 5.5%. Based on a 100% to almost zero test and looking at the energy used, I found my car has about 2.5% lost. That's quite a difference. Both methods are not 100% accurate.
What this graph shows and is encouraging, is that the degradation seems to be fast at first and then slows down with time. That is excellent news!
I'm not sure the charge to 100% and look at the 'rated range' number is a very good way to estimate battery degradation. We know from Tesla that the rated range numbers are based on a computer model and it has changed with each firmware version. So there is a variable that we don't know about. Based on just looking at the 100% charge range, my car has lost 5.5%. Based on a 100% to almost zero test and looking at the energy used, I found my car has about 2.5% lost. That's quite a difference. Both methods are not 100% accurate.
What this graph shows and is encouraging, is that the degradation seems to be fast at first and then slows down with time. That is excellent news!
You are completely right that there are multiple variables at play here. But we do see a overall decrease in range as time and mileage goes by.
The 5% is about what most people see when they've got the car for a longer time now. Did I really loose 5% of range? I don't know, but I do expect it to be somewhere between 3 and 5% in real life. But I never drive it that close to being empty.
Was it confirmed that the latest packs are using that new cell with the silicon doped anode?
I probably need to update the data on mine.
My first battery pack (rev A) failed at 18,000 miles (contactors). When it failed, a 90% charge would reach 228 miles and a 100% charge would max at 246-248 miles (down from 265 when I first received it).
The car is now at 45,000 miles. My second battery pack (a refurb rev D) has 27,000 miles on it. The other day, a full range charge brought 265 miles (down from 268 when it was installed).
Finally, keep in mind that all of this is also subject to Tesla's changes in the estimation software; there were some big difference swings as 4.x, 5.x, etc. rolled out. The only way you'd truly know is to create a very controlled environment and physically drive the battery pack to completely empty after a fresh charge completion.
My first battery pack (rev A) failed at 18,000 miles (contactors). When it failed, a 90% charge would reach 228 miles and a 100% charge would max at 246-248 miles (down from 265 when I first received it).
The car is now at 45,000 miles. My second battery pack (a refurb rev D) has 27,000 miles on it. The other day, a full range charge brought 265 miles (down from 268 when it was installed).
Finally, keep in mind that all of this is also subject to Tesla's changes in the estimation software; there were some big difference swings as 4.x, 5.x, etc. rolled out. The only way you'd truly know is to create a very controlled environment and physically drive the battery pack to completely empty after a fresh charge completion.
rpavlicek
***** Neophyte
My issue with using rated miles (US), is this:
Ever since new releases its hard to tell what this number really means. I assumed that X rated miles means I can drive X miles with efficiency of 300Wh/mile. This does not seem to be the case...its much closer to 275/280Wh/mi in my experience. I'm not sure if this is because the computer estimation is bad or Tesla is fudging the numbers for some reason.
Note: I live in the bay area where it typically does not get too cold or too hot. I hardly ever run the AC or heat. (In other words, my mileage efficiency alone is going to be fairly accurate related to battery usage)
Ever since new releases its hard to tell what this number really means. I assumed that X rated miles means I can drive X miles with efficiency of 300Wh/mile. This does not seem to be the case...its much closer to 275/280Wh/mi in my experience. I'm not sure if this is because the computer estimation is bad or Tesla is fudging the numbers for some reason.
Note: I live in the bay area where it typically does not get too cold or too hot. I hardly ever run the AC or heat. (In other words, my mileage efficiency alone is going to be fairly accurate related to battery usage)
That's how I'd really like to see it graphed. Or graphed by aging, and then a snazzy 3D graph by age and mileage.
Your blog isn't clear about it but remember the warranty does not cover degradation of the battery.
Mario Kadastik
Active Member
Ok, first off, are you guys reporting the degradation from comparable car battery states? Have you either validated with Tesla service that the pack is balanced or have you performed reasonable amount of balancing before taking the data point?
The reason I ask is that not all degradation in the number you see at range charge comes from actual degradation. I was getting about 394-396km on a range charge on Typical estimator in summer 2014 after about half a year and 15000km and was asking Tesla if that was normal. They told me to range charge, let the car sit a couple of hours at 100%, then drive to about 20%, then range charge and let the car sit for a couple of hours and repeat that ca 4-5x to actually balance the pack. It sounds counterintuitive as they normally recommend not to range charge too often or to let the car sit at 100%, but that's where the balancing actually occurs. So I took a long roadtrip weekend (covered 1500km in three days with no superchargers or any fast charging available and I have a single charger limited to 3x13A). The first time I started off I range charged and got 394km. I then drove 400km and had about 20km estimated to go when I arrived at my hotel. I range charged the car again and when I set off the next morning I had 398km of range. I repeated this 2-3x as I needed to range charge multiple times to get back home, but the last leg I started with 401km after the range charge. So I had recovered 7km of range based purely on doing what Tesla told me to do (well I also needed to do it because of my driving needs).
I have since then retried it now that I get 389km after ~30 000km and I think I only did one or two range charges and drove extensively and got to 394km so yes it does recover range. So after 30 000km I get 394km still and I think the most I ever saw in my car was 402km though I know newer cars that were delivered with 5.9 firmware (the one that changed substantially the range algo) have seen 408km or so as well. So if I take ~405km as the original, then I've lost around 10km of range over 30 000km or about 2.5%, but I can't be quite sure as I've not gone through the full balancing cycle as I've not had enough need to drive recently to have a good reason to do multiple range charges in row and then also drive 300+ km that day. Once I get that kind of conditions I'll do another pack balancing act and hope to recover some more of the range.
From what I have understood from people who've measured the roadster and now Model S battery degradation you'd expect about 1-2% the first year and ~1% afterwards. If we take 1 year is 20 000km as standard (that's what Tesla and others count for yearly services etc), then I've gone 1.5 standard years and have 2.5% degradation. This seems to indicate about on-par degradation curve though I think it's conservative. I think I quite recently saw 396km range on my car, but didn't document it.
The reason I ask is that not all degradation in the number you see at range charge comes from actual degradation. I was getting about 394-396km on a range charge on Typical estimator in summer 2014 after about half a year and 15000km and was asking Tesla if that was normal. They told me to range charge, let the car sit a couple of hours at 100%, then drive to about 20%, then range charge and let the car sit for a couple of hours and repeat that ca 4-5x to actually balance the pack. It sounds counterintuitive as they normally recommend not to range charge too often or to let the car sit at 100%, but that's where the balancing actually occurs. So I took a long roadtrip weekend (covered 1500km in three days with no superchargers or any fast charging available and I have a single charger limited to 3x13A). The first time I started off I range charged and got 394km. I then drove 400km and had about 20km estimated to go when I arrived at my hotel. I range charged the car again and when I set off the next morning I had 398km of range. I repeated this 2-3x as I needed to range charge multiple times to get back home, but the last leg I started with 401km after the range charge. So I had recovered 7km of range based purely on doing what Tesla told me to do (well I also needed to do it because of my driving needs).
I have since then retried it now that I get 389km after ~30 000km and I think I only did one or two range charges and drove extensively and got to 394km so yes it does recover range. So after 30 000km I get 394km still and I think the most I ever saw in my car was 402km though I know newer cars that were delivered with 5.9 firmware (the one that changed substantially the range algo) have seen 408km or so as well. So if I take ~405km as the original, then I've lost around 10km of range over 30 000km or about 2.5%, but I can't be quite sure as I've not gone through the full balancing cycle as I've not had enough need to drive recently to have a good reason to do multiple range charges in row and then also drive 300+ km that day. Once I get that kind of conditions I'll do another pack balancing act and hope to recover some more of the range.
From what I have understood from people who've measured the roadster and now Model S battery degradation you'd expect about 1-2% the first year and ~1% afterwards. If we take 1 year is 20 000km as standard (that's what Tesla and others count for yearly services etc), then I've gone 1.5 standard years and have 2.5% degradation. This seems to indicate about on-par degradation curve though I think it's conservative. I think I quite recently saw 396km range on my car, but didn't document it.
Mario's points are true and therefor a completely correlated graph nigh impossible to create. Nonetheless, despite the variables of rated miles algorithm changes, balancing, etc. I still find this graph useful and reassuring. It tells me that, except for aberrations, I can expect to get to 100k miles with less than 10% degradation. And, since this synchs with experience of roadster owners, that makes 2 reinforcing data points.
Mario Kadastik
Active Member
Btw so that you don't get discouraged. At times I've seen the first range charge sit on the 1 minute remaining for quite long (balancing) and then next time no change in the total number, but only on the 3rd or 4th full charge after driving the battery down to low SOC did the actual number change higher. So if you do one range charge and then don't see a change in the second it may or may not mean that the pack was balanced and it's true degradation. With the cold weather it's relatively safe to test this as high SOC and high temperatures are what's bad for the battery. Low temp and high SOC isn't as bad and hence it's safe to do the balancing in winter (definitely safer than during hot summer).
The second option of course is to go to Tesla and ask them to put the car into service mode and show the battery balance screen If it's all well balanced, then no point in doing this exercise, but if you regularly go months with only charging to 80-90% and drive it down to no lower than 50%, then it's highly likely your pack is out of balance at least slightly.
The second option of course is to go to Tesla and ask them to put the car into service mode and show the battery balance screen
If it's all well balanced, then no point in doing this exercise, but if you regularly go months with only charging to 80-90% and drive it down to no lower than 50%, then it's highly likely your pack is out of balance at least slightly.
I take it these are people reporting their "rated range" after doing a full charge?
I think the formula to compute rated range has changed over time, with different firmware releases.
Tesla is trying to make it more accurate over time, I think.
If that means rated range has been "falling" over firmware releases to be more accurate, then has the battery actually "changed" for that release and jump down 5% capacity? No.
Is this change in computation basis reflected in the graph? No. Because we don't know it.
Or did people reporting all driving their cars to zero and report range attained from a full charge?
Which of course requires precision of test that no normal person can reproduce two times, back to back.
Rated range and how I drive are two different things.
Even in ICE cars trying to estimate their range from a full tank of gas are only estimates, it doesn't know what octane rating gas I just pumped in.
Only lately has Tesla been showing % SOC on the dash. Because we were demanding to see "what's left this charge". So only now can we see the fuel gauge. And the car will always show 100% charged. Not, 93% charged after fully topping up, in a few years time.
I think a person can make up the loss of capacity over time by compensating how they drive to match.
I bet if you get 200 miles or something from a full charge today, and you drive slightly differently in 5 years, you can still get 200 miles from the same Model S.
Subtract two "zoom cycles" of gratuitous acceleration runs... per drive. And you'll be there.
Finally, people think of capacity loss over time as "something subtracted from the top of range".
When in fact, it is subtracted from the bottom of the "tank". Think of it as the tank shrinking in size from the bottom up.
In other words, the losses are not readily apparent because there is fuel sitting on top of it so you can't see the bottom of the tank shrinking on you.
If you're like me... I never leave charging until the last moment with 1 % SOC. And people don't typically cut it that close to get gas only when well below E and have driven twenty miles more.
They fill up sooner. I don't think I routinely see below 40% before I'm topping up charge. I do think ICE cars people get to know their absolute limits more, because it is a hassle stopping to get gas.
With E cars, it is so convenient to charge at home, fewer play the game of trying to get close to empty before "needing to fill up".
All of this to say, if you never see 30% remaining on your car then quite literally, "who cares?" if the battery is losing capacity over it's lifetime.
I know this all changes when you're on a range trip, every last km and mile you can eek out of a charge the better, right?
Is anybody buying model S today, that has a routine HARD REQUIREMENT that the car must achieve very close to its rated range (or actual achieved range), OR ELSE they are stranded on the road?
If you require actually achieving 95% rated range OR ELSE you are stuck in the dessert... then the E car is clearly not for you, today. Let alone in several years when the capacity has dropped.
Last edited:
Similar threads
- Question
