90D Range slowly declining

[NOLA_Mike]

221 with range mode OFF is equivalent to 224 with it ON. That's exactly where I was seeing my 90% charges for a while until this sudden and steep drop in the last month.

Mike, was your 249 number @100% with range mode ON or OFF?

I couldn't remember but luckily things are memorialized with posts here and I found my summary of that trip - 249 miles with Range Mode OFF

Not to derail this topic, but I'm still waiting for "redesigned rear shocks" to get rid of the noise in the back - you?

Mike

 

[brkaus]

Does the 70 have the same chemistry as the 90?

 

[MarcG]

I couldn't remember but luckily things are memorialized with posts here and I found my summary of that trip - 249 miles with Range Mode OFF

Not to derail this topic, but I'm still waiting for "redesigned rear shocks" to get rid of the noise in the back - you?

Mike

Thanks for the info, I'll read the thread you linked. As for the shocks, I'll PM you shortly with some info.

 

[NOLA_Mike]

Thanks for the info, I'll read the thread you linked. As for the shocks, I'll PM you shortly with some info.

Cool - and to satisfy my own curiosity I just started a full charge. I'll update what I get in about an hour when it's done.

Mike

Updated - it appears as though my current 100% charge results in 245 miles with Range Mode OFF.

 

[MarcG]

Cool - and to satisfy my own curiosity I just started a full charge. I'll update what I get in about an hour when it's done.

Mike

Updated - it appears as though my current 100% charge results in 245 miles with Range Mode OFF.

Nice, so that's around 248 miles with Range Mode ON. From my last 100% charge on 11/10, after the cold weather started, I got only 246. I was consistently getting 249 to 251 on 100% charges over the previous several months - see data & charts below.

My 100% charge will complete this afternoon, and I'm expecting something like 244 miles of rated range. So 241 if Range Mode was OFF. Quite a steep drop in such a short amount of time...

 

[liuping]

221 with range mode OFF is equivalent to 224 with it ON.

There used to be a 3-4 mile difference between the range at 90% when setting Range mode on vs off, but with 7.0, I'm not sure any more. I now only see 1 mile difference if Range mode is on.

That might be part of the reason some people are seeing more of a decline.

 

[MarcG]

There used to be a 3-4 mile difference between the range at 90% when setting Range mode on vs off, but with 7.0, I'm not sure any more. I now only see 1 mile difference if Range mode is on.

That might be part of the reason some people are seeing more of a decline.

Interesting theory, but I don't believe that to be true. As I mentioned upthread, I upgraded from 6.2 to 7.0 on 10/14/2015 and my rated range at 90% charge stayed very consistent (224 miles with Range Mode ON).

I have 10 total data points for charges done on 7.0 version 2.7.056, nine of which are at 90% charge limit and one at 100%. Seven of these show 224 miles and the last two 221. Those two happened after ambient temps dropped a lot.

I'll test it again after my 100% charge is complete, but I'm pretty sure I'll still see a 3-mile difference when toggling between Range Mode ON and OFF.

 

[FlasherZ]

@wk057 - I'm sure you've seen FlasherZs range numbers (266 rated at 100% for a P85 after 40,000+ miles). AFAIK, he has the best range numbers across the entire Model S fleet with equivalent mileage. Given what you have observed in cell cycle life, is there anyway that those numbers are accurate? IMO, it's pretty unbelievable that he's had 0 capacity fade with that many miles. I don't doubt the data he is collecting and reporting, but I do have serious doubts with the range algo and why it differs so greatly from what others report.

My fear is that I'll figure this out in some strange way - like the car shutting down in 25 degree (F) weather showing 25 miles or something, or a significantly higher Wh/mi number. But I hope not. My Wh/mi stats are the best they've ever been with 7.0 (averaging about 30 Wh/mi less than previous software versions).

For those who haven't seen my posts - I included more data here: Displayed Range and Seasonality

I'm currently in the valley of the seasonal trends, with max range charge going to 263 mi, and 90% charge stopping near 234-236 mi. Once Mar-May rolls around, I'll see how far it comes back up. That thread plots every data point I have in terms of range, so you see any outliers as well.

My data is showing a very, very slight downward trend using linear regression. Marc's data is definitely showing a continual and pronounced downward trend, something that I'm not seeing. I concur that the data plots do look quite different.

(EDIT: recalculated the miles per day number from the linear regression calcs, PEBKAC. When looking at linear regression of the data, I come up with a daily degradation of -0.0043 miles per day, or roughly 1.5 miles per year.)

 

[wk057]

Real quick, somewhat relevant: Tesla won't sell me a 90 kWh pack unless I give them my old pack for 12% market value

@wk057 - I'm sure you've seen FlasherZs range numbers (266 rated at 100% for a P85 after 40,000+ miles). AFAIK, he has the best range numbers across the entire Model S fleet with equivalent mileage. Given what you have observed in cell cycle life, is there anyway that those numbers are accurate? IMO, it's pretty unbelievable that he's had 0 capacity fade with that many miles. I don't doubt the data he is collecting and reporting, but I do have serious doubts with the range algo and why it differs so greatly from what others report.

My fear is that I'll figure this out in some strange way - like the car shutting down in 25 degree (F) weather showing 25 miles or something, or a significantly higher Wh/mi number. But I hope not. My Wh/mi stats are the best they've ever been with 7.0 (averaging about 30 Wh/mi less than previous software versions).

For those who haven't seen my posts - I included more data here: Displayed Range and Seasonality

I'm currently in the valley of the seasonal trends, with max range charge going to 263 mi, and 90% charge stopping near 234-236 mi. Once Mar-May rolls around, I'll see how far it comes back up. That thread plots every data point I have in terms of range, so you see any outliers as well.

My data is showing a very, very slight downward trend using linear regression (.068 miles per day). Marc's data is definitely showing a continual and pronounced downward trend, something that I'm not seeing. I concur that the data plots do look quite different.

@FlasherZ - Yeah, I find it hard to believe that you've suffered no real degradation after 40,000 miles. Almost certainly an algorithm issue at work here. At 40,000 miles you're at a minimum of about 150 cycles... that should be closer to 10% degradation at the worst case, and at a minimum about 4% or so based on some of my testing. It's either a screw up of the algorithm, or it confirms my earlier theory that Tesla hides some degradation in a small buffer on the 85 packs. At least with the latter you'll actually have the range available... with the former.... I'd be prepared next time you run the pack down below ~10% or so.

 

[FlasherZ]

@FlasherZ - Yeah, I find it hard to believe that you've suffered no real degradation after 40,000 miles. Almost certainly an algorithm issue at work here. At 40,000 miles you're at a minimum of about 150 cycles... that should be closer to 10% degradation at the worst case, and at a minimum about 4% or so based on some of my testing. It's either a screw up of the algorithm, or it confirms my earlier theory that Tesla hides some degradation in a small buffer on the 85 packs. At least with the latter you'll actually have the range available... with the former.... I'd be prepared next time you run the pack down below ~10% or so.

The only thing I can't figure out is why I don't see degradation occur, yet nearly every other 85 kWh owner does. You would think we'd see the same effects across the fleet, whether "reserved capacity" or not.

Last summer, I ran the pack down to less than 5%, so I know that I can do it (at least to 8 mi).

- - - Updated - - -

NOTE: I edited my post above after redoing the linear regression on my data. It's showing a degradation of 0.0043 miles per day, or ~1.5 miles per year...

 

[AWDtsla]

@FlasherZ - Yeah, I find it hard to believe that you've suffered no real degradation after 40,000 miles. Almost certainly an algorithm issue at work here. At 40,000 miles you're at a minimum of about 150 cycles... that should be closer to 10% degradation at the worst case, and at a minimum about 4% or so based on some of my testing. It's either a screw up of the algorithm, or it confirms my earlier theory that Tesla hides some degradation in a small buffer on the 85 packs. At least with the latter you'll actually have the range available... with the former.... I'd be prepared next time you run the pack down below ~10% or so.

Did you characterize how much cycle count matters when it isn't a full cycle? From the research I have done it seems to be more about wall clock time spent in very high or very low SoC and at certain temperatures, not the amount of cycles that matters. The battery test sheets from Panasonic, etc, don't seem to take this into account, they simply cycle as fast as possible not controlling for wall clock time.

 

[wk057]

Did you characterize how much cycle count matters when it isn't a full cycle? From the research I have done it seems to be more about wall clock time spent in very high or very low SoC and at certain temperatures, not the amount of cycles that matters. The battery test sheets from Panasonic, etc, don't seem to take this into account, they simply cycle as fast as possible not controlling for wall clock time.

Time spent at high and low SoC definitely has an impact. But the cycling itself, at any point in the cycle, has a discernible effect on the capacity of the cell as electrolyte slowly gets lost during the chemical process. This is pretty constant across the inner SoC range, and slightly more accelerated near the ends. My one test I've had running has been simulating "driving" by discharging for a bit, then charging (regen) for a moment, then a ramp up again followed by a level discharge rate, etc. It had about 4% degradation at what I would estimate would be 40,000 miles of "driving." (10% to 90% SoC range)

 

[andrewket]

Thanks for the info, I'll read the thread you linked. As for the shocks, I'll PM you shortly with some info.

PM me too?

 

[brkaus]

Does the 70 have the same chemistry as the 90? Is chemistry where they achieved the additional capacity or more cells?

 

[apacheguy]

Pretty sure the 70 has the same cells as the 85 and the former 60 kWh. The 60 had spacers in the pack and I'm guessing they just filled those in to get to 70.

 

[max35111]

Guys, on the same topic, I just received my 90d on Friday and the 90% is only 380km/237 miles!

It looks ridiculously low...
Is there a chance the mileage will ramp up with the first few charges or I should raise the concern right away?

 

[Oyvind.H]

I`m waiting for a 90D and find this thread a bit concering. If 85Ds have 400km at 100%, the 90D should hve 424km at 100%. My new S85 had about 400km at 100% back in 2013, but it seems to me as if new 85Ds have more than 400km at 100% when new?

 

[BertL]

Update: 62 days of S90D ownership.

• High 90% Rated Range of 255 miles at delivery; Low of 250; A few minutes ago it was 251 after topping-off to 90% at midnight. ...My 90% Rated Range has varied up-and-down 1-2 miles on any given day, and one time I've caught the number varying the same amount up-and-down within seconds via my Tesla App. I still don't have any correlation to temperature as others have suggested, although it remains moderate (thank you) here in SoCal compared to others.
• I've only trip-charged to 100% twice with a Rated Range High of 282; Low of 278.

Detail in attached: View attachment Rated Range.pdf

OPINION ON

I still believe Tesla is doing some sort of real-time guesstimate when it comes to Rated Range, and really needs to prioritize technical refinement of their 90 kWh-battery algorithm and how that number is presented, to remove some of this increasing concern with a number that varies as much as it presently does. 90kWh Owner's discretely paid $3K for the "Range Upgrade" option to provide 6% (aka 16 miles) more than an 85kWh model. As such, right or wrong, I believe we may be even more concerned with perhaps premature loss of range because it's easy to do the math as to what our decision may have cost us every time Rated Range at a particular SoC changes. Tesla needs to address this technically, as well as marketing-wise, to establish better owner expectations.
​

OPINION OFF

 

[jerry33]

OPINION ON

I still believe Tesla is doing some sort of real-time guesstimate when it comes to Rated Range, and really needs to prioritize technical refinement of their 90 kWh-battery algorithm and how that number is presented, to remove some of this increasing concern with a number that varies as much as it presently does.​

OPINION OFF

Ideal range appears to be much more stable (at least in the S85) than rated range because they don't seem to mess with it as much. I'm not sure why anyone uses rated range to determine degradation. Rated range is a reasonable number for driving, but not for battery state.

 

[msnow]

Guys, on the same topic, I just received my 90d on Friday and the 90% is only 380km/237 miles!

It looks ridiculously low...
Is there a chance the mileage will ramp up with the first few charges or I should raise the concern right away?

First you should triple check your slider to make sure it's set at 90%. Charge it again tonight after you have driven some miles. Keep a log and if it's still significantly less than what you expect (256/257) then definitely call Tesla.

@BertL, I'm exactly at the same point you are (249-250 @90% and 279-281 @ 100%) and I engaged Tesla. First they tried to convince me that the 6% range increase should only get me to where I'm at so I had to take the Service Technician through the math (270 rated miles for the 85D times 6% for the range upgrade equals 16.2 more miles before he agreed to escalate. I have a service appointment for 12/17 to check into this and a few other small issues. I'll keep everyone updated. I have to say it did piss me off that they tried to convince me it was within specs with fuzzy math.