Tesla's 85 kWh rating needs an asterisk (up to 81 kWh, with up to ~77 kWh usable)

[Electricious]

Here is the article from Electrek ... Tesla’s hacked Battery Management System exposes the real usable capacity of its battery packs

Jason Hughes, a Tesla owner and hacker tinkerer, was the first to reveal the discrepancy in Tesla’s advertised battery capacity versus the actual capacity in the pack and the available capacity. He did it through a tear down of the since discontinued 85 kWh battery pack of a Model S. He found that the 85 KWh battery pack actually only had a capacity of up to 81 kWh and ~77 kWh usable capacity. Now the data is directly from Tesla’s software and not a calculation based on the capacity of cells from a tear down of a pack. He gathered similar data from other Tesla models. Here’s a list he sent to Electrek:

• Original 60 – ~61 kWh total capacity, ~58.5 kWh usable.
• 85/P85/85D/P85D – ~81.5 kWh total capacity, ~77.5 kWh usable
• 90D/P90D – ~85.8 kWh total capacity, 81.8 kWh usable
• Original 70 – ~71.2 kWh total capacity, 68.8 kWh usable
• 75/75D – 75 kWh total capacity, 72.6 kWh usable
• Software limited 60/60D – 62.4 kWh usable
• Software limited 70/70D – 65.9 kWh usable

Thanks! Does anyone know the usable capacity for the 100 pack?

 

[lymex2018]

Thanks! Does anyone know the usable capacity for the 100 pack?

According to Scan my Tesla, the usable full pack of my car was 94.4kWh when it was fairly new. The full pack was 98.4kWh at that time.

 

[MP3Mike]

Can someone else explain better than me *why* there is no actual zero mile buffer?

Because if there was a zero mile buffer it would reduce the reported range and people would count on it always being there. It is better to have 0 actually be 0. (As best as possible to estimate.)

 

[Shaggy]

Because then people would use it and it meant that 0 wasn't zero. IMHO: it's better to know that the zero estimate is trying to be at zero.

Also, it's an estimate, I believe that anyone who got past 0, got there because the car's estimate was wrong and they were lucky.

 

[sorka]

Because if there was a zero mile buffer it would reduce the reported range and people would count on it always being there. It is better to have 0 actually be 0. (As best as possible to estimate.)

That sounds more like an explanation as to why you shouldn't count on a zero mile buffer rather than there not really being one at all.

 

[sorka]

According to Scan my Tesla, the usable full pack of my car was 94.4kWh when it was fairly new. The full pack was 98.4kWh at that time.
View attachment 286841

So in your screen above, how does that compare with "Pack" capacity in TM-SPY? What is nominal? Is that what it's supposed to be brand new or is that what it's reporting as total capacity of the battery including the 4kWh anti brick buffer?

I've seen other post TM-SPY screen shots of their P100D in the 98.something kWh range in the past yet everyone including the author states that the "Pack" value does not include the 4 kWh anti brick buffer. What app are you using scan your CANBUS?

 

[lymex2018]

I've compared SMT and TM-Spy many times back and forth before. This was my TM-Spy screen at similar time where "Pack" corresponds to "Nominal full pack" of SMT. Also, that SOC=72.0% corresponds to SOC UI of SMT.

One of my friend who also own a X 100D and his value of "Pack" is a little smaller than mine but I cannot remember the exact value. This makes sense since his car is little older than mine, both at less than 4000 mileage.

As for the 4kWh buffer, it seems always equal to Nominal full pack minus Usable full pack in SMT.

 

[sorka]

How do I get SMT?

 

[lymex2018]

How do I get SMT?

I bought it on Google Play.
It's for Android only.

 

[sorka]

I bought it on Google Play.
It's for Android only.

First place I looked but didn't find it hence my asking. Do you have a link?

 

[sorka]

First place I looked but didn't find it hence my asking. Do you have a link?

Whoops. You already have the link

 

[sorka]

OK, so the "nominal full pack" on SMT is the same as "pack". This morning both read 75.3. Interestingly that's down 0.8 from two days ago but the car has been sitting cold for several days so either I had a ton of new degradation or the BMS's idea of what full is can change based on temperature.

The "Usable Full Pack" is just the nominal minus the 4.00 kWH buffer that that SMT reports.

TM-SPY's SOC remaining is from the 75.3. The cars SOC remaining reported on the dash doesn't count the the 4.00 kWH buffer so it goes to zero 4 kWh sooner than TMSPYs SOC.

If rated range ONLY included the "Usable Full Pack" value, which in my case was 72.1 kWh two days ago, then rated range at 100% would only be 240.3 miles (0.3 kWh / mile).

Clearly the Tesla's rated range includes more kwh than what is being reported by "usable full pack" from SMT.

Yet the video:

clearly shows 4.0% SOC on TMPSY when the car shows 0% yet he's able to apply drive power all the way down to TMSPYs 0%.

So either TMSPY's "Pack" and SMT's "Nominal Full" doesn't in include the buffer(which if true, SMT is off by 4 kWH) or Tesla allowed the full 4.0 "Energy Buffer" to be used for driving.

There was one screen shot from someone on a nearly new 85 showing 78 kWH on "Pack" TMSPY (which would have been "Nominal Full" on SMT). Shouldn't a new 85 have shown 81 kWH, the point of this entire thread?

Interstingly, the difference between 78 and 76 (mine) is just about 7 miles which is about the amount of degradation I've shown since brand new according to Tesla's "rated" display.

It could also be that Tesla has dispensed with holding the 4.00 kWH buffer in reserve and the actual shutdown point isn't fixed but rather when the lowest voltage cell dips below some threshold.

 

[AWDtsla]

It could also be that Tesla has dispensed with holding the 4.00 kWH buffer in reserve and the actual shutdown point isn't fixed but rather when the lowest voltage cell dips below some threshold.

Only if they want to pay for full pack replacements. The buffer is there for a reason, because it is possible to "brick" the pack.

For a lot of these behaviors it is better to blame on incompetence than intention. We've all noticed trivial things that have been broken and brought back over the years. Yet another reason not to run your car down to 0 - you're depending on Tesla's software skills to keep the battery from eating itself.

 

[sorka]

Only if they want to pay for full pack replacements. The buffer is there for a reason, because it is possible to "brick" the pack.

For a lot of these behaviors it is better to blame on incompetence than intention. We've all noticed trivial things that have been broken and brought back over the years. Yet another reason not to run your car down to 0 - you're depending on Tesla's software skills to keep the battery from eating itself.

I honestly have to say I didn't understand a word of that

 

[lymex2018]

OK, so the "nominal full pack" on SMT is the same as "pack". This morning both read 75.3. Interestingly that's down 0.8 from two days ago but the car has been sitting cold for several days so either I had a ton of new degradation or the BMS's idea of what full is can change based on temperature.

The "Usable Full Pack" is just the nominal minus the 4.00 kWH buffer that that SMT reports.

TM-SPY's SOC remaining is from the 75.3. The cars SOC remaining reported on the dash doesn't count the the 4.00 kWH buffer so it goes to zero 4 kWh sooner than TMSPYs SOC.

If rated range ONLY included the "Usable Full Pack" value, which in my case was 72.1 kWh two days ago, then rated range at 100% would only be 240.3 miles (0.3 kWh / mile).

Clearly the Tesla's rated range includes more kwh than what is being reported by "usable full pack" from SMT.

Yet the video:

clearly shows 4.0% SOC on TMPSY when the car shows 0% yet he's able to apply drive power all the way down to TMSPYs 0%.

So either TMSPY's "Pack" and SMT's "Nominal Full" doesn't in include the buffer(which if true, SMT is off by 4 kWH) or Tesla allowed the full 4.0 "Energy Buffer" to be used for driving.

There was one screen shot from someone on a nearly new 85 showing 78 kWH on "Pack" TMSPY (which would have been "Nominal Full" on SMT). Shouldn't a new 85 have shown 81 kWH, the point of this entire thread?

Interstingly, the difference between 78 and 76 (mine) is just about 7 miles which is about the amount of degradation I've shown since brand new according to Tesla's "rated" display.

It could also be that Tesla has dispensed with holding the 4.00 kWH buffer in reserve and the actual shutdown point isn't fixed but rather when the lowest voltage cell dips below some threshold.

I believe that 4kWh buffer is reported by the BMS, and comes to existence long time ago (ref: wk057's decipher). However, Tesla modify the algorithms from time to time, and the internal parameters must be modified too resulting sudden range changes on the dash. Therefore, I think this 4kWh is just an old nominal value, no matter what it meant, it stays at 4kWh always. The actual anti-brick buffer or below zero buffer also varies according to factors such as time, pack size and degradation.

"Usable full pack" is of course nominal value too because it is the difference of two values of which at least one of them is nominal. As a matter of fact, this term is not Tesla's, It may be created by the author of SMT to roughly represent the energy gives out when dash SOC from 100% to 0%.

One other thing worth noticing is that the estimation of SOC (no matter which) is very complex and may well introduce errors (ref here).
My theory is, energy (Wh) estimation is based on capacity (Ah) estimation, and Ah is ultimately determined by the open voltage at long rest.

How the pack value is estimated? Many believe that it base on SOC estimation especially when near 0% and 100%. Therefore, pack estimation and SOC estimation are closely related. And, these values are related to temperature either because of the estimation error or the actual change, this is the reason of seasonal effect.

To summarize, energy parameters are not very accurate.

 

[sorka]

Yet another reason not to run your car down to 0 - you're depending on Tesla's software skills to keep the battery from eating itself.

What danger is there to the pack to running it down to 0 miles? As long as non of the cells are exposed to reverse polarity, they should be fine, right?

 

[scaesare]

If rated range ONLY included the "Usable Full Pack" value, which in my case was 72.1 kWh two days ago, then rated range at 100% would only be 240.3 miles (0.3 kWh / mile).

As an FYI for my S85, the rated miles indicator seems to be more like 275 Wh/mi, not the 333 your numbers work out to. (actually 273 seems right on the money).

Mine's not a P version, and I have 19 rims, not 21's, but I'm not sure those would incur a 22% energy usage penalty.

 

[sorka]

As an FYI for my S85, the rated miles indicator seems to be more like 275 Wh/mi, not the 333 your numbers work out to. (actually 273 seems right on the money).

Mine's not a P version, and I have 19 rims, not 21's, but I'm not sure those would incur a 22% energy usage penalty.

The P85D is 300 wh / mile (0.3 kWh/mile). Not sure where you're getting 333 from.

I've never seen any greater energy usage from the 21's vs my 19's. I have both and have driven many thousands of miles on each set. I have the square setup on the 21s rather than Staggered which I'm sure would incur a penalty given the staggereds have wider rear wheels.

 

[scaesare]

The P85D is 300 wh / mile (0.3 kWh/mile). Not sure where you're getting 333 from.

I've never seen any greater energy usage from the 21's vs my 19's. I have both and have driven many thousands of miles on each set. I have the square setup on the 21s rather than Staggered which I'm sure would incur a penalty given the staggereds have wider rear wheels.

I dunno why I read that as 3 mi/KW.

My apologies.

 

[supratachophobia]

Can someone else explain better than me *why* there is no actual zero mile buffer?

Degradation or change in software?

The general concensus is a better battery algorithm/years of tweaking that is able to make sure zero really is zero.