I think my car is lying to me...

[David99]

I did a little more testing recently while I was on a road trip and tried to optimize charge times. I was keeping close track of energy going in and out of the battery on the CAN bus. I noticed when discharging the battery (driving) some of the energy is lost.

Here is an example. The BMS reports I have 40 kWh usable left in the battery. I drive 100 miles and measure the amount of energy I have taken out when I arrive. It shows I have taken out 39 kWh from the battery. Now you would expect there is 1 kWh usable left, right? But in reality, there is 0 left. In other words, aprx 2.5% is lost inside the battery as you discharge it. This isn't surprising at all. A battery has a small internal resistance that causes some losses. 2.5% is actually really good. The efficiency is worse at lower temperatures. The fact that using Range Mode increases battery temperature supports that.

Tesla actually keeps track of this precisely. The BMS measures and keeps track of the entire amount of energy the car has use in it's life. But it keeps track of energy amount charged and discharged separately. This allows one to determine the round trip efficiency of the battery.

What does that mean for range? When the battery is charged fully, the BMS takes the total capacity and divides it by the rated consumption and this gives you the rated range. In normal driving you will have 2.5% losses that occurs over the entire time of the discharge. It means you end up getting 2.5% less range. This definitely confirms the data @supratachophobia has collected.

 

[ran349]

I did a little more testing recently while I was on a road trip and tried to optimize charge times. I was keeping close track of energy going in and out of the battery on the CAN bus. I noticed when discharging the battery (driving) some of the energy is lost.

Here is an example. The BMS reports I have 40 kWh usable left in the battery. I drive 100 miles and measure the amount of energy I have taken out when I arrive. It shows I have taken out 39 kWh from the battery. Now you would expect there is 1 kWh usable left, right? But in reality, there is 0 left. In other words, aprx 2.5% is lost inside the battery as you discharge it. This isn't surprising at all. A battery has a small internal resistance that causes some losses. 2.5% is actually really good. The efficiency is worse at lower temperatures. The fact that using Range Mode increases battery temperature supports that.

Tesla actually keeps track of this precisely. The BMS measures and keeps track of the entire amount of energy the car has use in it's life. But it keeps track of energy amount charged and discharged separately. This allows one to determine the round trip efficiency of the battery.

What does that mean for range? When the battery is charged fully, the BMS takes the total capacity and divides it by the rated consumption and this gives you the rated range. In normal driving you will have 2.5% losses that occurs over the entire time of the discharge. It means you end up getting 2.5% less range. This definitely confirms the data @supratachophobia has collected.

You are right about the internal losses in the battery during the discharge, but the same effect also happens during the charging cycle. So for your example, the total charge/discharge losses would be about 5%, which would be closer to what @supratachophobia was claiming he was missing. For my car, the total charge/discharge losses are about 6%, where charge/discharge efficiency is calculated as:energy extracted from the battery/energy delivered to the battery.

 

[hiroshiy]

Really an interesting discussion, but I think @supratachophobia is correct. I haven't done the whole logging there but:
- you can open TM-Spy (or SMT) on long trips and compare just percent left on the dash and TM-Spy. They are getting different and eventually about 5 percent off, dash being lower.
- you can definitely feel if you drive 15 percent to 5 percent in cold weather. The number on the dash is decreasing faster than usual, compared to 90 to 80 (same 10 percent).

 

[supratachophobia]

Really an interesting discussion, but I think @supratachophobia is correct. I haven't done the whole logging there but:
- you can open TM-Spy (or SMT) on long trips and compare just percent left on the dash and TM-Spy. They are getting different and eventually about 5 percent off, dash being lower.
- you can definitely feel if you drive 15 percent to 5 percent in cold weather. The number on the dash is decreasing faster than usual, compared to 90 to 80 (same 10 percent).

I keep confirming this on multiple trips btw. I can literally go on a drive, arrive at a destination with 11, 7, 12% (any number below 20 is a pretty much guarantee it will do it). Leave the car for a minimum of 20 min, and when I return, it will be 5% less than before. Again, with no HVAC or anything. Now if you try to drive all the way to 0%, you'll see the car just suddenly take away percent from you below 10%, that's super-scary. The service center keeps telling me the battery is healthy and the calibration is normal because the pack is out of balance. HOWEVER, I refuse to believe the pack can be out of balance by 5% after only 1-3 charge cycles. That's crazy.

No, the 5% is just too convenient for it to be random, it's always that 5% for me. And once again, the math adds up perfect for Nominal versus Usable kwh numbers. Tesla is intentionally doing this, and I have no idea how they have gotten away with it for so long. Unless this is just a certain revision of 90kwh batteries perhaps. And if their algorithm has always been "> 80% then Nominal" and "< 20% then Usable", I'm shocked that owners have put up with this, especially if they can do math and have access to the BMS info. It's literally cheating, and it's cost me some pretty stressful and frustrating trips recently.

 

[supratachophobia]

I did a little more testing recently while I was on a road trip and tried to optimize charge times. I was keeping close track of energy going in and out of the battery on the CAN bus. I noticed when discharging the battery (driving) some of the energy is lost.

Here is an example. The BMS reports I have 40 kWh usable left in the battery. I drive 100 miles and measure the amount of energy I have taken out when I arrive. It shows I have taken out 39 kWh from the battery. Now you would expect there is 1 kWh usable left, right? But in reality, there is 0 left. In other words, aprx 2.5% is lost inside the battery as you discharge it. This isn't surprising at all. A battery has a small internal resistance that causes some losses. 2.5% is actually really good. The efficiency is worse at lower temperatures. The fact that using Range Mode increases battery temperature supports that.

Tesla actually keeps track of this precisely. The BMS measures and keeps track of the entire amount of energy the car has use in it's life. But it keeps track of energy amount charged and discharged separately. This allows one to determine the round trip efficiency of the battery.

What does that mean for range? When the battery is charged fully, the BMS takes the total capacity and divides it by the rated consumption and this gives you the rated range. In normal driving you will have 2.5% losses that occurs over the entire time of the discharge. It means you end up getting 2.5% less range. This definitely confirms the data @supratachophobia has collected.

But isn't consumption, consumption? The car knows how much electricity it took to take you 1 mile. Why is that number one thing at one state of charge, but something different at another, lower, state of charge? That is my frustration. Again, with my S, using Tesla's EPA tested consumption, I can not ever drive what the dash tells me I can at 100% SOC. At the end of the discharge cycle, I will have 4kwh less energy available to me than I was told at the start or roughly 5% less available (whatever 4kwh is). And I suppose this would be totally fine if it didn't actually pull the rug out from under you at the end like it does. That's the big kick to the balls right there because the NAV is telling you that you have that range. The car is believing it's own lie!

 

[supratachophobia]

You are right about the internal losses in the battery during the discharge, but the same effect also happens during the charging cycle. So for your example, the total charge/discharge losses would be about 5%, which would be closer to what @supratachophobia was claiming he was missing. For my car, the total charge/discharge losses are about 6%, where charge/discharge efficiency is calculated as:energy extracted from the battery/energy delivered to the battery.

Would losses be calculated in the EPA consumption number?

 

[Sorabh]

Folks.

I am having a hard time following this so was hoping someone could let me know if I got it right.

Using my car as an example (90D) I would expect to get somewhere close to 90kwh of power out of the battery.

Reading this thread that number seems to be 86kwh or thereabouts.

Using actual numbers based on what the car is telling me I am getting between 72 and 75 kwh which is significantly less than the expected 86kwh. I am only looking at he kwh used since last charge on my trip grid.

So based on this my questions would be:

1. Do you trust the kwh used since last charge on the trip grid?
2. Does 75kwh of draw make sense for a 90D with less than 40000km on it?

Any other things I should be looking at?

 

[ran349]

Would losses be calculated in the EPA consumption number?

No, I don't think they are.

 

[supratachophobia]

Folks.

I am having a hard time following this so was hoping someone could let me know if I got it right.

Using my car as an example (90D) I would expect to get somewhere close to 90kwh of power out of the battery.

Reading this thread that number seems to be 86kwh or thereabouts.

Using actual numbers based on what the car is telling me I am getting between 72 and 75 kwh which is significantly less than the expected 86kwh. I am only looking at he kwh used since last charge on my trip grid.

So based on this my questions would be:

1. Do you trust the kwh used since last charge on the trip grid?
2. Does 75kwh of draw make sense for a 90D with less than 40000km on it?

Any other things I should be looking at?

I think I can help you understand what's going on because this thread alone might not give you the full picture. I started this thread based on my observation that I was "losing" about 5% of the battery in front of my eyes when the charge was low. I also was having a hard time achieving the rated range even though I was driving at watts per mile that Tesla was using to calculate rated range at full charge or 100%.

As 90kwh battery, from the factory, only has about 85-86kwh of power in it as you've noted. Subtract from that this "brick protection" of 4kw that Tesla builds into the equation that you are never supposed to use. In other words, at 0% on your dash, the battery still has roughly 4kwh of power remaining, but you can't use it (probably or safely). The two numbers are tracked in the BMS under the names "nominal remaining" and "usable remaining". And the numerical different between the two is always 4kwh.

Now, your total kwh is always decreasing due to battery degredation. I think last I checked, my "nominal capacity" was something like 77.5kwh and my "usable full capacity" was 4 less than that, or 73.5kwh. Quite the far cry from brand new which would have been 85.7 and 81.7 respectively.

My issue is that after observing linear consumption of the battery from 100% to 0%, I found that the numerical miles remaining on the dash isn't liner. The calculation starts off using your 100% "nominal capacity" divided by EPA consumption numbers to give you that numerical miles remaining on the dash. However, at some point on the way to zero (and after 80-90%), the car begins to calculate remaming range with the "usable remaining" number from the BMS, thus being actually accurate in the amount of miles remaining you can drive with EPA watts per mile consumption.

The bottom line here being, again, using the EPA advertised number (and thereby the consumption rate), a driver of a 90kwh battery car could NEVER actually get the EPA rated miles listed on the dash at 100% because the 4kwh brick protection isn't actually available to them to use, yet is advertised as being available to use.

 

[Sorabh]

I think I can help you understand what's going on because this thread alone might not give you the full picture. I started this thread based on my observation that I was "losing" about 5% of the battery in front of my eyes when the charge was low. I also was having a hard time achieving the rated range even though I was driving at watts per mile that Tesla was using to calculate rated range at full charge or 100%.

As 90kwh battery, from the factory, only has about 85-86kwh of power in it as you've noted. Subtract from that this "brick protection" of 4kw that Tesla builds into the equation that you are never supposed to use. In other words, at 0% on your dash, the battery still has roughly 4kwh of power remaining, but you can't use it (probably or safely). The two numbers are tracked in the BMS under the names "nominal remaining" and "usable remaining". And the numerical different between the two is always 4kwh.

Now, your total kwh is always decreasing due to battery degredation. I think last I checked, my "nominal capacity" was something like 77.5kwh and my "usable full capacity" was 4 less than that, or 73.5kwh. Quite the far cry from brand new which would have been 85.7 and 81.7 respectively.

My issue is that after observing linear consumption of the battery from 100% to 0%, I found that the numerical miles remaining on the dash isn't liner. The calculation starts off using your 100% "nominal capacity" divided by EPA consumption numbers to give you that numerical miles remaining on the dash. However, at some point on the way to zero (and after 80-90%), the car begins to calculate remaming range with the "usable remaining" number from the BMS, thus being actually accurate in the amount of miles remaining you can drive with EPA watts per mile consumption.

The bottom line here being, again, using the EPA advertised number (and thereby the consumption rate), a driver of a 90kwh battery car could NEVER actually get the EPA rated miles listed on the dash at 100% because the 4kwh brick protection isn't actually available to them to use, yet is advertised as being available to use.

OK now I get it. Thank you for the excellent summation.

I am not sure how I feel about this but definitely feel the voodoo behind the range calcs and battery capacity should be a lot more transparent.

In either case I am going to start some quantification as well just to boost my understanding. I'll share anything I find interesting.

Thanks again for this thread OP and HAPPY NEW YEAR TO ALL!

 

[BestHand]

The bottom line here being, again, using the EPA advertised number (and thereby the consumption rate), a driver of a 90kwh battery car could NEVER actually get the EPA rated miles listed on the dash at 100% because the 4kwh brick protection isn't actually available to them to use, yet is advertised as being available to use.

Actually EPA numbers are about how much you spend energy per mile and should be calculated in order with some conditions. It has nothing to do with how many miles you can get on full charge. I mean, you cannot calculate EPA just having or guessing your battery capacity and the mileage, so technically EPA numbers can be confirmed I think if you just make an EPA test as it should be done..

 

[supratachophobia]

Actually EPA numbers are about how much you spend energy per mile and should be calculated in order with some conditions. It has nothing to do with how many miles you can get on full charge. I mean, you cannot calculate EPA just having or guessing your battery capacity and the mileage, so technically EPA numbers can be confirmed I think if you just make an EPA test as it should be done..

EPA results can be replicated by any owner, with any car, given the correct conditions, and with certain caveats with regards to battery degradation. That's why we have EPA numbers, so that drivers can do apples to apples commissions. With electric cars, it's even easier because they're range number equates to an algorithm within the car, that never changes. It can't change, because that would result in different EPA numbers.

What does change, is that range remaining at 100%. That number goes down with battery degradation, but the algorithm won't change because that's how they arrive at the number.

The way the software is working now, miles at higher states of charge cost a different consumption rate than at lower states of charge. And it's intentional. All miles should be created equal at all states of charge

 

[supratachophobia]

OK now I get it. Thank you for the excellent summation.

I am not sure how I feel about this but definitely feel the voodoo behind the range calcs and battery capacity should be a lot more transparent.

In either case I am going to start some quantification as well just to boost my understanding. I'll share anything I find interesting.

Thanks again for this thread OP and HAPPY NEW YEAR TO ALL!

If they were transparent, then we would know they were lying for sure.... That would be hugely embarrassing for them.

 

[Brando]

If they were transparent, then we would know they were lying for sure.... That would be hugely embarrassing for them.

Physics - the worst of the liars. Donate to help me get "equal treatment" for all electrons.Time to end "Physical Law" oppression. It is not just electrons that are tired of being pushed around. Man, I got so tired when pushing my electric car.

Why isn't my gas gauge in gallons? or do newer cars have per mile usage and how many miles left calculations? [Mine is an '89.] When 1/4 of a tank shown, I start thinking of my next fill up. Can't just wait until I get home. Damn you electric guys.

 

[supratachophobia]

Physics - the worst of the liars. Donate to help me get "equal treatment" for all electrons.Time to end "Physical Law" oppression. It is not just electrons that are tired of being pushed around. Man, I got so tired when pushing my electric car.

Why isn't my gas gauge in gallons? or do newer cars have per mile usage and how many miles left calculations? [Mine is an '89.] When 1/4 of a tank shown, I start thinking of my next fill up. Can't just wait until I get home. Damn you electric guys.

Telling the owner they can go a certain number of miles under the right conditions, and then changing the rules halfway through is called.... lying.

 

[hiroshiy]

Let's look at this issue @supratachophobia is pointing out, with just two screenshots. I had a chance to go to 1 percent in my car and the first one is TM-Spy showing 6.1% SOC. The next one is my dash showing 1%. This means there's a discrepancy between TM-Spy report and the dash, which we usually see.

Guess what will happen if you charge above 80%? These numbers match. I will post it once I have a chance to make it above 80%, possibly tomorrow.

 

[supratachophobia]

View attachment 365667
View attachment 365668
Let's look at this issue @supratachophobia is pointing out, with just two screenshots. I had a chance to go to 1 percent in my car and the first one is TM-Spy showing 6.1% SOC. The next one is my dash showing 1%. This means there's a discrepancy between TM-Spy report and the dash, which we usually see.

Guess what will happen if you charge above 80%? These numbers match. I will post it once I have a chance to make it above 80%, possibly tomorrow.

And there is the issue I think. If you had this set to miles, above 80% would tell you one amount of miles toy could drive. But you can't actually do that because as you approach zero, that number is recalculated to leave out brick protection as being available.

Thank you for the extra data. It is much appreciated to have observations confirmed.

 

[ran349]

View attachment 365667
View attachment 365668
Let's look at this issue @supratachophobia is pointing out, with just two screenshots. I had a chance to go to 1 percent in my car and the first one is TM-Spy showing 6.1% SOC. The next one is my dash showing 1%. This means there's a discrepancy between TM-Spy report and the dash, which we usually see.

Guess what will happen if you charge above 80%? These numbers match. I will post it once I have a chance to make it above 80%, possibly tomorrow.

But the top chart shows pack =92.7 kWh with 5.3 kWh remaining which is 5.7%. Show I'm not sure what all this really means.

 

[supratachophobia]

But the top chart shows pack =92.7 kWh with 5.3 kWh remaining which is 5.7%. Show I'm not sure what all this really means.

I think it's different for each battery. The 90 should be 4kwh difference. Not sure what the brick protection would be for different battery packs.

 

[supratachophobia]

Just today, I had the car do another "recalculation". SOC dropped 8% to 2%. Luckily I was 1 mile from home and assumed it would pull this crap on me. Despite the service center telling me the battery is healthy and these were just random, infrequent instances of the battery being out of balance. Seems like the battery is ito of balance a lot these days......