Not saying that you’re wrong, but I read your post several times and could not understand any of your explanations. What are you talking about?
Well, I’ll explain
Let’s unwrap this statement 1st:
“Doing that test and reading the nominal full pack shows that even with 5% degradation or so, the service mode battery test says 99.X % battery health.”
If we begin with the battery capacity of a new Tesla model S (I use this example as I have a lot of data) the BMS reports the new capacity to 99.4 kWh.
Full Pack When New is the ”original capacity”
Nominal full pack is the BMS estimate of the capacity.
Nominal Remaining is the BMS estimate of the energy in the pack
The EPA test protocols are accessible via dis.epa.gov.
The plaid was tested with both 19” and 21”.
This is the 19” test, it delivered 99.371 kWh.
(21” delivered 99.3kWh).
It is safe to say that a brand new Model S Palladium pack has 99.4kWh capacity.
How do you arrive at this “math”? Are you saying that if a brand new battery is X kilowatts, the test actually stops when it fills up X-K kilowatts where K is some constant and if so determines health to be 100%. And if it only gets X-K-Y, then health is 100*(X-K-Y)/(X-K) % where in reality it should be 100*(X-K-Y)/X %? If so where did you get this information?
We can see that a car with for example has 93kWh nominal full pack, out of the 99.4 kWh (93.6% battery health =6.4% degradation) but the battery test in the service menu shows a much higher health value than 93-94%.
I only know about 1 test on the new S/X and the battery capacity is not really known (due to limitations in Tessie).
But for 3/Y I know several battery tests and all show a very high battery health.
Checking these by using the energy app/graph shows a much higher degradation.
The origin capacity in the 3/Y is well known from the same sources as described for the S above.
The energy app(below) is confirmed (by several checks on BMS data via CANBUS) to use the nominal full pack as the base for the calculations of remaining range.
The BMS calculates the range for the whole battery including the buffer. But it uses the displayed SOC (which is lower than the true SOC) so the displayed calculated range on the energy app is correct.
This means that the BMS ’nominal full pack’ can be calculated by backwards calculation.
Average x estimated range x 100 / SOC% = capacity in Watt hours.
(Setting down right must be normal range).
As most people do not have a possibility to check the BMS via CANBUS the way above is the most common way to check the capacity.
Several 3/Y has performed the battery test with for example 97, 98 or 99% as the battery health test where the energy calc was also performed and the energy calc for example showed 74.5 kWh and the health test showed 98%.
98% of 78.8 kWh original capacity for the LG battery should be around 77.2kWh.
So far I have not seen any battery health tests that is even close to the degradation calculated from the BMS estimated capacity vs the original capacity.
It is possible that Tesla use the degradation threshold (where the car still shows full range, but a further reduction causes the range to drop).
As it seems, (as we do only get a health value) the zero degradation capacity according to the health test is lower than the threshold.
One problem with doing a slow discharge is that the battery will perform better (deliver more energy) because of the reduced losses.
The nominal full pack value can most often be delivered by the battery.
For me it has done that every time I tested and did not drive too fast (90kph/55mph is a fair number).
My guess* is that one part of the to good result is the slow discharge in the test that gives a too good value.
( *)I know that a slow discharge makes the battery deliver more)