2021 Prius Prime traction battery: the first year degradation

Here is the first analysis of the traction battery degradation on my 2021 PP. I now have owned the current PP (2021 Ltd) for 14 months. This is my third PP. On this car, I have kept a daily record of kWh used to charge the traction battery for each charge session using a kill-a-watt meter on L1 EVSE. For each charge session, the % SoC charged was also recorded to extrapolate the kWh used for the "full charge" for the session. Using those data, I calculated the monthly average for the kWh used for a "full charge" (0 to 100% on the dash).

Below is the summary of monthly average data with the sample numbers and standard deviation. Plotting the value with SD for the error bar, I can clearly see the decreasing trend. Of note, one month (Dec) value 7.22kWh is higher than any other month with a large SD. This is because the traction battery had the "battery heater" on. For the following months, this "battery heater" function was turned off. No "battery cooler" function was used. Also, "Climate Preconditoing" was never used.

Comparing the Aug 2021 average of 6.71kWh+/- 0.16 and a year later Aug 2022 average of 6.41kWh+/-0.12, the difference is statistically significant (i.e. the error bars do not overlap). Accordingly, the rate of traction battery degradation calculated is ~4.5%/yr.



I don't know how this rate will change after the first year. I have anecdotal evidence to support the idea that the degradation is more pronounced in the first year and somewhat stabilizes after a few years. Here is the graph of similar data taken from my other two PPs. 2017 PP was kept for 30 months, and 2020 PP was kept for only 10 months. The 2020 PP (Red bars) shows a very similar rate for the decreasing kWh for a full charge for the first 10 months as the 2021 PP (Green bars). Unfortunately, I did not keep the record for the 2017 PP (Blue bars) between months 5-16. But after 24 months, the 2017 PP full charge kWh number is similar to 12 mo old 2021 PP data.

 

On current 2021 PP, ~10K miles in 14 months with an average of 2-3 charging sessions/week but not always 0 to 100%. On 2020 PP, I only drove ~3,600miles in 10 months and did very little charging maybe once a month. On 2017 PP, I put ~40K miles in 30 months with a daily full charge almost every night.

 

For this analysis, I used SoC displayed on the dash for all of the data points. The absolute value of SoC level in the traction battery is not important for this analysis, as long as consistently using the same L1 EVSE connected to the same outlet and reading the SoC off from the dash, the comparison from 1 year ago to now is valid. You are correct that the traction battery heater, traction battery cooler, and also climate preconditioning all use the wall kWh while the car is being charged. That's why I turned off the traction battery heater after seeing a big spike in the monthly average in Dec. Using any of those three functions will make the comparison very difficult.

As far as the comparison of the SoC displayed on the dash MID vs the SoC reported by HA, here is the copy of my comment from another thread edited for this tread content:

The MID SoC level and the absolute SoC level (reported by Hybrid Assistant) are perfectly linear relationships between MID 0-100% corresponding to HA 14-84%. Using the colinear points to generate a graph shown below (x) is the value of SoC reported by HA and is the SoC displayed on the dash MID.

This linear function can be written as

y=(100/70)x-20

So, the dash SoC is off from the absolute SoC, but since it is perfectly linear, it makes no difference in the analysis of the full charge which is always 70% of the absolute SoC.