Electric battery degradation

When i measured 2.95 kWh from the plug in October 2013 as reported in post #38 the car was 2 months old (well - new).
And my average annual EV distance/ratio is:

(Feb. 2014 to Feb. 2015)

 

I measured 2.97 kWh from the plug in March, but my 2012 was 2.5 years old when I got it. However, it only had about 700 km EV on it when I got it (2% EV), with over 29000 km total driving.

 

Was it March 2014? and if yes, what is your measured energy input these days for the same battery SOC span? What is the km EV reading now?
It is interesting and encouraging to see that your battery seems to not suffer from aging over 2.5 years of (almost) only HV running! Provided your 2.97 kWh was L2 charging, was it?

EDIT
I realize it cannot be March 2014, must be 2015.
The reported 2.97 kWh was L1 charging or L2?

 

I will make a small correction in the next overview I will post. Now I will take this 10437 km as your annual EV kilometers. I plan to post a next version when I have added one or two new cases.

 

Below an overview for 8 cases. Two new persons from the Netherlands provided their data.

The average of the estimated annual loss percentages is now 2.9%.
The numbers are:
1.3%, 2.5%, 2.6%, 2.8%, 3.2%, 3.3%, 3.7%, 4.1%

Below the new graph of the estimated annual loss percentages against the annual EV miles.

 

March 2015. Car was originally registered (by the dealer) in November 2012.

I now have about 3 thousand km on it since when I got it, and a little over half of that is EV. So approximately 1500 km EV. However, I don't know what it takes to charge it now, since I haven't measured it since.

L1. Good point, as I had forgotten to consider that L1 is less efficient than L2, although I don't know difference in efficiency.

 

Does capacity loss relate to the overall number of EV miles made or rather to time duration?

In labtests usually capacity loss of a battery is compared to numbers of charging cycli (charge+discharge). In our case the number of charging cycli has a direct relation to the number of EV miles made. Therefore it is reasonable to see whether we can obtain a coherent graph for overall capacity loss against overall EV miles, based on the data for 8 cases we have until now. The first graph below shows this. However, the coherence is not as good as we would hope. It turns out that the 4 points much lower than the trend line all concern shorter time durations from 1.3 to 1.9 year, and the 3 points much higher than the trend line all concern longer time durations from 2.6 to 3.2. So it seems that the incoherence here is caused at least partly by the effect of time.

The effect of time on degradation is much more difficult to test in a lab. If a test runs in a real time mode then surely the results will come too late to be relevant: in the meantime newer batteries will be available. And to simulate time in a test in an accelerated but accurate manner is very difficult. This means that the effect of time can only be analysed in reality based on data from practice as we can collect here.

The second graph below addresses this. Here again the overall capacity loss is shown but this time against total time duration. Maybe surprisingly, this graph is much more coherent than the other graph. It seems that the effect of time is much more dominant or more systematic than the effect of the number of EV miles. The trend line in the graph has been extrapolated to 5 years, with an estimation that at that point only 85% of the capacity is left. When this really happens, is Toyota offering to renew your battery?

Another indication for the role of time can be found in the graph for annual loss % against annual EV miles in post #85. If the trend line is extrapolated (speculatively) to the left until it reaches the vertical axis, it indicates a loss of more than 2% for cases with 0 EV miles. It would be interesting to know cases in which the number of EV miles is almost 0 to see in how far also then a loss percentage can be observed.

P.S. Note that the insights we think to get here are based on a low number of cases. It would be better when more cases can be included. So, if you have any data about this, please send them to me ([email protected]), or post them. I will handle it in a confidential manner unless you indicate that you want some personal information to be posted at this forum.

 

Post #82 by Eug can shed some light on time degradation:
After 2.5 years and only 700 km EV (2% of total) he purchased the car in March 2015 and measured 2.97 kWh @ 110-120 V (L1).
From many past reports I recall that PCers measured 3.15 kWh (average) on L1 when the car was new.
This can give some indication.

 

Yes, this is interesting. If you estimate the annual loss percentage in this way for this case, it is 2.3% per year. I can also include such data in the graphs.

Are there more cases like this?

 

I have included this new case in the graphs. To not overload the forum with updated graphs all the time I will keep an up to date version of the graphs at this URL:
http://www.few.vu.nl/~treur/PiPbatterycapacitylossgraphs.pdf

 

The charging capacity numbers I provided were solely as reported by L2 ChargePoint charging. I also charge daily using the 120V EVSE at home, but I don't track how much that is using. Is everyone in the chart reporting L2 charging capacity? L1 charging is less efficient, at least in part due to charging time-based losses in the charger (i.e., how much power the charging circuitry is consuming, cooling fans in the charger, etc). I suppose that as long as all of the numbers for a given car are for the same kind of charging, the % changes could still be meaningfully compared.

 

Yes, I assumed that for a given case at least the numbers were from the same type of charger.

 

Great stuff Jan!

I too am fascinated by the time vs use argument of battery degradation.

Example: How would a 2 year old battery with 15k miles on it compare to a 4 year old battery with 15k miles on it?

 

From the current estimations it seems that on the average of our 9 cases one year of time costs about 2.4% capacity loss, so 4 years instead of 2 years would cost about 4.8% extra capacity loss: almost 5%.

Note that such an estimation of capacity loss % due to time passing by may depend much on climate. So, for example, in the very mild sea climate at a latitude of 52 degrees as I have in my own area, the capacity loss will be (much?) less, whereas somewhere in the desert it may be (much?) more. This effect of climate is another challenge to address.

So, I would be interested in a measure for climate (one number) that reflects those aspects of climate that are relevant for battery wellbeing. Are there any ideas about this? Probably temperatures should be incorporated as a main element in such a measure. Would average annual temperature be adequate? Probably not, as the peaks in temperature may be important. So, maximal annual temperatures? Or are strong fluctuations in temperature more relevant?

 

Remember, it is not purely time dependent, the battery is not sitting idle when not in EV use. If you try to eliminate the EV use degradation you are left with degradation due to HV use.

 

Yes, you are right. In fact attributing loss just to time passing by is not a quite correct way of saying, unless the car is not used at all. So, probably in general this type of degradation can better be attributed to HV use of the car instead of just time.

A next question could be whether this HV use component of degradation can be separated from a component that really depends on time only, for example, by taking into account the number of HV miles (km's) and comparing.

 

Southern Californians might find the coast of the Netherlands to be anything but mild. I remember visiting Zeeland in July with daytime temps in the 50s. Brrr.

 

Yes, but batteries know better how to appreciate that!
They are happy that they don't have a sunny and hot car around them.

To avoid misunderstanding, for July day temperatures around 65 to 70 are more normal here, so you were not very lucky.

 

It was an amazing place, weather notwithstanding. Really enjoyed the country.

 

Very good Bisco, I missed that.