Not that new, I've already refered to that before

First of all, they might have a small point with the formula for calendar aging.

Second, other tests of calendar aging show other data, and in many cases it follows the "square root of time" very good.

Third, the difference between their formula and our "square root of time" is not that big. For our amateur calculations I think the square root works very well.

( I did a backwards BMS calib. recently to try to set my BMS from full range to to reflect the actual battery capacity. I think it worked, and now I have about 79kWh capacity according to my BMS. That is very close to my "square root of time" calcs. Using my logged data of average SOC and average temp I should have around 2.6% calendar aging now. The cycles should have worn around1% by now. This should put my battery capacity at around 79.1kWh if I calculate from 82.1 as the new pack number(once did show 82.0 nominal). My BMS calib put it at 79.0, and also a 0% to 100% charge one week ago indicated both 79.0 nominal and also the difference in nominal remaning was spot on. (79kWh -3.5(buffer))

I do not think the above is a coincidence.

There is a tendency for reserachers to use very high ambient temperatures and then draw conclusions like "we accelerated the calendar aging so each week was like one moth(or year).

In reality very high temperature (specially together with high SOC) kills the batteries. This makes the usual square root of time formula look bad, but when we look at researchers that use reasonable temperature, they do not get that behavoiur. The same is valid for using too high C-rate for charge/discharge cycles.

If you look very closely at the graphs you see that the linear part in their graphs actually show that the test points actually do show a not linear line but the curved line that fit the square root quite good. Also, from my point of view they stopped the test very very early in the "linear part". Way to early to make their statement clear.

There is an older report out trying to find a better formula than "square root of time". If I find it, I'll post the link later.