SOH increasing?

[Linus]

So I’m 2 weeks into ownership. I bought a 2016 SV with 21,000 miles on it but the battery has had a rough life. The car is a lease return from Hawaii so it had its fair share of heat. The SOH of the battery was 74% when I bought it. I never would have bought it but the price was such that I could replace the battery and still be into it for what they are selling for in my area with 85% SOH.

Now that I’ve fallen in love with the car I’ve started tracking the SOH daily (I’m an engineer so this is the kind of thing I do). To my surprise the SOH is steadily rising. It is rising at ~0.5% per week. I’m charting to see what will happen.

Has anyone heard of this? The theory I'm going on is it sat for an extended time and got some sort of battery atrophy and now that I’m putting 60-100 miles a day on it, it’s regaining some of its losses.

 

[LeftieBiker]

It happens with 24 and 30kwh packs, usually as the result of quick charging. It also can vary seasonally. Good for you.

 

[WetEV]

It happens with 24 and 30kwh packs, usually as the result of quick charging. It also can vary seasonally. Good for you.

Both driving and charging can change SOH in the 24kWh packs.

Remember that SOH is an estimate, the actual capacity of the battery is hard to measure directly.

 

[cwerdna]

Yep. On all of the above. It most definitely varies a bit on my '13 Leaf. I have no CHAdeMO inlet, so I can't give the numbers a "boost" via that.

These posts may help the OP, a bit.
https://www.mynissanleaf.com/viewtopic.php?p=516444#p516444
https://www.mynissanleaf.com/viewtopic.php?p=497564#p497564
https://www.mynissanleaf.com/viewtopic.php?p=460369#p460369

One can temporarily manipulate SOH upwards a bit by charging to 100% and running to pretty low. If one only charges 80% and doesn't run the battery that low before charging again, SOH tends to fall, eventually.

 

[Linus]

I'll keep monitoring and see where the trend leads.

 

[bobkart]

I've also been seeing SOH increase, but not so much steady as occasional spikes.

I have 110 days of LeafSpy data (also an engineer), and a daily activity log, and have started trying to correlate the SOH increases with activity around the time of the increase. Factors contributing to SOH spikes appear to include:

- driving a lot and/or driving fast
- charging a lot and/or charging fast
- high temperatures
- high SOC

Basically all things we're told to avoid to increase the life of the battery. And most contribute to higher battery temperatures. One of my theories is that this generation of battery is happier around 80F than 60F.

To give you an idea what 'a lot' is for driving/charging, my per-day average is between 10 and 11 miles. Most of my SOH spikes (there are 18 in those 110 days) are under 1%, but I have a couple of standout situations:

- SOH increased from 90.65% to 94.28% (a 3.63% gain) over the course of a three-day road trip that came to 208 miles, including a nearly-one-mile-vertical ascent (Hurricane Ridge in the Olympic National Forest). Turns out *no* L3 charging was involved, but it was the middle of August and I hit 6 battery temperature bars frequently.

- The largest single-day SOH increase was 2.77% (from 89.81% to 92.58%), after a 117-mile (round-trip) drive up to Snoqualmie Pass. Speeds were in the 70mph+ range most of the way up and down Interstate 90. There were two ~10kWh L2 charges and a 14.4kWh L3 charge *after* the ascent/descent, to 100%. I hit 7 battery temperature bars for the first time in nine months of owning the car (during that L3 charge), despite ambient temperatures in the 60's (and raining).

 

[coulomb]

I wonder if it's only battery temperature that causes the estimate of health to go up. The hard driving and hard charging do result in higher battery temperatures, which has the effect of greater degradation rate but also higher performance (lower internal resistance, hence lower losses, and just more available energy period). These effects would only last a few hours, but maybe during driving and charging is exactly the time that the health is calculated.

The high SOC being correlated with increased estimated health could be that the algorithm is a little pessimistic about available energy, and when it sees that a high SOC is actually achieved, that bumps up the estimated health.

 

[bobkart]

I wonder if it's only battery temperature that causes the estimate of health to go up.

I also suspect that . . . it's the only common part of all the ways I've seen that it can be made to happen.

Then the question becomes: is it actually better for your battery, or are the increases artificial and doing those things actually hurts the battery? Seems like it would take two of the same year/etc. Leaf to answer that.

 

[coulomb]

Then the question becomes: is it actually better for your battery, or are the increases artificial and doing those things actually hurts the battery?

My understanding is that heat is always bad for the battery. The fact that you get better performance is just a sad irony. Some heat is unavoidable. Enjoy it while it happens, but IMHO (if the heat only theory is correct), don't try to make the health estimate go up. It will only help short term, and will be harmful long term.

 

[LeftieBiker]

Then the question becomes: is it actually better for your battery, or are the increases artificial and doing those things actually hurts the battery?

My understanding is that heat is always bad for the battery. The fact that you get better performance is just a sad irony. Some heat is unavoidable. Enjoy it while it happens, but IMHO (if the heat only theory is correct), don't try to make the health estimate go up. It will only help short term, and will be harmful long term.

Exactly.

 

[Linus]

Well the SOH number has now dropped two days in a row by around the same amount as previous increases per day. My early hypothesis is SOH value is accurate to around +/- 0.5% and when I first measured it was at the low end. Long term graphing should give a better picture.

 

[nlspace]

Really now-- What the heck does SOH even mean?

Is it a real engineering measurement such as is found in the FSM, and thus has meaning?

or is it just some kind of guess or estimate, or some CAN buss calculation based upon some weird GID units and nebula theory?

There seems to be little or no technical rigor associated with much of the Leaf parameters.

How can anyone know the precision or accuracy of the so-called SOH without knowing what it is and being able to define it in technical or engineering terms?

 

[coulomb]

Really now-- What the heck does SOH even mean?

I think it's pretty clearly the BMS's estimate of remaining usable capacity, divided by the original nominal capacity (this is a LeafSpy setting), expressed as a percent.

That's pretty crisp and well defined. All the ambiguity then goes into what the BMS's estimate of remaining usable capacity is and/or means. However it's calculated, the BMS is presumably in the best position to estimate it. We have to take Turbo3's word that this estimate really does seem to be what various relevant figures are produced from, but I think Turbo3 is pretty confident of that.

Presumably, the BMS watches the battery voltage after long drives; that gives it an estimate of remaining SOC. It knows the Ah expended to get to that point, and from that it can calculate a remaining usable capacity. It must keep these figures, and every 90 days it averages all these and comes up with a new official SOH figure. In between, the best it can do is to use a fixed linear rate of degradation, adjusted for a few factors, which seem to include (possibly exclusively) battery temperature (since battery temperature has a fairly dramatic effect on instantaneous battery capacity).

 

[SageBrush]

Really now-- What the heck does SOH even mean?

I think it's pretty clearly the BMS's estimate of remaining usable capacity, divided by the original nominal capacity (this is a LeafSpy setting), expressed as a percent.

That's pretty crisp and well defined. All the ambiguity then goes into what the BMS's estimate of remaining usable capacity is and/or means. However it's calculated, the BMS is presumably in the best position to estimate it. We have to take Turbo3's word that this estimate really does seem to be what various relevant figures are produced from, but I think Turbo3 is pretty confident of that.

Presumably, the BMS watches the battery voltage after long drives; that gives it an estimate of remaining SOC. It knows the Ah expended to get to that point, and from that it can calculate a remaining usable capacity. It must keep these figures, and every 90 days it averages all these and comes up with a new official SOH figure. In between, the best it can do is to use a fixed linear rate of degradation, adjusted for a few factors, which seem to include (possibly exclusively) battery temperature (since battery temperature has a fairly dramatic effect on instantaneous battery capacity).

This is a very good summary but in the interests of trivia (or maybe not) I'll point out that capacity is also a function of C rate and internal battery resistance. The LEAF knows the internal resistance and I suspect folds it into the SOH reading, but it obviously cannot know how hard the car is driven so battery capacity will always have some measure of uncertainty outside of a standardized test.

All that said, SOH as a reasonable estimate of remaining battery capacity is fine.