Drive battery degradation and replacement.

[STS134]

I went up this climb in Charge mode. Started with 14/16 bars on the gauge. Finished with 8/16 bars on the gauge. The climb consumed ~3.3 kWh (it's about 0.55-0.6 kWh per bar) in about 8-9 minutes, which is approximately a steady state power drain of 23 kW or around 2C. In SoC terms, I think this was about 25% or so.
https://www.google.com/maps/dir/34.9394499,-118.9308095/34.8029275,-118.8772576/@34.859895,-118.926194,12.5z/data=!4m2!4m1!3e0!5m1!1e4

If I had started with the standard 30% SoC, I would have hit 0 bars (25%) VERY quickly. And would probably have dropped to 15% about half way up, at which point, the car would have gone into limp mode.

Not perse. As this was an Interstate, you were probably in Parallel mode most of the time (if not all the time). If so, the car would have reverted to Series hybrid mode as soon as you hit 20% SoC, allowing for more RPM (and thus more ICE power).

When towing a caravan, I often see SoC dropping to 20% in parallel mode and then climbing back up to 22% in Series mode. And so on. This is sort of a secondary hysteresis cycle. Of course, while towing speed is relatively low (60-ish MPH) and so is RPM in parallel mode. Therefor the difference in max power output of the ICE in Series mode versus Parallel mode is rather large. If you were going faster during your climb, the difference may in max power output may have been smaller. But still, it might have kept you out of limp mode.

Except the car literally CANNOT produce more power than it does in parallel mode @ 3750 RPM when it also draws 60 kW from the batteries. I've floored it from a stop before and it of course goes to series mode, then reaches around 83-85 mph, before the clutch engages and it switches to parallel mode. You can literally feel the car lurch forward as you get a surge of extra power (because it's no longer losing energy with the mechanical energy taking the electrical path). That's the real issue. Last time I went up that slope (known locally as "The Grapevine"), there was traffic in front of me that limited my speed to around 70-80 mph. Had there not been traffic, I would have tried to do 90 mph the entire way up, which absolutely means parallel mode only for maximum power.

 

[jaapv]

That last scenario is the only time in five years that triggered a low power warning on my dashboard.

 

[anko]

Except the car literally CANNOT produce more power than it does in parallel mode @ 3750 RPM when it also draws 60 kW from the batteries. I've floored it from a stop before and it of course goes to series mode, then reaches around 83-85 mph, before the clutch engages and it switches to parallel mode. You can literally feel the car lurch forward as you get a surge of extra power (because it's no longer losing energy with the mechanical energy taking the electrical path). That's the real issue. Last time I went up that slope (known locally as "The Grapevine"), there was traffic in front of me that limited my speed to around 70-80 mph. Had there not been traffic, I would have tried to do 90 mph the entire way up, which absolutely means parallel mode only for maximum power.

I did not say "more power from the car", I said "more power from the ICE". Anyway, you could have saved your breath as I already wrote:

Therefor the difference in max power output of the ICE in Series mode versus Parallel mode is rather large. If you were going faster during your climb, the difference may in max power output may have been smaller. But still, it might have kept you out of limp mode.

And indeed at those speeds that does no longer apply. But to be honest, if your use case is driving uphill at 90 MPH, personally I am not really interested in the outcome.

 

[Trex]

Now I stated previously in this thread:

Now my newest PHEV is due for its 1st service in early March 2019 ie a couple of weeks away and I am booked in to get a BMU recalibration (or "Auto Capacity Measured Procedure") and that BMU update done. Will let you know how I go.

Regards Trex.

The BMU update I wrote about above has not been released in Australia yet from what my dealer can see and after two tries at doing a BMU recalibration (or "Auto Capacity Measured Procedure") where the dealer was unsuccessful this has now been escalated up the chain to MMAL.

Now my local dealer has previously done a successful BMU recalibration on my oldest PHEV before, so I do find this interesting. I have my newest PHEV back to drive while they (MMAL) try to sort out a successful procedure which MIGHT include a full BMU reset followed by Cell Smoothing and finally a BMU recalibration (BMU recalibration or "Auto Capacity Measured Procedure").

Will let you know how I go.

Regards Trex.

 

[Trex]

Hi guys,

I bought my 2015 Outlander PHEV in March 2018 with 26,000kms and it read a battery SOH of 84%. Now less than a year later it reads 76.3% with less than 35,000kms. Does this seem acceptable

Gday plugev,

I suppose the first thing I would ask is are you having trouble getting to your destination in EV on your daily commute?

Now we have a transferable 5 year warranty, from memory, before anything else that may apply so you have time on your side IMO.

But only you can decide whether the lack of range is affecting you.

Now if you decide that you want to do something about your drive battery then we can talk about that later but I am pretty sure you will need to get your BMU (Battery Management Unit ) recalibrated to show the true capacity or SOH. Just look back earlier in this thread to see what I am talking about.

Also, the car is due for a 3yr service and I have been quoted $930, $850 and $680 to do the service (in Brisbane, Australia), do these amounts seem reasonable for what looks like a basic oil/filter change, a pollen filter, plus 'looking at' a list of things?

Ok to me those amounts sound excessive but my dealer is in a "town" in NSW compared to a city like Brisbane where they seem to know how to charge more for even a basic service like a schooner of beer.

Now I have just had a 5 year service done on my oldest PHEV (bought new) about 2 weeks ago so I will have a look around and see if I can see the price I paid for the 3 year service a couple of years ago.

Regards Trex.

Thanks Trex,

It doesn't affect my weekday commute which is only 14kms a day. But on the weekend and occasional longer weekday trip, I can't quite reach the destinations I could before without using the ICE. My warranty is until Dec 2020 so I do have time on my side, it is disappointing though to have such poor range when the car hasn't done much milage.

Using some petrol is probably no big deal since the car needs to use some regularly anyway, but I think it affects me because the car drives much better in EV only mode. The ICE is comparatively quite noisy and harsh.

Perhaps significant degradation is inevitable whether the battery gets replaced or not in these cars, as the battery is small and gets cycled so regularly and completely. I guess the problem is the way Mitsubishi sold the product claiming it will make 52kms on EV mode, when in reality after a couple years of use the car can only do 30km or less with ordinary use.

Gday again plugev,

Having a 14km commute I would, if I was you, wait awhile before thinking about a new drive battery and probably just get the BMU recalibrated for the moment to see the real capacity of your battery. The longer you can last with the existing battery the longer into the future a new battery will last.

Now my first PHEV's battery is "worse" than yours and we still get nearly 40 kms, if we drive carefully and not drag race at the traffic lights , in the 60kph zones around here ie EVing around town.

You said "The ICE is comparatively quite noisy and harsh" well this I do not understand unless you are talking about Series mode which is the only time I can easily hear the petrol motor running. If you are talking about Series mode I would suggest being in EV mode (Edit. as much as possible) for the slow bits ie under approx 65 kph and everything approx 70kph and over run the petrol motor in Parallel mode using the Charge or Save buttons on trips out or EV range.

The only proviso I should state on what I wrote above is to make sure the battery is "empty" when you get to where you can grid charge the PHEV.

Regards Trex.

 

[Trex]

Hi again Tytalus,

Just writing this here in my lunch break on a Friday so do not have very long to reply. I own and manage a smallish business that mainly designs and builds specialised machinery.

Now I just want to say this in as nice a way as I can but I think you, and others, are making too big an assumption about the supposedly 95% SOC the drive battery is charged to. I am actually in communication with MMC Japan to try and get answers to these sort of questions right at the moment.

Will let you and all others know how I go.

Now as for this:

"The more I look into this battery and how Mitsubishi have programmed its use, the more concerned I'm getting."

Other have said this sort of thing before and I am not sure why these people buy the PHEV or even keep it but I have far more faith in Mitsubishi's engineers.

Hell, for a very new product (first mass produced PHEV 4x4 or SUV) there have been IMO only some very minor recalls so far (touch wood). Probably just gave it bad luck by saying that. :lol: If something bad goes wrong with the PHEV now, everyone can blame me.

Got to go.

Regards Trex.

Ps There was problem with an assembly process from memory of the drive battery early on in 2013 which held up release of the PHEV to the rest of the world after the PHEV was released in Japan.

Hi Trex,

Similarly, not long to write; about to head off to South London and pick up my newish PHEV

Why am I buying a PHEV, and particularly, an Outlander? Here's are my list of requirements/wants/nices to have, and although now a mute point (as I should hopefully be driving back in one by the time you read this) I hope you'll come to the same conclusion as us:
- We've driven Prius's for over 10 years. Others were doubters of battery tech and saying things like: "You'll have to change the battery in 5 years, they cost more to make, they aren't as green as they say, the performance is terrible, you'll regret it." So my wife (also an electronics engineer) looked at what Toyota had done and realised it was the best of all the worlds at that moment in time (2008): Economy of a diesel without the rubbish (how people ever got convinced otherwise, I never will understand), size and price of an equivalent spec/age Ford Focus C-Max, geeky.
- 10 years later and apart from a terrible run in with 2 Toyota sales staff (the service staff were great) the cars have been great, especially the '04 plate we had for 8 years. Twice the economy of the V40 Volvo we used to own.
- We then started looking elsewhere for a newer car, especially as the Prius's don't like town driving in the cold (35mpg max, although on a good 40 mile run we got 70.1 mpg), each time we go anywhere (about 250 miles every couple of months, or camping in the summer) the Prius is packed and we cant take 3 bikes due to the limits put on the hatch.
- We've got 4kW solar on the house and 3kWh battery storage, so we could go full EV, and this would do us 90% of the time, but what of the bikes and those regular further trips?
- 80% of our journeys are fetching and carrying around town in the evenings after work (I either Work from Home or are on the train) or usual town/family stuff at the weekend, so the car's sat outside the garage anyway during the day.

The Outlander PHEV therefore ticks all the boxes, and quite frankly, at the price point, I can't see anything else that matches it. And with the tow-bar being fitted with a tow-rack for the bikes, it's just about perfect.

So to date (and in the 6 months before buying) I've been grateful for these forums and those on the Interweb and Tubes for all the useful stuff and I feel I know what I'm buying into.

The Prius was a 'Just Drive' car, infact, it says those exact words in the handbook. Yes you can 'hypermile' or press the nearly-pointless EV button, but frankly, you just drive.

However, the post MY16 Outlanders + Watchdog give the likes of you and me much greater control of what's going on. Between mains charging, battery saving, petrol charging, B levels, states of charge, states of health, etc, there is so much more scope. I was just looking for guidance on making the best of it all.

So when I started seeing comments about battery degradation, it peaked my interest as I had got over such comments 10 years previous. Hence why I came and registered here. To date I have:

- Got back up to speed on the latest battery tech
- Reviewed both technical and interest led articles on EV's uses of such batteries.
- Clarified with Daniel what the Watchdog does
- Looked for evidence of Watchdog reports to see if they stack up to what others are saying
- Stumbled across some interesting articles about other pressures on the EV industry, which I will get to in a moment.
- Spoke to an expert in this theatre, purely through timing as I was only aware of the EV Drive-train talk through my Institutes mail-shot last week.
- Got you upset somehow, for which I apologise, I'm just trying to follow a train of thought that I hoped others had followed. Which appears you have.
- Looking to see if there really is an issue, and if so, how to simply fix it with all of the tools at our disposal.

So, to date we have:

- Evidence of battery replacements under warranty due to low Capacities, and at least 2 that I know of. To put this in context, in Roger's talk (Ford EV) he said that there had only been 3 Nissan Leaf battery replacements under warranty in total.
- Lot's of other PHEVs on the Watchdog with worrying SoC/SoH's being reported. Now, I know it takes a lot of work to fully test a battery pack properly, these reports are guesstimates by the BMS and maybe a S/W issue, but when the driver is seeing 10-20km knocked off their range, the driver doesn't care the reason.
- A significant pressure from sales and advertising on engineers to push the cars. This was the interesting article I mentioned: It covered the case that back in 2012ish when EV's were being properly marketed the engineers used to set the cars to the 80% SoC unless the driver opted to add the extra juice to full nominal charge (lets say that 95% mark). But the sales brochures showed the full charge range. So the advertising watchdogs said that if the 80% charge is what you want drivers to go to normally, that's all the range you can advertise. So the engineers are under increasing pressure to remove the setting if it looked bad on advertising. And with the Outlander PHEV being a short EV mode, you can imagine what knocking 20-30% off the range would look like.
- So I have no disrespect to the Mitsubishi engineers, the product is a technical and mechanical marvel, and right up our geek-street and green credentials. But unless Yuasa developed a new type of LE40 that defies all the other Li-Ion chemistries and allows the battery to charge to the apparent 95% without the usual degradation at that level, it will, in my honest opinion, lead to increased battery degradation that would, mile for mile, be greatly reduced by only charging to 80-85%.
- In response to this 95% charge, I think the engineers have set software that is overly pessimistic when the car is driven in certain ways, certain conditions and the battery treated in certain ways. All I have to date is circumstantial I know, such as Andy's experience in Australia in hot conditions and lots of faster driving, in contrast to the experiences of Gary in sub-zero Canadian driving around town.

But that is why I am here. What is the truth (as best we can ascertain), is there an issue, can it be fixed, can we use them a little differently to avoid possible issues or is it all a panic about nothing and I should 'Just Drive'?

Why should I be bothered: This is a once in a lifetime chance where I have the money to buy a nearly new car with a nearly new battery that I want to last the warranty without needing it. Like the '04 Prius.

So I will wait in interest to see what you can find out from the MMC engineers.

Now I really must get that car, especially as this got a little longer than I anticipated!

Cheers,

Tytalus.

P.S. The last specialist machinery I think I built was a working K-9 to take to charity events as Tom Baker :lol:

Hi again Tytalus,

I brought this post here as most of what I wish to discuss in this post is about degradation,SOH etc which is more suitable to this thread's topic.

- Evidence of battery replacements under warranty due to low Capacities, and at least 2 that I know of. To put this in context, in Roger's talk (Ford EV) he said that there had only been 3 Nissan Leaf battery replacements under warranty in total.

Ok that bit about the "3 Nissan Leaf battery replacements under warranty in total" I do not know if it is true or not but I do remember the Nissan Leaf from memory having problems with the heat in Arizona because of no active cooling stated as causing accelerated degradation. Not sure I would be using the Leaf as a example of good drive battery design. Rodger (Ford EV) to me should look up the facts a bit more IMHO no matter what we think of the PHEVs problems.

- So I have no disrespect to the Mitsubishi engineers, the product is a technical and mechanical marvel, and right up our geek-street and green credentials. But unless Yuasa developed a new type of LE40 that defies all the other Li-Ion chemistries and allows the battery to charge to the apparent 95% without the usual degradation at that level, it will, in my honest opinion, lead to increased battery degradation that would, mile for mile, be greatly reduced by only charging to 80-85%.
- In response to this 95% charge, I think the engineers have set software that is overly pessimistic when the car is driven in certain ways, certain conditions and the battery treated in certain ways. All I have to date is circumstantial I know, such as Andy's experience in Australia in hot conditions and lots of faster driving, in contrast to the experiences of Gary in sub-zero Canadian driving around town.

Now that charging up to 95% by the PHEV is still getting a mention I see. So the PHEV charges to approx 4.1v per cell and the chemistry I think is in the cells can be charged to at least 4.2v per cell this is what Electropaedia (https://www.mpoweruk.com/) states under Rechargeable Lithium Batteries:



"The cell voltage is typically 4.2 Volts

Battery lasts longer with partial charges rather than full charges.

Charging to 4.1 Volts will increase the cycle life but reduces the effective cell capacity by about 10%."



Notice that last bit there Tytalus ie "Charging to 4.1 Volts will increase the cycle life but reduces the effective cell capacity by about 10%."

Sounds like we should be saying charging up to approx 90% SOC at the most IMHO.

Got to go.

To be continued.

 

[Darkflow]

OK, so latest update over at youtube unplugged EV is that the BMU calculates SOH by applying just two fixed degradation factors:

-0.3% for every month
-1.0% for every 6200 Ah in/out of the battery.

That's if the BMU is set to think that the starting battery capacity is 38 Ah.

Mine is, and that calculation closely matches my current BMU reported SOH.

Apparently someone has been looking directly at the BMU code, which is not sophisticated.

 

[anko]

OK, so latest update over at youtube unplugged EV is that the BMU calculates SOH by applying just two fixed degradation factors:

-0.3% for every month
-1.0% for every 6200 Ah in/out of the battery.

That's if the BMU is set to think that the starting battery capacity is 38 Ah.

Mine is, and that calculation closely matches my current BMU reported SOH.

Apparently someone has been looking directly at the BMU code, which is not sophisticated.

My battery is 65+ months, so should be down roughly 20% for this.
My Ah in/out is 170.000 Ah, so should be down another 27% down for this.
In total, I should be 47% down. I am 'only' 30% down.

When the battery is regularly charged using A/C chargers, the BMU wil lget a change to recalibrate itself and overrule the pre-programmed degradation params.

 

[zzcoopej]

When the battery is regularly charged using A/C chargers, the BMU wil lget a change to recalibrate itself and overrule the pre-programmed degradation params.

I agree with anko, and further I have observed that running the battery from completely full down to the point where the ICE cuts in (rather than using the ICE intermittently during a trip) obviously gives the BMU an "actual" which seems to help with the guesswork!

However the proof of more than a fixed BMU algorithm is that the BMU SOH value has actually INCREASED on a few occasions in some PHEV, mine included (Jan 18 2016, July 16 2016, Nov 2 2016, Mar 6 2017).

Having said that, since 45Kkm in Oct 2018 I have noticed a steeper decline in SOH in my PHEV, and no more upwards recalibrations since March 2017. Today my SOH stands at 30.3Ah or 80% SOH at 58Kkm. So I've recently lost 2.6Ah in 12Kkm which is about triple the previous rate of degradation, with no significant change in PHEV usage pattern. Not happy.

 

[Darkflow]

So I've recently lost 2.6Ah in 12Kkm which is about triple the previous rate of degradation, with no significant change in PHEV usage pattern. Not happy.

Urk. Hopefully not real either.

With the data set you have available, what happens if you plot all the cars with SOH on the Y axis and a suitable function combining time and +-Ah on the X axis?

 

[Trex]

OK, so latest update over at youtube unplugged EV is that the BMU calculates SOH by applying just two fixed degradation factors:

This is old news. I wrote this previously here http://www.myoutlanderphev.com/forum/viewtopic.php?f=10&t=3487&start=20:

Now anko asked me in this thread:

It has been discussed (if I am nor mistaken) how the Current Ah is calculated based on actual usage parameters and done theoretical model. Could these be 2 of the parameters?

Not sure still looking into it.

Notice that last sentence from anko "Could these be 2 of the parameters?" ie the Secular (calendar or age) and cycle deterioration.

Well I think they are used to "estimate" the current capacity of the drive battery (or SOH) by the BMU. I will bring in a image I found and put in another thread that I started:



Please note the Calculating total amount of battery (including capacity adjustment) using History of battery usage and History of discharge and charge
on the bottom left of the BMU.

History of battery usage I think is Secular (calendar or age) deterioration.

History of discharge and charge I think is cycle (or Cyclic) deterioration.

They are both stored on the PHEV as we use it and are fed into the algorithm that works out the "estimated" current capacity of the drive battery I think.

Regards Trex.

Hope they are not pinching my research.

 

[anko]

When the battery is regularly charged using A/C chargers, the BMU wil lget a change to recalibrate itself and overrule the pre-programmed degradation params.

Not sure I agree with the last part. The ICE start is triggered by the BMU when the SoC reaches a certain level. But that level is derived by the BMU from its current notion of SoH. ISo, the BMU would be learning from itself ...

 

[anko]

OK, so latest update over at youtube unplugged EV is that the BMU calculates SOH by applying just two fixed degradation factors:

This is old news.

In all fairness, I don't think we have ever seen actual percentages per month / Ah for these params, have we?

 

[zzcoopej]

Not sure I agree with the last part. The ICE start is triggered by the BMU when the SoC reaches a certain level. But that level is derived by the BMU from its current notion of SoH. ISo, the BMU would be learning from itself ...

Ah true, I see what you mean, hopefully the SOH would not diverge too far from reality if the chemistry is predictable (which is possibly why hot conditions away from Japan are proving to be unpredictable). Should the ICE be triggered by low cell voltages then?

 

[anko]

Should the ICE be triggered by low cell voltages then?

Would be nice, but impossible from what I understand. Voltage vary with load, which is why it is so difficult to establish real SoC while the car is being used. Instead, the BMU calculates an estimated SoC based on its notion of SoH, starting SoC and coulomb count.

And this is why a well aligned BMU is so important: if the notion of SoH is too low, the ICE will be started too soon and you mis out on range. If the notion of SoH is too high, the ICE will be started too late and you may be putting more strain on the battery than anticipated.

 

[Trex]

OK, so latest update over at youtube unplugged EV is that the BMU calculates SOH by applying just two fixed degradation factors:

-0.3% for every month
-1.0% for every 6200 Ah in/out of the battery.

That's if the BMU is set to think that the starting battery capacity is 38 Ah.

Mine is, and that calculation closely matches my current BMU reported SOH.

Apparently someone has been looking directly at the BMU code, which is not sophisticated.

Now I have already replied to some of this post above about it being old news from my previous research but I will now discuss the so called % values that is mentioned here.

It appears to me they are trying to get formulas to match the values for the Secular (calendar or age) and cycle deterioration as shown on the MUT-III reports.

Now I have at least 3 of those MUT-III reports (that I can find at the moment) for my oldest PHEV which I still own.

From what I can see from my quick appraisal of those reports those values are not always lining up with what I am seeing. For example:

1st report with BMU data list number 396 of 35 months I should see according to their so called 0.3% for every month formulae 10.5 %. Secular (calendar or age) deterioration on my report is 12.83% (.273%/month)

2nd report is showing 47 months with Secular (calendar or age) deterioration of 14.55% (.309%/month) their formulae 14.1% Ok closer

3rd report is showing 59 months with Secular (calendar or age) deterioration of 16.06% (.272%/month) their formulae 17.7%

Now lets get to the cycle (or Cyclic) deterioration and their 1.0% for every 6200 Ah in/out of the battery formulae.

1st report in the same order as above Charge and Discharge intergrated current 47,652 Ah actual report 9.25% their formulae 7.69%

2nd report Charge and Discharge intergrated current 65012 Ah actual report 10.76% their formulae 10.48%

3rd report Charge and Discharge intergrated current 75332 Ah actual report 11.89% their formulae 12.15%

Not sure I trust their formulas so far. though they got close a couple times I suppose. :roll: :lol:

Now as for adding those values together to get the so called BMU reported degradation of our batteries lets see what we get.

Now none of these reports have a BMU recalibration (or "Auto Capacity Measured Procedure") done to them so it will not screw up the measurement.

1st report Battery current capacity 100-(32.7 Ah/40 Ah) x 100 gives 18.25% add their formulas together 18.19%. Ok this is close.

1st report Battery current capacity 100-(32.7 Ah/38 Ah) x 100 gives 13.95% add their formulas together 18.19%. Not so close as seen by the apps.

2nd report Battery current capacity 100-(29.9 Ah/40 Ah) x 100 gives 25.25% add their formulas together 24.9%. Ok close again.

2nd report Battery current capacity 100-(29.9 Ah/38 Ah) x 100 gives 21.32% add their formulas together 24.9%. Not so close as seen by the apps.

3rd report Battery current capacity 100-(26.8 Ah/40 Ah) x 100 gives 33.00% add their formulas together 29.85%. Not so close.

3rd report Battery current capacity 100-(26.8 Ah/38 Ah) x 100 gives 29.47% add their formulas together 29.85%. Close as seen by the apps.

So all in all IMHO it is ok in some parts but it has very mixed results for "apparently someone has been looking directly at the BMU code".

I think that so called person might be having us on and is just guessing or "winging it" probably using just one report IMHO.

But I will check those calcs later to make sure I have entered them correctly as I was in a hurry.

 

[Darkflow]

Thanks for doing thoise calculations.

It does seem there is more to it than that, especially as noted earlier that SOH has been seen to increase at times.

I do wonder though if it is a moving percentage - so a percentage of the previous calculated result, rather than a percentage of the fixed starting capacity. That will give you curves.

 

[anko]

It does seem there is more to it than that, especially as noted earlier that SOH has been seen to increase at times.

I do wonder though if it is a moving percentage - so a percentage of the previous calculated result, rather than a percentage of the fixed starting capacity. That will give you curves.

Or maybe the BMU DOES have a learning capacity and they have missed it?

 

[Darkflow]

Or maybe the BMU DOES have a learning capacity and they have missed it?

If the SOH has been seen to increase sometimes, it must have. I'm wondering if the default degradation calculation is a moving percentage that then gets a step function correction from time to time.

 

[anko]

Or maybe the BMU DOES have a learning capacity and they have missed it?

If the SOH has been seen to increase sometimes, it must have. I'm wondering if the default degradation calculation is a moving percentage that then gets a step function correction from time to time.

My battery never showed an increase of SoH (apart from as a result of DBCAMs). But at the same time, mine was always / very frequently charged via AC. According to Mitsubishi, my BMU should have been rather well aligned all the time and there should be no reason to expect SoH to go up. So, looking at my battery, there seems to be no standard "correction from time to time".