Battery SOH - bragging rights or bullshit?

[mr.Magoo]

Could SOH just be tracking whether balancing is required?

There's dedicated data points for module variation, for both voltage, state of charge and capacity. There's also an extended battery report that can be pulled via FDRS (and maybe PTS - don't remember).

So using SoH as your "outta whack" indicator seems a bit redundant.

But, it's a valid question...
With all these data points, what's the purpose of SoH in the first place.

Perhaps the engineers thought it'd be cool to show it on the cluster just like SoC and they did some magic formula to take all parameters into account, but the warranty department found out about it before they had a chance to implement it because they didn't want to deal with people requesting a full battery replacement after two years just because their SoH showed 99.9%

Sponsored

 

[chl]

I wonder if we might be misinterpreting what is being measured. Could SOH just be tracking whether balancing is required? We have reports of people seeing the SOH drop into the nineties and then later coming back to 100%. This implies that there must be some treatment available for poor health. There’s really no way to bring back lost capacity so a rising SOH wouldn’t make sense if it was directly related to capacity.

The calculated number may change and go up or down due to recalibration, but the real capacity will never actually go up.

The SOH is the present usable capacity divided by the present original capacity times 100%.

So if your battery began new as 98kWh and is now 90kWh that means it's SOH is approximately:

90 / 98 x 100 = 91.8% The capacity is the usable energy expressed as kWh.

Because the truck BMS SOH calculation involves measurements, for example coulombs and voltages by the BMS, there can be errors resulting in 'drift', and as the errors are corrected the SOH calculation will change.

Strictly speaking, the BMS can only estimate the SOC and the SOH because to get the actual numbers, the battery would have to be charged to it's maximum level and then discharged to zero while counting the total energy discharged.

Since we never take the battery to zero, the SOH is an estimate using one of a number of analytical methods, such as the Kalman filter method:

https://www.mathworks.com/help/simscape-battery/ref/batterycapacityestimatorkalmanfilter.html

I don't think Ford has disclosed what method is used in their BMS to estimate SOC or SOH, but I think we assume it is a coulomb counting method with some kind of correction mechanism for drift.

The coulomb counting estimates the SOC (state of charge) of the battery by measuring the current going into and out of the battery during charging and discharging and integrating it over time.

Coulomb counting can be combined with the Kalman filter method for better accuracy.

Balancing is passive and happens during charging.
It involves discharging the strongest cells with the highest voltage until they match the weakest cells so they can reach their maximum capacity.

Some current that goes into the battery will be discharged through internal resistances during that process, producing heat as 'wasted' energy.

That wasted energy has to be accounted for by the BMS in getting accurate SOC and SOH numbers. The Kalman filter method can be used to correct for this.

https://ieeexplore.ieee.org/document/10471421

As the battery ages, the maximum SOC as calculated by the coulomb method counting will decline, and so the will the calculation of SOH as a result.

Inaccuracies in meaurements will occur and lead to an inaccurate SOC and SOH.

When the BMS recalibrates to correct for these errors, the SOC and SOH will change, which is what has been observed.

That does not mean the battery has somehow 'healed' however, only that the calculation has been made more accurate (we hope).

 

[lightning-spark]

I am currently in the process of having one module replaced. Before it went into the shop Car Scanner was showing 100% for the battery state of health. So I don't think it reliably tells you anything about the health of the HVB.

I ended up having two modules replaced. I love the truck but IMO HVB SOH doesn't reliably tell you anything about the health of the HVB.

HVB SOC module variation is what told me that one or more modules were bad. The first image is when the truck went in for repair back then and the second is from last week.

 

[chl]

Ideally the SOH tells you something about battery capacity degradation.

But it does not tell you about bad modules failing unless and until there is a huge drop in SOH.

Other module failure indications should show up before the BMS tells you you SOH has dropped significantly.

One or a few bad cells in a module is not going to change the 'capacity' of the battery that much so the SOH will not show that the module is in trouble.

As noted above the voltage different especially under load is the indication to look for.

From my experience reading about the Leaf battery, the best way to get a sense of the battery's well-being (avoid calling it 'health') is to monitor the voltages and delta in a dynamic test, that is when driving and especially accelerating. That is when problems can't hide.

There are ways to spoof the battery scan readings senn with LeafSpy for example, or any scanner, which unscrupulous Leaf sellers know about, but they fail with an acceleration test.

 

[StevenC56]

How do you obtain this battery SOH information?

 

[RickLightning]

How do you obtain this battery SOH information?

CarScanner app with OBD adapter. OBDLink MX+ or CX.

 

[abcut973]

I had one module replaced back in February. My SOH before that was 98%. When i got the yellow wrench I remember having my SOH down to 86% with voltage and module variations out of wack as shown above. After replacement my SOH went to 90% but my variations were within specs. Since February I've been charging to 100% a couple times per month in order to recalibrate the BMS. My current SOH is at 93.5% now. Car is charging and driving fine. I have 131kw capacity at 100%. Some members are reporting that it takes a while for the BMS to adjust. So I will see how long it will take to get my SOH back to pre module failure if it ever do so.

 

[Sam_oceans]

Nope. Both are NCM. SR starting mid 2023 is LFP.

Hey Rick ... I understand from my e-tech that if a 'V' is in the eighth position of the VIN number - the vehicle has an extended range NMC. Hope this helps.

 

[halfbakedenchilada]

My feeling is that it's probably not bullshit, in the sense that SoH is based on something real (or at least something estimated), but that, in general, EV drivers spend wayyy too much time thinking about it. What matters to me is the experience of driving it. I just got back from a ~2500 mile road trip where I hammered the crap out of it at DCFCs. I've got 30,000 miles on it, and the experience recently was no different than when it was new. What's my SoH? Who cares: my truck still handles days where I fast charge 7 times just fine, never thermally limited. The stops are the same length as they've always been, at least without measuring them with a stop watch.

I know OP wasn't asking about this in particular, but unless you have a problem I say forget it and drive, just like we used to do in our ICE cars.

 

[Grease Lightning]

I've noticed between the Mach-E and Lightning forums a consistent fact - Lightning SOHs are considerably higher, or perfect, as compared to Mach-E SOHs. Is this because of:

1) Better battery technology on Lightnings as compared to Mach-Es?
2) Better BMS on Lightnings as compared to Mach-Es?
3) Different algorithm calculating SOH?

Thoughts?

I will go your your title BS….

but in reality is I think they baked in the buffer to always make our SOH look close to …..

 

[chl]

Just looked at my 2023 Pro SR with CarScanner - 100% SOH and voltage variations was close to zero (0.001 or something even lower than @lightning-spark readings).

Fingers crossed!

 

[mr.Magoo]

Just looked at my 2023 Pro SR with CarScanner - 100% SOH and voltage variations was close to zero (0.001 or something).

Fingers crossed!

Hopefully this isn't a sign of what's to come, but my module variations made a significant shift for the worse a bit over a month ago...

 

[chl]

Hopefully this isn't a sign of what's to come, but my module variations made a significant shift for the worse a bit over a month ago...

Sorry my eyes have trouble reading the small features and numbers on your graph, what is the module variation quantity in volts - it looks like the numbers on the right say there is a 3.2mV delta?

I believe that is a good number.

Not sure what the threshold is for 'trouble' - I read somewhere it'd be around 40mV to 50mV delta.

Some small cell voltage variations can just indicate the limitations of our passive balancing, which happens during charging and even some time after charging they say, and the delta may vary from charge to charge.

I think if your delta continued to get larger rapidly approaching the 40mV-50mV threshold, that would be an indication of a failing cell or cells.

With over 3 years and 31k miles 3.2mV is not bad I think...but maybe others with similar mileage can weigh in and give us some more data points?

 

[chl]

PS: To put my numbers into perspective...

My truck is a 2023 Pro with an SR battery, manufactured in Oct 2023, bought in Jan 2024.
It has under 3500 miles on it.
I never DC charged it.
A couple times it got charged to 100% by accident after an OTA changed things, but for the first 1-1/2 years I charged to 80% at 48A max.
Now I usually charge to 90% at 32A max so I have a bit more in case of a utility outage.
It lives outdoors but I keep it plugged in to my FCSP in cold weather (below 40F).
Mostly do local driving, rarely over 45mph.
---
Always on a 12v battery NOCO GENIUS1 maintainer now since the summer of 2024 when an OTA failed.

Had a small screw in one tire that I patched successfully - other than that, no issues except with the FCSP software setup that was finicky until an update in March 2025 fixed it.

I am set up for PPOB to GENERAC transfer switch for house backup in an outage since June 2025 - but so far haven't had one. It'll be worth the $600 it cost me if it ever happens.

Did some FORscan tweaks for the headlights, chimes, etc. which improved my sense of satisfaction with the Lightning.

 

[mr.Magoo]

Sorry my eyes have trouble reading the small features and numbers on your graph, what is the module variation quantity in volts - it looks like the numbers on the right say there is a 3.2mV delta?

It used to be between 3-15mV but as of a month ago it's between 15-30mV.

Exceeds .030 volt threshold from average, Cell:

I ran that report in January so there was no deviation to speak of, but if I notice it going over 30mV I'll probably run another one just to see what it looks like.

Sponsored