Consequences of never charging to 100%?

[Just Lurking]

I know that charging to 100% allows the BMS to reset its understanding of battery charge level (i.e. accurately gauge the charge level of the high voltage battery). I also know that after repeated charge/discharge cycles that don't reach 100%, the battery charge reported by the car and the actual battery charge level will start to drift apart.

I have a 2022 GT which has the NCM battery chemistry and is supposed to be less susceptible to drift than LFP, but it still inevitably happens.

Has anyone quantified the level of drift over time? If I almost never charge to 100%, how bad can the drift get? If I go 6 months, or even a year without charging to 100%, will my battery charge levels be off 10%? 20%? More?

I realize this is probably quite hard to gauge, but I'm curious whether anyone has tried to test this or even if there are any good theories on what would likely happen in this scenario.

Sponsored

 

[hayhayday]

I charge to 100% a few times per year and notice it often charges 3-5kwh after showing 100% suggesting ~5% drift.

 

[Mach-Lee]

Yeah I would say up to 5% capacity error if you don’t change to 100%. Running it down low to less than 10% and letting it sit overnight is also important to do a couple times per year.

100% is important for balancing, but deep charges are also necessary for accurate capacity estimation.

 

[bshaw]

Has anyone quantified the level of drift over time? If I almost never charge to 100%, how bad can the drift get? If I go 6 months, or even a year without charging to 100%, will my battery charge levels be off 10%? 20%? More?

I also rarely charge to 100%. (couple of times per year)

However, the only time you're gonna care if your drift is significant is by going on a longer trip, and not having the capacity available you thought would be there.

You can prevent that situation by setting charge-to-100% about 12-hours before your long trip departure. The car will spend the additional time slowly leveling out the voltages and eliminating any drift.

But yea, if you have to leave on a moments notice, then drift may be misleading you on what's actually available. Plan a buffer and stop sooner for that first DCFC.

 

[azerik]

Yeah I would say up to 5% capacity error if you don’t change to 100%. Running it down low to less than 10% and letting it sit overnight is also important to do a couple times per year.

I've run down to ~10% twice since I've owned it. I just couldn't help but slap it on the charger as soon as I could. I felt like I owed it an apology for going that low lol. (damn that 'never run you tank less than 1/4 full' BS I was fed for years.)

 

[JohnFoxeSheets]

I've run down to ~10% twice since I've owned it. I just couldn't help but slap it on the charger as soon as I could. I felt like I owed it an apology for going that low lol. (damn that 'never run you tank less than 1/4 full' BS I was fed for years.)

Same! I didn't know one should actually leave it in that state a couple times a year...

 

[Mach-Lee]

Same! I didn't know one should actually leave it in that state a couple times a year...

Yeah it needs some time for the cell voltages to stabilize after driving so it can get an accurate read on all of them with no voltage drop. It will then calculate updated capacity. After a drive I was down to 4% and parked it, when it came back a few hours later it was up to 7% magically. That shows there was some capacity error.

Never going below 10-20% makes it really hard to estimate the SoC at the bottom of the pack. If you take a road trip and you're trying to stretch it below 10%, you may get some really odd behavior down low, such as suddenly running out of charge with some % left on the display. That's been a problem with Ford EVs (Kyle from OOS ran out in a F-150 2-3 times, he's never done that in any other EV). Ford has some work to do to make the capacity estimation at the bottom of the pack more accurate and more robust to eliminate sudden shutdowns and turtle modes for those that stretch it, especially when the pack is not run low on a regular basis. They had too many safeguards in place that end up causing problems with the way people actually charge and use their cars.

The "very occasional" road trip situation is difficult from an engineering perspective because you won't have accurate capacity data if all the pack charging deltas are very small. There needs to be more focus on dynamic capacity detection as the battery is run down, and there have to be sufficient bottom buffers to absorb let's say 3-5% of capacity error without inconveniencing the customer with an early power cut. I would handle that by adding a dynamic bottom buffer that expands when the estimated capacity error increases (and shrinks when you have accurate data so you get more range). Things also get worse with time as the cells age and diverge from each other increasingly more in capacity. You're limited in range by the weakest cell in your pack.

 

[Jeff-NoVA]

damn that 'never run you tank less than 1/4 full' BS I was fed for years.

Well yeah it's common sense; you don't want your motors picking up those contaminated electrons in the bottom of the battery pack!

 

[ack154]

Well yeah it's common sense; you don't want your motors picking up those contaminated electrons in the bottom of the battery pack!

As long as you swap your electron filters regularly, this shouldn't be an issue.

 

[Bruboy]

Very informative videos:
NMC:
LFP:

 

[VaporTrails]

Well yeah it's common sense; you don't want your motors picking up those contaminated electrons in the bottom of the battery pack!

The potential for a good joke was right there!

 

[Garyswar]

As long as you swap your electron filters regularly, this shouldn't be an issue.

I look for the premium electron filters (the capacity doubling kind). They can be tough to find but worth it.

 

[ralteredstates]

Yeah it needs some time for the cell voltages to stabilize after driving so it can get an accurate read on all of them with no voltage drop. It will then calculate updated capacity. After a drive I was down to 4% and parked it, when it came back a few hours later it was up to 7% magically. That shows there was some capacity error.

Never going below 10-20% makes it really hard to estimate the SoC at the bottom of the pack. If you take a road trip and you're trying to stretch it below 10%, you may get some really odd behavior down low, such as suddenly running out of charge with some % left on the display. That's been a problem with Ford EVs (Kyle from OOS ran out in a F-150 2-3 times, he's never done that in any other EV). Ford has some work to do to make the capacity estimation at the bottom of the pack more accurate and more robust to eliminate sudden shutdowns and turtle modes for those that stretch it, especially when the pack is not run low on a regular basis. They had too many safeguards in place that end up causing problems with the way people actually charge and use their cars.

The "very occasional" road trip situation is difficult from an engineering perspective because you won't have accurate capacity data if all the pack charging deltas are very small. There needs to be more focus on dynamic capacity detection as the battery is run down, and there have to be sufficient bottom buffers to absorb let's say 3-5% of capacity error without inconveniencing the customer with an early power cut. I would handle that by adding a dynamic bottom buffer that expands when the estimated capacity error increases (and shrinks when you have accurate data so you get more range). Things also get worse with time as the cells age and diverge from each other increasingly more in capacity. You're limited in range by the weakest cell in your pack.

I recently started the 40-70% maintenance charging. I'm a little confused with going below 10% for maintaining accuracy in range. I get the idea but if the accuracy is off, then I'd be very concerned about "running out" trying to get it to and below 10%. Playing chicken with the range isn't my idea of fun.

 

[rreddy3]

I look for the premium electron filters (the capacity doubling kind). They can be tough to find but worth it.

I found a bunch of them on sale at Tractor Supply in Manassas, go figure ….

 

[Reality Check]

Very informative videos:
NMC:
LFP:

The graphs in the LFP video comparing voltage vs. state of charge curves show how & why SOC can more easily and accurately be determined by voltage in NMC battery packs.

The voltage curve for LFP batteries has such a wide flat section of relatively unchanged/stable voltage across the middle range of SOCs that it’s difficult to determine SOC solely from voltage.

What the voltage curve for NMC shows is that voltage can easily be mapped into SOC without ever needing to fully charge or discharge. So NMC Mach-Es shouldn’t require doing that and suffering the accelerated battery life degradation that goes with it.

Sponsored