[Linwood]

I've read conflicting things.

Is there now a definitive conclusion? Will you get cell balancing charging to 80%, or must you periodically charge to 100%?

Let's presume I am recharging always from 30%, and not going really low -- there is a secondary recommendation I have seen that if you go below 20% to occasionally charge to 100%. Let's leave that one separately.

I drive locally. When I cross 30% and get home, I plug in. I never (so far) charge on the road, and don't need the extra miles of 100%. But do I need to charge to 100% anyway occasionally?

Linwood

PS. If it matters a July 2022 build AWD Wind with Tech in warm weather (Florida).

 

[alexw]

Cell balancing is a whole thing, and to the extent there's a related benefit to being at high or low SOC, nobody outside Kia's BMS design team will ever know.

What's more relevant is estimating the pack's state of health, which charging to 100% allows the BMS to do. You should probably heed the manual's recommendation and charge to 100% every so often. My theory is that DC fast charging peak power will eventually get throttled back for safety if the state of health estimation is too old, but I haven't seen any actual evidence this is the case.

My recommendation is that if you never need to charge to 100%, you should still do it occasionally. It probably doesn't need to be once a month, and less frequent is probably fine if you rarely or never DC fast charge.

 

[Lamrith]

Kia states to charge to 100% once a month. That tells me they, like many other mfg, use charging to 100% as an opportunity to balance cells.

I follow similar to OP, but run sub 30% as my daily is only 20mi total. Car is set to 80%, I override to 100% once a month. Follow their recommendations and don't worry about it further.

Sent from my Pixel 3a using Tapatalk

 

[Linwood]

You should probably heed the manual's recommendation and charge to 100% every so often.

That's not exactly what the manual says. It says (in my US version):

"If the high voltage battery charge amount is below 20%, you can keep the high voltage battery performance in optimal condition if you charge the high voltage battery to 100%. (Once a month or more is recommended)"

The prefacing "if" clause is rarely quoted by people recommending the once a month charge.

Is the implication if you recharge at (say) 30% and don't fall below 20% that it is not needed?

I get the issue with SOC calibration being possible only near 100%. But I suspect it would take a LONG time before it becomes so inaccurate that recharging at 30% is risky (and by then I will certainly take a longer trip and charge to 100%).

I'm much more interesting in general battery health, such as keeping cells nicely balanced.

 

[omniphil]

I charge at home all the time to 80%. The cells still get balanced if you leave it plugged in until it stops.
Some times it hits 80% at stops quickly, sometimes it charges to 80% and takes another 30-60 minutes drawing very low amperages as its balancing the cells... (Usually resulting in a 81% or 82% total charge)

 

[dc30307]

Kia states to charge to 100% once a month. That tells me they, like many other mfg, use charging to 100% as an opportunity to balance cells.

That doesn't necessarily suggest anything to do cell balancing. Case in point.

My personal expectation is that the EV6 can/does balance at any charge level. I can't think of any other logical explanation for the fact that it will frequently continue "charging" for a couple hours after reaching the charge set point, while only gaining anywhere from 0-2% during those two hours. I only charge to 100% when I think I might need it, and sometimes that might only be once every few months.

 

[Ucinn]

I have charged to 80%, but now in winter charge to 90% and have never had it trickle charge at any of those percentages. When I charge to 100%, it does slow down for the last 5% of charge and takes a bit extra to reach 100% which I figure is for the balancing. Maybe different charging methods on build date? since they do have a charging TSB out?

 

[Linwood]

Kia states to charge to 100% once a month.

Please, anyone... has Kia actually said that, or only said that in the context of if it is "below 20%"?

I hate to nit pick when the manual may well be a translation, but that phrase in the US manual is pretty clear, unless there is other guidance?

I have not looked at the cell balance directly (I do have a reader, just haven't, and not sure if I did I would know what constitutes good vs bad). I think inferring that charging slowing down (what do you guys do, sit in the car and watch?) is due to cell balancing is possible but hardly definitive.

I was hoping someone had either authoritative guidance from Kia, or actual hard data from measurements?

 

[alexw]

I have not looked at the cell balance directly (I do have a reader, just haven't, and not sure if I did I would know what constitutes good vs bad).

There's not much to see anyway - you can see all cell voltages and determine they're all within 40mV of each other, for instance, but that just tells you their voltages are aligned, not necessarily how their capacities or internal resistance are with respect to each other or what kind of balancing effort went into achieving that. I've read papers that suggest trying to actively balance cells is a fool's errand in the first place and can cause more harm than good. However Kia's BMS does or doesn't do it is opaque and will remain so.

I think inferring that charging slowing down (what do you guys do, sit in the car and watch?) is due to cell balancing is possible but hardly definitive.

Slowing down above 80 has nothing to do with cell balance. Each cell has an internal resistance. You can cram more current into them by increasing the voltage applied, but you can't increase the voltage applied (safely) above the cell's open circuit maximum of (in this pack's case) ~4.15V. 80% is the point at which the "constant voltage" charging phase begins: the pack is held at 4.15V/cell, and what current it draws is how fast it charges from that point on, with a roughly exponential decay until you say "eh it's done" at "100%"

This is a charge curve I measured back when it was warm enough to enjoy the car. We can see that the pack enters the constant voltage phase at 77% BMS SOC (so a little higher than that indicated), at a cell voltage of 4.13. It holds that constant 4.13V for like 10 minutes, when I cut the charge off at 90%.

It's not "balancing" - you just can't force more current in past that point by raising the voltage, without risking pack damage. This is why every lithium ion battery ever charges slower above 80%.

 

[azKiaUser]

I charge at home all the time to 80%. The cells still get balanced if you leave it plugged in until it stops.
Some times it hits 80% at stops quickly, sometimes it charges to 80% and takes another 30-60 minutes drawing very low amperages as its balancing the cells... (Usually resulting in a 81% or 82% total charge)

I routinely charge to 60% and found the odd time it'd be 61%. I thought it was weird...this may be the answer to that...

 

[t88dsm]

Please, anyone... has Kia actually said that, or only said that in the context of if it is "below 20%"?

I hate to nit pick when the manual may well be a translation, but that phrase in the US manual is pretty clear, unless there is other guidance?

I have not looked at the cell balance directly (I do have a reader, just haven't, and not sure if I did I would know what constitutes good vs bad). I think inferring that charging slowing down (what do you guys do, sit in the car and watch?) is due to cell balancing is possible but hardly definitive.

I was hoping someone had either authoritative guidance from Kia, or actual hard data from measurements?

From what I've seen, Kia has said this for other EVs (I think it was the Niro). But the US EV6 manual clearly only states the conditional "below 20%" bit.

I had the same question... see this reddit thread where I asked someone claiming to be an EV battery engineer. I can't vouch for it, but he/she seems more credible than most on the internet.

https://www.reddit.com/r/KiaEV6/comments/zjklew
"Good question!
I’m honestly not sure why a modern battery system is asking for a 100% charge. It’s probably to recalibrate battery management system parameters but the physics-based models and kalman filter lite-machine-learning algorithms that everyone else is using should be more than sufficient.
If they’re being super basic or redundant with their BMS algorithms, yeah, the lower the SOC before charging to 100% the better for prediction purposes.
If instead they’re limiting cell balancing to only above like 98% SOC, starting SOC won’t matter. You’ll maybe be able to recapture a couple miles of true range from being at a slightly higher average voltage without exceeding any cell voltage ceiling, while also not hitting any cell voltage floor as soon."

 

[dc30307]

I think inferring that charging slowing down (what do you guys do, sit in the car and watch?) is due to cell balancing is possible but hardly definitive.

Just to clarify, in my case, I'm not exactly referring to charge slowing down. Charging is essentially stopped, but the process is still active. I don't sit in the car and watch, but do occasionally check on it. I do most of my charging when I'm in the office (free!). I have the app set to notify me when the charge is 10 mins away from beign complete, and again when complete. The chargers at work provide roughly 10% charge per hour. If I open the app at any point in time while the car is charging, it will give me an estimated time of completion, which is reasonably accurate. If, for example, I have it set to charge to 70%, and the app estimates charging will be complete at 3PM, I will get a notification on my phone around 2:50 that charging will complete in 10 minutes. If I check the app, it will be around 68% charged. And it will be at 70% at ~3PM. Maybe 20% of the time, I'll get another notification around 3PM that charging is complete. If I check the car, the lights in the charging port are off. But most of the time, the charging session will continue for another two hours. I don't get the "charging is complete" notification until around 5PM, even though the car will report 70% @ 3PM. The lights in the charging port are still flashing as if it's charging, but if you open the car door and check the dash, it will indicate that it's charging at 0kw. I know from checking on it several times, this is not a gradual reduction: up until it reaches the charge setpoint, it will report charging @ ~8kw, then once it reaches the setpoint, it will drop to 0kw (unless it's one of the ~20% of times where it trully stops charging at the setpoint). During the next two hours of pseudo-charging, it might gain another 1%, or sometimes 2% (barely, if so it will usually drop back down to x1% as soon as I start driving). Or it might spend two hours and gain no reported % gain. This behavior is consistent regardless of setpoint (I've tried them all from 50% to 100%). Most of the time it will continue for two hours after the setpoint is reached, but sometimes it will truely stop at the setpoint. The timing is pretty consistent too... it's not like it goes 15 minutes over one day, and 1.5hrs another day. It either continues for two hours after the setpoint is reached (the majority of the time), or it stops when the setpoint is reached. No in between.

I can't say for certain what it's doing during these two hours. All I can say is, most of the time it does something for two hours after the charge setpoint is reached, but sometimes, for whatever reason, it decides the charging process can be completely shut down once the setpoint is reached.

 

[Linwood]

Thanks for that pointer, it is interesting, though the thread like all such degenerates into anecdotes, and a lot of "I do X and nothing bad has happened" is a lot like "my 2nd cousin's friend's uncle got the vaccine and has a big nose so watch out".

Honestly I only care what you (generic, everyone else "you", not aimed at anyone in this thread) do if you have some underlying engineering aspect fact to share. Anecdotes are just not that interesting to me (maybe it's my lack of social media addiction; yeah, I know, my bad).

I fail to see why Kia doesn't clarify. Though to be fair, if you trust it, the manual is fairly clear, it just does not speak well to what you should do if you do NOT start under 20%.

This whole world is getting more complicated as well as technology changes, and people apply an "everyone knows that" rule to new tech. A friend is getting a Model 3 with LiFePO4, and tesla is pretty pointed in recommending a 100% charge on that regularly. Now, are there downsides to it, is there more considerations, I do not know, but their recommendation is quite clear and unambiguous. Yet if you look around lots of people are saying "never charge over 80% unless on a trip" in response.

To me reports of very slow/no activity while finishing a charge are reassuring -- something is going on, must likely some kind of battery maintenance. That it is happening less than 100% seems reassuring as well.

But honestly, we ought to just get clear, unambiguous guidance from Kia. Unless ("if" predicate and all) we already have.

 

[alexw]

This whole world is getting more complicated as well as technology changes, and people apply an "everyone knows that" rule to new tech. A friend is getting a Model 3 with LiFePO4, and tesla is pretty pointed in recommending a 100% charge on that regularly. Now, are there downsides to it, is there more considerations, I do not know, but their recommendation is quite clear and unambiguous. Yet if you look around lots of people are saying "never charge over 80% unless on a trip" in response.

To that end, see @dc30307's post earlier - BMS Balancing - 80% or 100% needed?

TL;DR from that article: the LiFePO4 charge curve is much flatter, so state estimation integrates a lot more error and seeing the endpoints of the cycle is more necessary for nulling that integrated error.

 

[azKiaUser]

To that end, see @dc30307's post earlier - BMS Balancing - 80% or 100% needed?

TL;DR from that article: the LiFePO4 charge curve is much flatter, so state estimation integrates a lot more error and seeing the endpoints of the cycle is more necessary for nulling that integrated error.

word salad...no idea what this means

 

[Linwood]

To that end, see @dc30307's post earlier - BMS Balancing - 80% or 100% needed?

TL;DR from that article: the LiFePO4 charge curve is much flatter, so state estimation integrates a lot more error and seeing the endpoints of the cycle is more necessary for nulling that integrated error.

That makes sense. My point though was if you search, you see lots of people arguing the new Tesla recommendations are wrong because 80% is how they always did it.

Change comes hard.

 

[dscoti]

I may be going against the grain, but I charge at home, with L2, to 100% every time. It’s low kW (9.6 according to the dashboard and 8.5 according to Car Scanner Pro), and slow usually ~6 hours. I drive 2-3 weeks on that charge. So technically, I’m following Kia’s recommendation. 😜

When I travel farther, I only charge to 80%+/- on DCFC. I don’t think it’s harming the battery at all.

Yes, y’all can throw all kinds of comments my way, that’s fine and your prerogative.

I plugged my 2017 Volt in every single day for 5 years, drove a lot more as I was still working (retired now) and had no ill-effects, no degradation, no worries.I have to believe that the battery technology and maintenance has improved over these 5 years. Yes, the batteries still have the same general materials, but progress marches on.

Besides, with the EV battery warranty being 120 months or 100,000 miles, whichever comes first, I’m covered. We likely won’t own the EV6 for all 10 years, but if we do, we’ll cross that bridge when we get to it. I feel this is just one more unnecessary thing to worry about.

 

[KiaevCan]

Well here's on anecdotal data. Given it's using the same pouch Li-Polymer batteries, it gives us hope with the (I assume) better thermally managed e-GMP cars.

 

[Ucinn]

Please, anyone... has Kia actually said that, or only said that in the context of if it is "below 20%"?

I hate to nit pick when the manual may well be a translation, but that phrase in the US manual is pretty clear, unless there is other guidance?

I have not looked at the cell balance directly (I do have a reader, just haven't, and not sure if I did I would know what constitutes good vs bad). I think inferring that charging slowing down (what do you guys do, sit in the car and watch?) is due to cell balancing is possible but hardly definitive.

I was hoping someone had either authoritative guidance from Kia, or actual hard data from measurements?

Don't sit in the car, just like someone posted above, the car lets me know when I'm down to 10% left. I was a smoker up until a month ago, and could see my EVSE as it was charging and the rate it was charging at, the last couple percent once it hit 98% would slowdown from 9.7 to around 2.1... only when I would charge to 100%, if charging to 80%, it would stay at the same 9.7 until it made it to 80%.