Fastest 90% SOC short term “shore power” charging strategy?

[icyroads]

What are some good strategies for charging a battery bank to 90% SOC while connected to shore/utility power for a short time, eg a few hours?

I have Eve LF280’s in a 4s (12v) configuration with a Renogy 3000 inverter capable of charging up to 75 amps (ie 1/4C) and a JK-BMS managing the batteries with active 2A balancing set to start at 3.32v and a 0.05v cell difference. OVP is set at 3.55v.

This inverter has built in lithium profiles, which work, but they seem to leave quite a bit on the table when charging at this rate.
These profiles reach the top voltage at this charge rate and quit charging when the batteries are only at 50-60% SOC (using a coulomb counter instead of only trusting the BMS).

So, I’ve been playing with a custom profile.
I’ve noticed that to really get the amps going, the charging voltage must be at least 13.6v. But that still leaves a lot on the table since it gets to that voltage even sooner, leaving the batteries at only a 20-30% SOC. Adding on a float charge (the Renogy has a time limited float charge if I understand the manual correctly) at 13.2v helped a tiny bit. Bumping up the float charge to 13.4v helped a lot, but was “slow”. I only want to get to 90% SOC, but the bulk charge only gets to 20-30%, which leaves a lot for the float charge to slowly fill up (within whatever time limit Renogy uses).

So far, the best that I’ve been able to get is to bulk charge to 13.8v, which seems to get the batteries to 70% SOC (using a coulomb counter), then the float charge at 13.4v seems to be able to get to 90% SOC over the next few hours. 13.8v seems to be harder for the batteries to reach at 0.25C, and the BMS has no trouble keeping the batteries balanced. I’ve tried charging at 14v and higher, but at this voltage the BMS sometimes trips when an individual cell gets above 3.55v — it still recovers quickly enough and works, but I’d rather not let the batteries get charged that high, and charging will likely be fastest by not tripping the BMS.

So, is that a reasonable strategy to achieve 90% SOC as quickly as possible when on shore/utility power?

What other strategies/tactics/settings might work?
Eg, I have not yet tried lowering the charging amps in the hope that a lower rate will prevent the voltage from peaking above the threshold as quickly, allowing a larger amount of amps to flow in before reaching a max voltage?

 

[sunshine_eggo]

What are some good strategies for charging a battery bank to 90% SOC while connected to shore/utility power for a short time, eg a few hours?

Plug it in. Charge to a target voltage. Stop charging when you get to 90%.

I have Eve LF280’s in a 4s (12v) configuration with a Renogy 3000 inverter capable of charging up to 75 amps (ie 1/4C) and a JK-BMS managing the batteries with active 2A balancing set to start at 3.32v and a 0.05v cell difference. OVP is set at 3.55v.

Why? There is no reason to begin balancing at 3.32V. Balancing below 3.40V is essentially useless and can actually be counterproductive, though not typically with the 50mV deviation you allow. Regardless, you should set your balance start @ 3.40V and reduce to 20mV.

This inverter has built in lithium profiles, which work, but they seem to leave quite a bit on the table when charging at this rate.
These profiles reach the top voltage at this charge rate and quit charging when the batteries are only at 50-60% SOC (using a coulomb counter instead of only trusting the BMS).

So, I’ve been playing with a custom profile.

I’ve noticed that to really get the amps going, the charging voltage must be at least 13.6v. But that still leaves a lot on the table since it gets to that voltage even sooner, leaving the batteries at only a 20-30% SOC.

Yep. that's only 3.40V/cell.

Adding on a float charge (the Renogy has a time limited float charge if I understand the manual correctly) at 13.2v helped a tiny bit. Bumping up the float charge to 13.4v helped a lot, but was “slow”. I only want to get to 90% SOC, but the bulk charge only gets to 20-30%, which leaves a lot for the float charge to slowly fill up (within whatever time limit Renogy uses).

You need to be flexible with this target.

So far, the best that I’ve been able to get is to bulk charge to 13.8v,

Only 3.45V/cell

which seems to get the batteries to 70% SOC (using a coulomb counter), then the float charge at 13.4v seems to be able to get to 90% SOC over the next few hours. 13.8v seems to be harder for the batteries to reach at 0.25C, and the BMS has no trouble keeping the batteries balanced. I’ve tried charging at 14v and higher, but at this voltage the BMS sometimes trips when an individual cell gets above 3.55v — it still recovers quickly enough and works, but I’d rather not let the batteries get charged that high, and charging will likely be fastest by not tripping the BMS.

It is concerning that you're unable to attain peak charge voltage without OVP. Suggests either your cells are poorly matched, degraded, or your balancing scheme is counterproductive.

You should raise your OVP triggers to those required by the cells, i.e., 3.65V.

So, is that a reasonable strategy to achieve 90% SOC as quickly as possible when on shore/utility power?

Fast charge of LFP: 3.50V-3.65/cell, short absorption time.
Slow charge of LFP: 3.40-3.45V/cell, LONG absorption time.

Specifically targeting 90% is essentially impossible since LFP can be fully charged anywhere between 3.40-3.65V/cell with absorption time being the variable.

Recommend:

Fix your BMS settings
14.0-14.4V absorption
Absorption time as needed to reach desired SoC (you have to be flexible on your "90%")
13.5V float.

I'll say this explicitly, to make sure it's not missed... You can't target a specific voltage and achieve 90% SoC.

 

[icyroads]

I understand that SOC cannot be determined by voltage on LifePO4, which is why I took extra care to say that I was measuring with a coulomb counter.

Still, the chargers that I have access to charge to a voltage, so perhaps that’s a source of confusion.

 

[icyroads]

Fast charge of LFP: 3.50V-3.65/cell, short absorption time.
Slow charge of LFP: 3.40-3.45V/cell, LONG absorption time.

I think this helps to summarizes what I’m trying to do .
I want to fast charge when the inverter is on shore/utility power.
I want to slow charge (I don’t have much of a choice) when on solar.
I’m using the BMS to very conservatively protect the batteries.
3.55 really is as high as I want a cell to go.

 

[icyroads]

Specifically targeting 90% is essentially impossible since LFP can be fully charged anywhere between 3.40-3.65V/cell with absorption time being the variable.

And yet, I regularly get there in a few hours when I use 13.8v (3.45v) for bulk and 13.4v (3.35v) float.
But, are there other ways that it could it be faster, safer or easier?

 

[icyroads]

It is concerning that you're unable to attain peak charge voltage without OVP. Suggests either your cells are poorly matched, degraded, or your balancing scheme is counterproductive.

You should raise your OVP triggers to those required by the cells, i.e., 3.65V.

Yes, I worried about that too. But, it’s always a different cell, the cells are still all within 0.05v. Could be the balancing scheme, but then even down at 3v, no balancing and charging at 1/4C the cells are never more than 0.08v apart, and they typically stay within 0.02v, well with the balancer’s capacity when it finally does kick in at 3.32. Raising the balancer trigger to 3.4v shouldn’t really change things for me, but I’ll try it.

What I think is going on could be called the opposite of voltage drop under a load. Ie the voltage will temporarily raise above the actual voltage as it is being charged, depending upon how quickly it can adsorb amps and the rate that it is getting charged.
For whatever reason, it seems to raise more for me than expected, which seems to interfere with nominal LFP charging.

 

[icyroads]

ugh

 

[icyroads]

You should raise your OVP triggers to those required by the cells, i.e., 3.65V.

Isn’t that too late? Couldn’t the voltage get higher by the time charging is disabled?
I picked 3.55 because 1) it’s less than 3.65, 2) I tried 3.60, and 3) 3.55 seems to work well for me, plus I’ve seen it recommended as a conservative setting so it must be working well for others too, specific to their situation and goals. I understand that it’s not a universally “best” setting.

 

[400bird]

The BMS is supposed to be the cells last line of defense. Set it at the cell limits. If the cells end up hitting 3.66 or 3.68 for a moment, it's not the end of them. Have you calibrated the BMS cell voltage readings? The BMS reading could easily be more that 0.05 volts off from reality.

To charge quickly you need more amps (charge current). Not much else to it. Is the question on how to stop at 90% SOC?

 

[Suijkerbuijk]

My bms on 3.65 cell volt i have warning .
On the moment it hit 3.70 volt it shutdown the charging
On 3.40/3.45 volt the bms balance the cells .

 

[icyroads]

The BMS is supposed to be the cells last line of defense. Set it at the cell limits. If the cells end up hitting 3.66 or 3.68 for a moment, it's not the end of them. Have you calibrated the BMS cell voltage readings? The BMS reading could easily be more that 0.05 volts off from reality.

To charge quickly you need more amps (charge current). Not much else to it. Is the question on how to stop at 90% SOC?

Yes, the cell voltages match what the BMS reports.

I am able to get to 90% SOC now in a few hours charging at 75 amps to 13.8v, then float to/at 13.4v (seems to vary between 2-30 amps). That is acceptable to me. However, is it unsafe in some way, or are there faster or easier ways?

 

[icyroads]

If the cells end up hitting 3.66 or 3.68 for a moment, it's not the end of them.

Thank you, that’s good to know.

 

[Texas-Mark]

I am able to get to 90% SOC now in a few hours charging at 75 amps to 13.8v, then float to/at 13.4v (seems to vary between 2-30 amps). That is acceptable to me. However, is it unsafe in some way, or are there faster or easier ways?

The only faster way is to push more amps.

 

[sunshine_eggo]

Yes, I worried about that too. But, it’s always a different cell, the cells are still all within 0.05v. Could be the balancing scheme, but then even down at 3v, no balancing and charging at 1/4C the cells are never more than 0.08v apart, and they typically stay within 0.02v, well with the balancer’s capacity when it finally does kick in at 3.32. Raising the balancer trigger to 3.4v shouldn’t really change things for me, but I’ll try it.

It absolutely will change things for you. 3.40V is well into the upper "leg" of the charge curve and cell voltages represent SoC better and makes balancing more effective. Balancing isn't about voltage. Balancing is about SoC. You want all cells at 100% SoC at the same time.

What I think is going on could be called the opposite of voltage drop under a load. Ie the voltage will temporarily raise above the actual voltage as it is being charged, depending upon how quickly it can adsorb amps and the rate that it is getting charged.

You're not seeing the opposite of voltage drop, you're seeing voltage drop in the opposite direction, i.e., there is voltage drop from the charger to the battery. The battery is a load to the charger.

Isn’t that too late?

No. As has been pointed out, brief over-voltage is not harmful.

It helps to know that most "dumb" BMS don't engage protection to 3.75V. Some server rack batteries don't engage until 3.9V.

Electrolyte doesn't begin to break down until over 4.20V.

In the early days of LFP, they operated from 2.0-4.2V. When life was not as expected, manufacturers revised that to 2.5-3.65V, lost 10-15% capacity and gained substantial cycle life.

These things are not delicate flowers waiting to explode the moment you go out of spec.

Couldn’t the voltage get higher by the time charging is disabled?

Yes.

I picked 3.55 because 1) it’s less than 3.65, 2) I tried 3.60, and 3) 3.55 seems to work well for me, plus I’ve seen it recommended as a conservative setting so it must be working well for others too, specific to their situation and goals. I understand that it’s not a universally “best” setting.

You've already discovered that it interferes with your ability to fast charge to conservative voltages. This is precisely why you don't use conservative settings on your BMS. You get your battery balanced and then set your hardware to work inside the BMS limits, so that it's never triggered unless something is actually amiss.

I want to slow charge (I don’t have much of a choice) when on solar.

13.8V, 2-4 hour absorption
13.5V float

However, if solar is limited to lower current, there is no harm in using the "fast charge" settings you land on.

 

[Alkaline]

you will have a hard time balancing cells if you only charge to 90%...

 

[icyroads]

you will have a hard time balancing cells if you only charge to 90%...

True. The “fast charge” on shore power is to quickly top things up. I will use “Slow charge” via solar to get beyond that and engage balancing then when it happens to be a sunny enough day.

 

[icyroads]

These things are not delicate flowers waiting to explode the moment you go out of spec.

Thank you. This is a relief to know.

 

[icyroads]

Update: okay, I upped the “fast charge” voltage to 14v, upped OVP to 3.65v and OVPR to 3.60v, and dropped float down to 13.2v. These settings get the SOC up to 97% (measured via amps / coulomb meter, not voltage) even before float begins. So, more than I want, but one thing that I discovered is that one cell is unbalanced. Below 3.5v, this cell behaves like the others, and I get the same amps out of it without its voltage deviating from the rest of the pack, even below 2.8v. Above 3.5v and it starts to be on low side, regularly 0.1v below the rest of the pack.
So. I’m just going to leave if there for a while to see if the balancer can do its magic, and whether this cell is merely unbalanced or perhaps has a problem above 3.5v (I can live with that, I think, since I get more than 280AH out of it and the other cells even if they never go above 3.5). Either way, it helps to explain why OVP was being triggered.

 

[sunshine_eggo]

It's simply imbalanced.

To encourage balancing, you might want to set float to 13.8V. That will push you to 100% for certain, and you don't want to leave it that way for any longer than is necessary. Your active balancer should bring things into balance pretty quickly.

 

[740GLE]

If you want fast charging it’s key to have balanced cells.

No you don’t need to have the cells perfectly balanced each charge but if you hold voltage at 13.8v for few hours each week it allows to bring those bad cells into check little by little.

 

[icyroads]

Update: rebalanced the cells up above 3.5v. There was one cell that was way out of balance, even though it behaved the same as the other cells when below 3.5v.

After balancing, this pack would charge up to at least 14.2v (3.55v) without any OVP (>3.65v) at 70+ amps, and even (once) with balancing turned off.

So, even though I thought it was good, the insights and experience of this group led me in a better direction and gave me a much better understanding of what’s going on.

After more testing, the built in lithium profiles now work fine, but I’ve still settled back to 13.8v bulk (3.45v per cell) and 13.4v (3.35v) float for “ASAP fast charging” on utility/shore power. That’s all I need to get >=90% SOC (using a coulomb counter for more than one test) at a 70 amp charge rate in a few hours. The float at 13.4v doesn’t really do much other than let the pack soak up a few more amps and supply power to vampire loads for a time (instead of from the battery). Perhaps it’ll be enough to also top off the starting battery once I connect a dc-dc charger — another test for another time.

I’ll rely upon solar charging to periodically get the cells balanced above 3.5v. If that doesn’t work out I can simply change the charging profile of the inverter/charger when on utility/shore power. I’ll see how it works out this winter (at 45 degrees latitude).

Thank you for your help!