LFP memory effects thread

[Cpt Pat]

Originally Posted by tanglewood

So I can see how if a charge protocol terminates solely on voltage, and does so immediately after the target voltage it attained, you will have charged your battery to a slightly lower SOC than you thought.

The absorb time to achieve this could be strictly time based, or it could be return current based. Either way as long as it gets you past the hump, you are back on the normal charge curve.

What am I missing?

You're not missing anything except that, as the terminate-charge-on-target-voltage practice continues through successive cycles, the bump gets worse.

The "return current based" method: counting amp-hours taken out and amp-hours returned, is just another description for coulomb counting. (A coulomb is the equivalent of one ampere-second -- an amp-hour is simply 3600 coulombs.) This is the method I am using, and what is being vigorously opposed by common mythology - because the terminate on voltage method is cheap, conceptually simple, and conforming to the behavior of lead-acid batteries.

"For every complex problem there is an answer that is clear, simple, and wrong."
- H. L. Mencken

I've seen many reports that the loss of capacity caused by the flawed charging process is permanent. Since I've avoided the problem, I have no personal experience with it. It could be that those experiencing the problem are just repeating the flawed voltage-based termination method, over and over.

“The definition of insanity is doing the same thing over and over again, but expecting different results.”
- Albert Einstein

Considering the cost of LFPs and the pride of early adopters, I imagine there are some seriously bruised egos out there. No one enjoys a failed experiment, and its natural to blame the messenger -- but so goes scientific progress. It's a new technology, and there are bound to be failures.

[evm1024]

Originally Posted by Cpt Pat

You're not missing anything except that, as the terminate-charge-on-target-voltage practice continues through successive cycles, the bump gets worse.

That is a nice assertion - please cite the papers that show this to be a fact.

The "return current based" method: counting amp-hours taken out and amp-hours returned, is just another description for coulomb counting. (A coulomb is the equivalent of one ampere-second -- an amp-hour is simply 3600 coulombs.)

Yes, a coulomb is the transfer of charge. 6.242^18 units of charge. And as you note an ampere is 6.242^18 units of charge per second.

This is the method I am using, and what is being vigorously opposed by common mythology - because the terminate on voltage method is cheap, conceptually simple, and conforming to the behavior of lead-acid batteries.

"For every complex problem there is an answer that is clear, simple, and wrong."
- H. L. Mencken

I'm not sure where you see vigorous opposition to using SOC (coulomb counting) as one of the parameters in the end of charge set point. In my own case I know about how long it will take to restore the charge used and my BMV-602S reports SOC therefor if charge were to terminate before expected is would be obvious.

Perhaps, you are misunderstanding an opposition to using coulomb counting as the only parameter to define fully charged. Crusaders are often opposed.

I've seen many reports that the loss of capacity caused by the flawed charging process is permanent. Since I've avoided the problem, I have no personal experience with it. It could be that those experiencing the problem are just repeating the flawed voltage-based termination method, over and over.

“The definition of insanity is doing the same thing over and over again, but expecting different results.”
- Albert Einstein

It appears that you are caught in a positive feedback loop. And we all know what positive feedback does. I've bolded your most important statement so far. And underlined the key to your positive feedback. You assume that your actions have kept you from capacity loss.

In my case I bought discount cells that may have included "seconds". Who knows? In any case if as you so strongly state "mischarging" is to blame one could reasonably expect all 4 cells to suffer. That is not the case in my case and also not the case in a few of the other banks I am aware of that have 30% +- capacity loss. BTW - they won't post here in many cases because of the responses that we have seen. (see your last paragraph)

Considering the cost of LFPs and the pride of early adopters, I imagine there are some seriously bruised egos out there. No one enjoys a failed experiment, and its natural to blame the messenger -- but so goes scientific progress. It's a new technology, and there are bound to be failures.

I'm not sure why you come to this conclusion. It sounds like there is a lot of ego but not with the early adopters.

Pride goeth before a fall.

Lastly - this thread was to report on efforts to recover capacity by those who have banks that have reduced capacity. By your own report you do not suffer from reduced capacity. Thus, your posting here has not furthered the thread.

[john61ct]

Originally Posted by Cpt Pat

The "return current based" method: counting amp-hours taken out and amp-hours returned, is just another description for coulomb counting.

afaik, "return current" has nothing to do with your preferred coulomb-counting method, as in totalizing the Ah accepted by the battery.

I've only seen that term used for "current accepted by the battery", as in measuring amps, not Ah.

To me, a "return current" charging method means a charge extended into the Absorb/CV stage, where the stop-charge point is based on the current "tapering down" to a specified endAmps setpoint, usually somewhere between 0.03C and 0.005C

You are unable to use this method, since your low charge rate is usually within that range even during Bulk CC stage, correct?

And note, I do not see anyone advocating "terminate on voltage" charging here. Yes, it is a common method easily used for normal cycling, but afaik, no one claims it is optimal in terms of caring for a bank to optimize longevity.

[newhaul]

Originally Posted by john61ct

afaik, "return current" has nothing to do with your preferred coulomb-counting method, as in totalizing the Ah accepted by the battery.

I've only seen that term used for "current accepted by the battery", as in measuring amps, not Ah.

To me, a "return current" charging method means a charge extended into the Absorb/CV stage, where the stop-charge point is based on the current "tapering down" to a specified endAmps setpoint, usually somewhere between 0.03C and 0.005C

You are unable to use this method, since your low charge rate is usually within that range even during Bulk CC stage, correct?

And note, I do not see anyone advocating "terminate on voltage" charging here. Yes, it is a common method easily used for normal cycling, but afaik, no one claims it is optimal in terms of caring for a bank to optimize longevity.

I terminate on voltage with my solar . And now it seems that I need to do an equalization every couple months . At a higher voltage .
As of now I stop charge at 13.8 v so far I'm 2 years in and when I do a capacity test about every 6 months I still have 114ah in my 100 ah bank. I'm watching this thread carefully i know how not to loose capacity in the next 20 years .

[john61ct]

What is your stop-charge profile when doing your load tests?

[newhaul]

Originally Posted by john61ct

What is your stop-charge profile when doing your load tests?

I discharge via my solar controller load output at a rate of about 25 amp continuous to a stop point of 12.25 volts then I rehook the solar panels and the wind generator .shut off my load output and let the controller charge to the stop point I have programmed which is as stated 13.8v . The controller counts ah in and out .


400 watts wind and 400 watts solar .
So a potential of 60 amps in ( which is the max of my bms. In and out)
under average circumstances I'm usually charging at approx 25 to 30 amps .
Which gives me a rate of .25C to .3C

[tanglewood]

Originally Posted by Cpt Pat

You're not missing anything except that, as the terminate-charge-on-target-voltage practice continues through successive cycles, the bump gets worse.

The "return current based" method: counting amp-hours taken out and amp-hours returned, is just another description for coulomb counting. (A coulomb is the equivalent of one ampere-second -- an amp-hour is simply 3600 coulombs.) This is the method I am using, and what is being vigorously opposed by common mythology - because the terminate on voltage method is cheap, conceptually simple, and conforming to the behavior of lead-acid batteries.

"For every complex problem there is an answer that is clear, simple, and wrong."
- H. L. Mencken

I've seen many reports that the loss of capacity caused by the flawed charging process is permanent. Since I've avoided the problem, I have no personal experience with it. It could be that those experiencing the problem are just repeating the flawed voltage-based termination method, over and over.

“The definition of insanity is doing the same thing over and over again, but expecting different results.”
- Albert Einstein

Considering the cost of LFPs and the pride of early adopters, I imagine there are some seriously bruised egos out there. No one enjoys a failed experiment, and its natural to blame the messenger -- but so goes scientific progress. It's a new technology, and there are bound to be failures.

A few comments where I think at a minimum your terminology differs from most, and might be leading to miscommunication....


- Termination based on return/acceptance current and coulomb counting are different things. For charging based on coulomb counting one would presumably count coulombs drawn out of the battery, then when charging return that same number of coulombs, plus a few extra to account for inefficiencies. The actual return/acceptance current when you have finally "repaid the loan" could be anything, or more accurately, it's not something you are paying any attention to. I honestly haven't heard of anyone charging this way, but that might just be a matter of convenience.


- Termination based on return/acceptance current is stopping when terminal voltage is being successfully held at some level, and the return current has dropped below some threshold. No tally of coulombs in or out is required. Not to say this approach is necessarily any better or worse, but it is generally more convenient given the features available in most chargers.


So they are two different things.

[tanglewood]

BTW, it's still a bit unclear to me whether the capacity loss reported by the one or two people in this thread is indeed a cumulative memory effect, or a loss of capacity due to some other effect.

[nebster]

Originally Posted by tanglewood

BTW, it's still a bit unclear to me whether the capacity loss reported by the one or two people in this thread is indeed a cumulative memory effect, or a loss of capacity due to some other effect.

Yes, the two seem like they would be roughly indistinguishable from each other.

[john61ct]

Well in the original discussions on this memory effect,

what **started** as a minor "charge bump" in voltage, easily overcome by pushing past, no permanent capacity losses at first,

did apparently turn into irrecoverable damage over years of unknowing repetition.

[evm1024]

Originally Posted by john61ct

Well in the original discussions on this memory effect,

what **started** as a minor "charge bump" in voltage, easily overcome by pushing past, no permanent capacity losses at first,

did apparently turn into irrecoverable damage over years of unknowing repetition.

That is misleading or incorrect. Regardless of other threads. This thread is for those who have significant capacity loss and the steps that they are taking to try to reverse it. And wish to report on their actions and results. It has nothing to do with small memory artifacts from interrupted charging.

There is no indication that smallish (20 mv) bumps mid charge have become cells that charge with a normal profile and show a significant loss of capacity. In fact the cells the OP poster is talking about are cells that go to low charging currents at 3.65 VPC with a 30% loss of capacity. Their charge and discharge curves are for intents "normal".

I will add that there have be a few more folks who have seen such cells but are not posting on this thread. They have their reasons but I suspect that they do not want to bother with the off topic "chatter".


I'm out for the weekend see you all later - let's hope we can stay on topic.

[tanglewood]

Originally Posted by evm1024

That is misleading or incorrect. Regardless of other threads. This thread is for those who have significant capacity loss and the steps that they are taking to try to reverse it. And wish to report on their actions and results. It has nothing to do with small memory artifacts from interrupted charging.

There is no indication that smallish (20 mv) bumps mid charge have become cells that charge with a normal profile and show a significant loss of capacity. In fact the cells the OP poster is talking about are cells that go to low charging currents at 3.65 VPC with a 30% loss of capacity. Their charge and discharge curves are for intents "normal".

I will add that there have be a few more folks who have seen such cells but are not posting on this thread. They have their reasons but I suspect that they do not want to bother with the off topic "chatter".


I'm out for the weekend see you all later - let's hope we can stay on topic.

OK, I'm really confused.


The title of this thread, started by you, is "memory effects", and your opening statement is "This is a starting point for understanding the memory effects that a few of us with LiFePO4 banks have seen."


If the topic is cells that have lost a bunch of capacity, that's much broader since there are many causes for diminished cell capacity, and it's very different from the stated memory effect that has been referenced in one or more studies.


And I thought the assertion was precisely that this memory effect is what has lead to diminished capacity.

[CatNewBee]

You are right.

Cells are unique and different, depending position in the pack, charging and discharging regime, temp differences etc.

Cells drift over time, some accept more charge than other, simply because of voltage differences, especially near full. If you ignore this and do not use automatic balancing, there will be permanent loss of capacity from overcharging. The memory effect is one possible driver of the drift, while it is reversible in general, the damages caused afterwards when one cell drifts more than the other and there is no automatic mitigation, you end up losing capacity.

Manual maintenance, coulomb counting whichcraft, PSOC staying below the shoulders are all dark arts, usually worsen the issue than contributing to longer battery life. Use a good BMS.

[nebster]

Originally Posted by CatNewBee

Cells drift over time, some accept more charge than other, simply because of voltage differences, especially near full. If you ignore this and do not use automatic balancing, there will be permanent loss of capacity from overcharging.

This seems true. But, it sounds as if the OP and others may have automatic balancing capability in their BMS. I would expect substantial drift (which btw many of us have not observed, including sometimes after many years of use) to trigger the balancing hardware at some point... including when charging to a low pack SOC.

If anything, I would expect substantial drift and no opportunities for top-balance to create a problem at the bottom of the curve, when the pack is dangerously discharged too low.

But it doesn't sound like either of these are happening with these two batteries. ???

[guyrj33]

One of the confusing things about this thread is the title "memory effects".
Lithium cells when in a series arrangement, which is how we all use them, suffer from unbalanced charging, not memory effects.