LFP memory effects thread

[evm1024]

This is a starting point for understanding the memory effects that a few of us with LiFePO4 banks have seen.

It would be very nice if we could limit posts to those who have LiFePO4 banks. Of course if someone comes up with something that is directly applicable and not just opinion they should post that too.

This thread is to report observations and actions.

[evm1024]

My 5 year old Balqon 700 AH bank is showing 500 AH capacity. I've not been fully charging it for a number of years. In fact it has been sitting at around 50% for most of the time except when cruising.

The bank appears to "work" normally in that it charges up more or less normally and then discharges like you expect except that it has reduced (to 70%) capacity.

I switched over from a Xantrex inverter charger to a Victron inverter charger and added a 200 Amp alternator. These changes caused me to do a capacity test.

During charge the cells can get up around 100 mV out of balance with each other. Two cells voltage rise more slowly than the other 2 cells.

On discharge the cells stay within 10 mV of each other.

I presume that the 2 cells that reach higher voltages have less capacity than the other 2.

My current plan is to break the pack apart and charge all cells in parallel to 3.65 VPC. Then I'll take cells one by one and charge them again to 3.65 VPC until they have a very low tail current.

The individual charging is to verify that each cell is actually at 3.65 VPC. They "should" be but let's just check that.

Resting voltage will be measured after 24 hours.

Then the pack will be reassembled and a discharge cycle started. Target rate is 30 amps and a Coulomb counter will be used to measure capacity. Some reasonable VPC will be used for the cutoff voltage.

Rest again and measure the cell voltage.

Then do it all again to see what changes. Will this restore any capacity? Let's find out.

[tanglewood]

Easy for me to make work for you, but.....

It would be interesting after you charge each cell individually to also load and drain to mfg empty set point. That will get you the cell’s capacity without any influence from other cells.

Another possibility given what you are seeing is the the higher voltage cells are not reduced capacity, but operating at a higher SOC. The best was to differentiate, I think, is to individually discharge and measure each cell. Alternately, after discharging them all as a bank, break them apart and see if there is more that can be drawn out of individual cells. That would ID cells that are just cycling through a lower SOC range than the others.

[tanglewood]

Another thing that will be important to this discussion is to define what you mean by a memory effect. And how is it different from a reduction is capacity? Without this I think everyone will be talking across each other

[rgleason]

So LFP is in some respects similar to FLA, only "sulfation" or loss of capacity is neither so fast or permanent? Where LFP shines is in weight, ability to continue functioning in mid range states of charge, reduction in need to charge to 100% often, great acceptance and discharge rates, high efficiency?

So there is hopefully some easy way to counter a "memory effect".

When is this effect found generally?

[a64pilot]

I don't have a lithium bank.
I didn't think lithium chemistry was subjected to memory, thought that was a Nicad Nimh thing.
I was apparently incorrect, a quick search of LIfepo4 memory effect brings up several source of scholarly articles, not all nuts.

Here is only one
https://phys.org/news/2013-04-memory...batteries.html

it would appear its biggest effect is if your using voltage to determine SOC

[john61ct]

Originally Posted by rgleason

When is this effect found generally?

Start here, go through to the end.

http://www.cruisersforum.com/forums/...ml#post2909875

[john61ct]

It is a real thing, but not a significant problem if you are aware of it and proactively take steps to prevent it.

Thousands of cycles without any capacity drop below the mfg Ah rating is not difficult to achieve, nor is the necessary gear that expensive.

But you do need to **care** about properly looking after your bank and go to the trouble to learn a fair bit, especially if you're trying to DIY from bare cells, own parts sourcing and system design.

If you reckon, a decade's goof enough, I'll just replace when needed anyway, then no worries do that.

[evm1024]

Memory as I am using it in this thread is the change in capacity that is caused by specific usage patterns.

I do not disqualify non-reversible capacity loss provided it is caused by usage patterns (as opposed to abuse patterns).

Also, I would like to remind folks of post #1. Specifically:

This thread is to report observations and actions.

There is no need to educate people in this thread.

[Cpt Pat]

Here are some scientific studies on the topic:

1. https://www.nature.com/articles/nmat3623
   Memory effect in a lithium-ion battery
2. https://www.ncbi.nlm.nih.gov/pubmed/28657290
   Relaxation-Induced Memory Effect of LiFePO4 Electrodes in Li-Ion Batteries.

Note: in paper 1 above, there is a short mention that routinely using coulomb counting (counting amp/hours in and out) to determine charge termination can overcome this problem. The problem arises as a consequence of using terminal voltage to terminate charging.

However, if you switch to that method after a severe memory has been established, you must be careful not to go over cell voltage maximums. I'd stop at 3.65 volts per cell and float charge there while counting amp/hours in. Determing where 100% SOC has been achieved would require measuring open cell voltages after a long rest - and multiple cycles of that reigimine.

[NPCampbell]

Originally Posted by evm1024

Memory as I am using it in this thread is the change in capacity that is caused by specific usage patterns.

I do not disqualify non-reversible capacity loss provided it is caused by usage patterns (as opposed to abuse patterns).

Also, I would like to remind folks of post #1. Specifically:

This thread is to report observations and actions.

There is no need to educate people in this thread.

I'm assuming you are wanting to ignore usage that results in non-reversible capacity loss due to over/under voltage and high/low temperature?

[john61ct]

I have had great success with an oddball gadget

Tecmate / Optimate Lithium-Iron TM-291 charger

actually designed for those little 12V motorcyle starter batts.

Only puts out 5A, only suitable if you can put cells into (1P) 4S packs, and probable would just take too long once you go past 80-100Ah cells.

Ends up getting the individual cells **top-balanced** without any balancing leads, just doing an automated bulk charge on 4S wired bare cells.

Running it through a BMS may interfere with its special output algorithm. Rather than just putting out a steady current at a single static voltage, it pulses both factors, oscillating within a pretty wide range.

Speculating here, but I think that helps prevent the single lower-capacity "weaker" cells from actually rising in %SoC so quickly, letting them fallback during the lower parts of the oscillation, while more of the energy input is actually going into the "healthier" higher capacity cells.

In any case, no speculation here, the end result has been a well-balanced 4S set on a wide variety of LFP cell brands and sizes.

I usually stop-charge at around 13.8V when a larger bank size makes the 5A a very low C-rate, but if you want it will take it right up to 14.4V (3.6Vpc)

which although I never go that high in normal usage cycling

is no problem as an occasional maintenance routine, or as in this case desperate measures to restore capacity years later.

The bigger the cells in Ah capacity the lower the C-rate, the **higher** the finishing SoC at a given voltage, as @Cpt Pat so helpfully keeps reminding us this can quickly lead to damaging overcharge, as in losing lifecycles, but that is not so much an issue here.

Note that trying to get a good handle on endAmps doing a CV charge might require an ammeter with decent current averaging. Personally I just charge-to a setpoint, but in this scenario using it for capacity recovery it'd be worth trying letting it keep going in Absorb for some time, again depending on the cell size.

I'm not claiming this gadget **will** work to recover much capacity in this specific case, but I do think it's worth a shot.

And to be clear, I'm not recommending it as the daily use charger for a big bank,

although it may well work used for periodic maintenance, to help Prevent this "memory effect" **if** you notice lost capacity from chronic PSOC

it might not even do its balancing trick for cells too far apart from each other.

But IMO they're useful enough that if you try it and decide not to keep it I'll likely buy it off you for say $60, or you could probably get more than that on eBay if you were willing to wait a few weeks, there aren't a lot going through there.

Would also be useful I think for initial-commissioning top balancing, for those without an adjustable power supply at higher amps output.

Obviously if you do invest in that latter, then less likely this unit will be worth buying as well.

[evm1024]

Originally Posted by NPCampbell

I'm assuming you are wanting to ignore usage that results in non-reversible capacity loss due to over/under voltage and high/low temperature?

This thread is to report observations and actions.

There are at least 2 of us that have this effect in action on our banks. We are independently looking into reversing the effect.

The thread is not to theorize about what happens in general - we have a thread for that.

[guyrj33]

disclaimer the only lithium battery i own is in my phone.
But I have had long discussions with other engineers who were designing cell balancing chargers / monitors for lithium batteries.
The OP seems to be describing unbalanced cells, and the plan to individually charge cells to the same level should address the balancing issue.

[john61ct]

Well obviously doing a capacity test for this purpose, means on each cell individually.

The capacity loss caused by a BMS triggering cutoffs earlier due to unbalanced cells, would have nothing to do with any memory effect issues.