Is mantaining a cell at 3.6V overcharging?

[ltosolini]

Experts,
I built a 20S Wina50AH (LiFePO4)pack for my scooter Electra2500W.
Recharging the pack with a simple 20S charger just doesn't work because I have the first weakest cell that goes way above the maximum voltage of 3.6V and other cells remain below.
So I built this home made charge balancer based on TLV431 following this design:
http://m.electronicdesign.com/analog/high-current-low-voltage-shunt-regulator
Basically it is 20x shunt regulators tuned at 3.6V in parallel with every cells; this way when a cell reaches 3.6V the shunt drains all charging current from that cell which remains at 3.6V while other cells keep charging.
It seems to work pretty well, but I have the following question?
I do this balancing phase at pretty low current, between 200 and 400 mA, so it can take several hours; this means that some cells can stay at 3.6V for such long time. Can this produce overcharging or in any way damage the cell?
Thanks,
Luca.

 

[dogman dan]

3.6v is more charge than lifepo4 can hold. So technically yes, it's overcharged. But lifepo4 is different, and tolerates some overcharge easily. So many lifepo4 packs bms will have a balance point set to as high as 3.65v, and some even higher.

If your low capacity cell that fills first is going to 4v, that might be pushing things too far, but 3.6v is fine for lifepo4

 

[teklektik]

I do this balancing phase at pretty low current, between 200 and 400 mA, so it can take several hours; this means that some cells can stay at 3.6V for such long time. Can this produce overcharging or in any way damage the cell?[/attachment]

No. It's all good - what you are describing is a conventional charging situation for a pack with a BMS.

What you have built is essentially a very high powered BMS with a balance voltage of 3.6V per cell. This is a tiny bit lower than the usual recommended LiFePo4 termination voltage of 3.65V, but fine. Once a cell reaches 3.6V, all charge current will flow through the regulator reducing the cell charge current close to zero. So - this operation is identical to that of a conventional BMS - there are no issues whatsoever with your charging regimen or the termination voltage.

The issue of cells staying at the termination voltage for a long period is not a concern. Once charged, LiFePo4 may normally sit at this voltage on the shelf for many many days (depends on the construction/manufacturer). Since your regulator (BMS ckt) is essentially enforcing a constant voltage charge profile, the termination charge current is the same as the self-discharge current. That is, this is not a separate 'float charge' cycle enforcing a trickle charge as with special chargers for other battery chemistries. Since you are applying negligible charge current at this level, there is no issue.

This relates in part to questions of charging the battery to full charge, but LiFePo4 is unique among lithium chemistries in that it suffers no lifetime degradation from charging to 100% capacity. In any case, for this chemistry, determining SOC by voltage is very difficult because of the flat discharge (charge) curve so charging to 3.65V (100%) is a safe and convenient strategy.

A word on this matter from Justin:

I have a lifepo4 battery 48v and from what I have been reading is that not fully charging battery to full capacity will increase amount of cycles.

This is true for most flavors of lithium battery which normally charge up to 4.2V/cell, but it is not the case with LiFePO4 and we try to mention this every time we discuss partial charging (eg http://www.ebikes.ca/tools/charge-simulator.html#benefits-of-partial-charge ). So if you indeed have an iron phosphate pack, then simply charge it 100% to the full 3.5-3.6 V/cell. Because of the nature of iron phosphates very flat Voltage vs SOC curve, you can't reliably charge it to a partial charge point anyways, even if there was some benefit in doing so.

One of the key benefits of the Cycle Satiator is its ability to let you easily control the charge level of your battery. It is now well known that most lithium chemistries (with the exception of LiFePO4) can see drastic improvements in calendar and cycle life when they are not held at the nominal full charge voltage of 4.2 V/cell but are charged to a lower voltage instead.

FWIW:

• Although I am not a fan of positive anecdotal evidence as 'proof' of anything: I overnight charge a 20s2p Headway LiFePo4 pack (no BMS) with a bank of single cell chargers (2A/3.66V). It's always recharged after use so storage voltage is charge termination voltage. Five years and many hundreds of cycles later and the pack runs like new. The charge voltage and internal resistance are measured by CA on every ride - no material change. I would never build a big heavy LiFePo4 battery like this again, but the damn thing doesn't show the slightest sign of dying...

 

[999zip999]

I just bulk charge to 3.5 now - 902 cycles. The first 500 cycles I balance charged to 3.6 . After 4 burned out rc chargers later. I guess use a meanwell and a hp600 power supply .
Playing with the pack I got it to 3.8v twrice for a short time.

 

[wayover13]

Tek nailed it.
<snip>
LiFePO4 is very tolerant, 3.8V-4.0V can be tolerated by some short term without noticeable harm. They should not be stored at full voltage long term though. Time and temp degrades them more rapidly, including being stored at higher potential.

But:

LiFePo4 is unique among lithium chemistries in that it suffers no lifetime degradation from storage at full charge.

So did teklektik not nail it on this point but did on others? I stored my LiFePO4 battery for over a year at full charge (had a bike maintenence issue that took me forever to get around to fixing) and am now putting the battery back in service so am wondering whether it might have degraded or suffered any damage. What eTrike and teklektik are saying on the point of storing at full charge seems to be contradictory, despite one having said the other "nailed it.".

 

[teklektik]

Quite right - not sure what happened there - the sentence is incorrect, off topic, and even out of context with the referenced material that follows. Hmmm...
Good catch - fixed it.

AFAIK, the primary sources of capacity loss during storage are temperature and voltage-induced degradation. The recommended storage SOC is often stated as 50% - but I believe this is a general lithium battery recommendation that compromises voltage-induced degradation and self-discharge. This voltage-degradaton perspective and the unique flat discharge curve of LiFePo4 (where the cell potential is essentially unchanged from about 20% to 80% SOC) suggest that there should be no change in voltage-related cell degradation across the flat portion of the discharge curve. So, unlike other lithium chemistries, it seems that LiFePo4 storage from 20%-80% SOC would induce approximately the same amount of deterioration -- with the higher 80% SOC level offering somewhat greater protection from self-discharge. Anyhow, I was never able to locate any authoritative studies on the matter and all quasi-trusted sources which quoted the 50% figure seemed to reference Li-ion as a general group.
Just a little musing...

 

[wayover13]

With apologies if I am side-trackiing this thread, I would really like to know, as someone with a very limited comprehension of issues like battery chemistries and even electronics in general, to what level I should be discharging my LiFePO4 battery for winter storage. Sometime within the next couple of months I'll be taking the battery out of service, and it is likely to sit unused for at least 4 months. I'd earlier heard that it should be discharged to from 20%-80%. I believe previous to last year I drained it down to 80% for winter storage. But as I mentioned earlier, the last time I stored it I did not discharge it at all and it sat like that for over a year. I'm going to try and not let that happen again, but it would be nice to get some authoritative word on what is an acceptable discharge level for winter storage. Is there no authoritative stance on this question?

 

[teklektik]

Here is a tasty read for anyone with enough interest in obscure details of LiFe:

Ya - I'd seen that study some years ago and squirreled it away, but at the time was unsure of the general applicability of the study re: electrode type: graphite vs hard carbon, etc. for general commercially available cells -- just a little out of my area of expertise. Often these studies are funded to focus on new or experimental chemistries or concepts to evaluate their viability, but generalization to other widely used alternatives is left as an exercise for the reader....

That said, I do believe that graphite electrodes are in wide use. Based on that, this particular plot from the study is interesting in illustrating the differences in degradation from storage at 25degC at both 50% and 100% SOC. Not really much at all - maybe 3%... (although that would accumulate over the years for an annual storage regimen)

Just as an aside, the flatness of the discharge curve is frequently misrepresented as being inaccurate for estimating SOC during use.

I think not. Although it may be possible to tune up a discriminator for open circuit voltage for a particular cell manufacturer and model at a particular temperature, the curve is so flat that it is difficult to evaluate the SOC with confidence in the general case. The effects of age and temperature can have as large an effect on LiFe OC voltage as SOC. I can say with certainty that the CA mechanism for determining SOC for LiFe is very different from that used in all other Li-ion chemistries because of these issues.

 

[dogman dan]

Re winter storage,, if you can disconnect the bms, then store at 80% or less. If you cannot disconnect the bms, then some bms can run the pack down on oneor two cells over the winter. It may not hurt the battery, but by spring the thing is severely unbalanced.

So for some, the best approach is a balance charge monthly over the winter. Remember,, not likely you keep the house at 90F all winter,, so it's not both hot and full over the winter. If you have a really cool place to put it, I see very little harm in storing it full.

Lastly,, it is an option to charge and balance monthly,, but then discharge some to put away for the month. But in some climates, even a ride around the block is pretty brutal.