Care and feeding of LiFePo4

[markopolo]

We could discuss what it takes to care for LiFePo4 batteries in a variety of situations. (it would be fantastic if you can cite authoritative references whenever possible)

In another topic, Winston pointed us to a site about a marine LiFePo4 implementation where the owner is seeing great results.

A quote from that owners site:

Before you get too excited these cells have:

--Never been floated, they get charged, then discharged
--Only absorbed to a net 8A - 10A at 13.8V -14.0V
--Not charged above 14.0V unless for testing purposes (I now have a few other banks for that)
--Max charge rate at approx .3C
--Stored at 50% SOC when not being used or cycled
--Stored in 45-60F temps when not being used or cycled
--Only very rarely exceeded 80F
--Highest voltage they have ever seen was 3.8VPC while top balancing initially.

So, if your usage and care of LiFePo4 follows that method then you'll likely experience great results also.

What happens when your use and care differs from any or all of those points?

If you're often doing mini partial state of charge charging and discharging what should you do differently?

One possible answer looks to be performing a full cycle at least periodically.

How do you fully charge a LiFePo4 battery?

A combination of voltage and current is needed. 14.4V with current having decreased to 2% of the batteries capacity rating is at least one manufacturers recommendation.

How often should you do that?

What's the best balancing voltage? How often should you balance the batteries?

What's the best system standby voltage to support other loads so as to not be continually charging the LiFePo4 battery?


So many questions

[booster]

I had just put some similar stuff on the cold weather lithium thread, so will copy it here as this is probably a better place.


Quote:
Originally Posted by markopolo
I'd like to know more about the voltage difference between cells. From the photos it was 30mV max and 20mV but your latest report indicated the max spread was now 40mV.

I wonder at what point does cell voltage variance become a concern?

Would initiating top balancing now lessen the variance? How long would that effect last?

These are just general questions - I'm not in any way suggesting that there's a problem with Davyyd's battery bank.


I have the same type of questions as it is starting to appear that things like the balancing and kind of unclear memory things may all be kind of interconnected when it comes to some of the other things we have seen like narrowed operating ranges and lowered charge voltages. The balancing seems to in the middle of most of the things, though, and at this point there is not a lot of information on how is best to address it on an ongoing and in use basis, or even how bad is bad.



For instance, if a unit had automatic top balancing on each cycle, like some seem to do, would that also take care of some/most/all of the memory effect things?


Or the opposite, if you are careful about running a full charge and full discharge at some, yet unknown, frequency, would that eliminate the need to have separate balancing cycles?


We don't know how much imbalance is needed before it is a real issue, I think, and we have seen very little actual cell voltage information over time as most systems don't show it and parallel connections likely hide some imbalance unless disconnected to check.


Hopefully in the near future some clarity to what some of the interactions are between all these things really is, as it seems to be a bowel of spaghetti now, to me.

[markopolo]

Would balancing at other points on the SOC curve ever be beneficial - say 50% SOC around 13.2V for example?

Should you make sure you get a BMS that let's you control some of the programming? - I vote Yes to that!

[markopolo]

Should you go with a smaller capacity bank with an average SOC at say slightly less than 50% or a large capacity bank with an average SOC at 70% or more?

My current thinking on that is that a more fully utilized smaller capacity bank might be better. - use it or risk losing it due to other causes of degradation.

[Davydd]

I question the 50% SOC for storage and other comments about lithium batteries are happy at 60% or 80% SOC as I've read in the past. I suppose if you want to keep your B and batteries for 25+ years I guess it would be a concern. In practical terms I think this would be asinine especially when on the road.

When my batteries hit 100% SOC, the charging shuts down from all sources whether solar, shore power, idling or driving. When on the road driving my batteries replenish in under an hour.

When is storage I could safely disconnect the batteries over the winter as I have that capability. If I did so I would not handicap myself at putting them initially at 50% and of course disconnected they still drop but are not monitored. ARV reprogrammed my batteries so when first stored almost always at 100% since I drive and stop and park, it does not charge the batteries until they drop to 90% and then resume charging under shore power. I leave it plugged in because I use my B for a bathroom inside my garage and it trickle charges my AGM chassis battery. I could store it unplugged but it is strictly for convenience and security as I don't know how often I could get to my remote garage. For instance I didn't go there at all for a month this fall when I was in Europe.

The other safeguards. The garage temperature is set at 45 degrees F. so the temperature of the batteries are around 50 degrees always no matter how cold it gets because it doesn't get any colder than that. On the road, it is programmed to keep them over 41 degrees.

The top end temperature is controlled by common sense and practice. In storage the garage temperature has never gone over about 88 degrees as it stays cooler than the ambient temperature which doesn't get much hotter in Minnesota. Then our common sense travel of trying to follow the proverbial 70 degrees have proven to keep the batteries under the 110 degrees recommended for charging and I think they can withstand 150 degrees before permanent damage occurs. OK, I am luckier than someone living in the southwest in that regard. You will never find me there in the summer traveling.

I'm coming at this with observation and 5 years experience. The only other source of experience prior to mine was Technomadia with the same Elite batteries but with an older generation of batteries installed in 2011. They wrote extensively with their experience living in the southwest. They probably didn't need or encounter my cold weather concerns but they said they did store there Class A on asphalt for 6 months in Phoenix so you can imagine they had heat deterioration. I don't think they had the sophisticated controls I had either. They reported that after 4.5 years their batteries lost 20% of capacity.

I'm still not understanding what constitutes a cycle. Is 100% down to 90% and recharged a full cycle or is 100% down to a deep discharge to 20% considered a cycle. If deep discharge, at 2000 cycles that could mean about 40 years (I'm guessing based on my usage). I it means a 10% discharge is a cycle then it would be 5.4 years. So I doubt a 10% discharge is a cycle.

[Winston]

Originally Posted by markopolo

. . . How do you fully charge a LiFePo4 battery? . . . How often should you balance the batteries?

A very expansive inquiry to which, initially, we’ll offer responses to these two questions: How to charge and how often to balance LiFePO2 batteries.

CHARGING

We have railed in these forums before concerning the terms: Bulk, Absorption and Float - - terms we content are inconsistently defined within the industry, grew from the unique charging requirements of lead-acid technology and, therefore have no real place in the lithium world.

The beauty of lithium is that it ‘tells us’ when it’s discharged and when it’s charged. Look at a lithium SoC vs Voltage curve - - two nearly vertical segments connected by a nearly horizontal segment. The intersections of those lines - - the corners or knees - - define for all practical purposes full discharge and charge. When a cell reaches that point where it transitions from a place where ‘lots of current causes no change in voltage’ to the point where the “virtually any current causes dramatic increases in voltage”, voila - - you’re fully charged.

The definition of full charge, we contend, occurs when all of the cells have reached this second “dramatic increase” in voltage stage. The challenge, we suppose for those without the ability to monitor each cell’s voltage, is determining what ‘external’ conditions signify that this “dramatic increase” stage has been reached.

BALANCING

It is our recollection that the referenced Compass Marine article - - at least as of three years ago - - also noted that the owner’s test lithium bank had not been re-balanced in over 700 full cycles - - that it had not been re-balanced since its preliminary, initial top-balance. The conclusion we draw from this single test is that re-balancing is a rare, if ever, necessity.

In terms of the “millivolts” mentioned in some of the above responses, our experience (3 years of more or less continuous use) suggests that the cells stay (mostly) within 30 millivolts without balancing (excepting when you hit one of those ‘intersections’ - - the fully discharged or fully charged knees - - at which point very large cell voltage diversions take place. In this connection, as we cycle our batteries through a full charge to discharge cycle to measure capacity each year - - we have visited these two extremes.)

Thus, we see little ‘necessity’ for balancing unless you observe that one of your cells is hitting a discharge or charge ‘knee’ well ahead of the others.

[markopolo]

Originally Posted by Winston

....................We have railed in these forums before concerning the terms: Bulk, Absorption and Float - - terms we contend are inconsistently defined within the industry, grew from the unique charging requirements of lead-acid technology and, therefore have no real place in the lithium world............

CC Constant Current, CV Constant Voltage & Standby might work.

13.3V standby looks to result in at least 70% SOC being available when/if needed.
13.4V standby looks to result in at least 90% SOC being available when/if needed.

An accurate meter is necessary.

Originally Posted by Winston

..................The beauty of lithium is that it ‘tells us’ when it’s discharged and when it’s charged. Look at a lithium SoC vs Voltage curve - - two nearly vertical segments connected by a nearly horizontal segment. The intersections of those lines - - the corners or knees - - define for all practical purposes full discharge and charge. When a cell reaches that point where it transitions from a place where ‘lots of current causes no change in voltage’ to the point where the “virtually any current causes dramatic increases in voltage”, voila - - you’re fully charged. ................

I guess staying between those corners that would approximately equate to operating between 10% & 90% SOC (maybe 95%). I haven't found much research to support not going to 100% or close to 100%. With large banks there might not be a need to go to 100% but you might want to with a smaller bank. There's that memory effect thing that some folks are discussing to factor in. I assume that there's no real chance of that occurring if fully charging the batteries.

Originally Posted by Winston

..........The definition of full charge, we contend, occurs when all of the cells have reached this second “dramatic increase” in voltage stage. The challenge, we suppose for those without the ability to monitor each cell’s voltage, is determining what ‘external’ conditions signify that this “dramatic increase” stage has been reached.........

When charging, that dramatic increase in voltage event correlates to a dramatic drop in current. That can be pretty close to 100% if using low-ish voltage and low-ish current for the charge according to test results I've looked at online. If fast charging, it looks like you'd need to pause charging there then resume after the batteries have had a chance to settle if you want to get to 100%.

That end of charging should be automated based on at least voltage but better with a combination of voltage and current. Timer based chargers shouldn't be used IMO.

I'm currently looking into what is considered to be an overcharge by most. That would be in the ranges of 3.7 to 4.2 VPC. Observable damage like swelling, opened pressure relief valves happens after 4.2VPC.

3.7VPC looks to almost guarantee a 100% full battery.
3.65VPC will most likely result in a 100% full battery.
3.6VPC could result in a 100% full battery if held there for a short while.

-> https://www.powerstream.com/lithium-...ge-voltage.htm

I've been reading some interesting stuff on deliberate overcharging.-> https://avestia.com/EECSS2019_Procee...EE/EEE_113.pdf

I'd need to see a follow up or a longer term study before doing that. My assumption is that the capacity gains will result in shorter battery life in terms of cycles.

Originally Posted by Winston

..........BALANCING .........

Making sure the cells are well matched at the pack assembly stage looks to be critical.

It's interesting that a weak (low capacity) cell will reach the fully charged voltage quicker but it won't have the capacity of a healthy cell that takes longer to reach the fully charged voltage. I guess balancing lets you hold the voltage point so that healthy cell that can absorb more charge has a chance to do so.

[markopolo]

I've been reading the datasheets for the tiny controller chips used to recharge small LiFePO4 batteries. They all seem to use a combination of voltage and current to know when to terminate charging. Termination voltage is spec'd to be within 1% of 3.6vpc or 3.65vpc typically and held there until current is quite low. They check cell voltage first and if low, slowly bring the cell up to where it's safe to begin CC which transitions to CV with CV terminating when current is at a percentage (like 10%) of what CC ended at. They're amazing little chips with more features than I outlined above. Approx 5mm x 5mm in size!

There seems to be so much support for charging to or actually almost through 3.6vpc or 3.65vpc with LiFePO4 cells.

[@Michael]

Originally Posted by Davydd

I'm still not understanding what constitutes a cycle. Is 100% down to 90% and recharged a full cycle or is 100% down to a deep discharge to 20% considered a cycle. If deep discharge, at 2000 cycles that could mean about 40 years (I'm guessing based on my usage). I it means a 10% discharge is a cycle then it would be 5.4 years. So I doubt a 10% discharge is a cycle.

Some relevant info in this interview with the principle of Battleborn, starting @08:55.

https://www.youtube.com/watch?v=bmaIJvsoSig

I take this to mean that 100%-0%-100% is a cycle, and they claim 3000-5000 of those cycles and it would still have 80% or its original capacity.

Other bits:

@04:50 - Battleborn sorts cell by voltage/capacity and puts matched cells into modules, then voltage/capacity matches modules to within 'tens of milli-amp hours' and assembles them into batteries. So each battery will have a slightly different capacity, depending on the capacity of the sorted cells/modules.

By luck of the draw, you might get a battery that has lower or higher capacity, but always over 100AH.

@15:15 - Temperature degrades capacity - 32F == 85% capacity, 0F == 65% capacity.

[Davydd]

Thanks, @Michael,

So, 40 years is closer in my guess based on my usage if we go by Battleborn. I just conservatively went by 2,000 cycles which is a consensus low for lithium ion batteries and just 80% DoD which ARV sets on my batteries and calculated how much a would use at a maximum every day on the road and what I would use in storage and came up with over 28 years battery life but still have use. For all practical purposes the batteries can outlive the usefulness of the vehicle.

I know intuitively from experience this might not be the case, but I know from experience in Class Bs an AGM battery might not last a season with no BMS safeguards other than manually monitoring 4 LED idiot lights that I had on my previous B's.

What is this discussion again?

[RossWilliams]

I know intuitively from experience this might not be the case, but I know from experience in Class Bs an AGM battery might not last a season with no BMS safeguards other than manually monitoring 4 LED idiot lights that I had on my previous B's.

What is this discussion again?

The original idea was how tofeed and care for lithium batteries. But I think David's responses makes you wonder whether that is the right question. Perhaps the question should be how to allow someone to just use the batteries and avoid any special feed and caring. He identifies one of the key components, generally lacking in agm's, a bms that prevents it being damaged by being drawn down below 20% SOC. Another is that, where temperatures get below zero, there is some means to keep/get the battery warm enough first so that it can be used and then to warm them sufficiently to allow recharging. Those two things will help avoid accidentally damaging the batteries in ways that prevent them from lasting for the life of the vehicle.

There are others, starting with a charger that does not require intervention to ensure proper charging. And, of course, large enough capacity to handle your uses between charges. Any others?

[markopolo]

I'd like to hear from folks who do a full cycle capacity test each year like Winston does. That seems like a very good idea. It's the only way to know if the battery bank has lost capacity.

Does anyone else do periodic capacity tests? How is your battery bank holding up? Did you change how you care for the batteries if the capacity was not what you expected?

[booster]

Originally Posted by markopolo

I'd like to hear from folks who do a full cycle capacity test each year like Winston does. That seems like a very good idea. It's the only way to know if the battery bank has lost capacity.

Does anyone else do periodic capacity tests? How is your battery bank holding up? Did you change how you care for the batteries if the capacity was not what you expected?

I have ever couple of years in the past on ours, but that is AGM. I don't do full capacity only to 20% SOC and then let rest to see if the voltage was the same as previously and predicting the same 20%. No change in either of two tests to this point.

[Davydd]

Originally Posted by markopolo

I'd like to hear from folks who do a full cycle capacity test each year like Winston does. That seems like a very good idea. It's the only way to know if the battery bank has lost capacity.

Does anyone else do periodic capacity tests? How is your battery bank holding up? Did you change how you care for the batteries if the capacity was not what you expected?

I’m not sure what I can do to abuse my lithium ion batteries. I have the BMS and the reporting functions and controls to monitor my batteries at all times.

1. They never go below 41F degrees with applied direct heat off shore power, supplemental solar or just off the batteries themselves and if power outage they will survive at least three days I estimate off the batteries in time to get to them.

2. They have automatic prevention to shut down completely at 20% SOC.

3. High heat damage will not occur until 150F according to the battery manufacture specs. They can be safely charged up to 110F. I don’t know if charging would shut down when that temperature is reached because I never encountered it.

4. Charging from any source shuts down when 100% charged and doesn’t resume until about 90% SOC.

[markopolo]

Originally Posted by Davydd

I’m not sure what I can do to abuse my lithium ion batteries. I have the BMS and the reporting functions and controls to monitor my batteries at all times..........................

It's a good point. Most areas of concern are well covered.

It may turn out that operating lithium batteries in a limited SOC range is detrimental. That's if the memory effect that's been talked about is real and whether it's reversible or not.

I wouldn't think that a discharge from 100% to 20% (80% discharge) would be harmful other than using up 1 cycle from 1,000's of cycles. If the capacity delivered by the test is lower than what you expected (age and use adjusted) that would alert someone to look for the causes of the unexpected decline.

[Davydd]

I addressed that in an earlier posting and I think it is ridiculous to worry about trying to stay below 80% SOC or memory effect as the cycles conservatively analyzed for my usage profile would get me 28 years of good battery use. I don't believe that totally but it is reassuring that I can get full use of my batteries power and that worrying about a lab result of lithium ion batteries like a SOC somewhere midway is to me kind of an anal worry. Battleborn doesn't worry if giving a 10 year warranty. Just hope they aren't gaslighting like Roadtrek.

If you read that ARV white paper on lithium ion batteries it explains why they lock out 20% so you would never completely discharge your batteries. That makes more sense to me as it accounts for natural battery loss you wouldn't notice after 2,000 or so cycles and does double duty to keep you from totally discharging which is an AGM frequent failure.

[RossWilliams]

Here is Battle Born's warranty;

"Battle Born Batteries offers a 10 year manufacturers defect warranty from the date of purchase. This warranty does not cover negligence or misuse of the battery. If it is deemed that battery was used improperly, you will be subject to a $150 an hour repair charge plus parts and shipping."

Another slightly different description from the FAQ on their website :

"Our batteries come with an 8 year full replacement manufacturer’s defect warranty and 2 years prorated. They are designed to last 3000 cycles, at which point the battery will still hold 75 to 80% of its energy capacity."

Its not clear whether this warranty is assumable although it does not explicitly limit it to the original purchaser. And, as with Roadtrek's warranty from Hymer NA, it depends on the company still being in business. As far as I know, Hymer honored the Roadtrek warranties before they bankrupted the company.

[booster]

I have been continuing to randomly search different terms and descriptions for LFP charging parameters etc and ran across this article, and few others not as clear, about LFP memory and how it presents itself. The article is mostly about charge control so doesn't really address anything about the potential of longer term capacity loss because of memory. This article shows the most clearly that I have found how to clear the memory and how to generate it. The graphs also would allow you to actually see it pretty easily on a system, especially if you had a charting voltmeter on it.


https://makermax.ca/2019/08/lithium-...memory-effect/


What I seem to be seeing in this and other similar articles is that the memory effect is actually putting a bump in the recharge cycle of higher voltage at the memory point. It is often shown on graphs that also include SOC based on voltage, so my big question is if it is a real increase in SOC at that point or just an increase in internal resistance that is raising the voltage for a while? I would think that if it is a bump in internal resistance, you would see more heat and such which is never good, and if it was long term it may get irreversible like some of the articles that do talk about reduced capacity claim.


Of interest in one other article I saw, which was so hard to follow it was kind of not worth reading (Chinese translation and only charts unless you buy the access to the rest) did test capacity based on charge and discharge rates, giving the results as what appeared to be an energy total (like watt hours). Using energy it showed that the LFP batteries got more usable energy at lower charge and discharge rates. I think this kind of information based on energy rather than SOC or AH would give much better understanding for a lot of what we talk about, from LFP capacity to the Peukert equation. Even though voltage drop in lithium is smaller than lead, it still was enough to make significant differences in the overall energy stored and released.


https://pubs.acs.org/doi/abs/10.1021/acsami.7b05852

[markopolo]

Thanks for posting the links.

[markopolo]

Originally Posted by Davydd

........................................ The other safeguards. The garage temperature is set at 45 degrees F. so the temperature of the batteries are around 50 degrees always no matter how cold it gets because it doesn't get any colder than that. On the road, it is programmed to keep them over 41 degrees...............................

I like that Advanced RV keeps the battery temperature above 41F. in addition to no damage, the battery is nearer optimal operating temperature for best performance.

41F is is also the point where Trojan recommends halving the charging current.

reduce current at 41F and below.JPG

I was looking for info re: temperature slope & LiFePO4 & came across several images similar to this:

cold cool normal warm hot.png
Its basis is from Japan Electronics and Information Technology Industries Association guidelines and is not LiFePO4 specific but I think is a great representation of how we should think about temperature and charging slope or curve.

Instead of thinking something suddenly happens at 32F it would be better think of how to charge (or not to charge) in terms of cold, cool, normal, warm & hot.