LiFePO4 charging - the overcharging question

[booster]

Originally Posted by gregmchugh

I have tried on occasion to get people to not obsess about the 50% number with the idea that the discharges will average out and you likely have a lot of shallow discharges that compensate for the times you go past 50%. I guess I wasn't pulling that completely from where the sun don't shine...

I have had the same argument for quite a while about balancing the small and large discharge cycles, as it made perfect sense, while others have claimed that the deep discharges make the other discharges do more harm do to whatever. I even called Lifeline and Trojan and asked them about it, they both said they average, and on really techie guy said that it actually is an integration of all the points, area under the curve thing, which certainly would reinforce the graph showing a relatively fixed number of lifetime amp hours, as that is what the integration calc would give you.

I think folks thing the occasional deep discharge is killing their batteries because they don't last as long as the spec says for the discharges the do. It is far, far, more likely the reduced life is do to poor charging, mostly under, but also over, charging. Probably some do to heat, also, for some people.

[Davydd]

With AGMs you mention 50% or 80% DOD not mattering in regard to total amp hours. What about the effects of total 100% discharging an AGM battery? Does that do more harm beyond just the total amp hours used?

[booster]

Originally Posted by Davydd

With AGMs you mention 50% or 80% DOD not mattering in regard to total amp hours. What about the effects of total 100% discharging an AGM battery? Does that do more harm beyond just the total amp hours used?

If you look at the chart for the AGM lifetime amp hours, there appears to be a bit over 10% difference between 80% cycling and 100% cycling, so not as bad as we have been led to believe

What probably is important at 100%, and to a slightly lesser degree at 80% DOD is that the batteries get charged right away without sitting. More and more information is coming out about that being an issue.

[gregmchugh]

I would think the quality of materials and design and construction will effect the extent to which 100% discharges will reduce the life of a battery along with the other factors. The higher quality batteries from Lifeline, Trojan, Crown, Surrette, etc, may be more resilient than the average batteries from whoever is producing them for the rebranded market at Sam's Club, Costco, etc.

[ptourin]

Over the last few days I've been doing charge/discharge tests, tracking voltage, amps and % SOC against time. I'm trying to get the monitor configured the way I want it, and seeing how the Progressive Dynamics charger behaves. I'm graphing each charging run.

One surprise is that the charger doesn't seem to have a clear constant current stage and a clear constant voltage stage - I see the current start around 50 amps, then it starts dropping as the voltage climbs. There does seem to be a point at which the voltage climbs quickly and the current drops quickly, but it's not as clear-cut as I expected. I don't know if this is a characteristic of the charger or if it's an artifact of how the monitor measures the voltage, but it's interesting to graph.

The other thing is the surface charge issue. Once I turn the charger off and with very slight load, I do see the voltage drop over time. On my second run I charged to 14.5 volts and down to 4.3 amps. Ten hours after turning the charger off, and with only my parasitic load, I was showing 13.3 volts and 98% SOC. This seems in the ballpark. On my current run I charged to 14.53 volts and 2.2 amps (some of this is a bit random - the end of the charge cycle goes fast and I'm running back and forth from the house to the RV). After 90 minutes I was at 13.86 volts. We'll see where it's at tomorrow morning.

I'm guessing that none of the current data is going to be representative - my sense is that I need a few more cycles and that I need to discharge deeper before charging. I also am having some troubles with configuring the monitor. I started by defining 100% SOC as 13.4 volts and tail current of 10% (20 amps) for 3 minutes. I then changed to 5% (10 amps) tail current. Finally, I'm not automatically switching the charger off via the monitor's control relay - at this point I'm doing everything manually.

Once this is under better control I can post some graphs - not worth it yet, I'd say - still in the area of horsing around and getting used to the system.

[booster]

How big is your charger? It should run constant current, at whatever voltage that current can maintain do to load, until the load reduces enough to allow the current to match the charger capacity. At that point the charger will be at rated amps and at the voltage setpoint for the constant voltage stage. From there, it will hold setpoint voltage and current will drop as the battery fills and the batteries accept less. If the charger isn't holding rated amps until the rated setpoint, there very likely could be something wrong with it. The only exception would be if it is getting hot enough to lose output, but that should be less than 5-10% and quite consistent once hot, until the load starts to reduce.

[ptourin]

The charger is a Progressive Dynamics PD9160AL. Before winter, when I did my first tests, it did seem to stay up around 45-50 amps until the voltage climbed up in the 14's. Now the current starts to drop as the voltage climbs - I'm not seeing a clear-cut CC phase and CV phase. But these are only a few tests, I'm doing them manually (that is, no relay control of the charging), and so far I've only discharged to about 75% SOC. So I'll keep at it and see what I see.

My 3rd run is the best looking, in terms of deciding how to set up the monitor. I have SOC defined as voltage at or over 13.4 and tail current down to 5% (10 amps) for 3 minutes or more. Last night I started the charge cycle at 73% SOC and charged past 100% SOC, until I was at 14.53V with current down to 2.2A (didn't mean to go that far - the last bit went faster than I expected <g>...). Once I turned charging off, I dropped to 13.93 in an hour (with only phantom load), and this morning (10 hours later) I was down to 13.65. I'm expecting that when I get home from work this afternoon it will have dropped to somewhere near 13.4, which SmartBattery says is around the full charge resting voltage.

I'm making an assumption about monitor configuration - this makes sense to me at the moment, but I'd like to hear any thoughts. I'm trying to make the monitor's definition of SOC match the manufacturer's specs as close as possible. I assume I want the monitor to calculate based on 100% SOC being 13.4 volts. But if I turn off the control relay and stop the charging cycle when I reach the monitor's definition of 100% SOC, I'm already seeing that the battery voltage will drift down to below 13.4 as the surface charge dissipates - so in truth, I've charged to less than 100%. Luckily, there are relay control parameters in addition to the SOC definition - so if I really want to charge up to 100%, I can open the relay at 100% SOC plus current down to 5 amps or whatever I choose. That is - I try for a good definition of 100% SOC and how far up I charge is an entirely separate decision - I could charge from 30% to 80%, 70% to 100% (or "a bit past 100%"), or whatever I choose. But the monitor's reading of SOC should be pretty close.

Does this make sense - both my explanation and as a sensible way to set the monitor?

[booster]

You are coming up with a lot of good information, I think, even if it doesn't make perfect sense all the time.

We had discussion a while ago here on if the "recommended" charging parameters were now incorrect do to the, now accepted, desire to stop charging at less than 100% full, which in your case looks to be at 13.4 volts rested. At the time, one of the questions was how easy, or hard, it would be to stop the charging at less than the charger voltage setting accurately enough to have the resting voltage land on the target 13.4v consistently. The hope was that by reading the amps, at the voltage wanted, it would be possible to determine the right cutoff spot. I notice you mentioned being able to program hold time on the amps, which many can't do, so that is nice. Looking at the charge profile for the lithium, the voltage climbs very quickly once you reach the 95%is range of charge, so I think that needs to be taken into account. For instance, you hit your 13.4v and it looks at the amps, which may be under setpoint so the timer starts. By the time the amps timer times out, the voltage could be higher by quite a bit, maybe 13.8v, so you are at a higher state of charge. If the amps weren't at setpoint, the higher voltage will hold the amps higher and longer before the timer even starts. That will delay shutoff and give even higher SOC.

My guess would be that the amps hold timer should be set to a very short time, just long enough to cover any inherent variations in the charging amps. It might be a minute, or so, or even less. You would want to shoot for the amps themselves to be at the manufacturer suggested reading, when the timer is done, not when it starts, to be accurate, so it may need to be set a bit higher than you would expect.

Not directly to your settings and current tests, but I think this is really going back to the discussion about charging parameters.

What I think we know

* The 14.6v suggested charging voltage for lithium is based almost completely on getting them to 100% full, as that is what full cell voltage is.

* Based on the above, the charger is charging the batteries to desired voltage by being set at the same voltage as the batteries will be at when done, and rested.

* But now, the recommendations are to not charge completely to 100% full all the time, so a lower finished voltage is needed to be at the lower SOC. In your case the target is 13.4v.

* With the charger set at 14.6v it no longer gets to go to the finished charge voltage and hold and has to be stopped somewhere in the middle of the voltage climb, which is infinitely harder than having it just go to the setpoint voltage and holding.

* Going back to the premise of the recommendations not being appropriate for the lower state of charge desired, is the 14.6v actually the right voltage for the charger because it will never get a chance to run at that voltage during charging because you shut it off before it gets there. Why have a charger at 14.6v if you always shut it off at 13.4v and it never, ever, sees higher than that?

* From all of this, I get the more and more believable (at least to me) idea that the charger voltage should be set at the final rested voltage you want the batteries at with the appropriate amps setting for that voltage derived by watching it level off in testing. If the amps are right, the batteries should come off charge right at desired voltage and if the amps setting is right for some hold time, there should be no surface charge to deal with at all, and the rested voltage should be the same as at shutoff.

* The PD chargers are basically power supplies in most cases, so they shouldn't need a battery reference and are most likely voltage adjustable internally. I wonder if the 14.6v fixed voltage "lithium" charger could be adjusted down to 13.4v to try the above. Marko has been inside his PD, I think, so he may have some insight. A call to PD might also be in order. It would be extremely interesting to see how your system would charge and control at the lower voltage, now that you have the recommended voltage charging to compare to. I don't know where you are located, but it might also be possible to get you the Blue Sea charger, that I use as a bench charger, to test the theory. It has adjustable voltages so you would be able to set it to whatever you wanted to try.

Ptourin may be on the leading edge of actually learning what is really going on with some of the lithium charging issues. Very interesting.

[booster]

I just found this on the PD website sheet for the 9100L lithium charger.

Output voltage can be FACTORY ADJUSTED to meet OEM requirements for various lithium ion battery chemistries and Voltages. Standard “L” Series voltages are set for 14.6- Volts to meet requirements for lithium iron phosphate batteries. Other chemistries may require different voltages.

I also looked at some of the older schematics they have online as they don't seem to have the newer ones posted. All of them showed a 1K potentiometer, so that would probably be the voltage adjustment. The base voltage for the 9100 series is usually 13.6v, so I would think the range is pretty good.

Re the amps dropping before you hit CV stage. The PD specs show all of their 9100 series converters not being able to put out full rated amps above 12.6v, so they are really watt rated. I would expect you amps to start dropping as soon as you get over 12.6v of output.

[ptourin]

>Ptourin may be on the leading edge of actually learning what is really going on with some of the lithium charging issues. Very interesting.<

Ptourin is on the leading edge of wandering around in the swamp and being totally confused by all the gray areas <lol>...

Thanks for all the info and thoughts on the PD chargers. I need to read your stuff over again when I'm not at work and it's quiet, but my first thought is that it may not matter that the charger is set to 14.6. What I'm seeing is that it puts out a lot of current and tries to push the voltage up - as the voltage climbs, the current drops. I have a lot of flexibility with the Victron monitor settings that trigger the control relay. At present I can set it to look for 100% SOC (remember, that's 13.4v and 10 amps for 3 minutes - I'll watch it and see about shortening that) - then it also looks for a current setpoint and opens the relay when it reaches it. I can also specify a max voltage and let it also open the relay.

What this seems to come down to is that we're getting much fussier about exactly how we control charging than I ever did with my older chargers and flooded cells. That's all to the better! - and I expect that after watching how this particular charger works with these particular batteries, I'll end up with a nice charging algorithm - the trick is in understanding both the best practice requirements and the process. At the moment, the second thing seems easier than the first <g>... As I keep watching, it'll become clearer whether the PD's high max charge voltage setpoint is a problem or not.

I expect that if I can get things tied down with the charger, it'll be easier with the alternator, since it only puts out about 14.2 max.

Finally, it's of course easy to be safe by setting a charging goal that's below 100% until it's really clear how this all works, then inching it up. Until now, we often dry camp but we tend to travel a lot and not stay in one place for many days, so I can always alternator charge and rarely need to run the genny. And the 200Ah of lith batteries is so much more capacity than my old 180Ah of standard batteries that I'm not expecting any heavy need to charge all the way up - just trying to understand how it all works.

More later - I'll keep charging and graphing <g>...

[ptourin]

Booster, the only thing you said that I want to understand better is:

"From all of this, I get the more and more believable (at least to me) idea that the charger voltage should be set at the final rested voltage you want the batteries at with the appropriate amps setting for that voltage derived by watching it level off in testing. If the amps are right, the batteries should come off charge right at desired voltage and if the amps setting is right for some hold time, there should be no surface charge to deal with at all, and the rested voltage should be the same as at shutoff."

This doesn't make sense to me at first glance. Say I were to adjust the charger down to 13.4. Are you saying that as the charger/battery system approaches 13.4, the current will keep dropping as the voltage differential between charger and batteries gets smaller - so that finally there will be no differential and therefore no current? And that since you're not forcing a lot of charge current to flow, there won't be a surface charge - so that when you stop charging, the battery will stay at 13.4?

If that's true, I wonder how long it takes - I thought the reason for setting the charge voltage higher was for rapid charging - you push the battery over its rest voltage, and you attempt to disconnect the charger at the right moment so that the battery settles to its rated full charge rest voltage.

I think that if you use the high charge voltage, it's clear that if your charger shut-off setpoint is too low, you don't get full charge. But it's not so clear to me what happens if your setup is too high - discussion seems to be around either your running the risk of damaging the batteries or of moving cells or banks out of balance.

As always - I'm asking questions to see if I understand things correctly. With out the EE background, I keep finding that I thought I understood something but that I actually was missing some point or other - so this is my sounding box <g>...

[booster]

Originally Posted by ptourin

Booster, the only thing you said that I want to understand better is:

"From all of this, I get the more and more believable (at least to me) idea that the charger voltage should be set at the final rested voltage you want the batteries at with the appropriate amps setting for that voltage derived by watching it level off in testing. If the amps are right, the batteries should come off charge right at desired voltage and if the amps setting is right for some hold time, there should be no surface charge to deal with at all, and the rested voltage should be the same as at shutoff."

This doesn't make sense to me at first glance. Say I were to adjust the charger down to 13.4. Are you saying that as the charger/battery system approaches 13.4, the current will keep dropping as the voltage differential between charger and batteries gets smaller - so that finally there will be no differential and therefore no current? And that since you're not forcing a lot of charge current to flow, there won't be a surface charge - so that when you stop charging, the battery will stay at 13.4?

If that's true, I wonder how long it takes - I thought the reason for setting the charge voltage higher was for rapid charging - you push the battery over its rest voltage, and you attempt to disconnect the charger at the right moment so that the battery settles to its rated full charge rest voltage.

I think that if you use the high charge voltage, it's clear that if your charger shut-off setpoint is too low, you don't get full charge. But it's not so clear to me what happens if your setup is too high - discussion seems to be around either your running the risk of damaging the batteries or of moving cells or banks out of balance.

As always - I'm asking questions to see if I understand things correctly. With out the EE background, I keep finding that I thought I understood something but that I actually was missing some point or other - so this is my sounding box <g>...

What you describe is what I am suggesting might be a better method, setting the charger voltage at the finished voltage wanted.

My guess would be that it would not a whole lot longer than the normal charge you are seeing right now, because you are shutting off the charger when it hits a selected point that considerably under the 14.6v setpoint of the charger, and the batteries are taking as much as the charger can put out at that voltage or the voltage would go to setpoint. If you think about what happens at if you let the charger run to 100% full on the batteries, the batteries will be at 14.6v, which is considered the full charged voltage cutoff for them. They then drop back to 13.4ish and are considered full. I think it is reasonable to think that the amps that go into the batteries once they pass the 13.4 volts in the charging stage, is going to surface charge, with a possibility of some going to balancing depending on when they do it. The surface charge comes off when you lose the amps it took to go up to the 14.6v from the 13.4v. Now if you had a charger big enough to hold the 14.6v at the battery terminals (inside you would have the actual voltage of the cells), you would charge faster because of the resistance from the terminal to the cell. With the smallish charges in the real world, that voltage drop would be very small at the amps that are going into the battery by the time you get to the 13.4v, I think. It would be likely that the amps would go to zero after a short time held at 13.4v, and that would be 100% full. How long that would take is the big question, along with whether or not you need, or want, to go to that high of a SOC.

On the issue of not getting full, stopping lower is the whole point of doing that, I would guess, as they say not to go to full all the time. But, does that mean not have all the capacity, or not give it the 14.6v instead of the resting voltage. My guess would be the voltage to be more of an issue, but I have never seen any explanation of that.

I am not an EE, either, so we are both in the same electric boat.

Here is a chart I just found on the Cruiser's forum. Can't vouch for the accuracy of it but it is interesting. It shows the charge voltage going high after it passes 13.4v at the knee of the curve, and then dropping back quickly once charge is removed. The interesting part is the both points of 13.4 volts appear to be called out as 100% state of charge. The current knee appears to be at about 14.4v, but what isn't clear is if the amps from 13.4v to 14.4 are going to surface charge (I assume like a capacitor so would pull lots until full and then stop) or truly to capacity. The thing to remember would be that even if the batteries were not 100% full, that would be the goal as we are told totally full isn't good, and they like 95% area. The test would be to charge at the lower voltage and see what the amps would be at 13.4v charger setting, and how long they took to go to zero (which would be totally full and you don't want). Someplace in between or at the 13.4v point may be good. The proof would be in where it rests after disconnection, which should be just a little less than where they would rest after a 14.6v charge. I need to look back and see if davydd ever mentioned what voltage his charging stops at, as his is set to go less than 100%. Even if it is SOC controlled, the voltage should be a good indicator.

[gregmchugh]

ptourin, I followed your thread over on the Unity group on the Sprinter forum awhile ago but I don't recall which lithium battery bank you purchased and what type of battery management system is included in your battery bank. Aren't some of the risks of battery overcharging and excessive charge voltages prevented by the battery management system or is it just providing cell balancing of some type?

[booster]

Originally Posted by gregmchugh

ptourin, I followed your thread over on the Unity group on the Sprinter forum awhile ago but I don't recall which lithium battery bank you purchased and what type of battery management system is included in your battery bank. Aren't some of the risks of battery overcharging and excessive charge voltages prevented by the battery management system or is it just providing cell balancing of some type?

Good question. I think most of the BMS overvoltage shutoffs are at voltages considerably above the 14.6v, though.

[ptourin]

Greg, I bought 2 100 Ah SmartBattery drop-ins - that wouldn't be my choice now, but it's what I did back then. The company is very reticent about any technical info - their advertising is unclear and calling them multiple times doesn't really answer questions. For sure each battery has an internal battery protection system, with an internal protection relay that opens at over 15.8v or below 8v. They talk vaguely about battery management and balancing banks of cells, but nothing at all is clear about what they say.

They also say that during manufacture, all cells are carefully measured, and that each battery's cells are very carefully matched, so that they don't believe that balancing will be necessary over the battery's life. So at this point I can't say that I have much confidence in them as a company, since multiple calls have never gained me clear answers. They do have a 5-year guarantee, though who knows what it might cover.

[ptourin]

Booster, that chart is how I visualize charging - that you charge to a voltage higher than the full charge resting voltage, let the charge current drop, disconnect the charger, and (assuming no load, for the purpose of discussion) watch the voltage drop back down to the resting voltage (or lower, if you didn't charge high enough).

But this takes us to the issue of surface charge. What do we (meaning you <lol>...) know about surface charge? I'm not sure from what you wrote that I understand it at all. I've been visualizing it as a means of artificially pushing the battery voltage high - I haven't pictured it as something that can actually do much work, so to speak, but as something that pushes the battery away from its rest state during the charging process, and that the battery will move back to the rest state after some no-charge/no-load rest time.

Another way to say it - I pictured that any charger must supply a voltage higher than the battery voltage for charging current to flow - that no matter at what SOC we start charging, the charger must always be a bit higher than the battery. At some point the battery approaches the charger voltage, charging slows down and stops - and I've assumed that if you then disconnect the battery from charger and load and let it rest, the battery voltage will always drop somewhat. Am I misunderstanding surface charge? Do you know anything that will give a better picture of what it involves?

[ptourin]

A bit more on my simpleminded battery cycling. As I said, yesterday I charged up to 14.53 volts and down to 2.2 amps before I shut the charger off. 30 minutes after I turned the charger off and totally disconnected all loads, I was down to 13.93. After an hour, down to 13.86. After 11 hours, down to 13.65; after 21 hours, down to 13.6. So after almost a full day, I was still above what SmartBattery gives me as the full charge resting voltage of the batteries. I don't know if I've "filled the batteries too much" - they're not bulging and groaning <lol>... I assume I'm seeing the surface charge effect here, and I don't since I had no load on the battery except that of the monitor, it's hard to imagine that extra voltage is good for any real work.

This evening I'm working on another test cycle. I've got about a 50A load on the batteries. It's been on for about an hour plus a bit. According to the monitor, I've used 70 Ah so far - that seems about right. It's showing 62% SOC - that's somewhere close, as I have the discharge floor set at 20%, or 160 Ah down. Interestingly enough, under this load the battery is reading 12.73v - I didn't expect to see that much drop. I'll disconnect the load now and we'll see where things go - I'll let it rest for a while, then charge again.

[booster]

My understanding of surface charge in lead acid is that it is a delayed chemical reaction do to migration of stuff in the battery, or something similar. I haven't been able to find anything on what actually that little bit of extra voltage gives you in lithium other than the below, at least with an explanation of it.

What I did find was a chart that Powerstream made after testing batteries at different charge voltages (just what we are talking about).



The page it is on is here:

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

I haven't closely scoured the entire article, but it certainly looks like when you charge at 3.4v, you get to 96-99% full, which is higher than the 90-95% many recommend now, so could go even lower I would guess.

[ptourin]

I'm wondering if "fully charged" isn't a somewhat narrow range between "I can stuff a bit more in there" and "whoops, I pushed too hard and damaged it". That is, I wonder if the manufacturers have done a bunch of tests in which it's to their advantage to have as much capacity as possible, but to their disadvantage to have their product damaged by overcharging. Are we thinking it's more of an absolute concept than it really is?

I remember looking at several pages on that site. This chart interests me because they use "capacity at 3.65V" - or in the 12 volt arena, at 14.6V.

The chart shows that charging to 3.3 starts to get the chemistry going, and 3.4 is pretty close to rated capacity for all batteries. So moving to the 12 volt arena again, you'd want to charge at more than 13.2, but 13.6 gets you almost all the way there for all the batteries tested. Again, I'm wondering what that means - whether that "97% of capacity" isn't a somewhat squishy concept. And there's nothing here about charging time - I don't remember if there's anything about charging time in any of their other charts or discussions.

[booster]

As I mentioned earlier, I think the 14.6v came from early specs that optimizes capacity before the figured out that totally full and higher voltages weren't all that great for the batteries. From what we hear now, being lower than 100% is good, so 95-98% would be a good thing and could be charged at a much safer and easy to control charge profile at 13.4-13.6v maybe (Powerstream mentions each brand had slightly different specs they wanted). I think the 97%, or whatever you got at the voltage you chose, could probably be a pretty consistent as long as you had the control watching the amps to stop the charger at the same place all the time. I also did find some stuff on one of the electric vehicle sites where posters were pushing charging at the theoretical full voltage for their batteries of 3.33v per cell, claiming it would be best for battery life.

I have not been able to find anything on charging speed at the lower voltages, but that would be a very easy A/B test to do. I have done many similar tests on our AGM setup to see what various voltages and times do. The good thing is that if you did run out of time on some cycles, there really is no issue with the lithium batteries that don't need to ever get totally full. All that would happen is you might lose a few percent of capacity on that discharge cycle.