LiFePO4 Batteries: Discussion Thread for Those Using Them as House Banks

[evm1024]

Originally Posted by klaus53123

Ok, I tried to charge them higher and set the end of charge to 3.5V. Whilst batteries 1-3 reached that threshold with voltages very close together, battery 4 remained at 3.14 V.

Now the question is: does battery 4 just require a manual top balancing or is it damaged? Any idea how to find out.

Once again I would appreciate it a lot if you could advice on a charger suitable to top balance with.

Many thanks
Klaus

That is pretty odd - I've been doing some research with some smaller cells (8 ah headway cells) and in general beat the **** out of them. In one test I drew the cells down very low. To the point that one of the cells failed.

The failed cell got quite hot and ended up with 0 volts across the terminals. It gets quite hot if you try to charge it. This is expected for an under voltage failure mode.

The other cells likely have a reduced capacity. They still do meet spec and have 8 ah capacity. New they were about 8.15 ah so the loss is not great.

"Weak" cells tend to rise in voltage quicker that stronger cells (they have less capacity so they charge faster for a given current - series charging). And go to a lower voltage quicker than stronger cells.

Back to your cells - the cell that does not get to the same voltage as the others should be capacity checked. Isolate the cell and try to charge to 3.65 volts. If it still does not charge to 3.65 then I would say that is is bad and should be replaced.

If it does reach 3.65 volts and the capacity is lower than the other cells then it will be the limiter of your banks total capacity. This bank capacity may be OK for your use case or the banks capacity may be too low. Up to you.

I think that It would be a good idea to check the capacity of all of your cells and to decide from there what to do.

If you have more than one cell in parallel (and 4 of those in series) you can mix and match the cells so that each parallel group has about the same capacity as the other parallel groups. This allows you to maximize capacity.

For my house bank I have 4 700 ah cell is series (1P4S) so I do not get to do that. My weakest cell sets the bank capacity at 520 AH or so. (They were B grade cells to start with)

Hope this helps...

[Ontario32]

Talking about one bad cell in the bank...our Mewyeah(?) bms system shut things down as it should, after one of our 8 Winston 400s (in series to 24 V) dropped to ~2.6volts. Yes, I am new to lithium batteries, no I have not been keeping track of individual cells lately (too much other stuff seemed more urgent). Cells as follows: 3.302, *3.250, 3.330, 3.332, 3.336, 3.334, 3.333, 3.335V. I have been trying to raise that lowest cell by several methods. A 150W/38V unregulated semi-flex solar panel (lost overboard in a big wind gust in a temp location...yes, dumb), a small AC charger 450mA/4.2V, and now a 50W/18V unregulated solar panel. Very slow process, we celebrate each .01 V gained. Been at this for a week. The question is: Can I safely speed this process up by hooking up an AC charger for 12V systems with 6 amp output and watch it like a hawk with a voltmeter? This is all very new to us, not our install, bought the boat with it.

[CatNewBee]

Originally Posted by Ontario32

Talking about one bad cell in the bank...our Mewyeah(?) bms system shut things down as it should, after one of our 8 Winston 400s (in series to 24 V) dropped to ~2.6volts. Yes, I am new to lithium batteries, no I have not been keeping track of individual cells lately (too much other stuff seemed more urgent). Cells as follows: 3.302, *3.250, 3.330, 3.332, 3.336, 3.334, 3.333, 3.335V. I have been trying to raise that lowest cell by several methods. A 150W/38V unregulated semi-flex solar panel (lost overboard in a big wind gust in a temp location...yes, dumb), a small AC charger 450mA/4.2V, and now a 50W/18V unregulated solar panel. Very slow process, we celebrate each .01 V gained. Been at this for a week. The question is: Can I safely speed this process up by hooking up an AC charger for 12V systems with 6 amp output and watch it like a hawk with a voltmeter? This is all very new to us, not our install, bought the boat with it.

If you want to blow it too, yes. LFP cells have a very low resistance, means it is like a shortcut to your charger. Some survive blowing the fuse, some not.

But you can actually add a 55W 12V head lights light bulb or 2 in series to burn the extra energy and limit the current to the current they take. (about 5A) , if you take two in parallel and then connect them inline you are good for a 10A current limiter.

Redneck solution for the jungle, if you have no other options left.

[Ontario32]

I am getting 2.6 amps from the unregulated solar panel right now. Going into a 400 amp cell it takes hours just to gain a few tenths of a volt. Maybe this is due to the cell having actually failed and has much higher resistance than it should, but it seems like the AC 12V/6A charger may be worth a try. At 3.275V now. Anyone?

[newhaul]

Originally Posted by Ontario32

I am getting 2.6 amps from the unregulated solar panel right now. Going into a 400 amp cell it takes hours just to gain a few tenths of a volt. Maybe this is due to the cell having actually failed and has much higher resistance than it should, but it seems like the AC 12V/6A charger may be worth a try. At 3.275V now. Anyone?

Not a good idea at your 3.3v thats 24 amps and will definately spike the voltage above catastrophic failure point real fast .

[Ontario32]

Still hanging in there with the 50W solar panel. After days of overcast, we may finally hit 3.300V...the cell is probably failing, since it self/discharges much more than the others, which are pretty much where they were two weeks ago when I disconnected them from the house system.

[Ontario32]

Now that low battery is up to 3.300 volts I have series paired it with another cell which was at 3.300 volts and am bringIng both up to...? The other 6 cells are at 3.331 to 3.335. Should I go to ~3.333 for each of these 2 cells, or an earlier post suggestion of 3.65V?

[Wood]

IMO I would not. You risk bringing the good cell > 3.65V. You should just do it separately. You can get a variable power supply 10a, 0v-30v for less than $50US. I have a 30a and it was less than $90US

[Ontario32]

Thanks for the warning, disaster staved off. Good thing this method is so slow! The gain was in the healthier of the two cells. #1 is at 3.332V and #2 is 3.307. Can you recommend a source for the power supply you mentioned, please? Need to have it mailed to Culebra, Puerto Rico.

[Wood]

Here's a 10amp model: https://www.ebay.com/itm/30V-10A-Adj...EAAOSwz2VeevPi

This is the 30amp one I got for $85 [currently sold out]: https://www.ebay.com/p/2146314506

These are not really "lab" quality. You will want a reasonably good multimeter to double check your settings. Although the one I got tracks with my fluke meter.

[Ontario32]

Thanks, I ordered the 10A power supply from someone who would ship to PR. Looking forward to trying out the bank soon.

[rgleason]

Blog: s/v Gudgeon:LiFePo4 Battery Install



Design: Orion JR BMS, 400ah Lifepo, solar, alternator, ac charger.
Reference to Nordkyn Design Electrical Design For a Marine Lithium Battery Bank


Eight parts in total.


Has an excellent level of detail and shows the practical aspects of building a system.
Not so sure about the 70a relays used.

[DeValency]

All,
Quite amazing. This thread started in 2011! So much has been changed and developed since. Fascinating times for tech sailors.

*** interested in the experience of others in using DC-DC charger. AGM cranking/backup of 150-200A to a new LFP of 250-350Ah or higher house bank with BMS.***

1. as a way to maintain the existing alternator, no external programmed regulator etc.
2. no change of charger. Unfortunately, my Victron Phoenix doesn’t support LFP charging pattern

Overall, to simplify the system yet keep it safe with all the components involved.

I know this thread went way over the basics and intend to review as much as I can in these curfew days/weeks (who knows...) Any comments and practical suggestions with such an architecture considering the present state of technology and devices?
Thanks!

[Delfin]

Originally Posted by evm1024

I would suggest that you get a Victron BVM702 or the like and measure the AH used. You can then either recharge when you reach a specific consumed AH or build a table that equates SOC with voltage.

Given a sync at full the consumed AH is quite accurate.

It will also allow you to measure the actual capacity of your cells.

Take a look at these curves which show that capacity differences are quite minimal between charging to a limit of 3.5 to 3.7 VPC. The idea being that charging to a lower voltage will give greater cycle counts over charging to a higher voltage.

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


Be aware that Xantrex had 2 different models of their SW-3012. The newer model has user settable charging parameters.

I had the older model and was always fiddling with the charge parameters (mostly the max charge current limit). I sold that and went to a Victron Multiplus 12/3000/120-50 120V which allows me better control over charging and discharge parameters.

The only thing that I really liked about he Xantrex was the ability to just do pass through I.e. disable charger and inverter. This can be done on the Victron with the Two-Signal BMS assistant installed.

I also have Winston 700 AH cells in my house bank. I did buy seconds so one of the cells limits the capacity to 500 AH. I've been running them since 2014 and also love them.

PS in my experience getting accurate cell voltages is difficult with all of the "inexpensive" cell monitors. I could not say about your bms. But, I do check the accuracy of all measurements in my boat with a Calibrated Fluke 87V meter.

Ethan, in your opinion, is the memory effect enhanced when consistently charging to 3.5 vpc as compared to something higher? I recall this was extensively discussed earlier, but don't remember if a consensus on the issue was reached.

Totally agree with monitoring Ah used to trigger re-charge, when solar isn't part of the mix. I don't have solar, so re-charging at 25% SoC based on Ah consumed is usually what I do.

[evm1024]

Originally Posted by Delfin

Ethan, in your opinion, is the memory effect enhanced when consistently charging to 3.5 vpc as compared to something higher? I recall this was extensively discussed earlier, but don't remember if a consensus on the issue was reached.

Totally agree with monitoring Ah used to trigger re-charge, when solar isn't part of the mix. I don't have solar, so re-charging at 25% SoC based on Ah consumed is usually what I do.

Firstly I want to distinguish between memory effect and loss of capacity (that we have been calling memory effect).

Memory effect as the researchers is a "small bump in the charging profile that is an "echo" of prior charge terminations. During charge Li ions leave the cathode and work their way to the anode. A fully charged cell has virtually no Li on the cathode. This echo appears to be the results of a "logjam" of Li in the anode that blocks the movement of Li ions from penetrating deeper into the anode. The internal resistance goes up and thus the voltage bump forms. At some point the logjam is broken as Li ions move deeper into the anode.

This memory effect is reversible through complete charge/discharge cycles.

Oh, Don't expect me to be an authority on this!

The memory effect I and other were discussing way back then was actually capacity loss. Capacity loss happens many ways, loss of active area on the anode or cathode, loss of electrolyte in areas, loss of contact (between the anode, electrolyte region and cathode) due to expansion to name a few.

My investigations was an attempt to recover lost capacity by trying a number of charge discharge cycles at various rates (currents) and end voltages.

Basically I did not see any capacity recovery. I did see that with precise charging I could maximize the available capacity (really fully charged vs nearly fully charged).

To more directly answer your question: The presumption is that the memory effect "logjam" can grow over time to the point of significant capacity loss. I have not seen that myself or read of that in any paper. (which of course only shows the limit of what I know).

I have wondered if the thought that the "logjam" can grow to be responsible for capacity loss is a projection of Lead Acid sulfation into LiFePO4 chemistry rather than an observed reality. Again, just what I recall, no facts behind it.

Charging to 3.5 vpc does fully or nearly fully charge your cells in terms of capacity. Thus the memory effect bump (and logjam) does not happen or is at the very upper end of the capacity curve.

The consensus if there was one was to play it safe and to charge to 3.6 or 3.65 vpc every 3 to 6 months or so. That activity (charging to 3.6 vpc) would not hurt I suspect so long as the charge cutoff limit does not cause any of the capacity loss as defined at the beginning of this post.

Hopefully that charge to 3.6 vpc is not an echo of Lead Acid equlization...

One final thought is that the reason that Lead Acid batteries are toast at 80% capacity is that they (more or less) are just not able to take a charge and deliver what ever charge they have. The internal resistance is so great that they have a huge voltage drop at any current of note.

LiFePO4 on the other hand at 80% capacity still have low internal resistance and behave like a battery with overall lower capacity.

In my case my B grade 700 AH cells still deliver 500 AH at 100 amp discharge rate. So I have in effect a 500 AH bank.

Regards

PS (defensive hat on) This is just what I'm thinking - please do not confuse this with anything else. Always welcome more accurate information.