Voltage jumps as Lithium Iron Phosphate (LiFePO4) batteries warm up

[spk]

I set up my camping system as a semi-permanent fixture in my shed to experiment with this year, and up in Minneapolis temperatures have been less than balmy. I added some heating pads and wrapped my battery box with insulation, but as the days are shorter and the temps colder, I've gotten in the habit of shutting down the loads (including the Venus OS system monitoring things) overnight (using a Victron Battery Protect as a load switch) and then starting it in the morning when (if) it get's sunny.

Batteries are 4 Dakota Lithium 23ah 12v, for a 2p 2s 24v system.

It was colder today (-10C or so outside, battery maybe -4C) and I noticed the battery (at 70% SOC per the smart shunt) was a bit lower voltage than I expected when I started the loads (including the battery heat). I noticed one sharp jump on the bottom pair as they warmed, followed by another sharp voltage increase as the top pair jumped as warming continued. I'm guessing the heating is uneven, or maybe some slight chemisty differences. I'll see if I can post the graph from VRM:



The small jump on both sides is when the heater turned off when the batteries got to 7C or so. The MPPT won't charge below 4C.

Anyway, just curious if this kind of sharp voltage jump is normal with LFP, and if it's chemistry or BMS maybe? I haven't got data on the cooling side, as the computer doing the monitoring (BeagleBone Black with Venus OS) is shut down at night. I tried letting it run but we already had a couple of weeks mostly overcast and even the 8 watt system load drains it down over time.

 

[Ampster]

There is a big difference in voltage between LiFePO4 (LFP) and Lithium Polymer (Lipo) batteries. Please change the title to reflect the correct chemistry to avoid confusion. Lipos do get hot, so much so that some people charge them in fireproof bags. On the other hand LFP should not be charged below freezing.

 

[spk]

Thanks, updated. Also corrected the battery AH and mentioned low temp cutoff is set in the MPPT at 4 degrees C so I'm not charging until they're well above freezing.

 

[mikefitz]

The battery BMS is shutting down the charge path when the batteries are at low temperature. This introduces a volt drop across the fets of approximately 0.6 volts between the actual series cell volts and the battery terminals. When low temperature recovery temperature is reached the charge path is closed and the volt drop is removed. This is what you are seeing.
Due to BMS setting tollarance and perhaps different temperature of the batteries there was a time difference between the recovery points.
The smaller jump of 0.1 volts that occurred was perhaps a change in load, heaters?

Mike

 

[RCinFLA]

With no load, open circuit equilibrium, on cells, LFP cells have very little cell voltage change over temp.

Loaded with charge or discharge current, LFP cells have a significant change in their internal impedance over temp. Particularly at colder temps there will be much greater terminal voltage slump with moderate discharge current due to cell impedance degradation.

The internal impedance is also dependent on the aging of the cells. It increases 3 to 5 times near end of useful life.

Your chart makes no sense as it appears some cells are totally dropping out because they are depleted cells. Why the BMS is not shutting down is a mystery.

You need to report your load current on battery, but BMS should have shut it down with though voltages.
DISCHARGE SPEC
24 A max continuous, 60 A max pulse (<3 sec).

Since battery appears to be LFP your charts is showing some cells are totally discharged.

Fully charge to 14.4v and repeat test.

 

[spk]

I like the BMS low temperature charge cutoff theory as it fits the voltages of the jumps as the batteries warm, but I can't see any documentation at the vendor that states they have a temperature sensor or low temp charge protection.

Loads during morning heating are about 1.6 amps. The smaller jump shown on both bottom and top halfs of the battery bank are indeed when the heat turns off once it hits the temp set on my 24v thermostats.

I don't think I am dealing with depleted cells, as the Smart Shunt has the SOC above 70% when this happens, although it's been a few days since I was last at 100% and synced, sometime last week. Typically it hasn't been that far off that I'd have depleted cells and still show that high. ALso, since it happens on each half of the battery bank, just a few minutes later, it seems to me I'm probably unevenly heating the batteries, maybe one of the heating pads is under spec, or the insulation has a gap on one side or something.

I imagine I'll have more colder days in the future to observe this in more detail.