Self-heating battery: Problematic behaviour?

[NikR]

I am running test with a self-heating battery, and I noticed a certain behaviour when charging at low temperatures, and I am wondering if that could cause a problem down the road.

The battery is at -11°C (below freezing), and when I apply a 1 A charge current, the following happens in the BMS-monitoring app:

1. The BMS accepts the charge current
2. A few seconds later, the BMS throws a "Charge undertemp!" and stops the charge (as it should)
3. Another second later, the BMS registers a ~100 W load outgoing, and resets the charging alert (allows charging)
4. The voltage drops, the outgoing load stops, the cycle repeats at step 1.

All this happens in a couple of seconds, so that the error statistics for low temp charging accumulate very fast.

My reading of the situation is the following: the charge current is stopped by the BMS, and the heater kicks on, and is turned off right after because the voltage drops. This might be because my charging current is not enough to run the internal heater, but it's a realistic scenario with low solar input int he winter.

My questions are:

1. Is it problematic that the BMS allows a small charge current every couple of seconds at this low temperature?
2. Will the frequent switching between states cause any problems down the line?

 

[MaikaiLifeDIY]

The small current shouldn’t hurt the battery since it’s very minimal and temporary.

What is most likely to happen is the battery may continue to discharge to unhealthy levels if you can not get a warm enough temperature to charge the battery.

Since you’re in a loop situation you’re gonna have to figure out a way to break the cycle. Perhaps you need some other form of charging like wind.

 

[v_green57]

Yeah....this kind of "Catch-22" is why I looked very carefully at "self-heating" batteries but in the end went with naked and came up with a scheme that breaks the loop:

LFePO4 in remote cold/snowy install

Per subject line. I have a small wireless network in a remote area which uses solar power at all the sites. I have been using AGM batteries since I set things up in 2008, replacing them as they reach EOL. Recently, the AGM at one of the sites starting showing its age, so I thought this might...

diysolarforum.com

This has been working well.

 

[LakeHouse]

The battery is at -11°C (below freezing), and when I apply a 1 A charge current, the following happens in the BMS-monitoring app:

1. The BMS accepts the charge current
2. A few seconds later, the BMS throws a "Charge undertemp!" and stops the charge (as it should)
3. Another second later, the BMS registers a ~100 W load outgoing, and resets the charging alert (allows charging)
4. The voltage drops, the outgoing load stops, the cycle repeats at step 1.

I see a few issues with what the BMS is doing...
#2. 'A few seconds later...' I would suspect this should happen faster than a few seconds, but I don't have a good feel for how long it takes to cause low-temp charging damage. In my testing, my JK BMS disabled charging instantaneously (or rather, faster than I could register) when it recorded a temperature below my set point.
#3. This might look like an error, but it is at least sort of what should happen... If the BMS is going to allow discharge, it needs to enable the charge MOSFETs for reasons detailed here. I'm not sure I agree that it should clear the error, but enabling current flow in both directions is necessary.
#4. This is the real problem... Unless the battery is discharged to a point where the 100W load causes a low-voltage disconnect, the heating circuit should stay on until the under-temp problem is actually cleared. On the other hand, if the battery is fully discharged, I don't see a way out of this loop until whatever is powering the system is providing at least as much power as the 100W heater is requesting.

 

[NikR]

Thanks, very interesting points!
The configuration of the BMS says the "charging undertemp trigger delay" is 5 seconds, which checks out. What does not check out is the return condition 5°C, it clearly returns right away, maybe because of what you said about the MOSFETs.

I tried a more capable charger that supplied the 100W for the heating function, and the BMS did not switch state, and only run the heater (constant 8 A). I guess that makes it at least okay for generator charging, but I don't feel comfortable with continuous solar charging.

Oh well...