Be careful using LiFePo4 batteries

[Maple Dave]

About 4 yrs ago, I watched Will Prowse assemble LiFePo4 cells and I jumped in to the tune of about $2500 to get 16 cells. I have a grid connected/battery backup solaqr system at my sugarhouse (maple syrup).
I assembled the 16 cells in series, used a BMS that claimed to shut off at just above freezing and put a 65 watt heating pad under my 48V 100AH battery bank. I had 1" blue foam under and all around the battery bank to help contain the heat. The connecting cables were copper 4/0 high grade from a big name US based solar company, made to my specified lengths.
I am located in central New York State, about 25 miles east of Syracuse. The sugarhouse is not heated. I have only one small heated closet in the sugarhouse with 2 forms of heat, a propane 8000 BTU wall furnace which has a standing pilot, and for back-up a 1200/1500 watt electric heater. The furnace thermostat is set for 40F and the electric is set at 36F.
We had a major snowstorm which covered the solar panels, and I couldn't get into the sugarhouse and solar panels for 2 days. When I got there I first cleaned off the solar panels, then went into the sugarhouse. The grid power and the propane furnace both worked as they should, but the BMS failed, and apparently tried to charge the battery bank when it was frozen, the 65 watt heating pad had drawn the 48V 16 cell bank down to 17V. I tried to fix them, I unbolted the bank (it was 8x2 cells, each row of 8 had a long bolt to clamp them together. I then took 8 cells home and tried to charge them using an adjustable charger. I set it for 1.5V and 2A for a day. My hope was that it would accept the charge and I was going to gradually increase the voltage until I got each cell up above 3V, then I planned to bolt them back together and connect the solar to them to fully charge the whole bank When the low voltage didn't work, I stepped the voltage up about 1 volt at a time until I was at 14V, the highest my adjustable charger was capable of. The first cell never took a charge, I tried a few, each 1 at a time with no luck. $2500 worth of cells was now worth nothing, total scrap.
If you go with LiFePo4 batteries, be sure you don't make the mistakes I made. I've been out of the 1480 watt portion of my solar array ever since (actually 4 years next January) Next week, I'll finally get 4 12V AGM batteries so I can again send my excess electric thru my net metered system. I'm using AGM because they require no maintenance and can be charged at any temperature I'll ever get. I no longer use it for emergency back up heat, but my Xantrex 6048 inverter can't be run without the battery bank attached and charged. Then it is set to send power to the grid once the battery bank reaches 56V. That is then credited to my residence bill, I live 3 miles away.
By the way, we also have a 1770 watt fully off grid solar set up at the residence with 4@6v each Rolls Batteries each 853AH for an independent 24V system. It's not connected to the grid power at our house.

 

[Danke]

Be careful abusing LiFePO4 batteries and faulty BMS.

 

[OffGridInTheCity]

Sounds like the key issue is the BMS you used. It should not have allowed 1) discharge of a 48v battery down to 17v or 2) charge at cold temps.

 

[Bud Martin]

So the BMS failed to perform low Voltage shutdown and also failed to stop charging when the temperature is below freezing.
What is the BMS make and model?

 

[Bluedog225]

Paragraphs please. Sorry for your loss.

 

[Substrate]

A little knowledge goes a long way, but not apparently enough for this diy attempt.

Unless he has a remote temperature sensor on those agm's to rely upon for accurate temperature compensation, that can be a problem too. For example, if he relies solely upon the ambient temp sensor in the SCC in the shed, and the agm's are in the basement, the temp difference between the two will result in innacurate temp comp.

Add to the fact that lead does not acclimate to rapid temperature changes quickly - an agm can be degraded quickly if not using a remote temp sensor.

Ah, the details!

 

[nc73]

This is why I don't have my batteries in an unheated environment. If it's in the house, I can keep an eye on it. No frozen batteries either.

 

[OffGridInTheCity]

This is why I don't have my batteries in an unheated environment. If it's in the house, I can keep an eye on it. No frozen batteries either.

My understanding is that 'freezing' is not the issue - its trying to discharge/charge during out of spec temps that's the problem.

In other words, a LifePo4 (stilling idle - no charge/discharge) that goes thru say -20F for a week but is then allowed to warm back up to operating temps - will likely be able to resume normal charge/discharge operations.

I'd be interested in clarification on this.

 

[chess-equality]

Can AGM also be assembled DIY-style from cells? I think if you're assembling LFP cells without any experience, better to start small and with small loads, just to get enough experience before going 16s.

 

[rdhood]

Which BMS? Did it have a temp sensor, and did you actually test (by throwing it in a glass of ice water or something)? Did you ever discharge enough to test low voltage cutoff? Both of those would be prudent when putting a lifepo4 into service in below freezing temperatures.

 

[nc73]

My understanding is that 'freezing' is not the issue - its trying to discharge/charge during out of spec temps that's the problem.

In other words, a LifePo4 (stilling idle - no charge/discharge) that goes thru say -20F for a week but is then allowed to warm back up to operating temps - will likely be able to resume normal charge/discharge operations.

I'd be interested in clarification on this.

yes but if indoors in a conditioned space you won't have this issue.

 

[cs1234]

Can AGM also be assembled DIY-style from cells? I think if you're assembling LFP cells without any experience, better to start small and with small loads, just to get enough experience before going 16s.

I haven't seen them for sale.. a 12v AGM battery is composed of (6) 2v cells internally I believe. You can essentially make your own 12v flooded lead acid using 2v fla batteries(cells), but I haven't seen them come as naked as LiFePO4 or NMC cells.

 

[Substrate]

Can AGM also be assembled DIY-style from cells? I think if you're assembling LFP cells without any experience, better to start small and with small loads, just to get enough experience before going 16s.

YES! You have the right idea. Get experience with a small setup if you want to learn before you spend $$$.

Enersys / Hawker Cyclons cells - *PURE LEAD*. Very small in capacity, so makes it easy to cycle for learning. Build up what you need.

Amazingly, this was the very first format for agm. Sales guys used to walk around and melt a paper-clip with a single cell. Like from the 70's. And they still look like it!

So compared to conventional agm's, their pure-lead charge/discharge rates are mind-blowing. Like their bigger brothers.

The problem is getting fresh-stock, since the use-case for these small versions are pretty narrow. But what a trip to hammer them at 1C (provided your CV is tight) and have them just laugh at you. But you treat them just like you would a Hawker in an Abrams tank with the same specs.

Fun stuff to learn with, but obviously not cheap.

 

[Luk88]

We still haven't learned which BMS was in use (or did I miss it?).

Just building my first lifepo pack now I'd very much want to know which brand to avoid.

 

[seneysolar]

About 4 yrs ago, I watched Will Prowse assemble LiFePo4 cells and I jumped in to the tune of about $2500 to get 16 cells. I have a grid connected/battery backup solaqr system at my sugarhouse (maple syrup).

Do the batteries also charge from the grid?

I assembled the 16 cells in series, used a BMS that claimed to shut off at just above freezing and put a 65 watt heating pad under my 48V 100AH battery bank. I had 1" blue foam under and all around the battery bank to help contain the heat. The connecting cables were copper 4/0 high grade from a big name US based solar company, made to my specified lengths.
I am located in central New York State, about 25 miles east of Syracuse. The sugarhouse is not heated. I have only one small heated closet in the sugarhouse with 2 forms of heat, a propane 8000 BTU wall furnace which has a standing pilot, and for back-up a 1200/1500 watt electric heater. The furnace thermostat is set for 40F and the electric is set at 36F.
We had a major snowstorm which covered the solar panels, and I couldn't get into the sugarhouse and solar panels for 2 days.

Was grid power off for 2 days?

When I got there I first cleaned off the solar panels, then went into the sugarhouse. The grid power and the propane furnace both worked as they should, but the BMS failed, and apparently tried to charge the battery bank when it was frozen, the 65 watt heating pad had drawn the 48V 16 cell bank down to 17V.

Trying to understand the timeline here: First snow covers panels, no PV - gotit, but then what? Power goes out and the system acts as backup power for the furnace? assuming the electric heat didnt kick on being lower thermostat temp. How long can your system run this equipment? Then the batteries froze and power came back on? This all seems like it happened very fast.

I've been out of the 1480 watt portion of my solar array ever since (actually 4 years next January) Next week, I'll finally get 4 12V AGM batteries so I can again send my excess electric thru my net metered system.

Did this happen 4 years ago, and you were trying to revive batteries now or back then?

I'm using AGM because they require no maintenance and can be charged at any temperature I'll ever get.

Yes, here in Norther Michigan, I never recommend LFP for unattended cold weather sites. Too many variables, components to fail, specific parameters to be kept. I primarily use Trojan FLA or Full River AGM for cold operations. They simply work.

If you cant tilt panels to shed snow- Maybe add a single panel that is vertical on its own small charge controller. This acts as a trickle charger in the winter for roof mounted systems that get snow covered for long periods of time. Keeps batteries from freezing (assuming loads are removed).

I no longer use it for emergency back up heat, but my Xantrex 6048 inverter can't be run without the battery bank attached and charged. Then it is set to send power to the grid once the battery bank reaches 56V. That is then credited to my residence bill, I live 3 miles away.
By the way, we also have a 1770 watt fully off grid solar set up at the residence with 4@6v each Rolls Batteries each 853AH for an independent 24V system. It's not connected to the grid power at our house.

In your scenario it seems the batteries were ran down by something other than the 65W heating element? No matter what battery chemistry you have. AGM or Flooded lead acid, they work great in cold but will also freeze when drained in cold weather, the key is keeping the SOC high enough. So, winter storm - power outage - no PV - system running furnace until low battery cutoff will still result in frozen damaged batteries.

 

[Philip53]

It seems this was the OP's last post on this forum, in 2022.

 

[seneysolar]

It seems this was the OP's last post on this forum, in 2022.

Oh geeeez, thats what i get for not looking at all the dates

 

[Pi Curio]

Subsystem rendundancy. Why we use battery main OCPD on top of the BMS.

For Low temp charging protection, primarly, the control for charging from solar in my opinon should be at the source, e.g. victron smart solar + SBS or smartshunt with temperature sensor with set low temp charging cutoff.

Having both, the BMS and the SCC in control of low temp charging makes for a wise system design

 

[MisterSandals]

Why we use battery main OCPD on top of the BMS.

Exactly!
Obsessive–compulsive personality disorder is a cluster C personality disorder marked by a spectrum of obsessions with rules, lists, schedules, and order, among other things.

 

[Pi Curio]

I meant Over Current Protection Device, but okay