On Keeping LFP Warm

[Horsefly]

Unless I am missing something, this is not an issue. If you are just storing your batteries at the cabin-site without using them or charging them, they are fine just sitting there (supposedly down to about -40º F). Not really a risk to go with LifePO4... just bring them down to about 50-70% or something before you leave for the Winter and disconnect the solar panels. They can sit at 2/3 charge with no parasitic drain, and then in the Spring just charge them up. Only issue will be in the Spring when make it back up you will need to keep them warm still if you want to use and charge. I know in the spring at 9000' it is still going to be a challenge to use the batteries unless they are indoors or heated.

Even better, just keep your panels connected to your AGM batteries (or 1-2 of them) with a very low Watt heater that heats the battery space. That way you are still using the solar energy being generated but not putting it into the lithium batteries that won't need it anyway.

Well, the only thing you are missing is something I probably haven't said, or at least not in this thread. Of the three owners of our cabin, two (my sister and brother, and their families) are within about 35 miles of the cabin. I'm about 300 miles away. I'm not the first one there in the spring, and I'm not the last one there in the fall. Although it isn't common, sometimes my sister will trek into the cabin during the winter (snowmobile or cross country skiing). No one but me knows much about the solar. Everyone got used to knowing that when they get into the cabin they just take the inverter out of standby and flip the breaker for the DC loads and they can enjoy the electricity. The only way I could sell the transition from AGMs to LFP was if I promised it wouldn't get more complicated.

The box I've built with the heating logic seems to be working pretty well. It's gotten down below 40°F the last couple of nights, and isn't getting much above 40° today. I've got the thermostat set to keep the cells between 50° and 60° and it seemed to be working well. The BMS and the smart shunt don't show any decline in the SoC for the past 48 hrs.

 

[FilterGuy]

Well, the only thing you are missing is something I probably haven't said, or at least not in this thread. Of the three owners of our cabin, two (my sister and brother, and their families) are within about 35 miles of the cabin. I'm about 300 miles away. I'm not the first one there in the spring, and I'm not the last one there in the fall. Although it isn't common, sometimes my sister will trek into the cabin during the winter (snowmobile or cross country skiing). No one but me knows much about the solar. Everyone got used to knowing that when they get into the cabin they just take the inverter out of standby and flip the breaker for the DC loads and they can enjoy the electricity. The only way I could sell the transition from AGMs to LFP was if I promised it wouldn't get more complicated.

I hear you bro!!! I have built several systems for friends and family and ease-of-use has to be paramount in the design. If the system can't be brought online in a few easy steps.... it is going to be a problem. If your cabin is like my old one, there isn't even cell coverage that would allow them to give me a call. Consequently, they are on their own when working on the system. In addition, the process must be hard to screw up. There is no such thing as foolproof, but I have to strive for that. I have to assume that if it is easy to break the system, it will happen. It is not that they are stupid, it is just that solar and batteries are just not their thing... so they don't understand the inner workings.

 

[TennesseeMike]

yes the magnum inverter with its battery monitor system.

To those interested: I couch the following as I think, I have no proof as of yet, but once I get a wifi hotspot to land line conversion figured out so that I can export data in real time and record data I will be able to possibly answer this better.

Two weeks ago I added 12 panels for an additional 2.4kw to the roof of the shop (total 6.4kw) and since I fired them up, once a day for the last eight days has shown a fault history where it is at high voltage surge of 69-71ish volts. To date it has occurred between 1330-1430 and I think it represents when the morningstar charge controllers disconnect due to high voltage.

the SCC is set as 54.4 charge, with a 54.1 VDC 1 hour absorb, and then it drops to a 54.0 float for 3 hours. no temp compensation, a high voltage disconnect of 56 volts, and a reconnect of 54.6 . it also has a setting called maximum regulation limit which honestly I don't understand the need for as it acts as a total cap on voltage output. I set that to 56 the same as the HVD.

it has not shut down the inverter, nor the BMS so it must be very brief and with no actual current to speak of. I know it has not shut down neither of the two for two reasons the Magnum does not show a shutdown in conjunction with the fault, and if the inverter or BMS did shut down then my split pack unit in the house would have shut down as well. (and not restarted.) (split pack is run on timer from 1000-1400 to dry out the house during the damp seasons).

I think the sudden termination by the SCC causes a brief transient surge. this happened once before that I noticed and at the time I had just suffered a BMS failure so I thought it might have something to do with that. (now I do not think so). it has happened daily since I added the extra panels.

one other thought that just entered my pea brain is that it could be when the a/c shuts down at 1400 hundred and the panels are cranking out juice to charge the pack and power the hotel loads including the a/c. possibly the shutdown of the a/c on timer might for a brief microsecond allow the SCC to race up to hit the maximum voltage regulation set point??? too many questions on this one and not having internet I cannot monitor and log them unless I am there. (and last time I was there is was heavily overcast, batteries hovered between 60-70 percent the entire time.)

anybody have an idea on if there are any routers that can take the wifi signal from a portable wifi hotspot and send it out a normal ethernet cable? the magnum and the tristars need an ethernet cable to a modem. I figure there has to be a way to do this. a pocket wifi would cost me like $30/month and I could then monitor my solar charge controllers, my inverter and even my bms via the internet...

Regards

Ken

Look for GLI-Net routers. I used them to wirelessly connect to a tablet hotspot and then provide internet access via their Ethernet port.

 

[Horsefly]

I started this thread back in January of 2021 with a long-winded post that put out what I thought was the correct math for two aspects of keeping our LiFePO4 cells warm:

1. The amount of energy required to warm a pack of LFP cells 1°C, based on the mass of the pack and an estimate of the specific heat capacity of LFP batteries.
2. The amount of heat lost from the inside of a box to the outside, based on the size of the box and the insulation of the box.

Since then I have built my battery box complete with electronics to manage the heating (the build is described here) and I've also run some experiments logging the outside and inside temperatures, and how much the heaters had to run (one graph shown here and another shown below). I decided it would be interesting to look back at my original post's hypothesis and see how well it works.

Heating LFP Cells
The short version of my write-up about heating cells (full explanation is here) basically said that to heat LFP 1°C requires an amount of energy (in watt hours):

Wh = M * 1306 / 3600

where M is the mass of the LFP pack. In my case I have an 8S pack of EVE 230Ah cells. The Eve spec says that each cell weighs about 4.11kg, so the total mass of the 8S pack is 8 * 4.11 = 32.88kg. Plugging this in, my original equation would say that the energy required to warm my pack 1°C should be 32.88 * 1306 / 3600 = 11.92Wh. My system is set to start heating if the pack gets to 50°F, and stop heating when it gets to 60°F. That 10°F delta is equal to 5.556°C. So if my original equation was right, it should have required 5.556 * 32.88 = 182.67Wh to heat the cells that 10°F.

Here's a graph of my system working:

Here the green line shows when the heater was ON. The red line shows the temp at the top of the pack, and the blue line shows the temp at the bottom of the pack, where the heaters are.

Both times in this test the heater ran for almost exactly 3 hours to heat the cells (red line) 10°F. My heater is two 12V, 12W heaters in series, for a total (at 24V of 24W). Since my pack is a bit over 24V, what I actually saw the heaters use is about 1.2A at 26.5V, so 31.8W. Running for 3 hours, that meant that it took 3 * 31.8 = 95.4Wh. That's a little less than half of what my prediction would be. Since this heating of the cells was somewhat inefficient (it was also heating the entire interior of the box), it's even worse.

I've looked over my basic math from the original write-up, and can't find anything wrong with it. The only thing that I knew going into it was a complete guess was the specific heat capacity of LFP cells. Like I said earlier, it's really hard to find anything that talks about it. In addition, a cells is made up of lots of different materials (aluminum case, electrolyte, anode, cathode, etc.). So I'm going to guess that was wrong.

I'm keen to hear what others think, but for now I'm going to toss out that my original equation is right, but the specific heat capacity of my cells should be adjusted, and is really something closer to

1306 * 95.4 / 182.7 = 681.95 Joules / (kilogram x °C)

I'd be keen to hear what others think, but maybe this will prove useful to estimate how much heater and how much energy people need to plan on to warm their battery packs.

I'll follow up with another write up analyzing the heat loss of my battery box, which corresponds to item 2 of my original post in January.

 

[Horsefly]

Here's some additional thoughts, as I've gotten comments before questioning the logic of heating up the cells.

When it comes down to it, it is the LFP cells that I want to keep warm. By setting the ON temp to 50°F and the 60°F, the heat will run for a while, then turn off for a while. How long it takes to warm up is the item I just addressed, which is driven by how big my heat source is and the specific heat capacity of the cells.

Once the heater is turned off, it will stay off until the cells reach 50°F. When I did the first write-up (first post in this thread) I was assuming that the primary driver of how long it takes for the cells to cool down would be driven by the heat lost by the box. That is just a heat transfer equation that I talked about in the first post, which uses insulation of the box, the inside surface area of the box, and the temperature difference between the inside and outside.

I realize now that I didn't really think about this correctly. A very big consideration to how fast the cells will cool down is the thermal mass of the pack. That's the specific heat capacity combined with the temperature they were raised to. Turns out that once warmed to 60°F, they will take a fair amount of time to cool down. That is thermal mass. In the graph above, the outside temperature (which I excluded from the graph just to look better) fell below 40°F around 4:00pm o 17-Nov, and around 4:30pm is when you can see the cell temperatures start to fall. By 9pm the outside temperature was below 30°F and stayed there the rest of the night. The heater finally turned on again at 3:30am, which was 11 hours after the cells started to cool, and almost 18 hours after the heater started up last time.

Once you heat this large thermal mass called a battery, it will retain heat for quite a while. The better the insulation of your box the longer it will be. I could probably figure out and post some equations to describe this, although I'm pretty sure it would only be of interest to me , so I don't think I'll bother.

My somewhat quantitative conclusion:

1. Using around 32W of heating, my 8S 230Ah cells take 3 hours to warm from 50°F to 60°F. I've repeated this multiple times with different outside temps, and it is fairly consistent.
2. If the temperature outside the box stays between 25°F and 35°F, it appears that the heaters will not need to run more than twice per 24 hr period.
3. Therefore, for my cells and my insulated box keeping my battery warm will take no more than 2 * (3 * 32W) = 192Wh per day, or 192 / 25.6 = 7.5Ah from my 24V pack.
4. Based on the last four winters, there will never be more than 7 consecutive days with no solar (4 is the max we have seen), so my battery should have no problem keeping itself warm. With solar, the power to heat the cells will be coming from the SCC during the day, and the battery should be fully charged before sundown.

 

[wild01]

I just finished my shop battery. In my half insulated "I'll finish it one day I promise" shop I have built a 375ah 48v 16s15p topband 25ah cell battery. I assembled it in 3 layers in a small deep freeze. I wired the bms in with the battery and put a 20watt propagation mat between battery layer 2&3 hooked to a temp controller.

I then replaced the flimsy plastic sheet on the underside of the lid with a rigid sheet of luan, and attached a greenhouse wax motor vent to the inside of the deep freeze. I also removed the hold open springs from the hinges, so that I don't risk the lid not closing when the shop gets cold. I set the temp controller to 50F.

where I'm at, the temp at night can drop under -40f in the middle of winter. I use power daily, whether I'm home or not, so the waste heat from the bms's will always be radiating into the box, this is why I included the wax motor, I am just as concerned w/overheating as freezing. the bms should cut off charging before there is a problem(currently set at 2c) so heating pad should kick in at 50F off at 52F vent should open starting at about 75f. I'll try to remember to keep you all updated on how it's performing.

time for an update. It has been an incredibly mild winter, so I've only seen -20f so far, but the freezer battery box is working beautifully. It started out cycling the heating pad most nights and never dropping under 50f, then I realized I had enough spare power to run a small 500w space heater in my guest bath 24-7. this was ideal, because I usually struggle with a frozen bathtub trap in this bathroom when temps go negative. after I started running this, and thus cycling the battery deeper the temps started sustaining between 80f and 90f probably from the added heat of cycling the batteries deeper. wax motor raises the lid about 5 inches when I'm over 90 and closes back down around 78.

 

[wild01]

Latest build finally ready for install. couldn't get the built in battery bms's to function right, ended up removing them and wiring all 8 into one 200a jbd. Added aviation connectors to every battery attached to the internal balance leads. Used the same battery heating and cooling concept as the last system. Wired with no exposed wiring as it may need to pass an inspection.