Cooling LFP batteries

[tanglewood]

It's well documented that too much heat is bad for LFP batteries.


First, what degree of self-heating have people experienced, and at what charge rate? The math says that at max internal resistance and 1C charge (or discharge) rate, there will be a lot of internal heating. But in practice, my cells have an actual internal resistance way less than the spec sheet max (0.1 mOhms vs 0.6), and I get the sense that most boats charge and discharge at a good bit less than 1C.



Second, I'm curious what steps people have taken, if any, to aid the cooling of their batteries? Has anyone spaced them with an air gap around each cell for improved convective cooling? Has anyone provided forced air cooling, and have you found it needed? Has anyone provided AC cooling? (I know of one boat that has).


Thanks

[travellerw]

I was also worried about this. There have been a few RVers that have had capacity losses they attributed to heat. Here is one blog where they discuss just that.
https://www.technomadia.com/2015/02/...attery-update/

On our boat, we installed 2 large grilles in the compartment that batteries are in. One of the grills is right in front of the bank. I have actually been waiting for a fan to be delivered that I will install behind the second grille moving air outward (Amazon doesn't work the same down here in Grenada). Since the compartment is mostly sealed, air should be sucked through the first grille, past the batteries, and then exhausted out the second grille.

However, without the fan when we are dumping 100A into the batteries, they still feel "room temperature". So I'm not sure if it was required, but it can't really hurt. I plan to connect it to the load port on one of my Victron MPPT controllers. That way it will only turn on when there is sunlight.

[senormechanico]

I have not noticed any heating beyond what came from external sources except when my 200 aH bank gets a 20 minute, 100 amp load making hot water via inverter.
Even then, it's only a couple of degrees f.

[john61ct]

Yes, the ambient temps is an issue, but so far not internally generated. I plan to keep an eye at high charge rates over .5C, likely put in charge control / protections if I see issues then.

WRT ambient, I don't use active cooling, but do install in protected locations. For those camping in the desert I think necessary to expect shorter lifespans, but if an aircon living space is available that'd be where I'd put them.

I do have heating pads for sub-zero operations, but have not yet needed to use them.

[Delfin]

Originally Posted by tanglewood

It's well documented that too much heat is bad for LFP batteries.


First, what degree of self-heating have people experienced, and at what charge rate? The math says that at max internal resistance and 1C charge (or discharge) rate, there will be a lot of internal heating. But in practice, my cells have an actual internal resistance way less than the spec sheet max (0.1 mOhms vs 0.6), and I get the sense that most boats charge and discharge at a good bit less than 1C.



Second, I'm curious what steps people have taken, if any, to aid the cooling of their batteries? Has anyone spaced them with an air gap around each cell for improved convective cooling? Has anyone provided forced air cooling, and have you found it needed? Has anyone provided AC cooling? (I know of one boat that has).


Thanks

T, I've checked the container temp for my cells after charging and it is never more than ambient. I charge at .33C. Combustion air at outside temps is blown down on the batteries, which sit behind an insulated panel in the ER.

[nebster]

Originally Posted by tanglewood

First, what degree of self-heating have people experienced, and at what charge rate? The math says that at max internal resistance and 1C charge (or discharge) rate, there will be a lot of internal heating. But in practice, my cells have an actual internal resistance way less than the spec sheet max (0.1 mOhms vs 0.6), and I get the sense that most boats charge and discharge at a good bit less than 1C.

In my pack, interconnect contributes a large portion of the resistance. At one point I had a chart of the resistance contribution by component, but I seem to have lost it. The fuses were the biggest culprits, though.

Anyway, total resistance in my case is about 7mOhm, and that produces about 200W of heat at 0.25C (150A). Probably about half of that is actually dumped into my semi-insulated battery bay of about 0.5m^3, where I measure an ambient heat increase of about 1F per hour.

At 0.5C, I do see close to four times that heating rate.

I don't have instrumented probes on the cells themselves, and no doubt the cells' temperature lags the air by a fair bit. However, when the pack is under zero load, I can see the voltage rise and fall with the ambient temperature (just millivolts), and they are faster to respond than I would have thought.

Second, I'm curious what steps people have taken, if any, to aid the cooling of their batteries? Has anyone spaced them with an air gap around each cell for improved convective cooling?

I elected to use the newer aluminum-shell cells, because they are about 50% denser. They require spacing for isolation anyway, and about 5mm of air gap exists between every cell. Combined with the metal casing, the amount of passive cooling is substantial, and I suspect otherwise-identical quality cells made this way will last longer as a result.

Has anyone provided forced air cooling, and have you found it needed? Has anyone provided AC cooling? (I know of one boat that has).

I have forced air ventilation run to the battery compartment with a shutoff valve. I have had one day when I decided to use it, but it was a situation you will never have on your boat: I was parked on a black asphalt lot, midday sun, 95F outside, charging.

My battery bay is exposed to the sun depending on time of day and angle. On the hottest days with direct sun exposure, the warmest I have seen the bay get is about 105F while charging at ~0.3C.


I am comfortable with an occasional spike to 100F, but I would not be happy going much above that. I am happy with how the battery temperature has behaved itself without much intervention so far. It is much better than I was afraid it would be. (Now, the inverter bay, completely different story. I am really glad that my batteries are in a different location than all the other system hardware.)

[Thalas]

Out Chasing Stars just moved their LiPos into the salon for this reason. Internal heat was never mentioned as a factor, but external temperature variation was the reason given. I'd be surprised if you'd need active cooling unless you had a huge bank discharging quickly for things like ocean volt motors.

[tanglewood]

Thanks for all the feedback so far. A friend tracked the temp in his laz (same boat) over a bunch of hot weather and found the temp very closely tracked water temp. That's where the batteries will live, and also where the inverters and chargers are located. We are both switching to LFP, so both interested in the same issues. I guess with so much of the laz exterior in contact with the water, that becomes the dominant factor. So I'm hopeful that hot outside ambient won't be a huge issue, and as long a battery self-heating isn't a problem, we should be OK. Worst case I can just reduce charge rate when conditions are particularly warm.

[hellosailor]

CALB apparently makes two lines of lithium batteries. Blue case vs white case. Their explanation is that one offers higher energy density (a smaller package) while the other is bulkier due to the presence of internal heat transfer plates, designed to keep the battery running cooler and extend the number of charge cycles. And then at least one vendor offers the batteries with an aluminum case instead of plastic--claiming that also assures much great life cycles because the aluminum case sheds heat faster.

So, the concept of keeping the batteries cool--internally as well as externally--seems to be out there. But "bolt 'em up and hide 'em in a box" seems to be the standard they are being targeted for. If you added honeycomb or fluted aluminum plates between them, to promote heat transfer, and forced air to remove excess heat, that would be a personal investment, or gamble, or waste, all depending on your specific installation and wishes.

[NPCampbell]

First post here so bear with me.

My understanding is that the internal resistance (ir) of the battery is in series with the load. Therefore, the same current delivered to the load is also passing through the internal resistance. This makes the formula for power across the ir = Power(ir) = I^2 * ir. Assuming 1C, Power(ir) = C^2 * ir. Taking an arbitrary C of around 100 amps for a group 31 size battery (and to make the math easy). Total power at 1C would be 12.8 * 100 or 1280 watts. Of that 1280 watts, the power dissipated as heat across the internal resistance is P(ir) = 100^2 * .0001 = a whopping 1 watt

[tanglewood]

Originally Posted by NPCampbell

First post here so bear with me.

My understanding is that the internal resistance (ir) of the battery is in series with the load. Therefore, the same current delivered to the load is also passing through the internal resistance. This makes the formula for power across the ir = Power(ir) = I^2 * ir. Assuming 1C, Power(ir) = C^2 * ir. Taking an arbitrary C of around 100 amps for a group 31 size battery (and to make the math easy). Total power at 1C would be 12.8 * 100 or 1280 watts. Of that 1280 watts, the power dissipated as heat across the internal resistance is P(ir) = 100^2 * .0001 = a whopping 1 watt

That looks right, but where did the .0001 come from? Is that from my original post?


Actual internal heat generated for my cells at 1C woudl be 180a ^ 2 * 0.0001, or 3.24 watts per cell. With 32 cells, that's about 100W and that will definitely heat up over time. 0.5C would be 90 ^2 * 0.0001 or 0.81W per cell, or 26W for all cells.


So this would all seem to depend on how hard you charge and discharge the batteries. I'll generally be charging at around .5C for 2 hrs, so would expect some detectable warming, but should be an issue. Same for loads - in fact they should be a lot lower. Big difference compared to an EV.

[rgleason]

This looks like there is essentially no increase in temperature, but we know that batteries increase in temperature more than this.

Perhaps the formulas which are used well, do not include the difference between charging 100ah and what the alternator must put out to achieve that, all related to the Peukart value. This added energy going into the battery might increase temperature. Also the ambient temperature in the battery locker which is often next to the engine may increase, also causing increased battery temperature.

Originally Posted by NPCampbell

First post here so bear with me.

My understanding is that the internal resistance (ir) of the battery is in series with the load. Therefore, the same current delivered to the load is also passing through the internal resistance. This makes the formula for power across the ir = Power(ir) = I^2 * ir. Assuming 1C, Power(ir) = C^2 * ir. Taking an arbitrary C of around 100 amps for a group 31 size battery (and to make the math easy). Total power at 1C would be 12.8 * 100 or 1280 watts. Of that 1280 watts, the power dissipated as heat across the internal resistance is P(ir) = 100^2 * .0001 = a whopping 1 watt

[NPCampbell]

Originally Posted by tanglewood

That looks right, but where did the .0001 come from? Is that from my original post?

Yes, the 0.1 mOhms that you provided in your original post.

Originally Posted by tanglewood

Actual internal heat generated for my cells at 1C woudl be 180a ^ 2 * 0.0001, or 3.24 watts per cell. With 32 cells, that's about 100W and that will definitely heat up over time. 0.5C would be 90 ^2 * 0.0001 or 0.81W per cell, or 26W for all cells.

Good point. I wasn't taking into account that was per cell.

[nebster]

CALB makes three lines of LFP cells: CA, SE, and CAM. SE is the older, blue plastic one. CA has a white plastic shell, and CAM is an aluminum shell design.

[CharlieJ]

For general interest; one off the battery banks that Torqeedo offers for their inboard electrical propulsion system is the BMW i3; 30.5kWh @ 350VDC comprised of NMC cells that powers an 80hp motor. The i3 is cooled with an attached refrigerant system.