powerwall cooling

I want to ask a question to the Group on DIY power walls, is anyone planning for cooling?

I think that keeping the temperature of the cells below 40C would be pretty important.

I just watched this video from 2013 by Jeff Dahn.
The several takeaways that I got from the video were the following.

1.Heat Kills batteries.
2. Overcharging batteries to high Voltages above 4.2 VKills batteries
3. Leaving batteries in a high state of charge sitting at High temperature Kills batteries.

Secondary takeaways.

Lithium Cobalt batteries are the best battery chemistry for more charge cycles.

It seems pretty obvious to me that protecting your investment of time and money in building the power wall would be as important as building it in the first place.

I think most of us are just using the standard air cooling. We don't have enclosed units that can trap heat and cause thermal buildup (which would lead to thermal runaway). And Pete is installing fans in his battery shed to keep air flow up and temps down.

Since we are using so many cells, each cell under normal use would not use more than 1 amp draw, if even getting over .5 - .6 amps of draw. Even with a high IR, that won't create a whole lot of heat.

So, with that reasoning, logic, and justification, I think it's safe to bet most of us are not using any kind of pack level cooling. I think there is one person who was thinking of using water cooling, but not sure if that was just asking about it, or actually planning on it.

Now if you were expecting to draw more than an amp per cell continuously, then yes, some kind of cooling would probably be recommended, passive or active.

I personally, if I feel that I need to add a little extra cooling, I'll be tapping into my geo-thermal piping and add in a set of radiator coils and a fan to help cool the battery shed.

But good question, non-the-less

I was looking at a refrigerated water cooler at the thrift store yesterday thinking if could be adapted to cooling packs.. It will be there for a while no doubt so there's no hurry.

Some people seem convinced lithium ion cells are good to 50C, I think that's questionable from my own personal experience. I've cooked cells to death several times in the past including the cells in a laptop last summer that got left somewhere it shouldn't have been, one overheating incident killed a pack that had some good life left in it.

There are a variety of chemistries in cells and the manufacturers are constantly playing with the mix, I doubt all of them are exactly the same as far as temperature tolerance goes, better safe than sorry is the watchword.

egam said:

1.Heat Kills batteries.
2. Overcharging batteries to high Voltages above 4.2 VKills batteries
3. Leaving batteries in a high state of charge sitting at High temperature Kills batteries.

Click to expand...

2 and 3 are definitely correct ... but not 1.... batteries in some medical equipment is regularly sterilised at 120C ...

With current going through them our cells are happy up to 60C that would be too hot to touch ....I guarantee no one's power wall cells will get that hot ...even operating at 2A !!! Tesla has cooling because it sells cars to Saudi Arabia (50C in the summer) and quick charges/discharges

The main concern for usis keeping them warm ...they can't operate well cold ... and not at all near freezing ... the warmer they are the more efficient .... in practice most operators need to thermally insulate to some extent if ambient temperature is below 10C...internal resistance is much higher when cool , meaning efficiency is lower ....

This is a non issue for anyone operating at ambient temp around 20C and cyclingat less than1A per cell (C/2)

Lets remember powerwalls are meant to provide domestic power ..normally as off grid solar , even the smallest one should be able to provide 10hrs of power ... that means a discharge rate of C/10 .... 0.2A !!! this will increase cell temp by only a few degrees C .... if in a cold climate insulate your cells!!

ozz93666 said:

egam said:

1.Heat Kills batteries.
2. Overcharging batteries to high Voltages above 4.2 VKills batteries
3. Leaving batteries in a high state of charge sitting at High temperature Kills batteries.

Click to expand...

2 and 3 are definitely correct ... but not 1.... batteries in some medical equipment is regularly sterilised at 120C ...

With current going through them our cells are happy up to 60C that would be too hot to touch ....I guarantee no one's power wall cells will get that hot ...even operating at 2A !!! Tesla has cooling because it sells cars to Saudi Arabia (50C in the summer) and quick charges/discharges

The main concern for usis keeping them warm ...they can't operate well cold ... and not at all near freezing ... the warmer they are the more efficient .... in practice most operators need to thermally insulate to some extent if ambient temperature is below 10C...internal resistance is much higher when cool , meaning efficiency is lower ....

This is a non issue for anyone operating at ambient temp around 20C and cyclingat less than1A per cell (C/2)

Lets remember powerwalls are meant to provide domestic power ..normally as off grid solar , even the smallest one should be able to provide 10hrs of power ... that means a discharge rate of C/10 .... 0.2A !!! this will increase cell temp by only a few degrees C .... if in a cold climate insulate your cells!!

Click to expand...

I have some first hand experience with air cooling. In this case, it was a nickel metal hydride battery in a prius. The car had only 80K miles on it. one of the cell modules in the pack went bad which threwan error inthe battery management system. In addition, the fuel economy dropped.

It turns out the car spent a lot of its life in Texas.

I am thinking about charge and discharge of a diy power wall in my shed. In July, August and September, in the heat of the day when solarcharging happens, the ambient temp can reach 40C (104F). If the internal resistance in the cells is 100 mohm each, and each battery is charging 1amp,then each cell needs to dispose of .1 watts of power. For a 1000 cell pack, thats equivalent to a 100 watt light bulb. At high ambient temperatures, the cell temp could rise a lot higher than the battery box temperatures. For half warn out cells, the internal resistance can be 200 mohm.

I watched a video that measured internal resistance on new LG 3500 mAhr cells and the internal resistance was 35 mOhms. So you have to deal with the heat even with new cells. Thats why tesla puts liquidcooling in their power walls. It's just safer.

i would be interested to see the spread between ambient temp and cell temp. I guess I can watch some HBpowerwall vids hopefully there is some ambient comparisons.

egam said:

. In July, August and September, in the heat of the day when solarcharging happens, the ambient temp can reach 40C (104F). If the internal resistance in the cells is 100 mohm each, and each battery is charging 1amp,then each cell needs to dispose of .1 watts of power. For a 1000 cell pack, thats equivalent to a 100 watt light bulb. At high ambient temperatures, the cell temp could rise a lot higher than the battery box temperatures. For half warn out cells, the internal resistance can be 200 mohm.

I watched a video that measured internal resistance on new LG 3500 mAhr cells and the internal resistance was 35 mOhms. So you have to deal with the heat even with new cells. Thats why tesla puts liquidcooling in their power walls. It's just safer.

i would be interested to see the spread between ambient temp and cell temp. I guess I can watch some HBpowerwall vids hopefully there is some ambient comparisons.

Click to expand...

40C is extremely hot ... I can't imagine it stays that way for long ... andyour charge current is high... a full recharge in less than 2 Hrs... So you are anextremecase .... you say that's 100W in a 1000 cell pack that's 1/10th W per cell ....that's would be for a cell with an internal resistance of 100mOhms In a standard 80 cell pack that's just 8W ,,, that's very little heat , even at 1A....I bet if you put a thermometer inside the pack , you'd be surprised how cool it was....

In reality I think it would be closer to 60W/1000cells ... this graph shows that at high temp internal resistance drops greatly , so would the heat generated...I guess you probably measured int res.it at about 20/25*C

My batt's will be in enclosed space. That space will be able to circulate its air to the garage itself. The garage is heated during winter and during summer it stays naturally cool. If it isnt enough il just add an AC in that area. Lets see. First real test for my new place this summer.

daromer said:

My batt's will be in enclosed space. That space will be able to circulate its air to the garage itself. The garage is heated during winter and during summer it stays naturally cool. If it isnt enough il just add an AC in that area. Lets see. First real test for my new place this summer.

Click to expand...

Don't tell me your worried about heat , daromer !!! AC !!! You must have a few MWHr capacity by now... lol... that means your charge/discharge current will be very low per cell .......

I think people get carried away ... the basic physics says there is no problem here ...

Not heat from cells. That i dont care about.

But in summer that place could theoretically be 50c or more

ozz93666 said:

40C is extremely hot ... I can't imagine it stays that way for long ... andyour charge current is high... a full recharge in less than 2 Hrs... So you are anextremecase .... you say that's 100W in a 1000 cell pack that's 1/10th W per cell ....that's would be for a cell with an internal resistance of 100mOhms In a standard 80 cell pack that's just 8W ,,, that's very little heat , even at 1A....I bet if you put a thermometer inside the pack , you'd be surprised how cool it was....

Click to expand...

Some parts of Australia (and other parts of the world) easily get 40C ambient air temperature, it's a big problem for aircooled battery packs in general.
The Nissan Leaf had to get a new "lizard" battery to cope with hot climates that killed the original battery packs.
I did consider buying an entire Leaf pack cheap and shipping it to Australia, but it would never cope outside, and it's not something I feel OK with having inside.

Here is what I have to contend with:
"On 28 January [2009], the third day into the heatwave, the temperature reached 45.7 C (114 F)
https://en.wikipedia.org/wiki/Climate_of_Adelaide

I'm thinking there are only two option for me:
A) LiFePO4 batteries, which would be safe enough to have inside and should cope with some heat.
B) A watercooled battery pack. Perhaps with water chiller setup as well.
B is how the Tesla Powerwall is built, minus the chiller.

I don't feel a water cooled battery is an easy DIY project, look at all the issues people with water cooled PCs have with leaks.

I've got a peltier (4 blocks) coming in a few days, just to play with, to see how effective they are at cooling water. I wasn't thinking of using it to cool the packs, but who knows, you can buy small radiators to place near the batteries with a series of small fans. Shouldn't be too hard to set up.

station240 said:

Some parts of Australia (and other parts of the world) easily get 40C ambient air temperature, it's a big problem for aircooled battery packs in general.

"On 28 January [2009], the third day into the heatwave, the temperature reached 45.7 C (114 F)
https://en.wikipedia.org/wiki/Climate_of_Adelaide

Click to expand...

Hmmm..That is a problem ... I guess the temperatures drop a lot in the night ? ...One solution might be to build a small brick building , double walled,somewhere in the shade ,for the batteries .... fans switch on at night pushing cool air in to take the temperature down , switch off in the morning and the thermal inertia keeps it fairly cool all day ....

Evaporative cooling is another option if the humidity is low and plenty of water available ... the best plan must be to have lots of batteriesthen they charge at low current (per cell) , so don't get hot , operate at low current so last longer , low max voltage 3.9 is the most important factor for long cell life ....

If you have plenty of panels and sometimes power is not used , you could wire things so the power is diverted to a refrigerator or AC when batteries are full . There are refrigeration compressors that run directly from panels ... theyswitch on if they have a voltage over 22V

---

BobMuster said:

I've got a peltier (4 blocks) coming in a few days, just to play with, to see how effective they are at cooling water. I wasn't thinking of using it to cool the packs, but who knows, you can buy small radiators to place near the batteries with a series of small fans. Shouldn't be too hard to set up.

Click to expand...

...Yes , they're very cheap ...I could never get mine to work well!!! they have to have a very good heat sink on both sides .

But the inherent problem is they're not very efficient . The compressor in refrigerators and AC are 4 times more efficient at generating cold than peltiers .. But peltier's are used in car beer coolers so they can be made to work.

Peltier is last resort. So inefficient They would most likely need more electricity than you going to save from it.

Peltiers are very inefficient at anywhere near their rated maximums, turn them down to about 10 percent power and they improve quite a bit, even better at 5 percent or lower, just use more peltiers to make up the difference since they are fairly cheap in quantity. Also making sure you have a good thermal junction is important, if the heat sink surfaces on both the hot and cold sides aren't flat and solidly connected the overall efficiency plummets. Getting the heat to the cold side and away from the hot side is another thing that makes a big difference, water cooling the hot side and fan heating the cold side has increased the heat transfer a lot in my tests. Every place there is a thermal gradient you want something helping to minimize that gradient be it a fan moving air, a pump moving water or a solid flat connection with the right amount of the right thermal compound. I bought a "tub o' thermal compound" recently and it's not as good as the more expensive stuff I have been using, I can tell the difference in my cold side temps. A well insulated enclosure is a good testing device for peltier cooling, it doesn't have to be particularly large mine will hold two gallons of milk with a little space left, a variable power supply is a big plus.

After playing with them quite a bit over the years I'm convinced you are better off with many lower powered peltiers rather than one high powered one, the lower powered versions will have more area to pump heat through and getting heat into and out of the things is where it's easiest to run into problems. It's possible to burn one up in a couple of seconds with no or poor heat transfer on the hot side at higher power levels.

With peltiers it's more efficient to run them at a low continuous power setting than turn on and off a higher one, I^2*R losses are higher in the pulsed versus the continuous mode for a given average heat transfer, potentially much higher depending on conditions.


At the power levels per cell most of us envision air cooling of the cells seems more than adequate if the pack is insulated from the harsh environment in the first place. Keeping ambient environmental heat or cold for that matter away from the pack would be the aim in certain more extreme environments.

Theres a good air gap with the cell frames alot of us use,
Are you suggesting thats not enough?
how much draw per cell are you running?

What if you put the powerwall in a box with the top and bottom missing - would it cool itself just by the convection of air?

Most definitely. As long as the air coming into the bottom of the box is cooler. You could even add a couple Computer Fans to the mix to help it along. But the incoming air would have to be coming from somewhere other than just the storage building the box sits in.

station240 said:

Some parts of Australia (and other parts of the world) easily get 40C ambient air temperature, it's a big problem for aircooled battery packs in general.

Click to expand...

If I was faced with this problem, I think I would bury a water tank as deep as I could. That water would be passively cooled by the ground, which is why you need to bury it deep. Then I would use the water for a liquid-air heat exchanger, like a radiator in reverse. I would use that cool air to control the temperature of the battery pack. The goal is just to control the extreme temperatures. (I may even recirculate the air, depending on the external temperature vs the heat being generated by the batteries.)

The energy input to this system would be a water pump to circulate the water, and a small fan.

Bury some 100mm stainless steel tube several M into the ground in a U shape below foundations of Battery hut , Build hut out of cool room panel small fan the circulate air. Thermal mass of earth is heat bank , Here In country Victoria Australia (Southerly Australia , summer temps up to 45 deg c winter temps - 8 deg c) average ground temp @ 1m deep is 15 deg c year round.

With summer heat addition due to airflow it might rise a little around pipe but not a whole heap. Your cells can have cool air any time they need it If you fitted a thermostat to the fan for the fan to only switch at low or high events would be ideal to minimise power consumption.

Why water not oil