Dig a pit to keep battery temps stable?

2twisty

I live in far west texas and my batteries get rather toasty in the summer.  

I am considering digging a pit 3-4 ft deep to put the batteries in for the geothermal temp regulation, and then putting a hut over top that is insulated heavily.

What think ye? I'm only getting about 2.5 years out of a set of GC2s due to the high temps in the summer.

Estragon

I would.

Photowhit

That's awfully short life span, I would, actually I prefer bringing the ground up and around the batteries, making a berm shelter for them. Less worries about water collection.

WaterWheel

I also considered putting the batteries in the ground but ended up using a water bath with good results.

Here is a link to the thread.       Post #27 is where I start publishing the water bath battery tempature results.
https://forum.solar-electric.com/discussion/353721/cooling-the-batteries/p1

Keep in mind that with the batteries 3-4' in the ground maintaining the batteries will not be any fun.

jonr

I'd use a water bath with a chiller.

Wheelman55

2Twisty. Gentleman here in the Big Bend area of West TX puts his clients batteries in a hole, in water, outside on a north wall, with a simple cover. Very successful. 

He goes by the name of Desert Rat on the forum. Smart guy. Look up some of his posts. 

BB.

Here is Desert Rat's user page and search for posts:

https://forum.solar-electric.com/profile/Desert Rat

-Bill

mcgivor

Did the water bath a while back, using a 90W water cooler with a small aquarium pump, it was  very effective at keeping temperature at bay, the only problem I encountered was animals were attached to the water, often drowning, with predictable outcomes, but with  nylon mosquito netting these problems could be reduced/eliminated. The system only operated during the day using excess production which would otherwise been wasted, lead acid batteries generate heat whilst charging, so it's logical to to cool when most vulnerable, during the hottest period.

The change to LFP looks promising with regards to temperature gain, their inherent efficiency reduces the gain, albeit I'm in the transition from cool to hot seasons, today was 38°C or 100°F maximum, night time drops to 24°C /75°F, maximum battery temperature recorded is 30°C/86°F, but I wonder what the really hot period will reveal, when ambient temperatures don't drop below 30°C/86°F overnight, this would be a period of ~2 months, given historical data. The lead acid batteries by contrast never dropped below 36°C/96.8°F during this period , not good for the chemistry, will collect data to see if temperatures remain somewhat stable, so far so good. Having said that if temperatures drop significantly below zero for extended periods , LFP may not be the best choice.,

Horsefly

At the risk of hijacking this thread - still on the subject of temperature regulation, but on the opposite end.... 

We have our batteries (and all the electronics) in a small basement room on the north end of our cabin. That works well to regulate the temperatures in the summer, and even when the outside ambient air is 95°F the batteries stay in the 70s. Ok, great.

Our problem is in the winter. At some point our AGM batteries will be at end of life. Watching the progress on LFP (and noting the experiences here on the board, most recently from @mcgivor), I'd love to be able to at least have the option to consider that technology. Unfortunately at 9,000 ft elevation in Colorado, the outside temp can easily get to -20°F (I used -36°F for my hyper Voc design calculations). I see that LFP specs are pretty consistent that they must be at 32°F or higher for charging, and storage / discharge needs to be around 0°F or higher.   Although I'm not sure how cold it gets in the basement of the cabin, I'm sure it's below those temps.

My first thought was to simply take these lighter batteries out in the fall and put them back in the spring, keeping them in a warmer garage down range all winter. The problem with that is that the regular use of the cabin overlaps some with the cold temps both in the fall and the spring.  

The idea of pit or berm probably wouldn't go over well.  I'd be a bit worried about leaks in a large water bath container, especially during the months that no one is around. 

I periodically drag the idea back out and do some thinking about it, and then eventually fall back to having to stay with AGMs. Any ideas I'm missing?

Steve

mike95490

Horsefly said:

..... I see that LFP specs are pretty consistent that they must be at 32°F or higher for charging.

it's even worse than that,  40°F is about as low as the batteries can go and still be fully functional. Below that, recharge has to be throttled back, approaching no charging at 33°F.   And below 40°F, you start to seriously loose capability.   But using your BMS to incorporate some heater tape like water pipe heat tape, and some insulation, you can set up some timers to allow heating with surplus solar, and keep the pack around 50°F, where it should be really happy. 
 Some of the rigid 2" foam board insulation ought to do the job. 
 Watch your cable connection areas, copper is a excellent heat conductor, and your cables can suck heat away from the cells at the connection points, while other cells are cooking

Estragon

I also have the low temp issue with any sort of lithium batteries (ambient can be -40°).  My plan is to use a number of temp sensors incorporated into a DIY built pack and a microcontroller to make sure there's no charging at below freezing temps.  Pv output will go first to warming the pack (likely with heat trace as Mike suggests), and charging only started if/when the pack is warmed.

I'm playing around with sensors like these now:
https://www.amazon.com/Eowpower-Stainless-DS18B20-Waterproof-Temperature/dp/B0714588F8/ref=mp_s_a_1_17?ie=UTF8&qid=1549475377&sr=8-17π=AC_SX236_SY340_QL65&keywords=1+wire+temperature+sensor

With enough sensors, it should be possible to be fairly confident the pack has warmed consistently, including near terminals.

If they work out, I'll likely try making a small experimental pack before deciding on whether a full bank is worth pursuing.  Hopefully I have a good few years to experiment before needing a bank replacement

AFAIK, storage is ok at low temps, and unlike LA, partial SOC storage is actually better.

Horsefly

Thanks @mike95490 and @Estragon. You've both given me quite a bit to think about. 

@mike95490: I like the idea of building a rigid foam board enclosure. I don't think I know enough about what the BMS can do to know how to have it control some heating. But, you've given me something to go look at.

@Estragon: Your idea seems pretty cool. The engineer in me gets kind of excited to figure it out. For the microcontroller, are you thinking of Arduino, Raspberry Pi, or something else?  Would you redirect the charge controller power via relay?  Are you pretty confident that storage (with no charging or discharging) at below freezing is OK? The specs seem to indicate otherwise, but I've not seen anything about what will happen if you do go below freezing.  By the way: I've used DS18B20 temp probes for years in my home automation hobby. I've currently got (I think) 7 of them inside and outside our house, logging temps every 10 minutes 24 hrs/day. I'm guessing they have been working continuously now for about 10 years.

mcgivor

Both BMS that  I use have temperature limitations for charging/discharging, but no control output for external heating or cooling, though it wouldn't be difficult to make such control externally, perhaps incorporating a timer. 

Estragon

For my first go-around with the sensors, I'm using a pi zero with a 1-wire driver board and 12 sensors.  This will be recording temps at incremental depths from air down to lakebed, mostly so I can better understand freezeup and breakup water column changes.  Readings will be stored in a local db, and uploaded periodically along with solar monitoring readings.

For a battery pack project, I was thinking an arduino type board for lower idle power consumption if it's just redirecting pv output via relay.  A pi might be easier if controlling & monitoring charge controller state (modbus) though.  I'll also look at BMSs in more depth as well to see if some can do some/all of what I want.

I know EVs are a bit different chemistry, but they're sold in my climate, and sit in very cold weather.  If plugged in, they must use power initially to warm the pack before charging.  If not plugged in, I assume they use some battery power to self-heat before discharging to drive?  I don't think they'd be able to keep the pack warm for days at a time at -40° using battery power alone though, and they wouldn't find much interest around here if they went dead just sitting in the driveway or parking lot.

My understanding is the main problem with sub-zero charging is hard plating.  In storage, especially cold, there would be little/no chemical activity.  There may be an issue with physical changes though (expanstion/contraction, vibration, etc), which likely depends on cell and encapsulation type.

I've left small lithiums in cold temps to no apparent ill effect (once they warm up), but I'd want to test a small lifepo pack with the cell type I'd use in a bank. 

Maybe there will be commercial products designed for low temp off-grid use at some point, but I'm guessing the market is pretty small and a DIY will be needed.