Battery technology selection for harsh weather(below -20 Celsius)

[sharif]

Hello,

I am trying to make a remote sensor with a solar panel for Canadian weather.
As you may know, the temperature on some days goes below -25 in some regions. This is a significant problem for batteries.

The minimum battery capacity I need is 500mAh at 4.2V. The peak current I will need will be less than 80mA.
The space is tight, so I can't use an SLA battery or similar.

I would appreciate it if you can provide me with some information or any experience related to this issue.

[ejeffrey]

I think NiMH batteries can work this cold.  You can also insulate the batteries and provide a heater to keep them above freezing.

[trobbins]

NiCad may be your friend.

[Siwastaja]

All available battery technologies increase in ESR in very similar way in cold.

Additionally, li-ion cannot be charged at much below 0 degC, or needs very low charging current. Discharging works just fine, just with higher ESR but that only affects higher discharge currents.

How much current do you need?

Do you need to charge during coldest times?

Is thermal insulation an option? For example, if charging at -20degC is rare and you have enough external power available, you can just first heat up the battery and then commence charging.

Note that any of the batteries listed are not interested about the freezing temperature of water, because they do not contain water. Even SLA does not contain pure water but acidic mix with freezing temperature way below 0 degC.

[tszaboo]

You should use Lithium primary batteries, LiSoCL4. Buffer it with a large capacitance or supercapacitor depending on the length of the peak current.
LiMNO2 could work, but you will only get like 2.3V or so per cell in cold.
And just to state the obvious: Dont charge primary batteries.

[Siwastaja]

And just to state the obvious: Dont charge primary batteries.

It seems we both kinda missed the solar panel.

With solar panel, it will be charged, so a primary battery is not optimal; with didoe-ORring, solar panel would only double the battery life or so, it would make more sense to double the battery and not use solar panel at all.

I would just use any li-ion cell and prevent charging (by software, or dedicated li-ion charger IC with temperature sensing) below say -10degC and make sure the battery is large enough to last through the few coldest winter months. There will likely be snow on the top of the panel anyway so can you depend on gaining any charge for a few months?

Wireless sensors can be usually engineered to be very low power so to last for months without charging, but that obviously depends on sample rate and report rate.

[ejeffrey]

Note that any of the batteries listed are not interested about the freezing temperature of water, because they do not contain water. Even SLA does not contain pure water but acidic mix with freezing temperature way below 0 degC.

But watch out because lead acid batteries electrolyte becomes less acidic as they discharge and the acid converts to lead sulphate.  The freezing temperature is still below 0 C but a deeply discharged lead acid battery is more susceptible to freezing and will freeze above -20C.  So it's best to limit the discharge depth to ~50% for cold operation.

[Siwastaja]

But watch out because lead acid batteries electrolyte becomes less acidic as they discharge and the acid converts to lead sulphate.  The freezing temperature is still below 0 C but a deeply discharged lead acid battery is more susceptible to freezing and will freeze above -20C.  So it's best to limit the discharge depth to ~50% for cold operation.

Yes, but the deeply discharged LA battery will self-destruct anyway; it either sulfates beyond recovery, or freezes, or both. Lead-acid is good for UPS type of use where it is 100% most of the time, and will be recharged quickly after discharging. This is obviously out-of-question for the OP.

[Marco]

There will likely be snow on the top of the panel anyway so can you depend on gaining any charge for a few months?

It's only a 2 Wh battery ... a 200 Watt peak south facing vertical panel can fill that up just fine on a not terribly overcast day.

Then again, with a large solar panel why would the space requirements be such a problem? Put a box on the back of the solar panel with a Skelcap 3200F supercapacitor and circuitry. Rated to -40 Celsius, just a bit large.

[tszaboo]

And just to state the obvious: Dont charge primary batteries.

It seems we both kinda missed the solar panel.

With solar panel, it will be charged, so a primary battery is not optimal; with didoe-ORring, solar panel would only double the battery life or so, it would make more sense to double the battery and not use solar panel at all.

No, I didn't miss the solar panel. But you even answered the question about it.
I am even fairly sure that in a well designed system with primary battery, the system will last longer and requires less maintenance and battery replacement, than another system with a li-ion battery and a solar charger for it.

[Siwastaja]

Yes, I agree the first thing is to see if it can be made run with a primary battery.

With solar cell and snowy cold winter, you are going to need at least 2-3 month runtime without any power input. If you can make that happen, try to up it by order of magnitude to 2-3 years; that would be quite acceptable maintenance period (change the battery). Even if you use rechargeable cell, it might age fast in summer heat, and >5 year lifetime is not easy to guarantee.

The thing can be made to report its own battery voltage.

[floobydust]

OP you're system already won't make it if you want it small, cute and low cost. Most of Canada is easily colder than -25°C.
The remote monitoring systems systems I designed used in the Boreal forest have to survive the winter, which I've seen go below -40°C as well as the solid month of cloud.
Either use lithium primary batteries- common for satellite systems or I used SLA for private radio.

SLA's rated for -40°C Enersys Cyclon and they are expensive, some rated to -65°C.
At low temperatures batteries have less capacity as well as lower charge acceptance. You'll be putting in much more than their Ah rating when charging.
Some sites only had helicopter/Argo access (no roads), so saving a few bucks is a bad idea. The solar panel needs to be large, and I used tall mast clearing the tree line or it will catch little sun during winter due to the sun's low angle and shading.

[wizard69]

If you want rechargeable you will have to come up with a temperature control system for the battery or look into non mainstream batteries.   With enough insulation that can be accomplished with a few watts.   However you then need to consider how you cool them when the weather turns better.   Also -20 for Canada is pretty mild.   It would likely to be better to design for -30 degrees C.

Here is a cell that is actually spec'd for charging down to -30: file:///home/wizard69/Downloads/Data.Sheet.MP.176065.xtd.31109-2-0819.pdf.   Probably not cheap and finding the right electronics might be a problem.   This is just one example but honestly it is hard to find these alternative solution as low temp operation is a huge problem for lithium cells and your searches get flooded with 16750 articles.

[Daixiwen]

Your link got somehow changed to a local one, but from the reference I guess you meant this one: Data.Sheet.MP.176065.xtd.31109-2-0819.pdf
It's not exactly a small cell (19*61*69 mm) and its capacity is 10 times the requirement, but it should work well in cold weather. It's probably not going to be very cheap either.
In the same family Saft also has the smaller MP 174565

[sharif]

Thank you everyone for the provided information.
Let me add more details.
I'm using the BQ25570 chip. The chip is so efficient, that I can charge the battery indoors. As I mentioned the available space is so tight, something around 20cmx12cmx5cm maximum.
For the test, I tried This solar panel.
First, I tried a 5F supercap, but I saw the self-discharge is so high and I thought it would be better to use a secondary battery.
The battery I'm using now is a 105mA Li-Po. To test, I connected a load that draws 10mA each second with a pulse width of 8ms(on-time). In my room (considering low daylight and night with a roof lamp) I can say the daily voltage drop is about 0.005v.
If I put the solar panel in direct sunlight, it could charge the battery full in around 10 minutes(from ~3.9v to ~4.2). So, I'm hopeful it may work in shade for a long time if I choose a bigger battery by considering the winter.
But still, the main problem is not the sun, it's the temperature.

[sharif]

1- How much current do you need?
2- Do you need to charge during coldest times?
3- Is thermal insulation an option? For example, if charging at -20degC is rare and you have enough external power available, you can just first heat up the battery and then commence charging.

1- A pulse between 10mA to 80mA
2- Could be an option if the battery won't die due to the temperature.
3- I don't think be possible.

[sharif]

You should use Lithium primary batteries, LiSoCL4. Buffer it with a large capacitance or supercapacitor depending on the length of the peak current.
LiMNO2 could work, but you will only get like 2.3V or so per cell in cold.
And just to state the obvious: Dont charge primary batteries.

Thanks, could be an option if I use it as a backup only and if it lasts for more than 3 years at least.

[sharif]

Yes, I agree the first thing is to see if it can be made run with a primary battery.

With solar cell and snowy cold winter, you are going to need at least 2-3 month runtime without any power input. If you can make that happen, try to up it by order of magnitude to 2-3 years; that would be quite acceptable maintenance period (change the battery). Even if you use rechargeable cell, it might age fast in summer heat, and >5 year lifetime is not easy to guarantee.

The thing can be made to report its own battery voltage.

Thanks, do you think the primary battery can tolerate this weather? for at least three years?

[sharif]

OP you're system already won't make it if you want it small, cute and low cost. Most of Canada is easily colder than -25°C.
The remote monitoring systems systems I designed used in the Boreal forest have to survive the winter, which I've seen go below -40°C as well as the solid month of cloud.
Either use lithium primary batteries- common for satellite systems or I used SLA for private radio.

SLA's rated for -40°C Enersys Cyclon and they are expensive, some rated to -65°C.
At low temperatures batteries have less capacity as well as lower charge acceptance. You'll be putting in much more than their Ah rating when charging.
Some sites only had helicopter/Argo access (no roads), so saving a few bucks is a bad idea. The solar panel needs to be large, and I used tall mast clearing the tree line or it will catch little sun during winter due to the sun's low angle and shading.

Agree with you.
But it should be really low cost and small. So SLA won't be an option.

[NiHaoMike]

Some AA low self discharge NiMH would probably be the best choice. It has the widest operating temperature range of the mainstream battery chemistries.

[Siwastaja]

Thanks, do you think the primary battery can tolerate this weather? for at least three years?

Any primary cell, and most secondary cells (including li-ion) only enjoy the cold. Self-discharge and capacity fading are slowed down.

The problem is just with getting the charge out; in other words, ESR rises. And here's the big question - how long is your 80mA pulse going to be? Microseconds? Milliseconds? Seconds? If you can supply the single pulse with capacitors (note: electrolytic caps also increase ESR in cold!), so that the battery cell only sees a few mA of average current, then almost any battery will do.

But if you need to pull 80mA from the battery at -40degC, and battery cannot be much larger than say 1Ah, then we are talking about C/10. Not going to happen with almost any off-the-shelf battery.

[mzzj]

OP you're system already won't make it if you want it small, cute and low cost. Most of Canada is easily colder than -25°C.
The remote monitoring systems systems I designed used in the Boreal forest have to survive the winter, which I've seen go below -40°C as well as the solid month of cloud.
Either use lithium primary batteries- common for satellite systems or I used SLA for private radio.

SLA's rated for -40°C Enersys Cyclon and they are expensive, some rated to -65°C.
At low temperatures batteries have less capacity as well as lower charge acceptance. You'll be putting in much more than their Ah rating when charging.
Some sites only had helicopter/Argo access (no roads), so saving a few bucks is a bad idea. The solar panel needs to be large, and I used tall mast clearing the tree line or it will catch little sun during winter due to the sun's low angle and shading.

Agree with you.
But it should be really low cost and small. So SLA won't be an option.

2.5Ah Single D-size cyclon SLA should fit to your space constrains and cost about 10usd.
Or used to cost..now out of stock or out of production. 

[sharif]

Interesting!

[mzzj]

seem to be available
https://eu.nkon.nl/enersys-cyclon-2v-2-5ah-loodaccu.html?gclid=Cj0KCQjw2_OWBhDqARIsAAUNTTF_zJYmVysMU4tXjY5Z5sgwA0Ee9cxVCtaPEnHiyNKhwFT7Q6R8c5gaAsixEALw_wcB

[Faringdon]

Bascially you could benefit from  a prototype off Bramble energy...they are doing small "PCB hydrogen fuel cells" to make electricity. Not affected by temps down to -40degc. (can be charged down to -40degc)
https://www.brambleenergy.com/technology-be