Cell capacity at different temperatures

Oz18650

Member
Joined
Oct 2, 2017
Messages
209
I have been doing repeated capacity tests on 2 cells to look for loss of capacity over number of cycles the cells have done
This has been inspired by "generic" on here doing similar testing.
I am using a single tp4056 to charge and a single zb206 to test, so charging and testing should be as consistent as possible.
In Sydney at the momwnt emperatures get fairly cold during the night and warm during the day.
My testing is being done outside the house in an undercover area, so temperatures vary quite a bit depending on the time of day.
I have noticed that the capacity results seem to change with temperature. Tests started at night when it is colder seem to give lower capacity results and tests done during the day seem to give higher results.
Overnight temperatures probably get down to 5 degrees C.
Daytime temperatures are probably around 25 degrees C.
Today's forecast says 3 -19.
Overnight tests might give approximately 1950mah and daytime tests might give approximately 2050mah

My questions are.
Have other people noticed temperature affecting capacity results?
If so, do people try consider this and do testing in controlled conditions?
 
It's sure related.
That's why in every data sheet is a nominal ma at approx 18 degrees Celsius.
When it is 25 degrees Celsius you will get a higher number.
Some sheets have even data on temperature
 
Of course. Thats why 100mah Range is Good enough to sort cells in for most People...
 
Chemical reactions generally slow down at lower temperatures, so performance of batteries that convert chemical potential energy to electrical energy degrades at lower temperatures, and it's particularly noticeable as it nears its empty state.
So while a warm battery at 20% State of Charge (SoC%) might be able to supply 1A, the same battery at 3degC might only be able to supply 0.2A, making the tester/device think the battery is already empty.

Not 100% sure about this, but if you can test at say 100mA, you'd probably get much closer results regardless of temperature.

Some extra interesting info:
(Tesla?) EV cars sometimes warm up the batteries to get better charging/driving performance.
You can squeeze out some more power from cold batteries by rubbing (friction heating) them.
Cold car batteries usually perform better at the 2nd cranking try, because the first failed attempt warmed up the battery.
Cold weather camera(wo)men keep batteries inside their jackets to get more filming time.
 
daromer said:
Of course. Thats why 100mah Range is Good enough to sort cells in for most People...

Putting cells into 100mah range bins can make sense as the actual values will average out because of the large number of cells.
Temperature changes affecting WHICH 100mah range the cells end up in could result in different capacity packs being made.
My repeated tests for the 2 cells appear to show about 1950mah when tested while it is cold (about 5 to 10 degrees C / 41 to 50 degrees F) and show about 2050mah when warmer (about 20 to 25 degrees C / 68 to 77 degrees F).
Without temperature being controlled (ie lots of peoples workshop/shed areas) If you make a pack in Winter and another pack in Summer or measure a large number of cells at night then a large number of cells in the day your pack capacities may vary (by 5% from the testing I have done) even when you use the same test result numbers.
 
Oz18650 said:
I am using a single tp4056 to charge and a single zb206 to test, so charging and testing should be as consistent as possible.

Have you tested that consistency you have assumed ?

Same cell, tested 10 times, same temperature .....
 
Sean said:
Oz18650 said:
I am using a single tp4056 to charge and a single zb206 to test, so charging and testing should be as consistent as possible.

Have you tested that consistency you have assumed ?

Same cell, tested 10 times, same temperature .....

So far I have done 42 tests, so 21 on each cell.
Since I noticed that temperature seems to affect the results I have made a note of "cold" or "warm" at the start of the test being run and lower results are matching with "cold" and higher results are matching with "warm".
I might start recording temperature and time at tree start of a capacity test. A 9:00am test would mean that temperature is rising during the test while a 5pm yet would mean that temperature is dropping during the test.
I might also start a test with a another cell and try to test at consistent times/temperatures.
 
Oz18650 said:
Overnight temperatures probably get down to 5 degrees C.
Daytime temperatures are probably around 25 degrees C.
Overnight tests might give approximately 1950mah and daytime tests might give approximately 2050mah

I'd expect more difference with such a loosetemp range.
Ideal temperature for a Lithium cell is 23-27*C. Lower/higher temperatures assume lower capacity or reduced life and/or reduced charge/discharge current. That's all according to official datasheets.

Usually charging is allowed at 0-45*C and discharging at -20-60*C. For some high-drain cells there is an 80*C temperature cut requirement.

Oz18650 said:
My questions are.
Have other people noticed temperature affecting capacity results?
If so, do people try consider this and do testing in controlled conditions?

I have
I do
:)

P.S. a screenshot from Panasonic NCR18500 datasheet:

image_wborbd.jpg
 
Generic said:
How is your test coming along?

I have started recording the time that each test starts as well as the temperature of the cell measured with a laser pointer themometer when the test starts.
I am at 100 tests now and while I still have to do some analysis of the data, capacity results look pretty closely linked to temperature.
 
Its nothing new that temperature affects capacity.. it also affects the testers used...

This goes back to My first videos where i stated clearly that the envieonment you do the tests in need to be consistent.


And IF you have Read any datasheet or papper about batteries this should not come as a suprise but Still People tend to be supprised? :)

As an example a cAr starts easier the 2nd crank du to the battery heated Up..or when we fly quads or race cars we preheat thebatteries to gain run time and capacity.


IF you look for a Variable to calculate capacity derived by temperature you can find Them in datasheet and scientific papers :)

I understand people like to figure stuff out Them self but this is a thread that isnt anything new :)
 
While it is nothing new that temperature affects capacity, do you know if cycle life can improve in cooler, but not too cold, conditions, similar to how limiting depth of charge/discharge increases cycle life? Every data sheet I have ever seen only shows cycle life at 20 or 25*C. And we all know that 40*C will kill cells twice as fast or faster. But how about like 10 or 15*C? I had an idea of using a chest refrigerator with an external regulator set for like 15*C as the ultimate fire-resistant, low humidity, temperature regulated battery cabinet, assuming it could support the weight of the batteries, of course.

Also, for capacity testing, you need A LOT of cycles to see any results, unless you start with cells that are pretty worn out to begin with. In my test, I was expecting results right away because everyone says generic cells are bad without any sort of proof. Are you going to start a thread on your cycle testing? I'm curious about the cells you are using - age, % of capacity, brand, etc.
 
Not sure IF use during 10c yield longer life... But doing the math you get more cycles but less total capacity. So without data i would say No total gain in the end.

Thats One of the things i havent seen either nor tested since it takes years to test
 
I have not read many data sheets.
Having said that the only ones which have mentioned a temperature was just a single temperature - nothing about "W capacity at X temperature" and "Y capacity at Z temperature"
While everyone here probably knows that temperature can affect chemical reactions, this is a theoretical idea until it has a practical application.
Is the effect something big or small, will it affect what people are trying to achieve, or not?
I think what I have observed and posted about might change how people want to do thier testing to create balanced packs.
 
I cycle my power tool batteries sometimes three times a day.
Hilti 22v5.2Ah a pack.
Drill, jigsaw, circle saw, grinder and such, most important one, the radio
I use them in minus 15c and plus 40c. Charge them in the full sun, rain,ect
Two packs are running for two years know.
Maybe its how the bms is programmed for the soc, but i don't have the feeling that they are losing capacity.

For my powerwall i am not really afraid of sub zero temperatures, but i have to dig in to it more.
So get a second opinion on this.

As far for my testing endeavors, especially discharge, there is a huge difference with temperature.
15c 20 and 25c.
Winter is coming, it would be fun to see how known cells are going to react at 10c 5c 0 -5 and -10c

Best
 
There is a huge difference in - for Lithium Ion. You can loose easy 50% of the available capacity. (Loose as in useable)
The capacity as such is not "Lost" since as soon as they get warm again you can use it again.

Its like that for most of the chemistries. The reason EV cars have cooling/heating around their packs as example or why a car is hard to start during winter.
 
So basically Hilty packs are "limited" to 50% soc, sort a speak.
That's why i don't notice the difference between summer and winter.
And when in use, they will warm up a little.

But for the powerwall you want to have more out of it, so you will notice.
And must apply some heat to them.
I assume in Sweden it would be more difficult to ride a EV?
The average temperature is a bit lower.

Would LTO not be more efficient in a car, or it does not matter?
 
@ajw22 in the case of Li-Ion that is not valid. Li-ion works great in low temperatures and it's a disaster at high temperatures.
This is valid for anything from smartphone batteries to 18650 cells.
 
Powertool batteries tend to heat up yes. You can notice it in beginning or on cheap ass gear but most powertool brands will work due to high current creating inside the cells :p
 
Can you identify the cell manufacturer and part number?
If so, please share that information since it seems you're committed to life-cycle testing these cells.
Also, could you include the charge time, the rest time, and the discharge time - in otherwords, what your cycle looks like?
And how often you cycle - how many times a day?
As to your question on how temperature affects cells, the "aging path" is affected by cell temperature, cell nominal or average state-of-charge, and cycling severity.
Even daily thermal cycling of the cell - similar to the "cold start" of an automobile on the driveway needing time to "warm up", affect cell aging.
 
Back
Top