Why you cannot charge LiFePO4 below 0 degrees Celsius

[Will Prowse]

This study summed it up nicely:

With the complex material system used in LIBs, the performance degradation at low temperatures can be attributed to several different sources. First, the low temperature will affect the property of electrolyte. With the decrease of temperature, the viscosity of the electrolyte will increase, which will reduce the ionic conductivity. The internal resistance will subsequently rise due to the increase in the impedance of the directional migration of chemical ions. To counter such effect, electrolytes with low freezing point were explored, and different electrolyte additives were studied [51], [71], [72], [73], [74]. Bugga et al. [51] presented a guideline for developing formulations of low temperature electrolytes during their research of LIBs for space applications. The improved electrolytes can be used in an extended operating temperature range (Fig. 2A) and were proved to be effective in aerospace batteries. Li et al. [75] also reported an optimized electrolyte formulation of 1.0 M LiPF6 in ethylene carbonate(EC)−propylene carbonate(PC)−ethyl methyl carbonate (EMC) (1:1:8 by wt) with 0.05 M CsPF6. Such formulation enabled a capacity retention of 68% for the batteries tested at −40 °C, while the ones with conventional formulation only showed a capacity retention of 20% (Fig. 2B). Specific electrolyte additives, such as lithium difluorophosphate (LiPO2F2), were also proved to be effective in improving the performance of LIBs at low temperature (Fig. 2C) [74].


The increase of charge-transfer resistance in LIBs is also an important factor that contributes to the performance degradation at low temperatures. The charge-transfer resistance of LiFePO4-based cathodes at −20 °C was reported to be three times higher than that at room temperature [76]. Such high charge-transfer resistance largely affects the kinetics in batteries. The study of LIB performance at low temperatures by Zhang et al. [77] demonstrated that the charge-transfer resistance significantly increased when the temperature decreased. The charge-transfer resistance of a discharged battery normally is much higher than that of a charged one. Charging a battery at low temperatures is thus more difficult than discharging it. Additionally, performance degradation at low temperatures is also associated with the slow diffusion of lithium ions within electrodes. Such slow down can be countered by altering the electrode materials with low activation energy. For example, Li3V2(PO4)3 (LVP), which has an activation energy of 6.57 kJ mol–1, showed a 200x improvement of apparent chemical diffusion coefficient of lithium ions over LiFePO4 (LFP) with an activation energy of 47.48 kJ mol–1 at −20 °C [78].


Another typical effect that occurs at low temperatures is lithium plating [79], [80], [81]. The cold condition will trigger the polarization of anodes and lead to the approach of the potential of graphite and other carbon based anodes to that of lithium metal, which would slow down the lithium-ion intercalation into the anodes during charging process [82]. The aggregated lithium ions are thus deposited on the surface of the electrodes, which causes the reduction of the battery capacities. Furthermore, the lithium plating exists in the form of dendrite, which may penetrate the separators, and result in the internal short-circuit [83].

Source: https://www.sciencedirect.com/science/article/pii/S1002007118307536

 

[nwillitts]

thank you.

 

[Mark.Jackson]

thank you.

 

[grizzzman]

"Cold weather LFP charging."

 

[Steve_S]

@Will Prowse I believe this type of info should be STICKIED because we all know, it will be regurgitated, over & over again...

It may also be prudent to have a Sticky Post related to EMI/RFI and cable routing and management to prevent / reduce such. Too many folks are suffering from Ripple effects and more, many related to having battery cables separated and/or with long runs.

 

[HRTKD]

"Cold weather LFP charging."

I can hardly understand what the narrator is saying in that video. It's like he's talking into a toilet at warp speed.

 

[grizzzman]

I can hardly understand what the narrator is saying in that video. It's like he's talking into a toilet at warp speed.

That's too bad, he has "real" info.

 

[toms]

The article quoted is saying that with the correct electrolyte and anode construction, charging at minus 40deg C is possible.

Not sure how much access the author has to the better manufacturers, but anode construction is continually evolving.

I guess the post should be “why your cheap LiFePO4 cells can’t be charged when frozen”

 

[Deleted member 9967]

Some parts were clear as mud and other parts I actually understood. Like the dendrites piercing the foil and causing a short.
Scary stuff.
Well I have been saying to never charge below zero, and now everyone knows the "why".

Thanks Will for the post.

 

[Dzl]

"Cold weather LFP charging."

That's too bad, he has "real" info.

I recently brought this up in another thread on low temp charging. If I recall, you were the person who first showed me that video.

Since its easier to show a chart than a video (and many people won't click through to watch a video or read an article) I made (well adapted) this chart to illustrate the authors main basic point on low temperature charging but the video has a lot more useful info.

edit: wording on the right of the graph may be confusing or misleading, if it is, here is some clarification:

"This chart shows the relationship between temperature, charge rate and battery degradation. Any point along the blue line represents roughly equivalent degradation per cycle"

I also highly suggest anyone struggling to understand the graph watch the video, it has much more context, the graph is just a screen grab from the video with some added context.

 

[Alber C]

well good to know this.

 

[HaldorEE]

The article quoted is saying that with the correct electrolyte and anode construction, charging at minus 40deg C is possible.

Not sure how much access the author has to the better manufacturers, but anode construction is continually evolving.

I guess the post should be “why your cheap LiFePO4 cells can’t be charged when frozen”

Possible is not the same thing as wise.

Based on Dzl's chart, -20 Degree C charging should be limited to 0.1C rates. You are off the scale (0.0C) at anything -30 and below.

With my 120AH battery pack I plan on a max of 0.5C charge rate. If I use the 280 AH battery pack, the max charge rate would be more like 0.2C. The batteries should last forever at that charge rate.

This might make up for the degradation caused by charging at elevated temperatures. I almost said I was glad to be living in Phoenix. After this summer, I doubt if I will ever say that again.

 

[Steve_S]

A reminder for people with battery packs in Parallel.
If properly paralleled, the battery packs will split the charge coming in, essentially dividing the amperage between them. There is a GOTCHA lurking in this as well. The individual packs will vary somewhat and the amps will float around a little while charging is occurring. One battery may take more or fewer amps than the others at any given point in time because of the cell absorption rates.

Also a Critical Reminder about LiFePo4.
LFP is a "Chemistry Family" there are sub-categories with slightly different chemistry components, for example, LFP with Yttrium which is suitable for Cold Temps (also more $$$). At this time I believe there is roughly a dozen sub-chemistries for special applications and for energy handling capability. Remember that basic ESS level Cells are typically rated for a Max of 1C Discharge rate but there are LFP chemistries which allow for 5C Discharge (way more $$$)

Keep these details in mind when looking at the Temperature Ratings for cells to ensure you are comparing the same apples (McIntosh to MacIntosh, not Red Delicious to Crab Apples). As usual, the Devil is in the Details.

 

[Reed Cole]

I recently brought this up in another thread on low temp charging. If I recall, you were the person who first showed me that video.

Since its easier to show a chart than a video (and many people won't click through to watch a video or read an article) I made (well adapted) this chart to illustrate the authors main basic point on low temperature charging but the video has a lot more useful info.

I think there could be some improvements in the wording. Maybe something like:

"The blue line represents the rate of charge needed to maintain a consistent rate of degradation per cycle as temperature changes."

The way it's currently worded makes it sound like the wear and tear is a variable when it's actually the only thing being represented that is constant. It was a little confusing to me.

 

[Dzl]

I think there could be some improvements in the wording.
The way it's currently worded makes it sound like the wear and tear is a variable when it's actually the only thing being represented that is constant. It was a little confusing to me.

You are right, looking at it again without the commentary from the video being fresh in my mind, the graph and the wording is a little confusing and difficult to interpret.

Though I think the proposed wording below doesn't quite capture it either:

"The blue line represents the rate of charge needed to maintain a consistent rate of degradation per cycle as temperature changes."

The blue line does not represents the rate of charge, the y axis is rate of charge, the x axis is temperature, the blue line I'm struggling to define precisely, but I think your language works well: "roughly equivalent degradation per cycle"

Maybe: "This chart shows the relationship between temperature, charge rate and battery degradation. Any point along the blue line represents roughly equivalent degradation per cycle"

Does that sound more clear to you?

 

[HRTKD]

Maybe: "This chart shows the relationship between temperature, charge rate and battery degradation. Any point along the blue line represents roughly equivalent degradation per cycle"

It took me a while when I first saw that graph, but I eventually figured it out. Maybe one of the subsequent posts helped too. The wording suggested above would have helped a lot.

 

[Dzl]

Updated the post with a clarification (but not the image, maybe later). Thanks for bringing this up @Reed Cole and @HRTKD

 

[bexamous]

I made (well adapted) this chart to illustrate the authors main basic point on low temperature charging but the video has a lot more useful info.

I like that graph being continuous. I feel like too much emphasis is put on 0C as being this magic number where charging below is awful, but above is just fine. There is no water in a lifepo4 battery, or any(?) other lithum chemistry batteries. Nothing special happens at 0C yet its always referenced as this temperate at which bad things happen.

In OP's linked article:

Temperature effect and thermal impact in lithium-ion batteries: A review

Lithium-ion batteries, with high energy density (up to 705 Wh/L) and power density (up to 10,000 W/L), exhibit high capacity and great working perform…

www.sciencedirect.com

I'm ont sure it really identifies 0C as special in any way, it does have this quote:
> The performance of LIBs will degrade at temperatures below 0 °C

Statement has two citations, one is to:

A comprehensive review on hybrid electric vehicles: architectures and components - Railway Engineering Science

The rapid consumption of fossil fuel and increased environmental damage caused by it have given a strong impetus to the growth and development of fuel-efficient vehicles. Hybrid electric vehicles (HEVs) have evolved from their inchoate state and are proving to be a promising solution to the...

link.springer.com

Look at it I see no reference to 0C, the other is to:

https://iopscience.iop.org/article/10.1149/2.1221802jes/pdf

Here too I see no mention of 0C. Not that statement is incorrect, but I could say performance degrades below 5C and that is true too, eg:
Long link to pdf

That pdf is kinda nice with graphs easy to see, haha, eg slide 12 showing dendrite growth.. its not fine and then some magic number and then awful.. 5C growth is already multiple times greather than 25C. Yes -5C gets even worse.. but 0C isn't some special cutoff.

I mean at some point you need to have cutoffs, and 0C is often in spec sheets so its pretty reasonable cutoff to use. Then again if you want to extend life of battery you can avoid 100% DOD and use less of its capacity to increase number of cycles.. you can also choose 5C or something as cutoff to extend life. Or if using heater don't try to keep batteries >=0C but >=5C.

 

[Dzl]

I like that graph being continuous. I feel like too much emphasis is put on 0C as being this magic number where charging below is awful, but above is just fine. There is no water in a lifepo4 battery, or any(?) other lithum chemistry batteries. Nothing special happens at 0C yet its always referenced as this temperate at which bad things happen.

In OP's linked article:

Temperature effect and thermal impact in lithium-ion batteries: A review

Lithium-ion batteries, with high energy density (up to 705 Wh/L) and power density (up to 10,000 W/L), exhibit high capacity and great working perform…

www.sciencedirect.com

I'm ont sure it really identifies 0C as special in any way, it does have this quote:
> The performance of LIBs will degrade at temperatures below 0 °C

Statement has two citations, one is to:

A comprehensive review on hybrid electric vehicles: architectures and components - Railway Engineering Science

The rapid consumption of fossil fuel and increased environmental damage caused by it have given a strong impetus to the growth and development of fuel-efficient vehicles. Hybrid electric vehicles (HEVs) have evolved from their inchoate state and are proving to be a promising solution to the...

link.springer.com

Look at it I see no reference to 0C, the other is to:

https://iopscience.iop.org/article/10.1149/2.1221802jes/pdf

Here too I see no mention of 0C. Not that statement is incorrect, but I could say performance degrades below 5C and that is true too, eg:
Long link to pdf

That pdf is kinda nice with graphs easy to see, haha, eg slide 12 showing dendrite growth.. its not fine and then some magic number and then awful.. 5C growth is already multiple times greather than 25C. Yes -5C gets even worse.. but 0C isn't some special cutoff.

I mean at some point you need to have cutoffs, and 0C is often in spec sheets so its pretty reasonable cutoff to use. Then again if you want to extend life of battery you can avoid 100% DOD and use less of its capacity to increase number of cycles.. you can also choose 5C or something as cutoff to extend life. Or if using heater don't try to keep batteries >=0C but >=5C.

From what I've seen (from spending too much time looking at datasheets of cells I don't even own...) cell manufacturers usually consider 0*C a hard limit, below which charging should not happen (probably a reasonable but somewhat arbitrary number, its conveniently round and easy to remember).

But they also usually identify soft limits at higher temperatures. With reduced maximum current up to 15*C in many cases. As you noted, its not a black and white limit, Its a gradient, and its a relationship between a few factors (Temperature, C-rate, SOC, duration of charge).

Here is a screenshot from a Ganfeng Datasheet (the most detailed data on temperature and charging by a cell manufacturer I've found) that shows the relationship between temperature, c-rate, and soc:


Even with the above, they still consider 0*C the hard lower limit that should not be exceeded.

 

[Terrapin]

...I glazed over...Soooo far past me...... but, does this mean that the claims of the -20F chargeable LiFePO4 cells could be true?

 

[DerpsyDoodler]

...I glazed over...Soooo far past me...... but, does this mean that the claims of the -20F chargeable LiFePO4 cells could be true?

Based on what i’ve read in this forum, not for any practical use. I’m no expert on the matter, though.

 

[Deleted member 9967]

Here is a screenshot from a Ganfeng Datasheet (the most detailed data on temperature and charging by a cell manufacturer I've found) that shows the relationship between temperature, c-rate, and soc:
View attachment 28852

Even with the above, they still consider 0*C the hard lower limit that should not be exceeded.

So I am going to assume that the numbers IN this are the C rate at which they recommend charging? Is this correct?
thank you DZL for this.

 

[Dzl]

...I glazed over...Soooo far past me...... but, does this mean that the claims of the -20F chargeable LiFePO4 cells could be true?

Based on what i’ve read in this forum, not for any practical use. I’m no expert on the matter, though.

I am not an expert either but my understanding is:

The simple answer is just don't charge under 32*F

The complicated answer is... its complicated, but still no..
There are a number of variables to consider with low temperature charging (1) temperature (2) C-rate (3) SOC (4) degree of acceptable damage, (probably more). But @DerpsyDoodler is correct, practically speaking there is really no situation where you should be charging LFP at anywhere near -20*F / -30*C outside of some sort of emergency maybe.

 

[fafrd]

From what I've seen (from spending too much time looking at datasheets of cells I don't even own...) cell manufacturers usually consider 0*C a hard limit, below which charging should not happen (probably a reasonable but somewhat arbitrary number, its conveniently round and easy to remember).

But they also usually identify soft limits at higher temperatures. With reduced maximum current up to 15*C in many cases. As you noted, its not a black and white limit, Its a gradient, and its a relationship between a few factors (Temperature, C-rate, SOC, duration of charge).

Here is a screenshot from a Ganfeng Datasheet (the most detailed data on temperature and charging by a cell manufacturer I've found) that shows the relationship between temperature, c-rate, and soc:
View attachment 28852

Even with the above, they still consider 0*C the hard lower limit that should not be exceeded.

I’m not understanding the meaning of that table.

At 10degC and starting with a cell at 0% SOC, what is it saying, that charging rate should be limited to 0.5C to achieve 100% SOC or that SOC at that temperature can never exceed 50% SOC when starting from 0% SOC?

 

[DerpsyDoodler]

I’m not understanding the meaning of that table.

At 10degC and starting with a cell at 0% SOC, what is it saying, that charging rate should be limited to 0.5C to achieve 100% SOC or that SOC at that temperature can never exceed 50% SOC when starting from 0% SOC?

I read it as “at x temperature, and for y as the existing state of charge, only push z amount of charging current.”