Science based discussion on charging LiFePO4 below 32°F

[SeaGal]

It wouldn't be hard for an engineer designing a BMS to scale charge current with temperature and state of charge. Seems like it could be done in firmware. For example: 10A@-10°C, 15A@-5°C, 20A@0°C, 30A@5°C, 40A@10°C, 100A@20°C. This would allow them to better protect the battery from damage at cold temperatures. But I'm not sure if any of the manufacturers have figured that out yet or implemented it.

This is what I have done myself, but also using SOC info too, it's not complex! ?‍? Have a Solis Inverter and Overkill BMS and an ESP32 based system that sends the BMS info to the Solis over CANBus. Using EVE LF280K charge vs. temperature recommendations (see attached) it just tells the Solis what charge current to use, based on cell temps and SOC info from the Overkill BMS. A few lines of code adjust the charge current per battery temp degree from 0 to 10 degrees with different mappings for 3 different SOC settings.

The good thing is that the BMS will just do its job and protect the cells if my code fails in an unlikely way (if it hangs, the CANbus messages will fail to be sent and the Solis stops using the battery anyway). And my battery heater should mean this code is rarely used.

Obviously you could just program your own charge vs SOC and temp mappings depending on your needs / battery type etc.

 

[eemarty]

Also here's some interesting info I found in the 2022 Nissan Leaf Owners Manual. The leaf uses a different chemistry, not LFP so take all this with a grain of salt.

I couldn't find an actual charge rate vs temperature table, but we can infer some things from what is said. It looks like they are only using their heater below -17°C!

For models with 40 kWh battery: The Li-ion battery warmer automatically turns on when the Li-ion battery temperature is approximately -1°F (-17°C) or colder. The Li-ion battery warmer automatically turns off when the Li-ion battery temperature is approximately 14°F (-10°C) or higher.

For models with 62 kWh battery: The Li-ion battery warmer automatically turns on when the Li-ion battery temperature is approximately -4°F (-20°C) or colder and outside temperature is approximately -11°F (-24°C) or colder. The Li-ion battery warmer automatically turns off when the Li-ion battery temperature is approximately 0°F (-18°C) or higher, or outside temperature is approximately -8°F (-22°C) or higher.

When the ambient temperature is 32°F (0°C) or less, charging time may be longer than normal and the level to which the Li-ion battery can be charged may be less than at higher temperatures.

Nissan's quick charge looks like about 1.2C to 1.6C charge rates, but obviously it's lower at lower temperatures.

Quick charge uses public charging stations (up to 50 kW of power [for 40 kWh battery models] / 100 kW of power [for 62 kWh battery models]) to charge the battery in a short period of time.

It may take more time to charge the Li-ion battery using the quick charger if the temperature of the Li-ion battery is high or low

And this hilarious one:

If the charge port is frozen, melt the ice using a hair dryer. After the ice has melted, charge the Li-ion battery

 

[SeaGal]

Yes, but guess the -17 degree limit is for discharge protection, rather than for charging - it wouldn't make sense for a car to use energy to heat for charging unless it was imminent it was about the be charged. I read somewhere that Telslas turn on heaters if the charge is low and you select a fast charging station on the Sat-Nav.

Re the 'charge time may be longer if < 0', it could be the batteries are heated up for 15 mins or so, before charging starts - just a thought.

 

[eemarty]

Yes, but guess the -17 degree limit is for discharge protection, rather than for charging - it wouldn't make sense for a car to use energy to heat for charging unless it was imminent it was about the be charged. I read somewhere that Telslas turn on heaters if the charge is low and you select a fast charging station on the Sat-Nav.

For Tesla, They want to be able to fast charge at up to 3C rates so they probably pre warm the battery to enable a faster charge. I would love to see the table Tesla uses of safe charge rate vs cell temperature and state of charge though. And I'd love to see the test data they got to create that table. I highly doubt that's public though.

As far as I can tell the Nissan LEAF warmer only keeps the temperature above -20°C then shuts off at -18°C for the 62KWh battery. If plugged into a charger at home, it will maintain that temperature while the car is parked. They talk about this a lot in the manual and I can't find anywhere that it says it will heat above that to charge. I believe it charges at that temperature.

pg5-173:

To prevent damage to the Li-ion battery: Do not store the vehicle in temperatures below -13°F (-25°C) for over seven days. If the outside temperature is -13°F (-25°C) or less, the Li-ion battery may freeze and it cannot be charged or provide power to run the vehicle. Move the vehicle to a warm location. NOTE: Connect the charger to the vehicle and place the power switch in the OFF position when parking the vehicle if temperatures may go below -4°F (-20°C). For models with 40 kWh battery model, this provides external power to the Li-ion battery warmer (if so equipped) when it operates and does not discharge the Li-ion battery. Vehicle driving range is reduced if the Liion battery warmer (if so equipped) operates (Li-ion battery temperature approximately -4°F (-20°C) or colder) while driving the vehicle. You may need to charge the Li-ion battery sooner than in warmer temperatures.

pg EV-6:

The Li-ion battery warmer does not operate if the normal charger is not connected to the vehicle. To help prevent the Li-ion battery from freezing, do not leave the vehicle in an environment if temperatures may go below -4°F (-20°C) unless the vehicle is connected to a charger.
The Li-ion battery warmer helps to prevent the Li-ion battery from freezing when the temperature is cold. The Li-ion battery warmer automatically turns on when the Li-ion battery temperature is approximately -4°F (-20°C) or colder and outside temperature is approximately -11°F (-24°C) or colder. The Li-ion battery warmer automatically turns off when the Li-ion battery temperature is approximately 0°F (-18°C) or
higher, or outside temperature is approximately -8°F (-22°C) or higher. The Li-ion battery warmer operates when the normal charger is connected to the vehicle, and it automatically uses electrical power from either the external source or from the Li-ion battery.

 

[sunshine_eggo]

It's very possible that EV cells may be Yttrium doped. IIRC, this lowers the safe charge threshold to around -20°C.

My limited understanding is this adds cost and is not typically done unless specifically needed.

 

[justgary]

At least around here, I almost always recommend an underground vault.

Where is "here", and what vault do you recommend?

 

[RCinFLA]

You need to keep in mind cell electrode thickness. It changes a lot of cell attributes.

Cells capable of high current, for both charge and discharge currents, are made with thinner electrodes. To get similar AH capacity from a cell with thinner electrodes you need more layers, which also adds more copper and aluminum for each additional layer, which also reduces current collection resistance of cell which is also good for high currents. But all this cost more money per AH capacity and added volume and weight to cell.

If AH's is all the customer cares about, the cheapest approach is to screen print down thick electrode material to copper anode and aluminum cathode current collector foils. There is an upper limit to manageable electrode thickness when rolled up into a cylindrical cell. Eventually the electrode material just cracks and crumbles when trying to bend the thick coating into a roll. Generally, cylindrical cells have thinner electrodes than prismatic cells, but nothing technically stops a manufacturer from making a thin electrode, high peak current, prismatic cell. Cylindrical cells are better for dissipating internal self-heating.

Thinner electrodes require lower overpotential for given current per layer area. Overpotential drives battery kinetics, both good and bad aspects of operation. At cold temp for a given cell design, the overpotential must increase to get the same cell current, but a thinner electrode design will handle lower temperatures better than a thick electrode design.

Big mistake if you take data from a lithium-ion cell of a given electrode thickness and apply it to another cell of significantly different electrode thickness.

 

[HighDesertOffgrid]

If only there were someone who started a DIY solar forum and constantly gets free batteries sent to them for review that would donate some Guinee pig batteries for real world testing of such matters... ?

 

[fastline]

As for mine, I have decided to hold off charging even though the temps were showing about 35F today on the batteries and will probably be 40F in a couple days. I am going to expedite another other project which will enable me to get them out of storage. They have not been charged in over a year and if something happens, I would have no way to stop it.

As for cold weather charging, it really sounds like at the very least, reduced charging rates may apply, in which that is a compounded issue for offgrid solar stuff as you likely would be working with reduced sun hours anyway so max charging would be a good goal.

One of the reasons I advocate for an underground install is the safety aspect. I have about 100kwh of storage so that could present some risk.

Just a thought but the fire proof insulated room works too! I just see so many solar systems built on wood with no thought to "what if".

 

[sunshine_eggo]

@fastline

For the love of all that is holy, will you tell us how these cells are all sitting at 2.7V?

 

[fastline]

@fastline

For the love of all that is holy, will you tell us how these cells are all sitting at 2.7V?

I thought that was apparent above. Just "sitting". Like ignored and other stuff to do. Not a good strategy but I stay very busy.

 

[sunshine_eggo]

I thought that was apparent above. Just "sitting". Like ignored and other stuff to do. Not a good strategy but I stay very busy.

That shouldn't cause that. I let several dozen CALB cells sit for a year to the day, and they were all at 3.30V and retained >95% of their charge. Many of those same cells have sat for another year and all still measure 3.30V.

I let my 280Ah Eve cells sit for 5 months, and none of them lost more than 3Ah.

Cells received at 3.29V equating to 30-50% charge should not completely discharge for a very very long time.

"just sitting" isn't a reason. If it's because the BMS drained them over an extended period of time, I could see that, but even my cheapo JBD doesn't drain the cells just sitting there. Last charged not less than 4 months ago, the little 4S navitas battery has all 4 cells at 3.33V.

If I were you, I would be VERY concerned about the viability of the batteries not because they need to be charged, but why they're at 2.7V in the first place. Recommend you conduct capacity and IR testing ASAP.

 

[curiouscarbon]

have left 4 LFP cells fully charged for at least a year. they settled at 3.333 volts.

the BMS had charge/discharge mosfet disabled for this entire time. 100Ah 4S.

MOSFET BMS consume power with charge and discharge function enabled. keeping both disabled should reduce idle consumption, but the microcomputer will still consume power.

some mosfet require power to stay connected. other types require power to stay disconnected. the BMS i use seem to have mosfet that require power to stay connected

 

[Sverige]

Highly dependent on electrode thickness of given cell design.

For the typical thick electrode prismatic 'blue' cell used by DIY'er, should not be charging above 0.5 C(A) at any temperature.

For thick electrode cells, electrode lithium-ion local transfer rate starvation begins to happen on discharge above about 0.5 C(A). This drives up the overpotential required to move lithium ions (cell terminal voltage slump). Should avoid getting into ion starvation region as it drives up the voltage gradient across electrolyte increasing potential for damage.
View attachment 120880
View attachment 120881

You mentioned the electrode lithium ion local transfer rate starvation begins to occur above 0.5C for thicker electrode cells, is this only a problem for charging or is discharging above 0.5C also beginning to get into a region where this damage occurs?

 

[RCinFLA]

You mentioned the electrode lithium ion local transfer rate starvation begins to occur above 0.5C for thicker electrode cells, is this only a problem for charging or is discharging above 0.5C also beginning to get into a region where this damage occurs?

Anything that elevates the required overpotential voltage (also called polarization potential) to drive lithium-ion migration rate to support demanded cell current has some detrimental effects. Greater overpotential voltage accelerates negative parasitic chemical processes and increases the cell internal losses and heating.

Very large cell currents demands a lot of ion migration. The thickness of the electrodes (neg graphite and positive LFP) creates more obstacles to ion migration.

The 0.5C point is approximate and is typical for thicker electrode design cells. Cell heating and losses goes up significantly when starvation effects come into play.

For cooler temps, lithium-ion migration gets sluggish. Overpotential voltage rises to push the lithium-ion migration harder to satisfy the cell current demand.

For charging, the greater the volume of lithium ions bunched up in the graphite proximity trying to find a safe 'parking spot' in the negative graphite, the greater the chance they will mingle with electrons from graphite electrode increasing the chance they will chemically bond creating pure lithium. Any lithium metal created is permanently out of circulation for normal cell operation resulting in capacity loss. It also provides the lithium metal to begin growing metal dendrite shorts.

 

[Sverige]

Anything that elevates the required overpotential voltage (also called polarization potential) to drive lithium-ion migration rate to support demanded cell current has some detrimental effects. Greater overpotential voltage accelerates negative parasitic chemical processes and increases the cell internal losses and heating.

Very large cell currents demands a lot of ion migration. The thickness of the electrodes (neg graphite and positive LFP) creates more obstacles to ion migration.

The 0.5C point is approximate and is typical for thicker electrode design cells. Cell heating and losses goes up significantly when starvation effects come into play.

For cooler temps, lithium-ion migration gets sluggish. Overpotential voltage rises to push the lithium-ion migration harder to satisfy the cell current demand.

For charging, the greater the volume of lithium ions bunched up in the graphite proximity trying to find a safe 'parking spot' in the negative graphite, the greater the chance they will mingle with electrons from graphite electrode increasing the chance they will chemically bond creating pure lithium. Any lithium metal created is permanently out of circulation for normal cell operation resulting in capacity loss. It also provides the lithium metal to begin growing metal dendrite shorts.

Ok, so it applies equally to discharging as for charging? I asked because I’m pushing my lishen prismatics at around 0.5C discharge, sometimes a little over and their temp at this time of year is around 10C, so I might have two factors combining against me. Maybe I need to give them an easier life. On charging they don’t get more than 0.3C.

The lishens are rated for 1C discharge, so I was surprised to learn of this damage mechanism beginning from 0.5C and above, but if it’s applicable to discharge as well as charging then I will be careful.

 

[RCinFLA]

Charging is worse than discharging.

 

[On_The_Road]

So, what about this scenario.....

Rig is on shore power and no solar...and SOC is at 100 percent. Rig is drawing less than 1A at 13.4 volts and of course maintained by shore power. Bank is 560AH.

Temps dip to the mid 20s F. Is there any risk of cold weather damage to the batteries?

 

[sunshine_eggo]

So, what about this scenario.....

Rig is on shore power and no solar...and SOC is at 100 percent. Rig is drawing less than 1A at 13.4 volts and of course maintained by shore power. Bank is 560AH.

Temps dip to the mid 20s F. Is there any risk of cold weather damage to the batteries?

If you're referring to battery temp, not just ambient AND the BMS has low-temp charging protection, i.e., it doesn't allow charging below freezing, then no.

LFP can discharge down to -20°C.

If the battery has no charging protection, it may be okay due to the extremely low (near 0A) charge current.

 

[RCinFLA]

Although LFP's can be discharged at colder temperatures, their performance is greatly reduced. The overpotential voltage slump greatly increases at cold temp reducing cell voltage under discharge current. The greater the current demand, the worse the cell voltage slump.

It limits the maximum discharge current where battery voltage drops too much to set off inverter under voltage shutdown and significantly reduces the extractable capacity from cells.

At -20 degs F you will get only about 25-40% of the extractable capacity you get at +25 degs C, depending on how high the load current demand is.