LFP batteries – revised advice

Let’s start with a little background. Many electric vehicles on the road have batteries that use a NMC, or Nickel Manganese Cobalt, chemistry. NMC is a type of lithium-ion battery that uses a cathode made of nickel, manganese, and cobalt. NMC batteries are popular for their long lifespan and high energy density, which allows them to store a lot of energy in a small, lightweight package. And the best advice for extending the life of a NMC battery is:

Don’t let it get too hot or too cold
Don’t store your car’s battery at 100% for long periods of time
Don’t wait until the battery is low to recharge it unless absolutely necessary
Don’t regularly charge to 100%, better to charge often to 70 to 80% SOC (State of Charge)

That’s a lot of Don’ts. Then on the Do side of the ledger, there’s DO keep your NMC car plugged into your home charger whenever the car is not in use. We call that the ABCs (Always Be Charging).

So, when I got my 2023 Tesla Model 3 RWD with an LFP battery (lithium iron phosphate, aka LiFe or LiFePO4), I thought the ABC rule applied to it as well. It turns out it doesn’t, and I’ll explain why. But before going there, ’cause that part gets really complicated, I’m going to summarize how you should treat your LFP battery:

Do charge to 100% at least once per month so your vehicle accurately displays the charge percentage (Ford recommendation), or (at least) 100% once per week (Tesla recommendation)
When storing your car for extended periods of time store it at a lower state of charge (ex. Ford recommends 50%, probably the same for Tesla)
Do operate at lower state-of charge ranges when possible – charge up to 100% only on occasion, or as recommended by your car’s manufacturer
Only plug your car in when you need to*

What? That’s right! No ABCs for your LPF… But why?

*Small charge cycles are really important to extend the life of a NMC battery, but not for LFP. And here’s why… I told you this would get technical!

Sources:

Explanation:

Transcript from Your Battery Questions Answered with Dr. Jeff Dahn, starting at 17:56

“If you look at the voltage versus state of charge for the NMC 811 which is a very common vehicle battery (the black curve), you can see the voltage varies significantly and almost smoothly with the state of charge. So by measuring the voltage per cell of an EV pack you pretty much know where you are. For example, if you measure 3.75 volts you know that you’re about 62.5% state of charge it’s very simple, but for LFP, the red, imagine that you’re in this region between about 70 and 100% state of charge where the voltage curve is very flat. Then a voltage measurement doesn’t tell you where you are because the voltage is the same across this large region of state of charge.”

“So what the battery management system does is rely on charge counting, so when it’s discharging it’s counting the number of electrons that are being used to discharge the cell. When the cells are charging it’s counting the number of electrons and the accuracy of these charge counters is not perfect. And after a week or so without resetting all the way to 100%, they lose range accuracy. So you need to reset the charge counters by charged 100% from time to time. Now this is not a problem with LFP batteries at all. You don’t have to worry about it. Just to go ahead and charge 100% happily and it’s not going to matter. So the point of this is by operating the battery in your EV in a way to extend battery life you can keep more value in the battery.”

Quote from “The Operation Window of Lithium Iron Phosphate/Graphite Cells Affects their Lifetime“, page 11

“We propose a degradation pathway shown in Fig. 7 (not included) to explain why the LFP cells cycled across a higher average SOC fade faster than the lower SOC cells. At higher SOC, there is more exothermic heat flow produced from faster SEI growth and electrolyte reduction. This leads to VC (vinylene carbonate) being depleted at a faster rate in high SOC cycles. Additional factors that accelerate VC consumption are high temperature and higher surface area graphite (AG2 in this work). Once VC is consumed, then new SEI growth cannot come from VC reduction, so lithium alkoxides form from linear carbonate reduction and DMOHC (dimethyl-2,5-dioxahexane carboxylate) begins to form on the imperfectly passivated graphite. At the same time, iron dissolution from the positive electrode and deposition onto the negative starts to occur.

We propose that the Fe dissolution is caused by lithium alkoxides, and this chemical mechanism is verified in Fig. 8 (not included). After the Fe dissolves from LFP and transports in the electrolyte, Fe deposition onto the negative electrode is accelerated at high SOC (on lithiated graphite). New electrolyte reduction at those Fe sites on the graphite leads to more rapid capacity fade through lithium inventory loss. Therefore, cycling LFP cells over high SOC has two mechanistic reasons for accelerating failure. First it causes faster electrolyte additive consumption due to the increased reactivity of highly lithiated graphite which accelerates the production of lithium alkoxides. Second, these lithium alkoxides cause Fe dissolution and subsequent deposition on the graphite negative electrode which accelerates Li inventory loss.”

And page 12, “Based on these results, we would recommend that LFP cells for long lifetime applications operate at low states of charge on average, with charging up to 100% only on occasion. This raises several questions: how practical is it to cycle a battery cell in only low SOC
ranges? There is clearly a tradeoff between useful capacity and capacity retention. It is not realistic to recommend cycling LFP cells between 0%–25% SOC only, because that is a waste of capacity. However, we propose that LFP cells cycled between 0%–80% (or 0%–60%) would have a reasonable capacity and a longer lifetime than cells cycled between 0%–100%.”

Pages 12-13, “Overall, this work shows that there is a trade-off between cycling LFP/graphite cells at low average SOC for improved cycling stability at the expense of a lower capacity output in the limited range cycles. Understanding this trade-off is important for EV users and grid storage operators to optimize their usage of current LFP/graphite cells. For example, in a Vehicle to Grid (V2G) scenario, electric vehicle owners could choose to operate their LFP cells between 20 to 50% SOC when they are supplying power to the grid, to maximize the cells’ lifetime for their vehicle. If there is a network of vehicles connected to the grid, the lowered energy density from any given vehicle is less of a disadvantage. Therefore, this tradeoff between energy density and lifetime is use-case dependent. The focus of this work was to show the mechanisms behind this tradeoff, especially how lithium alkoxides cause Fe dissolution and how operation at high SOC increases Fe deposition and capacity fade.”

Pages 13-14 (Conclusions), “Cycling near the top of charge (75%–100% SOC) is detrimental to LFP/graphite cells. Our results show a correlation between the average SOC of battery operation and capacity fade rate, meaning that the lower the average SOC, the longer the lifetime, in these 2500 h of testing. The average SOC was found to be the most critical factor influencing capacity fade for LFP cells, over the factors of temperature, depth of discharge, electrolyte salt choice or graphite choice. Cells cycled in the conventional 0%–100% SOC window showed capacity fade rates intermediate to 0%–25% and 75%–100%. Therefore, the time spent cycling at high states of charge is critical to minimize. The degradation mode of LFP/graphite cells can be summarized as lithium inventory loss from electrolyte reduction on the negative electrode, also called shift loss.

Using isothermal microcalorimetry to measure parasitic heat flow from lithiated graphite pouch bags reacting with electrolyte, we show that elevated SOC causes incrementally higher reactivity. This degradation mode at high SOC most likely affects all lithium-ion cells that use a graphite negative electrode. Specific to LFP cells, iron dissolution and deposition is another degradation mode, accelerated by high temperature, imperfectly passivated negative electrodes, and time spent in high SOC cycling rather than in storage. The chemical mechanisms for LFP/graphite cell degradation, which are accelerated at high SOC involve: (1) faster electrolyte additive depletion, (2) lithium alkoxide generation from linear carbonate reduction once VC is consumed, (3) lithium alkoxide migration to the positive electrode to cause iron dissolution by ion exchange with lithium into the LFP, (4) deposition of this iron onto the reactive lithiated graphite electrode surface, and (5) electrolyte reduction on the deposited Fe sites causing additional lithium inventory loss.”

I told you this was complicated! But to conclude, how does this information impact me as a the owner of a car with a LFP battery? Well, this changes EVERYTHING… and nothing+.

No more ABCs for LFP.
Without plugging in, you’ll just need to monitor the SOC. I do have a charge limit of 50% (which is as low as Telsa will allow) set during the week and will plug in if/when a quick charge is needed.
I plan to charge my car to 100% once per week or so to make the BMS happy, or before starting out on a long road trip.
When road tripping, charge to whatever level Tesla’s navigation system (or ABRP) recommends to limit time spent at chargers. And you do NOT need to charge your LFP battery to 100% when visiting a supercharger.

Just to be clear, Tesla recommends that you keep your charge limit set to 100%, even for daily use, AND that you also fully charge to 100% at least once per week.

According to the battery engineers referenced above, this doesn’t make sense and WILL reduce the longevity (and value) of your car’s battery. Yes, charge to 100% ‘periodically’ to calibrate the BMS. That you should do. But keeping the charge limit at 100% for ‘daily use’ is crazy talk! Don’t do that!

+OK, so why doesn’t this change anything? If you’re coming from a ICE (internal combustion engine) vehicle you fill your tank (to 100%) when the fuel indicator gets to about a quarter tank (25%). That’s what we’re doing here. Treat your LFP car like an ICE car. When the tank says 25% stop and fill to 100%, and not before. Constantly ‘topping off’ your EV battery is just not what it wants or needs.

PS – During the winter months in PA, there will be a question about preconditioning, which I do recommend when the ambient temps get below about 50° F. In this case, I set my max charging level during the week to 50% and plug in when the car is not in service.

LFP batteries – revised advice

One Response to LFP batteries – revised advice