LFP v’s NCA… and the winner is…

[tgrobbo]

I’ve now had an LFP RWD MY since 2022 and now I’ve also got an NCA MY LR. Although I’m not a battery guru by any means, I’ve been tracking a lot of high mileage cars with both chemistries and have read the study on LFP BYD taxis and pretty much any other info that’s out there on them.
I’ve come to conclusion that NCA is still the vastly superior battery, for an electric car that does an average mileage - maybe 15,000km or 10,000 miles a year. I decided to post this because I see a lot of folks going on about how amazing LFP is, but perhaps they’ve not owned an LFP car, or maybe they have assumed that superior resistance to ‘Cyclic Aging’ means that the battery is better in every way.
I don’t think it is, here is why:
LFP Pro: - superior cyclic aging - great if you have a taxi or are doing a heap of charge cycles in a short time, like 60,000 miles a year or more.
LFP Pro: less likely to catch on fire apparently.
LFP Pro: A bit cheaper.
But that’s it really for Pro’s, the cons (compared to NCA) far outweigh those pro’s in my opinion for the average user.
Lower discharge and charge speeds - this sucks in multiple ways, I swear my LFP is hobbled below 20mph to save the battery - it has no guts at all and is slower than a Yaris across an intersection. It also doesn’t like regen very much because of the charge speed - so you lose regen down any decent sized hill. Plus, obviously it’s slower to supercharge on a road trip, but this is a minor thing for me.
Weight - it’s heavy! My RWD LFP car is about the same weight as my dual motor NCA car. Boo! The NCA car has a much bigger battery and a second motor! (LFP doesn’t like discharging rapidly so dual motors are out of the question)
Cold weather - LFP more affected.
Degradation - oooh touchy subject this - but from what I can see, and I’ve looked at a lot of data, LFP does something called calendar aging - loses capacity over time - it doesn’t matter too much what you do with it, (unless you only ever charge it to 50%) it will lose capacity at a steady rate. At this point, perhaps due to the 100% charging requirement, LFP batteries seem to be calendar aging more than NCA batteries - I make this statement based on fairly limited data of course. It does remain to be seen if LFP cars will fare worse than NCA cars over the warranty period, but from what I can tell thus far, I feel like my NCA battery will have more of its original capacity at the end of my warranty period than my LFP battery with the mileage I’m currently doing. Just a feeling, and based on limited data, but I’ve certainly looked into it more than most.
Charging to 100% kinda sucks too - the last 1% (calibrating) takes forever and obviously it slows down for the last 5% or so. I just don’t like my car sitting at 100% for ages, but with LFP it ends up doing that quite a bit. With NCA I like the higher charge speed right up to 80%, then I don’t mind so much if my car sits overnight at that %. (I don’t/can’t charge at home)
Anyway, that is all, as I said, I have both batteries now, so I have a foot in both camps, I don’t want to be hating on LFP, just thought I’d enlighten a few people on it, as it’s a bit of unknown and misunderstood chemistry due to not being prevalent in the US.

 

[fhteagle]

Thanks for sharing your observations.

Calendar aging affects all lithium based battery chemistries, not just LFP. The amount varies mostly by chemistry type, some by secret sauce additives, storage SOC, and temperature.

 

[tgrobbo]

Thanks for sharing your observations.

Calendar aging affects all lithium based battery chemistries, not just LFP. The amount varies mostly by chemistry type, some by secret sauce additives, storage SOC, and temperature.

Yep I just wonder if LFP - especially with the 100% charging - is calendar aging a little more than NCA. I mean it almost certainly is right, if NCA is aging from both charge cycles and time, but LFP seemingly only from time, but the loss of capacity is much the same between the two over the first three years, then LFP is obviously calendar aging more. Will this continue at the same rate, maybe not. But it doesn’t seem to be slowing down as much as you’d hope from the data I’ve seen.

 

[LostVector]

Yep I just wonder if LFP - especially with the 100% charging - is calendar aging a little more than NCA. I mean it almost certainly is right, if NCA is aging from both charge cycles and time, but LFP seemingly only from time, but the loss of capacity is much the same between the two over the first three years, then LFP is obviously calendar aging more. Will this continue at the same rate, maybe not. But it doesn’t seem to be slowing down as much as you’d hope from the data I’ve seen.

LFP batteries are widely understood to experience less calendar aging than NCA batteries ... something around 40-50% of the degradation over time. You can google through various studies to confirm this.

 

[tgrobbo]

LFP batteries are widely understood to experience less calendar aging than NCA batteries ... something around 40-50% of the degradation over time. You can google through various studies to confirm this.

I’ve read a few studies and not found this. They’re either the same or LFP is a bit worse. Are you sure you’re not getting confused with cyclic aging? (Seems like everyone does this when extolling the virtues of LFP lol)

 

[tgrobbo]

In this graph LFP is degrading 1% or more than NCA at higher states of charge. During the warranty period this might mean 5% more capacity loss for an LFP car with the same number of charge cycles as an NCA car.(not for high mileage cars perhaps) This calendar aging might be made worse by consistently charging to 100% because that will lift the overall average state of charge over a long period for the LFP car. Time will tell I guess but I don’t think it’s definitive that LFP is a clear winner. Not all all.

 

[LostVector]

I’ve read a few studies and not found this. They’re either the same or LFP is a bit worse. Are you sure you’re not getting confused with cyclic aging? (Seems like everyone does this when extolling the virtues of LFP lol)

In this graph LFP is degrading 1% or more than NCA at higher states of charge. During the warranty period this might mean 5% more capacity loss for an LFP car with the same number of charge cycles as an NCA car.(not for high mileage cars perhaps) This calendar aging might be made worse by consistently charging to 100% because that will lift the overall average state of charge over a long period for the LFP car. Time will tell I guess but I don’t think it’s definitive that LFP is a clear winner. Not all all.
View attachment 1046141

Ah I see the oft quoted graph you are using. Well that specific study does appear to contradict what I said, but a meta study of multiple chemistries and studies does conclude that LFP is generally more stable. There are differences at various operating ranges that may result in different graphs, etc. Just a starting point.

Assessment of the calendar aging of lithium-ion batteries for a long-term—Space missions

Energy availability is a critical challenge for space missions, especially for those missions designed to last many decades. Space satellites have depended on various combinations of radioisotope thermoelectric generators (RGTs), solar arrays, and batteries for power. For deep space missions...

www.frontiersin.org

 

[tgrobbo]

Ah I see the oft quoted graph you are using. Well that specific study does appear to contradict what I said, but a meta study of multiple chemistries and studies does conclude that LFP is generally more stable. There are differences at various operating ranges that may result in different graphs, etc. Just a starting point.

Assessment of the calendar aging of lithium-ion batteries for a long-term—Space missions

Energy availability is a critical challenge for space missions, especially for those missions designed to last many decades. Space satellites have depended on various combinations of radioisotope thermoelectric generators (RGTs), solar arrays, and batteries for power. For deep space missions...

www.frontiersin.org

Yeah I think that study is ‘oft quoted’ because it’s nice and easy to understand. I have skimmed the study you’ve listed there and the conclusion was reached for batteries at 10% SOC, so probably not mega relevant for EV’s. Studies designed for space travel will probably never be that relevant for EV’s I reckon. In the real world I’ve been tracking a fair few high mileage LFP cars and they do seem to losing more capacity than the NCA cars - like 12% in 3 years - the NCA cars with the same mileage are older and only around 8%. There is often a large discrepancy between a battery test and the displayed range though. If you disregard the battery test results, they’re pretty similar in terms of displayed range loss.

 

[evrepair]

We have disassembled both types of batteries and the LFP pack is just a better newer simpler design. All cell connections are on the top, so internal cooling leak won’t damage the cells. There are much fewer cells and connections, so fewer things to break. Cooling is on the bottom of the cells there is more cushion in case the car runs over something in the road. If I had to pick a car to drive for 10 years I would pick one with an LFP battery for these reasons.

 

[tgrobbo]

We have disassembled both types of batteries and the LFP pack is just a better newer simpler design. All cell connections are on the top, so internal cooling leak won’t damage the cells. There are much fewer cells and connections, so fewer things to break. Cooling is on the bottom of the cells there is more cushion in case the car runs over something in the road. If I had to pick a car to drive for 10 years I would pick one with an LFP battery for these reasons.

Good intel. I note Shanghai built cars apparently have NMC batteries for the dual motor versions. I wonder how they compare in design terms…. Cheers

 

[AAKEE]

LFP’s suffer from calendar aging about in the same way as NCA and NMC does.
No big differences, just like that picture above with all three types.

The calendar aging might differ slightly from test to test (different test approach and different cell brands or batches).
The big picture is = its about the same.

Several tests shows that the calendar sging is less above 80% compared to 80%.
This for both LFP and NCA (in dome cases, NMC as well).

We ”should/could” think that the calendar aging is relatively flat above 75% for all types, except in extreme temperatures.

A common misunderstanding in forums/facebook threads is that LFP’s can be charge to 100% while NCA can not.

What differs is that LFP’s can do several thousand large or full 100-0% cycles without very much degradation.
NCA will get a much shorter cycle life if they do very large cycles like 100-0 or so compared to smaller cycles (they still hold up OK, but the added calendar aging plus higher cyclic aging might not be that good long term.
LFP’s matches the need for a car with a shorter range that need to use larger cycles on a daily basis.

For LFP’s, the degradation we see is For the absolute most part calendar aging, like this:

 

[jsugarsd]

I was wondering about the calibration part with LFP. I just had the upgrade recently as my 21 M3 SR+ NCA failed at ~24k miles (wouldn't charge but luckily I was near a Tesla service center). Anyhow, covered by warranty, they replaced it with the heavier LFP and installed a stronger new suspension while they were at it. I signed the waiver and I'm about a week in with it. So, when I am charging it gets to 99% and says calibrating. I waited for over 10 minutes but had to get going. How long would it take if I just sat there? What are the repercussions if any by stopping it before it's done? I'm guessing the BMS will not be as accurate? Thanks in advance

 

[tgrobbo]

LFP’s suffer from calendar aging about in the same way as NCA and NMC does.
No big differences, just like that picture above with all three types.

The calendar aging might differ slightly from test to test (different test approach and different cell brands or batches).
The big picture is = its about the same.

Several tests shows that the calendar sging is less above 80% compared to 80%.
This for both LFP and NCA (in dome cases, NMC as well).

We ”should/could” think that the calendar aging is relatively flat above 75% for all types, except in extreme temperatures.

A common misunderstanding in forums/facebook threads is that LFP’s can be charge to 100% while NCA can not.

What differs is that LFP’s can do several thousand large or full 100-0% cycles without very much degradation.
NCA will get a much shorter cycle life if they do very large cycles like 100-0 or so compared to smaller cycles (they still hold up OK, but the added calendar aging plus higher cyclic aging might not be that good long term.
LFP’s matches the need for a car with a shorter range that need to use larger cycles on a daily basis.

For LFP’s, the degradation we see is For the absolute most part calendar aging, like this:
View attachment 1046431
View attachment 1046433

Nice, so in your view calendar aging is about the across the different types, I don’t disagree. I was just highlighting that there are a few downsides to LFP too. I would also guess that calendar aging makes up a much bigger portion of their overall capacity loss than other chemistries. Cheers

 

[tgrobbo]

I was wondering about the calibration part with LFP. I just had the upgrade recently as my 21 M3 SR+ NCA failed at ~24k miles (wouldn't charge but luckily I was near a Tesla service center). Anyhow, covered by warranty, they replaced it with the heavier LFP and installed a stronger new suspension while they were at it. I signed the waiver and I'm about a week in with it. So, when I am charging it gets to 99% and says calibrating. I waited for over 10 minutes but had to get going. How long would it take if I just sat there? What are the repercussions if any by stopping it before it's done? I'm guessing the BMS will not be as accurate? Thanks in advance

I don’t think anyone knows how much accuracy you lose by not finishing the calibration, but from experience it can take up to 30 minutes as I said in my original post. At times I haven’t charged to 100% for months, I didn’t notice any loss of accuracy but I wasn’t doing heaps of road trips so how can you tell?
People have noticed their displayed range dropping a bit if they haven’t let it finish calibrating for a while, especially when they haven’t let it run down to a low SOC either. Once they do these things, the displayed range tends to go back up again. Cheers

 

[Nack]

I don’t think anyone knows how much accuracy you lose by not finishing the calibration, but from experience it can take up to 30 minutes as I said in my original post. At times I haven’t charged to 100% for months, I didn’t notice any loss of accuracy but I wasn’t doing heaps of road trips so how can you tell?
People have noticed their displayed range dropping a bit if they haven’t let it finish calibrating for a while, especially when they haven’t let it run down to a low SOC either. Once they do these things, the displayed range tends to go back up again. Cheers

Requiring to charge to 100% for calibrating sounds like a crappy BMS more than anything. Generally the curve for LFP starts looking good at over 80-85% (where voltage changes enough with SOC increase to know where you are on the curve).

For cell balancing it depends on a lot of factors. But going to 100% is not required generally as the system can support some delta between the least voltage to the most voltage cell and you don’t need things exact (or as close to as possible). Really you just want to keep the delta from getting really big as it will take much longer to bring it into balance.

Honestly I wouldn’t sweat getting it to 100% as there are diminishing returns and the difference in balance is not huge in benefit vs time.

I am not familiar with the Tesla BMS though so maybe they do things differently. This is just generic info for LFP automotive systems.

I’m looking forward to more vehicles having LFP. I don’t really want an NMC/NCA battery.

 

[AAKEE]

I don’t think anyone knows how much accuracy you lose by not finishing the calibration, but from experience it can take up to 30 minutes as I said in my original post.

Even if Tesla calls it ”Calibrating” it is balancing the cells in the end of the charge.
The cells with less capacity and/or highest voltage reach 100% first (a set voltage held when charging to 100%.).

The cells with less voltage continue to charge to reach about the same voltage during the time the ”calibration” occurs.

There is really not much to calibrate at the top end, so the term is slightly missleading, I would say.

 

[Nack]

Usually “calibrating” is confirming the starting point for the Coulomb-counting method of determining SOC. If you don’t get into the extreme low or high <20%/>80% the voltage is essentially the same so you don’t really know if you are 50% or 65% SOC as an example as your starting assumption may be out of whack.

Not sure what SOC method Tesla is using on these. Very well could be something more advanced.

Sorry if you/others already know this but many do not so its helpful to read and learn either way.

Cell balancing is different and is bleeding off the high voltage cells (relatively) to align with the lowest voltage cell. Charge for a bit, then rinse and repeat. This is a lot of the reason charging >80% takes much longer as they have to make sure not to overshoot, rebalance, charge, rebalance…

 

[Darmie]

Cell balancing is different and is bleeding off the high voltage cells (relatively) to align with the lowest voltage cell. Charge for a bit, then rinse and repeat. This is a lot of the reason charging >80% takes much longer as they have to make sure not to overshoot, rebalance, charge, rebalance…

I would think a Tesla battery charges using constant current / constant voltage. If so, when charging switches to constant voltage (CV Charging) is when charging speed really slows down.

 

[Skavatar]

degradation and range anxiety used to be my biggest reasons for not going EV. but since Tesla has built so many SC along major highways range anxiety doesn't bother me anymore. and after watching the Out of Spec cross country 4 EV truck race I've learned the lower you drain the battery during a road trip the larger/faster the charging curve, and you only charge enough to get to the next SC instead of trying to charge all the way up to 80% or 90% at each stop.

my wife's 2023 lfp M3 RWD uses 20% round trip for 40 miles to and from work, mainly city roads. my non-lfp 2023 MYLR also uses 20% round trip for 40 miles to and from work, half city half highway. even if we lose 50% battery capacity its enough for everyday driving. we use my MYLR for fishing trips about 90 miles away. There are several SC locations along the way so making one or two SC stops (round trip 180 miles) is easily doable even with 50% battery degradation.

 

[Nack]

I would think a Tesla battery charges using constant current / constant voltage. If so, when charging switches to constant voltage (CV Charging) is when charging speed really slows down.
View attachment 1046853

Pretty much any battery in an EV has CC and CV charging. The reason why you switch to CV is so you don’t overcharge and because its balancing so it can’t handle all the juice as its bleeding off the high cells so there is no need for a huge inrush of power.

When the battery is empty depending on cooling capability and battery temp you can apply multiple C. As the battery tops up the C rate drops as the cells can’t handle the power at the same rate.

I usually use an analogy of filling a large barrel to the brim where you can’t overflow. When its empty you can give max flow because there is no chance of overfilling. Once you get about half full you start paying attention and near the top you start to taper back. Its oversimplified, but similar principle.