Why do lithium batteries die and how to improve them?

My Nissan Leaf Forum

Help Support My Nissan Leaf Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
RegGuheert said:
What are you referring to as "voltage fade" and what are you referring to as "electrolyte degradation"?

voltage fade is when the voltage goes down but the capacity in amps stays constant. Its not a behavior exhibited by conventional li ion (Co, NCA, NMC, Mn2O4). So an example of voltage fade would be a cell that maintains its capacity in amps, but lost Wh due to average voltage dropping from say 4.5 to 3.5 Its mostly only relevant to exotic Envia style cathodes.

Electrolyte degradation is what the Dahn lecture is focused on, and it is what Nissan LEAF owners have personal experience with, it causes severe amp capacity reduction, but the voltages are basically maintained.

3M owns the international patents for NMC, its no surprise that they fund work to expand the usefulness of NMC.

Getting economic 4.5V/5.0V class electrolytes is a very big deal. All else being equal, with the appropriate cathode, it can allow a simple 30% cost reduction per kWh. So GM's LG cells would drop from $145/kWh to $101/kWh (and increase capacity by 30%) if they could boost voltage sufficiently.

(Or another way to put it, LEAF IDS would increase from 60kWh to 85kWh, for the same weight/cost)
 
ydnas7 said:
RegGuheert said:
What are you referring to as "voltage fade" and what are you referring to as "electrolyte degradation"?
voltage fade is when the voltage goes down but the capacity in amps stays constant.
Thanks!

So I'll guess that the collapse in voltage is due to loss of structure at the cathode. What I've never quite grasped is why one cathode structure and chemistry gives a different voltage than another. I've found this post on researchgate regarding this question, but it appears to me that asking a bunch of chemists that question is akin to asking a bunch of electrical engineers to describe noise: everyone has a (seemingly) different idea of how things work.
 
Here is an extensive new study on calendar aging. Aging was studied as a function of both SOC AND temperature. It answers a few questions we have had around here:

- Nissan was correct when they decided that 80% mode did not significantly reduce capacity fade.
- OTOH, storage at 100% tends to increase battery resistances significantly more than storage at 80%.
- OTOOH, 80% mode may be a good idea for the 30-kWh (and larger) LEAF batteries.
- Storage below 50% SOC results in lower capacity fade than storage above 70% SOC.

A surprise:

- Storage at 0% SOC is best! (Anyone up for arriving home at turtle every evening?)

Here is a plot showing calendar effects for NCA (like the 24-kWh LEAF), NMC (like the 30-kWh LEAF) and LFP:

F2.large.jpg
 
RegGuheert said:
...Here is a plot showing calendar effects for NCA (like the 24-kWh LEAF), NMC (like the 30-kWh LEAF) and LFP...
24 kWh Leafs use LiMn2O4/LiNiO2 (LMO/LNO mix) cells, not NCA. Tesla uses NCA in Model S. Voltage/SOC ratio is very different between those.

I guess I was not wasting time trying to keep my Leaf at around 30% SOC when parked overnight. I set charge timer to start charging 2 - 4 hours before departure next day, so that I would start driving as soon as battery charges to minimize time spent at high SOC. When I arrive home from work at 20% SOC I would leave it plugged in but on a timer to start charging next day.
 
RegGuheert said:
Nissan was correct when they decided that 80% mode did not significantly reduce capacity fade.
I think that's hard to say absolutely without testing the actual cells used in the LEAF, but it sure does seem like there's a couple step changes in capacity loss due to SOC - somewhere around 55-70% SOC and 90%+ SOC. It would be interesting to see what voltage corresponds to SOC in these tests.

RegGuheert said:
OTOH, storage at 100% tends to increase battery resistances significantly more than storage at 80%.
Which is never good as any significant discharge will lose efficiency.

RegGuheert said:
OTOOH, 80% mode may be a good idea for the 30-kWh (and larger) LEAF batteries.
Don't follow here?

RegGuheert said:
Storage at 0% SOC is best! (Anyone up for arriving home at turtle every evening?)
I'm not surprised, I've been suggesting that lower SOC for storage is better for a while now based on another calendar life study that's floating around.

As a general rule, lower SOC and lower temperatures extends battery life. What's interesting is to see how big (or small) the differences are.
 
drees said:
RegGuheert said:
OTOOH, 80% mode may be a good idea for the 30-kWh (and larger) LEAF batteries.
Don't follow here?
I'm pretty sure the newer LEAF batteries have NMC cathodes. According the study, the calendar degradation drops off a cliff between 90% and 100% SOC, at least at 40C and above. Also, 80% mode is more useful in more situations when you have a larger battery.
 
Seems to me that you want as low a SOC as possible, especially if it lets you spend significant time below 50% where rate of capacity loss slows significantly for all chemistries. Doesn't matter how big the battery is, just how much of the battery you usually use.

And for sure, do not worry about letting the car sit at low SOC, unless you're talking about turtle and then sitting for months.

Best practice would be to charge just enough to make it to your destination, finish charging just as you're ready to leave and charge at a quick rate - the quickest rate possible that does not unduly heat up the pack.
 
drees said:
Best practice would be to charge just enough to make it to your destination, finish charging just as you're ready to leave and charge at a quick rate - the quickest rate possible that does not unduly heat up the pack.
No argument. In fact, this is precisely how we have used our LEAF for the past three years, except that we do not target getting home at turtle, but with fewer than three bars. But as the capacity drops, we charge to full more frequently than we did in the past.

Looking at your signature, I see the following:
9th 8/12/16 43.3Ah 186/152 GID
10th: 10/23/15 @ 47.4Ah 203/167 GID
11th: 11/11/14 51.6Ah 220/181 GID
12th: 9/23/13 55.3Ah 228/194 GID
Doing a little math, I calculate that the degradation of the battery in your LEAF is accelerating as it ages:

New->12th: 806 days to lose 11 Ah or about 73 days/Ah or 5 Ah/year (66.3 Ah assumed for initial capacity)
12th-11th: 414 days to lose 3.7 Ah or an about 112 days/Ah or 3.3 Ah/year
11th-10th: 346 days to lose 4.2 Ah or about 82 days/Ah or 4.5 Ah/year
10th-9th: 294 days to lose 4.1 Ah or about 71 days/Ah or 5.1 Ah/year

It appears that your rate of capacity loss initially dropped to a lower value around the loss of the first bar, but now it has accelerated to a rate faster than the average for the first bar loss.

Looking at our LEAF, I see the following:

11th: 6/11/16 51.2Ah
12th: 4/6/15 54.0Ah

New->12th: 1292 days to lose 12.3 Ah or about 105 days/Ah or 3.5 Ah/year (66.3 Ah assumed for initial capacity)
12th-11th: 432 days to lose 2.8 Ah or an about 154 days/Ah or 2.4 Ah/year

Like your LEAF, on average it took about 50% longer to lose an Ah between bars 12 and 11 than before bar 12. It will be interesting to see if we experience the same acceleration in degradation rate that you have seen.

Currently our LEAF is a bit below 50 Ah, but I don't expect to lose much capacity over the wintertime. When the 10th bar drops, we'll see if there is any obvious acceleration in degradation.
 
About a year ago in the Model 3 thread I commented on a new paper coming from Dr. Jeff Dahn:
RegGuheert said:
RegGuheert said:
Dahn’s research focuses on increasing the energy density and lifetime of Li-ion batteries in order to drive down costs of Tesla’s automotive and grid energy storage products.

He works mostly with NMC Li-ion cells, Tesla’s preferred chemistry for battery cells, and his keynote address titled “Surprising Chemistry in Li-Ion Cells” will discuss how they could stop harmful reactions in those cells in order to increase their capacity:
It is important to increase the operating voltage of NMC Li-ion cells to obtain higher energy density. However, the electrolyte reacts with the positive electrode at high voltage. Using simple experiments involving only pouch bags, we show that the products of these reactions are extremely harmful to the positive electrode. This talk demonstrates how these harmful reactions at the positive electrode can be virtually stopped, leading to superb NMC Li-ion cells that can operate at high potential.
Dahn’s presentation will follow Kelty’s on March 22 at the International Battery Seminar & Exhibit in Fort Lauderdale.
I finally found that presentation on YouTube:

[youtube]http://www.youtube.com/watch?v=5WpQh4kZ_MU[/youtube]

My takeaways from watching the video:

1) Dr. Dahn is focused on increasing the capacity of NMC by finding ways to operate that chemistry at higher voltages without rapid capacity fade. He leads off by showing the following graph from Linden's Handbook of Batteries, 4th Edition:

Jeff_Dahn_Plot_From_Mc_Graw_Hill_Handbookof_Batteries4th_Edition_Figure261.png


2) Dr. Dahn demonstrates that there is an unknown reaction between the electrolyte and the NMC electrode and that the effects of this reaction is mitigated by the products of this reaction being absorbed by the negative electrode.

3) He goes on to demonstrate that he had developed a new formulation of NMC that does NOT exhibit this problem. He has built batteries using this new NMC that cycle well up to 4.4V and very well up to 4.2 V.

4) Dr. Dahn is working with NMC622. I saw no mention of NMC811 in this presentation.

Some thoughts based on what I have seen:

- NMC can exhibit very poor cycling characteristics if this parasitic reaction at the positive electrode is not addressed.

- I will go out on a limb and guess that this may be the issue which is plaguing the 30-kWh batteries in the Nissan LEAF. If not addressed, it might also be a real problem for the new 40-kWh versions.

- Tesla now has access to NMC technology which will allow them to dial in either very long life or very high capacity (and decent life) by adjusting the maximum voltage they use for charge termination. In fact, they could do this within a single product. For instance, their cars could cycle to a low voltage like 4.1 V normally and then have an option to go to 4.4 V for very long range trips. (They already do this, but I think with the new NMC, this would mean quite a bit more capacity in both cases.)

- There is more capacity to be accessed in NMC if/when Dr. Dahn or others are able to tackle other damaging reactions that occur as the termination voltage moves up to 4.6 V and beyond.

It seems clear that many of the problems which have plagued Li-ion batteries are systematically being addressed as researchers focus on solving them. As a result, these batteries continue to get closer and closer to the ideal.
 
Back
Top