Li-ion cells cycle ageing

[mighty82]

Cool! Will you sell me your testing equipment for cheap then? Since you won't be needing it anymore I mean :wink:

Sad to loose you as a resource though. Your work is really interesting. I'm sure a lot of people would chip in a few $ for your work if you had some sort of donation thing going.

Wrong, it´s not easy, however it´s possible.

Can't see how though, without access to everyone's equipment and information about lots of other factors probably long lost by now.

 

[Dak77]

I'm sure he's going to continue testing cells in the future anyway.

Very probably not, cells testing became boring. Vegetable growing is much more interesting and enjoyable.
1.jpg
3.jpg

Comparing cells tested at different times, by different people with different equipment is impossible.

Wrong, it´s not easy, however it´s possible.

Any idea why the cells have lost more capacity percentage wise inside the cycling interval than total capacity degradation?

No.

Have you tried retesting the cells with the previous parameters after it has been fully cycled to see if some of the capacity within the cycling interval has been regained?

No.

I'm incredibly jealous . My life goal is to buy a small farm.

 

[Dak77]

If there's one thing I learned in the barrage of data I've seen over the past year or so here, it's what NOT to buy. That alone has saved me a lot of time and money. Huge thank you ! to everyone involved .

 

[docware]

If there's one thing I learned in the barrage of data I've seen over the past year or so here, it's what NOT to buy. That alone has saved me a lot of time and money.

Well, just be aware that cycle aging is only part of the whole picture. It doesn´t include calendar aging.
Interesting fact : capacity of SONY VTC6 after 700 cycles is the same as initial capacity of LG HG2 before cycling.
LG HG2 initial capacity 2792 mAh 4,2 – 2,5 V / 1 A
SONY capacity after 700 cycles 2 797 mAh 4,2 – 2,5 V / 1 A

 

[mighty82]

Vtc6 seems like one of the very best cell to consider for high amp applications. You really can get both high power and high energy these days. What's the drawback?

From what I can see from my own experience and what I can read online, calender degradation only seem to be a problem if stored fully charged and/or at high temperatures.

I have a few unused cells of both nmc and nca chemistry from 2010-2013 and they still test like new. These are stored at roughy 3.5v.

I even found a pair of old unused lipo packs a while back, 7-8 years old. Stored at 30% charge at room temperature they didn't have any capacity loss worth mentioning. The IR was also good.

Might be a bigger problem with some chemistries though.

 

[john61ct]

Yes, a scale of decades is normal for all mainstream LI chemistries if stored properly.

From experience quality LFP loses nothing significant at all over ten years, actual capacity still well over rated, and even SoC% / voltage not measurably self-discharged, at all from how they were delivered from the factory.

That is not to say calendar life won't be significant, once you get past say 30 years in the event cycling patterns also are insignificant, e.g. sub C rates, cycling only a few times per week, average DoD under 30% and centered around the midpoint

then indeed temperature becomes the biggest deciding factor, in Norway might last 60-80 years.

 

[docware]

Let me remind LG HG2 No2 :
LG HG2 No2, production date November 2017. Purchased fom nkon in September 2018. Cell stored at aprox 25 – 40 % SOC, 15° - 20°C.

In December 2018 capacity 2 909 mAh at 3 A discharge, measured at ZKETECH EBC-A10H.
In September 2019 : 2836 mAh at 1A, 2804 mAh at 3 A, measured with Maynuo M9712/ GW Instek GDM 8351.
In February 2020 2748 mAh at 1 A (Maynuo M9712/ GW Instek GDM 8351).

IMO we miss real reliable precise data on calendar aging for the particular cells. Otherwise it is only discussion without any real basis.

https://endless-sphere.com/forums/viewtopic.php?f=14&t=103092&p=1528453#p1528453

 

[Dak77]

Let me remind LG HG2 No2 :
LG HG2 No2, production date November 2017. Purchased fom nkon in September 2018. Cell stored at aprox 25 – 40 % SOC, 15° - 20°C.

In December 2018 capacity 2 909 mAh at 3 A discharge, measured at ZKETECH EBC-A10H.
In September 2019 : 2836 mAh at 1A, 2804 mAh at 3 A, measured with Maynuo M9712/ GW Instek GDM 8351.
In February 2020 2748 mAh at 1 A (Maynuo M9712/ GW Instek GDM 8351).

IMO we miss real reliable precise data on calendar aging for the particular cells. Otherwise it is only discussion without any real basis.

https://endless-sphere.com/forums/viewtopic.php?f=14&t=103092&p=1528453#p1528453

I've seen quite a few vendors selling 2 year old cells. I learned to read date codes of Samsung and LG and ask before I order anything. Obviously , you know this, but some assume they're getting fresh cells , and often they are not.

 

[mighty82]

Let me remind LG HG2 No2 :
LG HG2 No2, production date November 2017. Purchased fom nkon in September 2018. Cell stored at aprox 25 – 40 % SOC, 15° - 20°C.

In December 2018 capacity 2 909 mAh at 3 A discharge, measured at ZKETECH EBC-A10H.
In September 2019 : 2836 mAh at 1A, 2804 mAh at 3 A, measured with Maynuo M9712/ GW Instek GDM 8351.
In February 2020 2748 mAh at 1 A (Maynuo M9712/ GW Instek GDM 8351).

IMO we miss real reliable precise data on calendar aging for the particular cells. Otherwise it is only discussion without any real basis.

https://endless-sphere.com/forums/viewtopic.php?f=14&t=103092&p=1528453#p1528453

That's a fairly rapid decline. In many applications, calender ageing will render it useless before cycle wear.

That might be why tesla is sticking strictly with NCA chemistry for electric vehicles? NCA has always been touted to have superior calendar life.

According to this study the difference is minimal though, exept for at high temperature where NCA excels:
https://iopscience.iop.org/article/10.1149/2.0411609jes/pdf

 

[john61ct]

Would it be possible to please confirm that

Sony VTC6
Samsung 50E
LG MJ1 & M36

are all NCA, correct?

 

[mighty82]

Would it be possible to please confirm that

Sony VTC6
Samsung 50E
LG MJ1 & M36

are all NCA, correct?

VTC6 is definitely NMC, I don't think the LG cells are NCA either.

 

[ridethelightning]

subscribed

 

[Pajda]

That's a fairly rapid decline. In many applications, calender ageing will render it useless before cycle wear.

That might be why tesla is sticking strictly with NCA chemistry for electric vehicles? NCA has always been touted to have superior calendar life.

According to this study the difference is minimal though, exept for at high temperature where NCA excels:
https://iopscience.iop.org/article/10.1149/2.0411609jes/pdf

This test show comparison between three particular cells, so the only one conlucison is valid for direct comparison of this three cells. I'll just repeat myself that it is unfortunately uselles for generalizations.

When we talking about calendar ageing it should be also considered an impact of cell internal design (HE vs HP). Docware tests show us that samples of popular HP cells (HG2, VTC6, 30Q) shows significantly worse cycle life than some HE cells (MJ1, M36,50E). Based on this observation it will be nice to run calendar ageing test to see if there is some similarity in calendar ageing or not.

 

[docware]

This test show comparison between three particular cells, so the only one conlucison is valid for direct comparison of this three cells.

No, it is not comparison of three different cells, it is measurement on the same cell (LG HG2 No2).
Frankly, I am surprised by such enormous capacity decline, moreover, the cell was stored at proper SOC and temperature. Maybe it was cell from lower quality batch, who knows.

 

[mighty82]

This test show comparison between three particular cells, so the only one conlucison is valid for direct comparison of this three cells. I'll just repeat myself that it is unfortunately uselles for generalizations.

You are right about the comparison of course. Although I recognise the trend of nmc cells having less rise in IR etc. They have some basic differences in chemical buildup.

When we talking about calendar ageing it should be also considered an impact of cell internal design (HE vs HP). Docware tests show us that samples of popular HP cells (HG2, VTC6, 30Q) shows significantly worse cycle life than some HE cells (MJ1, M36,50E). Based on this observation it will be nice to run calendar ageing test to see if there is some similarity in calendar ageing or not.

But then again, some HE cells show even worse cycle life than mentioned HP cells, like panasonic PF and sanyo/panasonic GA.

 

[mighty82]

No, it is not comparison of three different cells, it is measurement on the same cell (LG HG2 No2).
Frankly, I am surprised by such enormous capacity decline, moreover, the cell was stored at proper SOC and temperature. Maybe it was cell from lower quality batch, who knows.

I think he referred to the 3 cells in the study I linked to.

In the study all 3 cell types lost about 3% in 10 months at room temperature under 60% soc. So it's not too far off their results.

 

[serious_sam]

Docware tests show us that samples of popular HP cells (HG2, VTC6, 30Q) shows significantly worse cycle life than some HE cells (MJ1, M36,50E)

...at a relatively low rate of discharge.

What it doesn't tell us is how sensitive to discharge rates are the cells.

I suspect that the HP cells cycle life doesn't drop dramatically with a reasonable increase in rate of discharge. Whereas the HE cells cycle life will drop much more rapidly with increased rates of discharge.

And at a certain rate of discharge, the HP cells will have better cycle life that HE cells.

 

[Pajda]

No, it is not comparison of three different cells, it is measurement on the same cell (LG HG2 No2).
Frankly, I am surprised by such enormous capacity decline, moreover, the cell was stored at proper SOC and temperature. Maybe it was cell from lower quality batch, who knows.

I think he referred to the 3 cells in the study I linked to.

Exactly I referred to the linked study and then start talking about another idea, sorry for the confusion..

 

[Pajda]

But then again, some HE cells show even worse cycle life than mentioned HP cells, like panasonic PF and sanyo/panasonic GA.

Yes, exactly it only confirm my approach "carefully with a generalization". By the way from my personal point of view Panasonic PF or PD is not representative of HE category but something which i personally called "BP - Balanced Parameters". Others representatives are Samsung 29E and LG M29.

HE category representatives are Sanyo GA and Samsung 50E, both NCA chemistry cells and both shows totally different cycle life. So If you evaluate NCA chemistry based on parameters of only one representative you can make many confusion.

 

[Cephalotus]

VTC6 is definitely NMC, I don't think the LG cells are NCA either.

There is way to much wrong information around, so it is very difficult to know.

Actually VTC-6 is a NCA cell, while HG2 is (seems to be) a 811 NMC cell, both in combination with a C+Si anode, as spectroscopic analysis shows:

"...The NCA in the Sony VTC6 cell appeared to have a surface coating of aluminium oxide. This made it difficult to estimate the actual particle composition, but the Ni:Co ratio appeared to be higher than in the VTC5A cell. Surface coating of cathode particles has been proposed as a method to increase the charging voltage limit, and hence the capacity of the material..."

https://www.mdpi.com/2313-0105/5/4/64/htm

 

[Pajda]

What it doesn't tell us is how sensitive to discharge rates are the cells.

I suspect that the HP cells cycle life doesn't drop dramatically with a reasonable increase in rate of discharge. Whereas the HE cells cycle life will drop much more rapidly with increased rates of discharge.

And at a certain rate of discharge, the HP cells will have better cycle life that HE cells.

You are 100% right, I was mesured this dependency for almost all small cylindrical cells on the market, particularly the trio of 30Q, HG2 and VTC6 (they all shows very good results at 3C continuous and 100% DoD, where HE cells dies very soon under this test), so if any generalization from me then it is this one. "Use HP cell only in apps where you need high continuous discharge rate like 3C or more, if your averege is well below 1C with short peaks, which is the case of the most traction apps, then use HE cells".

 

[mighty82]

There is way to much wrong information around, so it is very difficult to know.

Strange that all reputable sellers lists vtc6 as a nmc cell. In fact the study you are pointing to is the only result I can get from searching vtc6 plus nca.

It just goes to show one can't be sure about anything. It was my understanding that nca was not ideal for high discharge cells.

 

[Cephalotus]

..., so if any generalization from me then it is this one. "Use HP cell only in apps where you need high continuous discharge rate like 3C or more, if your averege is well below 1C with short peaks, which is the case of the most traction apps, then use HE cells".

OR if you want to charge your battery faster than 1/4C.

(regen could still work even if quick charging does not, I assume that you have to test that for each cell)

There is also the issue of using (charging and discharging) cells in cold conditions. Using high charging currents from regen with a cold battery could be "interesting".
The M36 data sheet is a good indication for that problem: http://www.nkon.nl/sk/k/m36.pdf

 

[multifrag]

Wanted to add some of my data that I've gathered for LG M29.

I was using ZKETECH EBC-A20 with a 4-wire BF-2A fixture.

500 cycles of charging to 4.1V at 1.5A, waiting for 5 minutes and discharging to 3.1V at 3A.
The full discharge was done according to datasheet. Charge to 4.2V at 1.37A, wait 60 minutes and discharge to 2.5V at 0.55A. First discharge gave me 2852mAh, after 500 cycles it went down to 2754mAh. That's 96.5% of original capacity.
*Note that datasheet requires 50mA charging cut off, while ZKETECH EBC-A20 can only go as low as 100mA.

This data tracing was not done in a temperature controlled environment. I had power outages, fast temperature drops, aircon left overnight for no reason... Because of that the CycleGraph has spikes.

Overall I'm very happy with the cells performance. My main characteristics for the cell were cycle life and price to capacity. I bought 200 cells from nkon and getting 2850mah for £1.86 a cell with this cycle life is just what I was looking for.

 

[Pajda]

Great work Multifrag :thumb:

I can confirm that your test results are valid. Your setup, methodology and test equipment are correct. Do I understand well to the yellow course "100% Dis. Datasheet" that it is a simple linear extrapolation of datasheet value, where the cycle life should be >70% of initial nominal capacity after 500 defined cycles?

Only small and not essential remark to the presented test values. I think that EBC-A20 can set 100 mA as the lowest value of current. Datasheet says that you should charge to 50 mA cut-off for nominal cycle. It is a good practice to state this value as well.