Li-ion cells cycle ageing

[goatman]

Here is Samsung 30Q No 3 condition before the cycling :

https://endless-sphere.com/forums/viewtopic.php?f=14&t=103092&p=1513064#p1513064

Cycle No 1 had 2180 mAh, average at cycle 535 is 1880 mAh.

I find this interesting on whether to baby a 30q or not to baby it. if I look at the ah produced by the cell on your tests after 550 cycles, starting with 2180mah and now 1880 mah. take average mah 1980x 550 cycles, it has produced 1089 ah's

okashira 95% DOD at 7amp discharge/2amp charge started at 2800 mah and after 540 cycles was about 2100mah. average 2450mah x 540= 1323 ah's produced. also watt hours were 8wh at start and 6wh after 540 cycles. average 7wh x 540= 3787wh.

if im weed infesting let me know and ill stop after this post. if I average 20wh/km, that battery could have gave me 189.35kms so if I had a 20s10p pack/ 200 cells I would've been able to travel 37870 kms

after 700 cycles the average mah was 2375= 1662.5ah or 47587.9kms.

your cell so far is 1089ah or 31171.9kms travelled.

the pack is charged twice as fast and discharged 3 times faster.

I guess this is more like a race between a turtle and a rabbit, at 7 amps I could've travelled 37870 kms probably 4 times faster than 2amps for 31171kms, are you going to test the 30q to 700 cycles?

 

[Pajda]

Pajda, could you please comment on my attempt ? Do your 50E tests have similar trend of the capacity ?

Continue with the test Yes I measured similar trend with my 50E samples. My closest settings is 4.1-3.2V at 0.5C-1C rate. The 50E cycle life is outstanding. Slightly better are only M36 and M29. Unexpectedly, the M50T cycle life is significantly worse, but on the other side M50T have lower 10s DCIR than 50E and it keeps it lower for the whole life.

 

[docware]

if im weed infesting let me know and ill stop after this post.

I accept your suggestion, especially after reading your thread „100% discharge rate capacity“ , so thank you for no more posting here.

https://endless-sphere.com/forums/viewtopic.php?f=14&t=104568

Just few citates from ES members to your posts there :

LuboN : „I am sorry to inform You that You got it totally wrong. Please read some battery life related threads to get the basics right first.“
„…. I must say it is unbelievable how far can one misunderstandig go... Your problem is simple. You totally misunderstood and misrepresented the data in datasheet. You have no idea what it is about and for what.“

Ohbse : „Each time you clarify, it makes it significantly more confusing...“
„Completely, absolutely 100% wrong. This is my point - your fundamental understanding is utterly incorrect.“
„No, again, your graph is meaningless. The % figures you reference have ZERO bear“

Madin88 : „Sorry you are wrong“

Hillhater : „You Know... offering up a pile of regurgitated mess that has already been chewed over and spat out.... doesnt make it any more palatable the second time around !“

Ianhill : „Goatman are you electricgod playing a trick on John cuz surely No one can get it wrong this much.“

 

[docware]

..... Yes I measured similar trend with my 50E samples. My closest settings is 4.1-3.2V at 0.5C-1C rate. The 50E cycle life is outstanding. Slightly better are only M36 and M29. .........

Thank you Pajda.

 

[zacksc]

Does anyone have an idea where Samsung 40T cells might be in the cycle aging spectrum? I am thinking perhaps similar to the 30Q data shown above in docwares post.
And thanks very much, docware for sharing that awesome cycle aging data.

 

[Pajda]

Does anyone have an idea where Samsung 40T cells might be in the cycle aging spectrum?

Yes we have

 

[zeccato]

My question is: how much percentage of the new cells lose each year?
If I buy cells of 21700 5000mAh or 18650 3000mAh, and use them at 1c of continuous discharge.

for example: (I don't know how he treated them) one guy said:
a 10s4p battery consisting of cells LGDBMG118650, are 2850mAh with 10A discharge,
after about 4 years of non-intensive use, they now take only about 4000mAh.

 

[docware]

I was curious about DCIR after 700 cycles through the whole SOC range. Initial DCIR measurement (green line) and after 700 cycles (yellow line) was done on different cells.

Conclusions:
- Panasonic PF DCIR is really very different from others cells
- the general shape of DCIR through SOC is confirmed for all 4 cells
- M36 and 29E7 are really excellent cells regarding DCIR
- it´s very difficult to properly determine SOC in the area 0 – 10 %

 

[Honk]

Wow, extremely good and informative measurement.
I can't imagine how you manage all these tests in such short timeline.
Keep up the good work, it is so very great of you taking the time doing this.

 

[john61ct]

Yes a tremendous service to us all, thank you

 

[hemo]

As John & Honk have said this really is fantastic testing by Docware, so kudos for the time consuming work being put in and aired here for all to view the results .

 

[zeccato]

Good test, but to be more real I'd like to know how many amps the new cells lose each year. (used normally).
Because if they lose more than 10% every year, I don't convert: I stay with my lifepo4 (8 years with 75-80%).

 

[john61ct]

Not enough information, too many variables.

C-rate range, cycles per week, avg DoD%, charge profile, temperatures have a huge impact.

There is no "normal use". . .

But generally speaking everything else being equal, LFP lasts 4-8x longer

 

[zeccato]

Good test, but to be more real I'd like to know how many amps the new cells lose each year. (used normally).
Because if they lose more than 10% every year, I don't convert: I stay with my lifepo4 (8 years with 75-80%).

is that right, do you think?
is it the same with the usual e-bike batteries?

(car) Lithium-ion battery degradation over time
Temperature 40% charge 100% charge

0 °C (32 °F)
40% charge: 2% loss after 1 year
100% charge: 6% loss after 1 year

25 °C (77 °F)
40% charge: 4% loss after 1 year
100% charge: 20% loss after 1 year

40 °C (104 °F)
40% charge: 15% loss after 1 year 100% charge: 35% loss after 1 year

60 °C (140 °F)
40% charge: 25% loss after 1 year 100% charge: 40% loss after 3 months

Batteries slowly lose their ability to store energy over time even when unused.

 

[john61ct]

Not all chemistries.

I have seen quality LFP cells lose less than 0.01% capacity stored cool and isolated well over a decade.

And cycled daily, average 80% DoD, but gentle C-rates, less than 5%, again over a decade.

LTO goes much **much** longer.

The chemistries selected for EV / propulsion for higher density

e.g LCO, LMO, NMC, NCA

may only go 300-500 cycles.

 

[Dak77]

Pajda (or anyone else with the data) , how does the M29 compare to 29E in DCIR at around 500 cycles at 2C discharge and .5C charge rate at 70% - 80% state of charge ? The reason I ask such a specific question is to get an idea about what kind of voltage sag one would expect at a 2C load . I'm close to ready to build a new pack with M29 , but I want to make sure 29E wouldn't be a better option. I'm not really asking for precise numbers as much as "are they in the same neighborhood" ?

 

[zeccato]

N................

If I were to take as an example: (LG INR21700-M50T 5000mAh - 7.3A) or (Samsung INR21700-50E 5000mAh - 10A) to make a battery of 15s 35Ah, and i use them with maximum discharge 1c, and charge 0.5c, 2 cycles per week, with winter temperature + 10-15  °C, and in summer +30-35  °C,
theoretically how many years will they last? or how many amps will they lose each year?
 
Thanks for all the suggestions.

 

[john61ct]

theoretically how many years will they last? or how many amps will they lose each year?

Too many variables to predict specifically, IRL 500 cycles is a lot

The max 1C discharge / 0.5C charge rate is excellent, go slower or pre-warm in winter.

How many cycles per year?

Discharge to what avg DoD%, what charge profile?

What is avg ambient temperature?

All these factors have an impact, buy good cells for high energy use like the 50Es and do the best you can do.

 

[zeccato]

Sorry about the translation but I didn't understand: IRL 500 cycles is a lot.
I don't need specifically, more or less.
100 cycles per year.
I'll discharge the battery at 1c, and i charge it at 0.5c.,
winter temperature + 10-15  °C, summer +30-35  °C,

theoretically how many years will they last? or how many amps will they lose each year?

Too many variables to predict specifically, IRL 500 cycles is a lot

The max 1C discharge / 0.5C charge rate is excellent, go slower or pre-warm in winter.

How many cycles per year?

Discharge to what avg DoD%, what charge profile?

What is avg ambient temperature?

All these factors have an impact, buy good cells for high energy use like the 50Es and do the best you can do.

 

[john61ct]

Should last at least 5 years

IRL = "In real life"

as opposed to bench test data

This is my last post responding to this diversion, start a new thread if you like

 

[zeccato]

Replies from another forum (translated):
But it's true that a 36V 10Ah loses something every year, and after three or four years you have to change it, but it's also true that if you make a 25Ah battery it will last long enough since it's not used like the classic 10Ah batteries.

If you want to have a LiIon battery that really lasts seven years, you will have to make a careful choice of cells to ensure that they are not stressed at all by the use and remember that they should never be discharged more than 80% of the maximum charge and in case of non-use should be "hibernated" to 50% of the charge and keep them in the lower part of the refrigerator, otherwise the hell you get what you want.

The LiIon batteries of electric cars are guaranteed for 1800 charging cycles and this thanks to some charging and discharging devices. they control the temperature etc. etc.
So if you make yourself a brand and first choice LiIon battery and then treat them very well as I have written above 500 cycles you double them great

 

[docware]

As John & Honk have said this really is fantastic testing by Docware, so kudos for the time consuming work being put in and aired here for all to view the results .

Yes, DCIR mapping is quite time consuming. Curiosity was driving force here, as usually. I just, as others, share information which are interesting for all of us, hopefully. For that purpose was ES created, I believe.

New data : Samsung 30Q 636 cycles, LG MJ1 600 cycles, SONY VTC6 358 cycles, Samsung 50E 300 cycles, Samsung 35E 277 cycles, LG HG2 230 cycles.

 

[flippy]

what is your indication why the 30Q suck so much?
spec sheet wise it should behave better.

 

[docware]

what is your indication why the 30Q suck so much?
spec sheet wise it should behave better.

I have only this official 30Q specification. Do you have any other spec promising better lifetime ?

 

[Honk]

From the chart it looks like 35E could be the new GA in terms of capacity decay.
I also wonder how a 18650 cell of LifePO4 would come out in tests like these.
It might reveal something we never thought of, or just imagined how it should behave by myth.