Testing the impossible (nr of cycles)

[Frank Blue]

Hi everyone,
6000 cycles @ 0.5C means 4h per cycle or 24000h, or 1000days, or almost 3 years of continous testing.
Latest semi solids datasheets now claim 12000cycles@0.5C that needs almost 6 years.
I wanna know, are they selling us a 3/6 years old product claiming its the latest tech available or is there a time warping device to make those tests?
We are busy discussing impact of clamping on cycles and all the world is betting big money assuming they tell us meaninful numbers (think about return of investment dependent of those nr of cycles).
They may test for a period and project in the future assuming a certain curve..but who knows if after some nr the cells just fail?
A cell with 6/12000cycles cycled once per day needs 20/40 years to reach end of life..how do we test something like that to prove them wrong or just make a comparison between brands to make an educated choice?
They can claim whatever number they want, we have no way to really compare cells lifetime in cycles/years, that is the more important factor when investing in a battery.
Who cares about €/kWh..what really is important is €/(avg kWh*cyclesnr).
NR of cycles vs capacity is the test we really need but it is IMPOSSIBLE because result will come too late.

Am I missing something? How do you make your choice when buying cells?

A side example..my pv panels was certified 25years @75%...after 17years they are @55%, my friend bought another brand certified the same..he is @80% today.
The inverter, expected to be the best at the time was negative hearted that ruins the panels production in the long run so those who invested in cheap less efficient inverters with worse datasheet did the best thing at the time because money invested in best brand never returned.

 

[chrisski]

I've always been curious about how a battery type us tested.

I am also curious about how each lithium manufacturer has a curve for their product.

 

[HighDesertOffgrid]

Am I missing something? How do you make your choice when buying cells?

The odds of any particular manufacturer even existing in 20 years are almost nil. Especially considering the global economic climate.
I base the cost risk against the competitive technology (lead acid) which will give me 2+ years of "deep cycle". LFP is nearly the same price now, so it's a clear choice.
Products that demonstrate unacceptable short to medium term defect rates, I avoid.

Most importantly: Let someone else be the beta tester. There are plenty standing in line for that privilege.

 

[Q-Dog]

They test by running X number of cycles, then measuring degradation and extrapolating out from there. Only time will tell if their numbers hold up.

 

[AntronX]

The cycles are extrapolated by literally drawing a line on a chart from couple hundred tested cycles. I would not trust those 12k cycle claims unless they publish their methodology.

 

[BarracudaBob]

Motorola's factory in Plantation, FL had a complete lab to preform accelerate aging test. It was all environmental of warm salt mist, physical dropping, a shake table, on/off button testing. Charge and discharge limits itself to exactly that. You would have to do it at a accelerated current, temperature, charging/discharging repetitive method.

Send me some batteries and I'll torture them!

 

[Frank Blue]

So Motorola owns a time machine, we finally prove that :D
btw I have exaclty that renogy batt in my rv..is it a video or what? I want to know more ot that torture. I love your harvesting phrase, very nice (technically you are converting but I love the concept).

Back in topic..now we agree that there is no real complete test and numbers are just the "extrapolated best guess" they can afford to tell, given that the company economy and existence depends on that numbers. I think we can agree that as buyers we cannot simply rely on what manufacturers says..we need experience from others.

But long term experience is not available for new products..that are the cheapest.
I am torned between proven cells like winston that costs a lot but are the same since 20y and known to work for decades, or normal lfp or semisolids.
My experience tells me to spend more and have something proven..but damn it is more than double the cost..this kind of things make me think we are all scammed by all the companies that ask 80$ for a 280Ah cell because they will be quickly useless while proven brands costing double will be a better investment in the long run.
I personally feel like I am blindly betting..no problem if its an hobby or passion but as an investor..I would never put my money on lfp today.

ps: Nothing personal with winston, its just a recent long term review available in this forum.

 

[wpns]

My neighbor discovered that two of three LFP 48V 30AH golf cart batteries from Allied Lithium are open circuit after 2 years and Allied Tech support is <ahem> not very responsive. Which is a great business model. For Allied.
Twice the cost of the standard FLA pack which usually lasts 3-5 years depending on maintenance.

 

[J.H.]

I have seen numbers from 3000 cycles to 15000 cycles used in advertising..it seems the cheaper the battery then the more cycles they promise in many cases…

It is my “professional ,high paid and very scientific opinion” that many companies just make this shit up…!

Most of us won’t still be here to even know if this was BS.. if they are here they will be to old to remember who to sue.

J.

( P.S.) .Anyone wanting their very own FREE “ professional, high paid and very scientific opinion” on any other highly technical or complex product in the world , order NOW while supply’s last..

All you pay is $9.99 US..for SHIPPING..
If you act in the next 24 hrs we will include a 2nd Bonus scientific opinion for only $2.99 US

Operators are standing by….
Credit Cards accepted..

Call 0011-86 - 867-5309 ……Today..!

(Toll fees for international chineese calls may apply)

EDIT( geeez guys) I’m dissapointed .. I thought 867-5309 would give away the whole sick humor of this parody on 80’s info-mercial BS….. that song “ Jenny” still gets airplay today on classic rock stations..
J.

 

[Frank Blue]

We everybody know someone who had bad cells..btw golf carts are very hard load for cells, I would never put cheap cells there, only proven long lasting cells.

Ok, from now on I propose:

1: to call those pdf datashit until the lfp producer prove to own a time machine.
2: stop arguing about things supposed to affect cycles nr because there is no real data about it.
3: create a place to report annually the result of our battery with a simple "cells 280Ah eve v3 cycled 560 times owned for 24 months, now @265Ah". That would probably be better than any datashit around and force manufacturer to move from feeling safe in theyr diy marketing strategy of sending free samples to youtubers that make a short test to be afraid of a real db with real long term data about them. This stop the phenomenon of new brands that rebrands the same mfg because the newest has zero records, sticking with a single brand that force them to fulfill guarantee promises.
All for the better of the world an humanity..sure, with a loss for chinese that are dominating this market with ignorance. In a sense could be a political tool.

ps the long term db is a serious proposal, I hope the forum owner would consider it and give me credit, I pretend it to be called fdb (frank data database) in the sense that data is frankly presented :D

 

[wpns]

golf carts are very hard load for cells, I would never put cheap cells there, only proven long lasting cells.

Yes, I know. What manufacturers are making drop-in LFP replacements for FLA golf cart batteries, that have 8,000 cycles experience in the market?

<crickets>

 

[BarracudaBob]

Already been working on that:

Capacity: 99.5 Ah
State of health: 99.5 %
Cycles: 139


So doing linear math at 5560 Cycles I'll be at 80% ?? This is on a LifePower4 V1 by its BMS. Assuming it doesn't accelerate. It almost seems like programmed obsolescence

So I lost 0.5% in 139 cycle.

 

[Frank Blue]

I am pretty sure they make calculation exactly in this way, linear projection (there is no time to make a binomial or whatever curve and is not convenient to them). The problem is..out of 100 cells, how many will fail before the claimed 5000 cycles? As I said the only important number is avgCapacity*nrOfCycles, this is the nuber we should care when buying amd IMHO exactly here is the culprit:

they are givin number about the depletion on capacity vs cycles nr (they should say 0.004% lost per cycle instead they arbitrarly express it as 20% lost in 5000cycles) in a way that is interpreted as a life expectancy (cell will survive 5000 cycles @ 80% cap).
New semisolids are taking advantage of that claiming 12000@70%
I am waiting for some one who claims 40000 cycles at 0%, teh next will be 80000 cycles @-100% (you put back small current in reverse polarity with zero efficency, number perfectly valid).

golf carts:
I have no experience in doing this with golf cart but I did the trick in various situations: reduce the number of cells until the cart has acceptable performance, than put the biggest cells you can. In this way the current density of the battery will drop a lot and battery lasts more. Clearly you should trick the protecion of the cart (usually there is a wire that reads the tension) and you lose the capacity meter on the cart (who cares, use the BT bms).

 

[wpns]

Why on earth do we think the degradation is linear?

 

[robby]

Three years ago before all of the exaggerated claims started to come out the A+ batteries like EVE's were rated at 5,000 to 6,000 cycles. The 5K being if you charged to 100% and discharged to 20%. The 6000 was based on 90% to 20% and all of them I remember stated around 54V as the charging voltage.

The Warranty was 10 Years under these numbers and that sounds about right given that 5000 cycles would mean 13.6 Years of life. It's enough Margin to make a 10 year warranty sound credible, like the battery manufacturer expected to be in business in 10 years and is making sure the return rate would be low.

When I see 10,000 cycles I know it's just a scam, they are protected by the same 10 Year warranty, they know you will cross that 10 year marker before you realize that the 10,000 cycles was a scam.

 

[Frank Blue]

I agree with your point of view.
Where do I claim warranty anyway? the reseller usually just disappear, do eve and the others make anything after say 9 years? I mean is there anyone here that has been refunded after years because a cell failed?

 

[wpns]

I agree with your point of view.
Where do I claim warranty anyway? the reseller usually just disappear, do eve and the others make anything after say 9 years? I mean is there anyone here that has been refunded after years because a cell failed?

If Eve even wanted to honor the cell warranty you'd have to prove how many cycles to what DOD and under what conditions of charge/discharge rate, OVP, UVP, temperature, etc you used them, and no-one has that kind of data. I mean I'm obsessive, and I have a lot of that for the 6 months since my batteries were installed, but it doesn't mean EG4/SS/<battery mfr> would take my word for it... At the end of the day, EG4/SS are successful enough and have a low enough returns rate that they can afford to honor their warranty(*) and get a good reputation in the marketplace that they are able to make me happy, or they went out of business and I'm on my own(+) for dealing with the issue and selecting new hardware. That's where we are today with everything. Look at Rivian...

(*) See the other threads on how poorly defined their warranty is, and how easy it might be for them to weasel out of it. Not to say they _will_, as their service has been increasingly robust lately, but they _could_. But all the warranty clarity in the world doesn't help if they go down the tubes. And I don't want a prorated credit for a bad battery, I want the problem fixed... Actually, I want 20 years of uninterrupted service from my hardware without worrying about the warranty at all, but "What's it like to want?"

(+) Plus of course, this forum, which is how I've been so successful so far in my solar journey.

 

[hwy17]

I just assume a 3-5 year lifespan with a 10% chance of early individual cell failure.

Don't try to arbitrage the grid with a narrow financial payoff, you'll probably lose.

Just stick to what's worth it for your backup comfort and convenience.

 

[J.H.]

Where do I claim warranty anyway?

Shucks ,that's simple … ya fly to China to the big city whose name you cant pronounce .

You hire an interpreter and a taxi to drive you around looking for an address you can’t pronounce, located in a warehouse behind a row of other warehouses and search for the company owners home address as won’t answer the phone or meet with you…

Remember, you are DUE that prorated refund ..badger him till you get it..
be persistent …don’t accept no as an answer…!

P.S. avoid the bat soup if possible…
jus sayin….

J.

 

[Frank Blue]

That has been a good thread, full of info and humor
Thank you for all the replies.
I guess the final word is..they sells us cells that mostly works in the first period, with some early fail, than people disappear without reporting long term data so we can't predict the economy with precision.
Clearly not the time to reopen my little company, what I should tell to my clients, make a bet hoping for the best and please don't sue me if it fails? LOL

 

[BarracudaBob]

Now they have the mandatory mitigation clause so you can't even sew them. in 9 years that company changes its name or other ways out of their warranty long term.

 

[doctron]

Whats best way to get proven original grade A cells ?


Trusted seller or Qr code data?

 

[BarracudaBob]

I found this thread:

F

Thread 'Collection of studies about LiFePo4 degradation'

Nov 17, 2020

For making proper purchase decisions at LiFePo4 cells, I collected studies about battery degradation and put all together in a PDF,
each picture linked to the study.

• founder
• Replies: 13
• Forum: DIY LiFePO4 Battery Banks

• 

 

[BarracudaBob]

Whats best way to get proven original grade A cells ?


Trusted seller or Qr code data?

I would say both but having a deformed, sanded, or bogus QR code is pretty bad.

 

[BarracudaBob]

Why on earth do we think the degradation is linear?

Unless you give us the formula my brain will stick with linear progression