A123 B grade info

[wb9k]

I'm still trying to get some movement on the small inventory of B grade 20 Ah pouch cells here in Livonia. No news there, but I can pass along some info on the traits of B grade cells.

A B grade 20Ah pouch will have between 18.0 and 19.49 Ah, is allowed up to 5% self-discharge per month, and will have a max DCR of 3.5 mOhms. This is in contrast to A grade which will have 19.5 to 21.0 Ah of capacity, is allowed up to 3% self-discharge per month, and has a max cell DCR of 3.0 mOhms.

I don't have data for grading of cylindrical cells, but I would expect them to be derated in similar proportions.

Not a lot of info, I know, but hopefully that's at least useful information for some folks.

 

[docnjoj]

How could you tell a Grade B cell from the serial number?
otherDoc

 

[999zip999]

I can live with it. But have taken 18ah out of my pack and the cells where still close in voltage. Just seeing so many bad cells with liver spots ect. I would feel better if there was some way of knowing the story of how the cells came for sell at 18.50 ea. A, B or trash cells ?

 

[wb9k]

I really can't share info on how to decode markings on our cells. Nobody does that because it's one of the only good safeguards a company has against counterfeiting of their product. That might seem far fetched, but I've seen major manufacturers of electronics find that they have counterfeit microchips in their products, and it was a wacky date code format that proved it. It happens with all sorts of parts.

I share this info to shed some light on what makes a B grade cell (several have asked) so that if you have an opportunity to buy cells from a reputable seller who indicates that the cells are B grade, you'll have some idea of what that actually means. Grey market is the same crapshoot it ever was, no matter what marks are on the cells.

 

[davec]

thanks for the info w9
it would be interesting if you could get your hands on these b grade cells... im sure you could give us a deal
overall it seems like they still have decent IR and and the capacity shortage dosnt seem like a big deal
i would put my money on your legitimate b grade than grey market anyday
i have no confidense in those fake cells- every grey market cell i've delt with had wrinkles, pin holes,leaks, scratches,deformed, were all lacking AH - i truly never felt safe around it they are a hazard- but it's not the first time china sold us garbage

 

[dnmun]

be careful that you don't draw attention to yourself at work because of this blogspot and end getting fired for something you don't see coming.

 

[liveforphysics]

thanks for the info w9
it would be interesting if you could get your hands on these b grade cells... im sure you could give us a deal
overall it seems like they still have decent IR and and the capacity shortage dosnt seem like a big deal
i would put my money on your legitimate b grade than grey market anyday
i have no confidense in those fake cells- every grey market cell i've delt with had wrinkles, pin holes, scratches,deformed, were all lacking AH - i truly never felt safe around it

According to my contact who used to work at A123, all those cells we get sold from China were supposed to be destroyed. >99% of them being the Fiskar modules that statistically each modules has an average of ~3.5 defective internal foil to tab weld joint cells (if I remember the numbers right). Bad ju ju.

What's equally spooky to me, is hearing this extremely wide range of acceptable tolerances before it's B-grade.

Most EV cell providers I've talked with cull to B-grade on +-0.5% cap and +-0.25% Ri. If you have something with say 5% difference in capacity or Ri, you had something go very badly wrong in assembly.

 

[wb9k]

thanks for the info w9
it would be interesting if you could get your hands on these b grade cells... im sure you could give us a deal
overall it seems like they still have decent IR and and the capacity shortage dosnt seem like a big deal
i would put my money on your legitimate b grade than grey market anyday
i have no confidense in those fake cells- every grey market cell i've delt with had wrinkles, pin holes, scratches,deformed, were all lacking AH - i truly never felt safe around it

According to my contact who used to work at A123, all those cells we get sold from China were supposed to be destroyed. >99% of them being the Fiskar modules that statistically each modules has an average of ~3.5 defective internal foil to tab weld joint cells (if I remember the numbers right). Bad ju ju.

What's equally spooky to me, is hearing this extremely wide range of acceptable tolerances before it's B-grade.

Most EV cell providers I've talked with cull to B-grade on +-0.5% cap and +-0.25% Ri. If you have something with say 5% difference in capacity or Ri, you had something go very badly wrong in assembly.

Your contact is right about the provenance of the grey market cells, and I've said exactly the same here many times already. Those bad tab welds are what's responsible for the black spots that show up around the perimiter of the cells. OUt of 21 cells in those modules, you can expect ~10% or so will have that problem.

I'll worry about how other makers sort their cells when they outperform ours.

 

[bigmoose]

Luke can you go a bit further on where the "bad tab welds are" and why/how they result in the attack upon the foil? ... or did I miss it in a previous thread?

Trying to not ask something we can't of w9bk...

 

[RedLine19K]

http://endless-sphere.com/forums/viewtopic.php?f=14&t=38068

 

[bigmoose]

Thanks redline, don't know how I missed (or forgot :x ) the information about the effect on the pouch.

Copied again in case others missed it (info from ohzee and LiFe and flathill)

This defect was undetected by our standard visual and electrical inspection yielding cells which initially met specification. When the prismatic cells with the undetected defect were subsequently compressed as part of the standard module assembly process, a mechanical interference was created between the misplaced component and the foil pouch which contains the cell. In certain cases, this interference can breach the foil pouch electrical insulation, causing an electrical short which can cause premature failure of the battery module or pack, including a decrease in performance and reduced battery life.

25% of Livonia production material was affected. One of four stations that prealign and weld the tabs was misaligned by approximately 2mm, resulting in possible pouch fusing, or pouch separation (seal) problems.

If a123 is telling the truth, "only" 1 in 4 cells might be bad out of all the reject packs (1 of the 4 machines was not calibrated right). What you might want to try is fully charging the cells, clamping in a vice with moderate force, and then let sit for a few days. Those with internals shorts would exhibit higher self discharge I would think.

I remembered the "misalignment" issue, but didn't get how it connected to pouch corrosion... good info!

 

[dnmun]

you could see it on cwah's A123 pouches which puffed up and had the black areas where the aluminum film in the mylar sleeve had shorted to the electrode and corroded. i assume that was where the gas formation initiated.

 

[wb9k]

you could see it on cwah's A123 pouches which puffed up and had the black areas where the aluminum film in the mylar sleeve had shorted to the electrode and corroded. i assume that was where the gas formation initiated.

It has nothing to do with gas formation, though that could come later after the cell has lost capacity if you overcharge/overdischarge it due to lack of proper controls.

The black spots are the result of (previously cyclable) Li alloying onto the Al pouch, causing the discoloration as well as embrittlement and cracking of the Al pouch. Eventually, the pouch leaks electrolyte.

All of this is a thing of the past. Processes have been dramatically improved since that time.

 

[liveforphysics]

In my experience, when pouches develops spots (which typically first are transparent-ish white, then turn grey, then turn black), it's because the inner foil surface was not handled as delicately as it needs to be (bad pouch stamping process, etc), or it was originally manufactured from defective with pin-holes. Like wb9k said, the aluminum can be lithiated among other more ugly processes

The bad news is, once you have the electrolyte in contact with the middle aluminum layer inside the laminated foil composing the pouch, you are just playing the waiting game until pouch failure, as nothing can passivate or halt the destruction of the aluminum layer once it's begun. One of the key functions of the aluminum barrier is serving as a non-permeable humidity diffusion barrier. Why do you care about keeping humidity out of your electrolyte? Because on contact it decomposes into HF (Hydro-Fluoric Acid).

The devil is in the details with pouch cell mfg, and more companies have under-estimated it's difficulty than gave it the care and consideration it requires. There are pretty grand opportunities for methods to fail at every single step of the many step path involved in making pouch cells. Getting it right means discovering all of them and creating processes/tooling to ensure none of it can happen, which is a lot easier said than done.

Back to the topic of B-grade cells.
To have a difference in capacity or resistance after formation in the multi-percent range, it takes something being different by a pretty substantial amount. Either you blended an inconsistent or defective batch of cathode or anode slurry material by starting with bad materials, or combining them at the wrong ratios, or inadequate mixer design to create a homogeneous slurry. Or your coater under or over coated and after baking and rolling it wasn't rejected, but modern lines are sampling coating thickness ultrasonically thousands of times a second as the foil leaves the roller, so this seems a little unlikely. It could also be caused by rough coated foil handling causing the binder to fail and have substantial amounts of active material flake off, which makes for a dangerous cell. Or it could be impure or inadequate electrolyte causing less active area to develop in formation (which also seems unlikely unless they are using very sketchy, like RC-lipo grade sketchy electrolyte.)

If you have uniformly mixed slurry, and you apply and rolls it at the correct design thickness, and you have the same coated foil area and same number of sheets of foil, and they all got properly wetted and formed, you will not see beyond tiny fractions of a percent variance in capacity or resistance. If you are down by say 1Ah on a 20Ah pouch, this may mean something has caused you to loose as much active area as an entire foil sheet missing or un-connected etc. That would seem concerning to me personally and they normally are destined for a recyclers tub-grinder, outside the RC world where they just get sold down stream to ebay vendors.

Unless you know of some unique way a pouch ends up being off by 5% in design capacity yet didn't have something go badly wrong to cause it?

 

[arkmundi]

I share this info to shed some light on what makes a B grade cell (several have asked) so that if you have an opportunity to buy cells from a reputable seller who indicates that the cells are B grade, you'll have some idea of what that actually means.

Thanks for the post of additional info. So that's an interesting question, "a reputable seller who indicates that the cells are B grade". I'm assuming that buya123batteries.com is selling A grade? Are there similarly A123 sanctioned resellers for B-grade cells? How does one go about finding & verifying a reseller and their stock? Best!

 

[docnjoj]

Thanks WB9K. I thought it was worth a try to ask.
otherDoc

 

[Hillhater]

Mabe WB9K can comment but....
Unless there are some lax manufacturing practices being used, any modern mass production process with multi station (parallel processes ) being used will employ various means of "traceability" ..to be able to identify from the finished product which exact machine it was processes on at each stage of production.
That can be anything from laser etched codes, to old world UV ink marking, or even mechanical ID marks..
If Livonia has these systems , it should be possible for someone with the background info, to easily identify and differentiate the defective 25% of pouches from the unaffected ones.

25% of Livonia production material was affected. One of four stations that prealign and weld the tabs was misaligned by approximately 2mm, resulting in possible pouch fusing, or pouch separation (seal) problems.

 

[999zip999]

You would think, but this misstakes happen alot. Plus they have to get caught. Look at all the car recalls ect. After yrs. on the road.

 

[miro13car]

I would repeat Arkmundi question,
does "buyA123batteries.com" sell only grade A cells??
I would think so , they charge $70 , for that price I would expect A grade.

 

[Hillhater]

You would think, but this misstakes happen alot. Plus they have to get caught. Look at all the car recalls etc. .

Exactly !.. there is always the possibility of defects, and QA procedures that should detect those defects can fail, ..but that is why "Traceability" features are important, such that when a fault is eventually uncovered , the manufacturer can both tell precisely, when , and where the product was made, and also where it was dispatched to and which other components (car models etc) it was installed in.
Failure to employ effective traceability systems can be extremely costly to a manufacturer.
I bet A123 wish they had employed both better product tracking , and better QA testing procedures for those pouches !

 

[whatever]

at wb9k
you can tell something about a cell from its serial number? how much info can you get? only where/when it was manufactured?

 

[John in CR]

B grade I can handle, but defective is another matter. I have 31 cells that quite likely were made during the 1 in 4 are potentially bad time period, so my bottom line is how best can I force an early failure of the bad cells before I build a pack. If a cell is defective, what is the best way to force the issue...cycle under extra compression, cycle under no compression, cycle at high rates, ???

 

[t3sla]

You would think, but this misstakes happen alot. Plus they have to get caught. Look at all the car recalls etc. .

Exactly !.. there is always the possibility of defects, and QA procedures that should detect those defects can fail, ..but that is why "Traceability" features are important, such that when a fault is eventually uncovered , the manufacturer can both tell precisely, when , and where the product was made, and also where it was dispatched to and which other components (car models etc) it was installed in.
Failure to employ effective traceability systems can be extremely costly to a manufacturer.
I bet A123 wish they had employed both better product tracking , and better QA testing procedures for those pouches !

They did and they do.
The problem is the cells sanctioned for destruction entered the grey market.
Make no mistake, A123 handled the situation accordingly, all those cells are contractually marked for termination. The company has zero reason to inform the public which cells salvaged from oem packs passed through misscalibrated machinery.

I don't know why people think otherwise, just because you gave money to someone with another companys product markings doesn't entitle you to customer support.

 

[RedLine19K]

You missed the point completely. Of course A123 wouldn't tell buyers of grey market salvage which cells were defective. Actually, they couldn;t if they wanted to because.. they have know way of knowing themselves.

Any reasonable QA should include a way to tell which cell was made on which machine. It didn't. If A123 had been able to identify which line each cell was made on, it might have lessened the $56M oops that killed the company.

And if each "line" had unique identifier code, it would be possible (but not guaranteed) that *we* could identify the bad cells also.

 

[t3sla]

I believe they were able to know, but the end solution was to scrap all affected packs, not repair. The situation could have been handled differently.

But that is a separate discussion, sorry for aiding in going off topic.

I also would really like to know what production variation causes such a massive difference. My reference point would reject that variation as unacceptable, faulty not B grade.