Why 18650 makes things worse for EV.....

Jolly Jumper

Jolly Jumper

100 W
Joined
Sep 21, 2013
Messages
108
Location
Germany
As i have seen the http://www.buya123batteries.com/APR18650M1A_Lithium_Ion_Cylindrical_Cell_p/apr18650m1a.htm i calculatet the more wight against the 19.5Ah pouches.

Surprise Surprise :mrgreen:
they are wayyy more heavy
AMP20M1HD-A 495g with 19.5Ah = 130W/kg
APR18650M1A 39g with 1.1Ah = 93W/kg * 39% = 130W/kg 39% more :mrgreen:


Lets take a "would be" look with the currently aviable

INR18650-24R 42g with 2.35Ah = 201W/kg *39% for "would be 500g pouch" = 279W/kg almost double that from turnigy or zippy RC pouch cells !




So my question is now why they not produce 300g or 500g pouches with the current aviable technology/chemestry ?
 
Jolly Jumper said:
So my question is now why they not produce 300g or 500g pouches with the current aviable technology/chemestry ?
Click to expand...
the Chemistry is "Similar" but the "RC pouches" are designed to give high discharge rates ( 20, 40, 60+ "C" ).. much higher than the round cells.
Different use needs different product !
But the "Turnigy cells are up at 166 Whr/kg pouches ( 114gm for a 5Ahr cell)
 
I think pouch cells lose their energy density advantage when you have to add space for cooling, or racks for compression. Still, why not 1c cells with the 230Wh/Kg that the 18650s have reached?

Tesla uses of billions of cells, which could enable them to create a larger format for the next gen. The savings in pack assembly would hopefully offset the change. But the proprietary approach has cost other EV companies severely. Tesla chose the commodity 18650s with a very good reason, lowest cost, rapid improvement and multiple suppliers. As much as like the idea of a 30ah 4.2v cell in 38140 or some other cylindrical format... 8 18650s for 27Ah would fit in the same space and cost less because of volume. Cutting the cell count by 8 would have to offer some advantage. Will they continue to take risks, or stick with what has worked before?
 
High C-rate 18650's have lots of foil edge joint contact points to get a low resistance joint to the can. You don't want to be conducting down the length of the long spiral of super thin Cu/Al foils much at all before hitting the can again in a cell rated for 30A. It unfortunately eats both gravametric and volumetric specific energies. Pouches are a little more forgiving at scaling up current densities for this reason.

If you pick a bit lower C-rate cell, you will find 18650's stop taking as much of an energy density hit to pouches. (NCR18650B etc)
 
Didn´t knewed they need compression. RC Licos don´t have it. But made off aluminium with smart engeniering its maybe only 2-5% more wight.

Pannasonics 3400mah 1C have >250W/kg

I would instantly sell my old battery when there would be a 10C constant pouch cell 24R ~300g format with ~266W/kg
 
liveforphysics said:
High C-rate 18650's have lots of foil edge joint contact points to get a low resistance joint to the can. You don't want to be conducting down the length of the long spiral of super thin Cu/Al foils much at all before hitting the can again in a cell rated for 30A. It unfortunately eats both gravametric and volumetric specific energies. Pouches are a little more forgiving at scaling up current densities for this reason.

If you pick a bit lower C-rate cell, you will find 18650's stop taking as much of an energy density hit to pouches. (NCR18650B etc)
Click to expand...

Does that mean that pouches are additionaly to the lower wight easier to produce?
 
Jolly Jumper said:
liveforphysics said:
High C-rate 18650's have lots of foil edge joint contact points to get a low resistance joint to the can. You don't want to be conducting down the length of the long spiral of super thin Cu/Al foils much at all before hitting the can again in a cell rated for 30A. It unfortunately eats both gravametric and volumetric specific energies. Pouches are a little more forgiving at scaling up current densities for this reason.

If you pick a bit lower C-rate cell, you will find 18650's stop taking as much of an energy density hit to pouches. (NCR18650B etc)
Click to expand...

Does that mean that pouches are additionaly to the lower wight easier to produce?
Click to expand...


Definitely not. For under $200k you personally could ebay enough used equipment to start making your own 18650's. This is because it's production is refined to a commodity science with many competing vendors pushing new advantages and higher speed higher reliability tech for 18650 production. For pouch production, you could do it on a dirt floor with scissors and an iron for laminating if you wanted to make some very shitty sketchily made cells (like almost all chemistry test pouch cells made by chemists look like lol) However, if you want to make EV grade pouch cells (which only a handful of companies in the world can even make), it takes a level of automation that you couldn't touch setting up for under $10million, and since there is no size standards on pouches, you pretty much have to have everything custom designed for you as 1-off parts. 18650 making machines roll out of factories by the truckload and the top quality ones are dead-nuts reliable and have crazy fast cell assembly throughput. You need warehouses of formation machines setup just to handle the cells post assembly from a single modern highspeed 18650 line, and with good materials coming in they see <0.01% QC fallout.
 
Its need more robot skills to roll the layers and connect all these layers to the can poles than just lay them like sheets of papers of each other and easy put this in a film welding machine.
Can´t comprehend your 200k vs 10m estimation.

A standard for pouches would be easy to find. Take the size from A123 15Ah and 20Ah cells and only vary the thickness or make a CEO meeting to find new format i would vote for 120mm*80mm thikness vary :mrgreen:

From who you know about the <0.01% QC fallout?
 
But is there really a technical reason pouches couldn't be improved and made cheaply? seems easier. The whole Enerland/A123 fisco that killed A123 and wounded Fisker is one glaring example of pouches gone wrong. It's all demand and that ship has sailed. So this basically precludes any larger format displacing the 18650...even if Tesla will eventually consume most of the demand? i think manufacturers would oppose a standard size, simply because they'd want to prevent customers from leaving if someone else found a way to jam another 1Ah into the battery, or sell .10 cheaper.

http://www.torquenews.com/1075/teslas-electric-car-plans-set-swamp-worldwide-lithium-battery-manufacturing

http://hardware.slashdot.org/story/13/09/03/1254229/at-current-rates-tesla-could-soon-suck-up-worldwide-supply-of-li-ion-cells

http://www.teslamotorsclub.com/showthread.php/12709-18650-Batteries/page3

There's a contradiction in that, by the time production is big enough to make a new format feasible and necessary, the risks are too great to attempt it. Instead, panasonic will just keep building new plants http://www.autonews.com/article/20131021/OEM06/310219870/panasonic-to-increase-its-battery-capacity#axzz2luqr8JjI

The contract for 1.8B cells is to 2015, Tesla could actually prepare to make a huge change then
http://www.mercurynews.com/business/ci_24531612/tesla-motors-may-make-its-own-batteries

I'm not sure any of this matters. NCR18650B are still $2 for Tesla and $10 for us which is the same $ .80/Wh that A123 and other "good" batteries cost. So it will always be $700for A grade, $500 for B grade or 300 for China LifePO4....A new higher density format won't likely be cheaper or more available than any existing type, if the last 10 years is any guide.
 
Jolly Jumper said:
Its need more robot skills to roll the layers and connect all these layers to the can poles than just lay them like sheets of papers of each other and easy put this in a film welding machine.
Can´t comprehend your 200k vs 10m estimation.

A standard for pouches would be easy to find. Take the size from A123 15Ah and 20Ah cells and only vary the thickness or make a CEO meeting to find new format i would vote for 120mm*80mm thikness vary :mrgreen:

From who you know about the <0.01% QC fallout?
Click to expand...

First I would advise learning the various steps in learning how a high energy density pouch cell is made (accordion fan not a spiral wrap). Second, the tab joint process doesn't work well past about 1cm thickness. A few places figured out methods to go slightly thicker, but it's not going to be a very good idea just to deep draw the corner of the foil material deeper, because most deep pouches already have pinholes on the outer PE laminate material already (never make a design that doesn't assume your outer pouch surface isn't pin-holed somewhere).

As far as making a new standard size, I wish somebody would just settle on SOMETHING for a standard size for large format pouch cells. The foil current density and thermal optimization for a pouch cell isn't something overly tricky. I would love to see standards made for an optimal aspect ratio/thickness standards. Something like small, medium, large formats. Then automated pouch cell foil die cutters and separator roll production, and tab/hot-melt strips, Ultra-sonic welder anvils, vacuum purge and de-humidification ovens and metered solvent setups and formation racks and post formation pouch die-cutters and post formation vacuum sealing equipment could all be off-the-shelf equipment like it is for 18650's instead of every piece needing to be engineered and developed and refined with many $$ prototypes like it currently is to setup each new automated line for each new pouch cell. The smallest fully automated pouch cell line I've seen would have fit a football field inside its perimeter, and even then I've never seen one that didn't just use humans to handle mounting the cells to formation charging racks individually by careful human hands with little cell carrying trays and slick double kelvin clip tab clamps to get an incredibly precise ability to evaluate Ri during formation. 18650 equipment comes in every size and flavor you could want, from a tiny setup made for a battery lab to do test runs on something bigger than a coin-cell, to plants doing tier1 production where they have arrays of Costco-sized buildings all making 18650's.
 
liveforphysics said:
..with good materials coming in they see <0.01% QC fallout.
Click to expand...

I would like to see /meet the plant manager who claims he can mass produce ANY commodity item ( even paperclips, ..let alone batterys !),.. with 0.01% qc failures.
I know they will state that kind of figure for public consumption, but its impossible to verify unless you have access to all commercial data.
The only way that those levels could be achieved is if you set ridiculously wide tolerances and standards.
....( or none at all and call yourself "Ultrafire" !)
 
I feel Luke is spot-on with his comments.

Secondary Li-ion battery tech has been primarily driven by the portable electronics industry. 18650's are used in huge numbers in laptops and (I suspect a knock-on from laptops) power tools. They are a mature technololgy because they are placed in their millions into the hands of retail consumers.

Evs must be such small fry by comparison.

Yes, pouches are used in phones and other portable devices, but not in a standard size. I suspect one of the reason manufacturers use pouches is because they can fit a battery into an awkward space, as a result they tend to be application-specific = relatively low volumes and non-standardised production equipment.
 
Unless you come up with a simple, yet high current capable means of terminating cylindrical cells, such as a spring-loaded compartment like the back of your Sony Walkman :)lol:) I'm personally inclined to go for 20 Ah or so pouches. With any number of paralleled cells more than 2, it's not impossible to swap them out.
 
That makes absolute sense for the hobbyist or perhaps the prototyper. For a mass-manufacturer a battery tab welder is peanuts and quick/cheap/consistant compared to other methods like crimping or soldering
 
Hillhater said:
liveforphysics said:
..with good materials coming in they see <0.01% QC fallout.
Click to expand...

I would like to see /meet the plant manager who claims he can mass produce ANY commodity item ( even paperclips, ..let alone batterys !),.. with 0.01% qc failures.
I know they will state that kind of figure for public consumption, but its impossible to verify unless you have access to all commercial data.
The only way that those levels could be achieved is if you set ridiculously wide tolerances and standards.
....( or none at all and call yourself "Ultrafire" !)
Click to expand...

Well you better get in touch with any one of the companies that have achieved six sigma.

.01 is a huge failure rate when we're talking numbers like Panasonic's supply contract to Tesla.
 
But .01% is like90x better
liveforphysics said:
<0.01% QC fallout.
Click to expand...
A six sigma process is one in which 99.9999998% of the products manufactured are statistically expected to be free of defects :shock: 1 reject per 500,000,000? srsly?

Six Sigma reads like scientology http://en.wikipedia.org/wiki/Six_Sigma

It's based on statistics; I find it beyond belief the actual defect rate is 1/300,000. That would mean of the 1.8 billion cells Tesla ordered, the junk ones would fit on a pallet. But .01% would fill a small warehouse..so the reality is somewhere in between, eh?
 
0.01% fallout is a very easy spec to hit for a dialed 18650 line. For a pouch cell line its an exceptionally low fallout rate (but achieveable). For 18650s its common.

For the joke/comment about ultrafire, remember, thats not a cell mfg at all, just shrink wrapped and laser marking equipment. The cells are sourced from real mfg's B-stock fallout, or more sadly, disassembled lappy packs that failed QC.
 
High C-rate 18650's have lots of foil edge joint contact points to get a low resistance joint to the can
Click to expand...

Luke, the latest "buzz" cells right now are the Samsung INR-29E, INR-20R, and the newest 24R. cell_man has begun carrying the 29E and 20R as options for his triangle packs.

I'm not even sure what questions to ask, but...do you have any thoughts on these cells?
 
spinningmagnets said:
I'm not even sure what questions to ask, but...do you have any thoughts on these cells?
Click to expand...

Yes, I think they are awesome, along with all the other Tier1 reasonably high drain and reasonably high energy 18650's getting made. Nickel Cobalt Aluminum based cells are my favorite flavor of the week to look for if you're hunting.
 
liveforphysics said:
0.01% fallout is a very easy spec to hit for a dialed 18650 line. .
Click to expand...
Sorry, but a battery line could not even make the cases at anything like that failure rate,..let alone adding all the internal battery components and assembly.
6 Sigma ! .. is like ISO Certification,...a term invented for use for "selling" your products !
And lets just think where "B" grade ( Ultrafire /Trustfire etc) cells come from if we have a real 6 Sig industry ?
 
Hillhater said:
liveforphysics said:
0.01% fallout is a very easy spec to hit for a dialed 18650 line. .
Click to expand...
Sorry, but a battery line could not even make the cases at anything like that failure rate,..let alone adding all the internal battery components and assembly.
6 Sigma ! .. is like ISO Certification,...a term invented for use for "selling" your products !
And lets just think where "B" grade ( Ultrafire /Trustfire etc) cells come from if we have a real 6 Sig industry ?
Click to expand...


You are mistaken my friend. I can't comment on the amount of fall-out in deep drawing the can's or coater/rolling related fallout, but I can tell you with 100% confidence that 0.01% is an easily achievable fallout rate for 18650's to get assembled at a Tier1 supplier facility. The fallout in the QC of the materials to feed the cell assembler machine may be higher (I don't know, I'm sure it depends on the material suppliers used), but the assembly machine itself is incredibly fast, refined and consistent.
 
RedLine19K said:
..I find it beyond belief the actual defect rate is 1/300,000. That would mean of the 1.8 billion cells Tesla ordered, the junk ones would fit on a pallet. But .01% would fill a small warehouse..so the reality is somewhere in between, eh?
Click to expand...

Keeping it in reality... those 1.8bn cells will be spread over 4 years, which actually means production rate of about 1.3 million per day,
So even a 0.01% reject rate would only mean 130 cells per day ..not even a bucket full of rejects from this massive manufacturing plant ! :shock:
your 1 ln 300,000 rate would mean only 4 cells rejected in 24 hours of continuous production !
..That is all fantasy even in the real world of modern automated manufacturing!
PS ..The Japanese are particularly skilled in generating these kind of "fantasy" performance figures ! :wink:
 
Ypedal said:
the production of soda cans was one of the most impressive fast rate production items i've watched ( that i can recall atm ) .. when things move this fast, you don't want frequent production stops !
Click to expand...

Ah!..now you are in my back yard. :wink:
I have spent my working life in that "beverage can" manufacturing industry, designing / modifying equipment, improving process controls, planning entire production plants, recruiting, training and managing production crews.
They are some of the fastest manufacturing processes in industry. 3000+ units per minute from a single line, with multi line plants producing 10+ m units per day 24/7 non stop continuous, automated, production 355 days per year.
The cylindrical battery case manufacturing is a direct "spin off" from the production equipment developed for the can industry
 
Hillhater said:
Ypedal said:
the production of soda cans was one of the most impressive fast rate production items i've watched ( that i can recall atm ) .. when things move this fast, you don't want frequent production stops !
Click to expand...

Ah!..now you are in my back yard. :wink:
I have spent my working life in that "beverage can" manufacturing industry, designing / modifying equipment, improving process controls, planning entire production plants, recruiting, training and managing production crews.
They are some of the fastest manufacturing processes in industry. 3000+ units per minute from a single line, with multi line plants producing 10+ m units per day 24/7 non stop continuous, automated, production 355 days per year.
The cylindrical battery case manufacturing is a direct "spin off" from the production equipment developed for the can industry
Click to expand...

So what is the failure rate on a can line.
I assume most faults are at the labeling end.


Don't correlate failed cells in a battery pack to badly produced cells.
That's like blaming the plant when when you open in a soda can it's all over your shirt.
There are A LOT of steps from production to consumtion.
 
You must log in or register to reply here.
Back
Top