Pack building business advice - What do I need?

katou

10 kW
Joined
Jul 22, 2009
Messages
791
Location
Toronto
I would like to set up a business in Canada building and selling battery packs to Canada, US and Europe. I know that most kind souls will try to dissuade me from doing so, with the best of intentions.

That said, I'm hoping to find a way to comply with regulations, and minimize liability to the point where a company will insure the business.

Can anyone point me in the direction of an existing pack manufacturer that might advise me?

I also need to track down the relevant regulatory information, but I'm not sure where to start.

Katou


Ps. I also need a business name that doesn't suck. Battery Baron? Lightningbox?
 
Wow, crickets!

Someone out there must know someone who builds packs for pay?

Heck, I'm contacting Ezee and Allcell just to get someone to talk to.


Katou
 
first thing is investigate the cost of your liability insurance and the canadian shipping regulations. see how much it cost for import duties to other countries.

then find a source for cells that is less than it cost to buy them in a manufactured battery already assembled from china. then find a supplier of the BMSs who will sell you product at dealer cost and not the reseller cost.

then guess how much you can sell them for and make a few cents/hour for your labor and post it up on ebay and see how many people will buy them at your price. if you can sell one then go ahead and order everything and build it if there is a buyer.

then after you sell it and the buyer refuses to pay because they decide it is not what they expected and uses paypal to take their money back you can see if you can then relist it on ebay and sell it again. if they even send it back.
 
dnmun pretty much nails it. Get ready for massive headaches and customers who will spit in your face at the first sign of any trouble at all. OEM's are demanding, but they're generally rational and will listen to reason supported by data. Hobbyists, not so much. Sad and frustrating, but true.

If you plan to build packs that people can't build for themselves, I would think you would need a welder of some sort first off. What type(s) of cells are you working with? This will drive many other decisions you need to make. Allcell seems like a pretty good outfit that works with lots of cell types. Looking their stuff over may be a helpful primer. Make sure you fully understand the liability and regulatory problems before getting to far along--this is daunting stuff, especially for somebody starting from scratch. Are you authorized to ship hazardous materials? Do you have UNDOT-approved packaging for the packs you plan to sell? You're in for $5,000 right there, just to get a box defined and tested before any boxes can even be made or sent to you. You may need to do this procedure for every unique product you ever ship. These things are a big deal and stymie even the big players at times.

Good luck!
 
Good stuff, thanks.

The BOXES have to be tested? Wow. I'm surprised that companies don't just choose already tested boxes, instead of testing new ones.

I'm not terribly surprised though. I've had some dealings with a friend trying to manufacture exterior lights. Each design has to be tested, about 1k$ each.

Dnmun, you're plan is very scalable, I like that. I am about halfway through your list.

If either of you can refer me to the actual regulations (or their names) I would be grateful.

Katou
 
Look up the various regulations for handling class 9 hazmat materials. UNDOT is the main body governing international shipments. Most nations also have their own regulations, making for a nightmare patchwork of discontinuous rules that will make your head spin. Do realize that some countries forbid the export of ANY cell or battery system deemed to be defective. Germany and China are among them. A123 had to set up remote warranty labs in those countries for this reason. You can be subject to fraud under this regime if you don't have local investigators or people to receive the "defects" in countries that have this rule. You can ship almost anything within North America by truck or boat. Air is more restrictive, but still very usable.
 
katou said:
Good stuff, thanks.

The BOXES have to be tested? Wow. I'm surprised that companies don't just choose already tested boxes, instead of testing new ones.

Oh, the box is custom to your product. So, unless your item can be shown to be identical to somebody else who already has approved packaging, you're pretty much screwed. I'm not sure off-the-shelf UNDOT packaging even exists. Very small packs (like for an RC car) might be easier in this regard, but that doesn't sound like where you're headed.
 
I really appreciate your contribution wb9k! You seem very familiar with this stuff. Do you work in the industry?

It will take me some weeks to process what you have given me so far. I would love to hear any further ideas or advice you might have, if anything crosses your mind over the next weeks.

Katou
 
katou said:
I really appreciate your contribution wb9k! You seem very familiar with this stuff. Do you work in the industry?

It will take me some weeks to process what you have given me so far. I would love to hear any further ideas or advice you might have, if anything crosses your mind over the next weeks.

Katou

Yeah, I run the warranty lab at A123 in Livonia. It could be that there is off-the-shelf UNDOT-approved packaging for small packs meeting certain conditions, so do some research there before just taking my word for it. Even if it exists, it wouldn't have been an option for what I have needed in the past.

For more insight into the "spit in your face" comment, see this recent thread: http://endless-sphere.com/forums/viewtopic.php?f=14&t=67941

Are you sure you want to get into this biz? :|
 
Well WB, let's just say that my ongoing reading has REALLY explained why people sell individual cells!!!

Warranty centre eh? Well, you are uniquely positioned to help me with negotiating one of the many regulations I've been reading about - specifically, the prohibition on shipping cells that have not passed QC testing.

Cells from used sources is one idea that am exploring. I have access to a large supply of used tool batteries.

According to the "reject", definition, there should be no problem, they are not defective. (They already passed QC to get sold right?) On the other hand, of course they're defective, that's why the pack got chucked!

Like Dr.Bass's findings, most of the cells I have run into do not have major problems. There are cells that have been damaged by corrosion, and some have decreased capacity, but they seem fine in other ways. That said, I am learning new ways to test them, so my understanding of their shortfalls will grow over time.

Some of the packs failed because the internal BMS faulted, DOA, so the cells are untouched other than shelf wear.

I will need to start load testing soon, to ascertain if C-rate still meets spec. If a cell meets all relevant manufacturer specs, then what is it? New old stock? Refurbished? Remanufactured? Used?

Granolaboy, great find! I have found a bunch of stuff myself, I will post what I've found for all to share when I get a chance.

Katou
 
you can make more money working on a roofing crew or pouring concrete. there is no profit in this stuff and only nagging buyers who demand perfection or else they will trash you all over this place.
 
dnmun said:
you can make more money working on a roofing crew or pouring concrete. there is no profit in this stuff and only nagging buyers who demand perfection or else they will trash you all over this place.

I think that's a bit of an exaggeration, but there's some truth in it. Grading used cells of unknown provenance is time consuming at the very least, not to mention all the teardown and rebuilding of packs, overhead needed in the equipment to build and test, shipping concerns, etc.. This is not a "streets of gold" kind of opportunity...it takes commitment, planning, and some level of serious engineering. I wouldn't expect to make much money soon. People are skittish of the nicest folks in the world in this realm, unfortunately that's for good reason in some cases. Really good buyers are hard to come by, though everybody and his brother seems to be interested.

I do think there's a big future in repurposing of big packs that are at end-of-life for automotive purposes. Processing will be complex, but the scale (and ongoing flow, as time goes on) of the incoming pieces (as opposed to tool packs, for example) means larger residual value with less engineering and busy-work input from you, the re-conditioner/re-seller.

This discussion could go on for a very long time....
 
katou said:
Well WB, let's just say that my ongoing reading has REALLY explained why people sell individual cells!!!

Warranty centre eh? Well, you are uniquely positioned to help me with negotiating one of the many regulations I've been reading about - specifically, the prohibition on shipping cells that have not passed QC testing.

Cells from used sources is one idea that am exploring. I have access to a large supply of used tool batteries.

According to the "reject", definition, there should be no problem, they are not defective. (They already passed QC to get sold right?) On the other hand, of course they're defective, that's why the pack got chucked!

Like Dr.Bass's findings, most of the cells I have run into do not have major problems. There are cells that have been damaged by corrosion, and some have decreased capacity, but they seem fine in other ways. That said, I am learning new ways to test them, so my understanding of their shortfalls will grow over time.

Some of the packs failed because the internal BMS faulted, DOA, so the cells are untouched other than shelf wear.

I will need to start load testing soon, to ascertain if C-rate still meets spec. If a cell meets all relevant manufacturer specs, then what is it? New old stock? Refurbished? Remanufactured? Used?

Granolaboy, great find! I have found a bunch of stuff myself, I will post what I've found for all to share when I get a chance.

Katou


Sorry, realized I didn't answer you on this. Once you test a used battery and deem it to be functional, it is "used". As I understand the regulations, there are limits on the shipment of cells deemed to be "defective". For the purposes of shipping, "defective" means you are either leaking electrolyte, or have an imminent elevated risk of thermal excursion, up to and including fire. The limits on shipment of defective batteries vary widely from country to country, but generally speaking you cannot ship defective cells by air at all. Most times you can ship by ground, but some countries forbid the export of defective cells altogether--Germany, for example. Bad cells can be shipped within Germany by ground, but a single battery can cost up to $2,000 to ship within the country because of all the required precautions--so you see just how onerous this can get. Cells that work OK are far easier. Not sure how you would have to handle shipment of cells you may buy for testing, given that their condition is probably "unknown". I would think ground will be the only option most times. Hopefully you've got channels set up from which you will regularly receive cells so this doesn't get too daunting. "New old stock", btw, means an unused product that is a few years old or more--I've sold "NOS" items that were upward of 100 years old (vacuum tubes, not batteries). Used cells/packs may fit the descriptions of "repurposed", "refurbished", "remanufactured" or "used"--whatever makes the most sense. Hope that helps.
 
Calibike appears to be one individual who sells battery packs from California. Since he's doing it, maybe you can too.
 
Thank you WB, that's really good info.

In the warranty centre, how do you ascertain the remaining life left in a cell?

Thanks GB, I will reach out to Calbike and see how they handled the issue.

Katou
 
We rarely see cell problems. Mechanical and electronics issues are far more common. When we do get cells, capacity is rarely the concern, so we don't often find ourselves measuring capacity. When we do see cells, we're usually either quantifying a high self-discharge situation or going straight to teardown of a cell that's already toast.

When we were in full swing with a well-established program a couple years ago at our third-party warranty center, we had cyclers that would test both capacity and impedance of battery modules. You could do the same testing at the cell level, but that would be rather expensive for the potential return. Look up the Hymotion module description (an MS Word document) I have posted on ES a couple times. That doc details the test I'm talking about and shows graphical results.
 
I've built packs for e-bikes, flying machines, go karts and electric superbikes. I think I've made about 35 kWh worth of LiPo battery since joining this board. It's possible, but there are a few things you absolutely must get right.

Packaging. Find yourself a good plastics fabricator or come up with an easy injection mould and build to that specification. Outside of a failed component or connection, the next most common issue is physical trauma to the pack. I use pouch cells which are very fragile and need to be protected from even the most mild of elements (fretting against a plastic surface for example).

Idiot-proof. The battery package needs to be so easy to use that an 12 year old can work it out. Ideally there will be a BMS which stops it from being drained too far, and prevents it from being over charged. Idiots will do all sorts of stuff, like go for a 30 km ride and park it up for the winter... Folks will have issues, and expect to eat that sandwich when a pack fails.

Repairable. If the battery is set in epoxy or has all the critical components underneath something that takes two hours to pull apart, you're in for a lot of difficulty. Use connectors between the pack and the BMS, and the BMS and the power outputs. If the BMS fails, have a spare ready to swap in.

And yes, you'd make more money and have less stress by flipping burgers, but that's not what motivates us :)
 
Thanks WB, for the Hymotion reference, I will look that up after I post this.

I've been reading about graphing cell impedance during charge and discharge as a predictor of cell life. Now, how to do that? I've been looking for a description of how to do a true 4-wire measurement of internal resistance, but I haven't found one yet. The postings I've read either suggest using a RC charger (but accuracy seems to suffer) or explain the process incompletely. I'm not an electronics idiot, but when people start talking about rds on and replacing mosfets, I get lost.

I wish I understood how to get a truly reliable measurement for internal resistance. Until I figure that out, I'm investigating using an infrared thermometer to find cells that deviate from the norm by a degree or more. It appears that the temp increase is most reliable as an indicator of cell life during charging compared to during discharge.

Given that injection molding is so expensive on the front end, I've looked into vacuum forming. Acrylic won't form worth a toot, but polycarbonate (my choice for strength anyway) vacuum forms beautifully. Which parts were you thinking of using the injection process for? The case overall?

The weakest link seems to be the BMS. I read a lot about the various units for sale, and all of them seem to have big problems, most of which seem awful technical. Am I just hearing all the bad stuff?

Jonescg, what BMS's did you put in your packs? There HAVE to be good options out there!

Thanks all, really good info.

I have 200 cells that I will be assembling as a proof of concept. Most are charged and resting, waiting to see if they hold their voltage while I peel more packs and finish charging the last of the 200.

Charging the cells has been a bit interesting, I set the CV power supply to 3.6 and it overshot 2x to 3.7v. I read a battery document about A123 that suggested that the energy capacity is very low between 3.5-3.65v. This suggests that the overshoot may be a consequence of setting the power supply to 3.6v. I am now setting it to 3.4 to get the majority of the cell full, then re-setting it to 3.6 to slow down the charge.

Work continues at a snail's pace, but no step is wasted if you are moving in the right direction.

Katou
 
Yeah, whatever you make the battery pack from, it will need to be protected from fretting, rubbing, road rash, shaking and the like. Best idea is to find a way of including the BMS inside the same enclosure, as a separate compartment. If you want to make batteries for the mass market, an injection mould is the only way to go.

Lots of BMS options out there. I've been using 12s Modules from EV-Power.com.au however these don't have a high cell voltage trimming feature. Which is fine cause the cells need very little balancing, but I do it from time to time to be sure. The next revision will have a 100 ohm resistor for trimming a high cell down over a few hours.

Bestech in China makes a great variety of complete BMS packages which trim high cells, prevent over discharge, prevent over charge and the like. Fairly affordable units too.
 
Back
Top