DIY Electric Car Forums banner

1 - 5 of 5 Posts

·
Registered
Joined
·
2 Posts
Discussion Starter #1
Howdy folks! I live in Portland, Oregon, and am considering a bargain-basement EV conversion. You know, buy a Metro off of Craigslist with a blown engine, drop in a cheap DC motor, and go! I was expecting only an errand-runner with a range of 20-30 miles and a top speed of maybe 40 miles per hour. My back-of-napkin calculations incline me to believe that doing this much is a) plausible, and b) cheap. (I thought I was the first to think of scrapping an old forklift, ha ha!)

So, the seeds of this idea come from when I found one of those DIY home battery bank forums, a year or two back. It seemed like it was right up my alley - scrounge dead laptop battery packs, do some careful fiddling and some math, get Free Energy.

Yeah, right.

But I was too busy with a three-year-old, and shelved the idea for later. But now I have a five-year-old! He is in kindergarten, so I have a chance to restart my freelance career, and maybe scrounge up some Cheap Energy Storage.

So I have been pricing out low-cost batteries for aforementioned EV conversion, and I am having some problems running the numbers comparatively. When I look at y'all's battery forum, everyone seems to be in agreement that taking apart a battery from a commercial EV (a Volt, right?) from a scrapyard is the price/performance sweet spot. I was kinda expecting to find what I'd found in the DIY powerwall forums - a thriving market in rescued-from-the-landfill 18650s.

So, here's what I am currently stuck on. If I was only working for an hour or two a day, I would have plenty of time to go on the battery-pack-scrounging routine. But I'm not - I am actually getting decent amounts of work, at what I think of as good hourly rates. So I have to decide how much my time is worth, right? If I can assemble a battery for my EV conversion by spending an hour a day (pick up scrapped battery packs for the going battery-scrap rate, take the packs apart, test, recharge, reassemble) then why would I want to instead work for an hour, get paid, then turn that money around and buy the scrapped Volt pack? What if I was buying tested known-good recycled 18650s of, say, two amp-hour capacity at, I dunno, fifty cents each? What about eighty cents each? How can I compare those numbers against my hourly rate, and against the cost of the scrapped Volt battery?


For the purposes of this question, let's assume that my time is worth $20/hour. It's worth rather more than that, but I have an old client with easy, fun work who I've not raised the rate on since 2004. I'm getting maybe four hours a week of that work, and while I am not super eager to fire that client, I have been warning her since like 2010 that I'm all grown up now and that she should look for some other young person to do her data-filtering for her. ("I'm a middle-aged dad!" I say. "You're a kid!" she responds. She's somewhere in her eighties, so I guess she's right.)

If that $20/hour is better spent sunk into a cookie jar marked "Chevy Battery" then I'll keep doing her data-filtering. But if those four hours per week could net me a stack of 18650s that would be enough to power my Metro, then I will start scrounging straightaway.

But I don't want to grow up to be a creepy battery hoarder. I am kinda hoping that buying tested 18650s from a creepy-yet-reputable battery hoarder might be the cheapest and fastest way to go. How should I start this calculation? Dollars spent per watt-hour?
 

·
Registered
Joined
·
4,791 Posts
Sorry, I don't have any suggestions for your value-per-hour estimation; however, there is another factor that you may not have considered. Once you have a pile of one to two thousand 18650 cells - at whatever cost in cash and time - you still don't have a battery. You need something to hold all of those cells, plus a few thousand connections to make - what will that cost in cash and time?
 

·
Registered
Joined
·
1,106 Posts
Please talk me out of being a creepy battery hoarder
*eyes several hundred pounds of 18650s and several hundred more to-be-tested 18650s*

I... umm... hrm. What are you implying?

I don't want to grow up to be a creepy battery hoarder.
You take that back!

How should I start this calculation? Dollars spent per watt-hour?
Yes, dollars per watt-hour (kWh is probably a more sensible range) is the metric you'd use to measure energy.

The power (how fast you can get at the energy) is a different story, if you only want to go 20-30 miles, that means at highway speed your pack is going to be drained in 20-30 minutes (else you'd travel farther), which is 2-3C. 2-3C for laptop batteries. That's aggressive for those cells. Anything above 1C for those cells might be pushing it.

That said, a piece-of-shit Geo is only going to use maybe 300 watt-hours per mile at highway speeds. So you only need 300Wh/mile * 30 miles = 9000 watt-hours. With a whole Volt pack you'll have nearly double that. When you double capacity, you double the power along with doubling the energy (because there's twice as many cells, so each is only getting emptied half as fast). So, with 16kWh of 18650 laptop cells, you're on the edge but it should be possible. Also you're going to have like, 60 miles of range as a consequence. Bonus.

A Volt pack is ~16kWh and you can buy one basically any time you want for $1600. So, $100/kWh, dats quick mafs. That's new watt-hours, but, presuming watt-hours degrades the same, we'll stay consistent and use that for both types of used cells.

An 18650's nominal voltage is 3.75v, and cells range from 1ah to maybe 3ah on the upperer end. Average of 2-2.5. That gives you around 8 watt-hours per working cell.

To match the 16kWh Volt pack, you'd need 2000 cells.

If you can buy a volt pack for $1600, then you'd need to buy 2000 working cells worth of 18605s for $1600. $0.80 each. I stress "working cell" because every pack is going to have at least one bum cell.

Also, the Volt packs can pretty much be used as-is. The 18650s you will probably want to buy plastic grid holders for, maybe heat-shrink sleeves, and need to buy copper wires/bus bars and several pounds of solder to wire them up.

I am kinda hoping that buying tested 18650s from a creepy-yet-reputable battery hoarder might be the cheapest and fastest way to go.
It's not.

It's not cheapest or fastest.

You probably cannot purchase 18650s for less than you can purchase Volt packs, and you sure as hell can't process them anywhere nearly as quick, you're looking at months for 16850s.

I had a free bottomless source of 18650s (source ran dry due to operational changes after a year just now), and *I* think it's not worth it. I'm still doing it, because I'm at least 2000 cells and 6 months deep into it, but, yeah.

Many people talk about making EV batteries from 18650s (or used to, 3 or 4 years ago), but I haven't seen anyone actually finish one I don't think. Not even from new cells, let alone salvaged ones. I saw someone make a booster cell, but that's only a couple hundred cells. Far as I know (please anyone link me if I'm wrong, I'm curious to see their approach), no one has built an EV battery from 18650s. Bicycles only. I've only ever even seen motorbikes built using the fatter cells (mini-coke bottle-ish sized).

If the DIY Powerwall (secondlifestorage or whatever now) community had any clue, they'd be buying up EV packs instead. But, the bro-science permeates and people "get" 18650 recycling, it seem like an approachable methodology. Also don't tell them, they'll bid up the price of OEM EV packs.


Some numbers:

I think I collected for 2 months, processed for 6 months, and then have let the rest accumulate for 4 more months now.

In 6 months (maybe less, no fewer than 4 though) I processed around 2000 cells, and that's cutting every corner. I built a pre-charger and post-charger, so that my capacity testers are only doing the discharge capacity test (they're the bottleneck), not charge and recharge. That sped things up by 300%. I bought 2 capacity testers that have 4 independent 18650 chargers each. They're around $40-50 each I think, I might've got them on sale. Maybe $75 invested, plus the DIY charging rig, to do 8 at a time.

They discharge at 0.5A, so an average cell is a 4-hour discharge cycle. 24 hours in a day, so if you set an alarm to wake up and swap out while you sleep, and miss no cycles, you'll process 6 batches in a day, 8 at a time, so 48 batteries a day.

Absolute minimum, just in time elapsed with 2 of those chargers, you're going to take 41 days. That's if you happen to have a job or lifestyle where you can have your batteries handy to swap out, perfectly, every 4 hours.

I did pretty good, I didn't set an alarm to do through-night processing, but only occasionally missed cycles or lost some because they finished an hour after I went to bed and then sat for 7 hours doing nothing, and it took me about 4x the absolute minimum, Which was around 6 months, and I felt like a slave to my capacity rig the whole time. I ended up with the amount of batteries you need.

Now, not a lot of my day was spent actually doing work on the batteries, so I don't know how much it counts as "work", but, I couldn't stray from them for more than 4 hours.

Actual disassembly I probably spent an hour or two a day (less at the end when I was a salvaging machine), and I was doing tool packs which are much more difficult to salvage than laptop packs. Call it 250 hours. For me, this was about the most time I could squeak into my schedule here and there, so it didn't make too much sense to invest in more capacity testers to increase my throughput. I couldn't have kept supplying them much faster than I was.

I should've kept better track of my salvaging metrics and I'm sure I could dig them up (I still have all the bum cells), but, expect something like 10-30% dead cells. (so 2200-2600 cells to process to get 2000 working ones).

A dozen of friends once processed a bin full of laptop cells (just to yank out cells), we ran the numbers if you had a bottomless free source and how much time that took, and, it's actually quite poor of an hourly wage. I think it broke minimum wage but not by much.

If laptop packs were as slow as tool packs to disassemble, and if you're as slow as me, suppose it takes you 250 hours to do the same 2000 cells I have. And your actual capacity testing time and time writing capacities on the cells is not counted in that.

When you're done you'll have to lay out and solder all of them together, probably dozens more hours we're not counting.

$1600 for a Volt pack, divided by 250 hours to do the processing = $6.40/hour.

Which is a lot less than firing the octagenarian currently given a sweetheart deal from you.

HOWEVER...

This is work you can do literally any hour of the day, in as small as maybe 3 minute chunks. It has very little mental load so, there's nearly zero ramp-up or multi-tasking penalty. So, if time-shifting gives you new opportunities, maybe it's still worth it to you.

Anyway, you have the metrics you require and your error bars to figure out whether this is worth your time or not. (HINT - As above, IT'S NOT).

The DIY EV community generally has 2 types of people:

1 - Cheap people who drool at the idea of never paying for gas.

2 - Performance minded people who drool at the aspect of VROOM.

There are plenty of people as cheap as you and I, who've been doing this for years. No one to my knowledge has actually completed an 18650 EV project on a full-scale like this. If the kind of people that'd pick peanuts out of poop to save a buck aren't doing this, there should be red flags for you.

But whatever, like anything, if it sounds like fun to you, do it anyway.
 

·
Administrator
Joined
·
6,120 Posts
What an awesome reply!

I did it the other way - a Chevy Volt pack - and it's not only easier and more vrooom

But the engineering is absolutely excellent! - and there are a lot of other goodies in that battery pack
 

·
Registered
Joined
·
2 Posts
Discussion Starter #5
Yeah, Duncan, I have to agree, that was a top-notch post. Consider me impressed. Also: as per your request, Matt. I take it back. I- I- I want to be a creepy battery hoarder! I do I do! You make it sounds so... compelling.

Basically, you told me that you amassed 2000 cells in six months of not quite obsessive-compulsive work done in little five-minute dribs and drabs of effort. And then you told me that, in order to amass the kilowatt hours to power an old crappy Geo (or, more to the point, to equal one Chevy Volt cell), I would need: 2000 cells. In fact, I did write the question "So, if I had an infinite supply of decent scrapped cells, and worked in a diagnosably obsessive-compulsive pattern, and furthermore was a freelancer who was home five days out of seven... how long would it take me to amass enough cells to do an el-cheapo repower of a Geo?" But I felt that my post was too long, so I deleted it from my draft. So yeah, thanks for reading my mind and answering questions that I didn't even ask. Top notch posting there, buddy.

That said, a piece-of-shit Geo is only going to use maybe 300 watt-hours per mile at highway speeds. So you only need 300Wh/mile * 30 miles = 9000 watt-hours.
Yeah, well, I don't need highway speeds, so that's a non-issue. Also, while the extended range would be nice, I think that you've specced out a battery that is more that enough for my expected range and speed.

I had a free bottomless source of 18650s (source ran dry due to operational changes after a year just now), and *I* think it's not worth it. I'm still doing it, because I'm at least 2000 cells and 6 months deep into it, but, yeah.
I will keep this fragment of your posting in mind, should I ever encounter a well of infinite recycled lithium batteries to process. As it is, I think I'm more into drumming up new clients than I am into drumming up sources of toxic trash to recondition. I am still interested in the whole battery-scrounging thing but maybe I will set my sights lower (e-bike, e-skateboard, portable audio, other tasks requiring less time investment and, most important to me, less fiddly soldering). So in order to buy the Volt battery, I will instead invest my time into drumming up another corporate client that I can bleed at my normal full rate, and let them pay for my Real Battery.


Many people talk about making EV batteries from 18650s (or used to, 3 or 4 years ago), but I haven't seen anyone actually finish one I don't think. Not even from new cells, let alone salvaged ones. I saw someone make a booster cell, but that's only a couple hundred cells.
It was, in fact, a booster pack made out of 18650s I saw on YT that motivated me to post my question. I looked at it and said "Yeah, in an old VW it's a booster pack, but put it into a stripped Geo and it should be enough pack all by itself."

If the DIY Powerwall (secondlifestorage or whatever now) community had any clue, they'd be buying up EV packs instead. But, the bro-science permeates and people "get" 18650 recycling, it seem like an approachable methodology. Also don't tell them, they'll bid up the price of OEM EV packs.
EXCUSE ME DUNCAN, PLEASE DELETE THIS POST










;)
 
1 - 5 of 5 Posts
Top