Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Some cell harvesting questions
#1
I have harvested my first 50 cells or so, and I have some questions. 

 
First, I pulled apart a battery bank, it had already had the front, and back panels, and BMS removed, it was basically a 3S4P cell pack in an aluminum case with all cells at 2.62V.  If I didn’t know any better I would say somebody tried to harvest it and couldn’t figure out how to get the cells out.  I ran a knife between the cell wrapping and the case to break the glue, and pushed the cells out.  I took one 4P pack apart and tested the cells, all 4 were perfect with 100% capacity and .04V self-discharge after a week.  I charged the other two 4P packs for testing while still assembled.  The packs have no self-discharge issues; they test the same as the individual cells.  So, the question is should I take the other 2 packs apart, or do capacity testing as a 4P pack?  I don’t have a spot welder, and removing the strips just to replace them with solder seems to invite trouble.  I’ve read here before of others leaving parallel packs intact if they pass self-discharge testing, and I’m curious about the groups experience with this is.
 
Second, I’m curious as to whether there is benefit to disassembling the pack ASAP.  I have noticed people with hundreds of cells, removed from packs and awaiting testing. 
 
Third, when you go to pick up packs from the recycler, and he drops a 55Gal drum of misc. packs in front of you, how do you pick which packs are worth the effort to harvest and test?  My first check is to look at SOC meter if it has one, if it still responds, it goes in my harvest bin.  Next I look for dates that would give me a rough age, but beyond that it is kind of hard to tell what to pick.

Thanks again for your help.
Reply
#2
I pick all packs i get that have 18650 cells. It takes to much time to sort at that point and i sort them when i disassemble
I like to do larger ranges aka dissassemble alot and then go for testing...

I capacity test ALL cells no matter. If the tester runs them they runs them. If not I bin them. I do not care about initial voltage nor IR because if the capacity is good enough at 1A discharge current that shows me IR is low enough and they will work for me.

No second chance, nothing... Just one test. Fail or pass.
I have 29 000 cells running now where over 6000 of them are from laptop packs!
The Ultimate DIY Solar and build place
YouTube / Forum system setup / My webpage  Diy Tech & Repairs

Current: 10kW Mpp Hybrid | 4kW PIP4048 | 2x PCM60x | 83kWh LiFePo4 | 10kWh 14s 18650 |  66*260W Poly
Upcoming: 14S 18650~30kWh | Automatic trip breakers, and alot more
Reply
#3
When I have many pack to work with I do like this:

1. Measure cells/pairs to exclude 0-volt ones (less disassembly work). This can save quite a lot of time if there are many bad cells in the packs.

2. Disassemble everything.

3. Initial triage considering existing cell voltage: 3V+ - tier 1, 2-3V - tier 2, 1-2V tier 3, under 1 V initially ignored (but I do keep them to be processed later).

4. Full charge of cells depending on their tier. T1 can be charged with any charger while T3 requires chargers that can charge even if the cells is over-discharged.

5. During charging, I filter out the overheating ones. Those are given a 3-day break, than recharge is retried (this can repeat up to 3 times but in most cases they end up recovered and usable).

6. Capacity measurement. I put the charged ones in capacity testers and give that a go. Since the Opus testers require additional cooling to continuously work, I can safely put fans on them because I already filtered the overheating ones in a previous step.

Now this procedure is highly efficient since I do not keep my capacity testers busy with initially charging the cells. After the capacity reading is displayed, I remove them and charge them back at a normal chargers so the Opuses are available for the next lot.

After all of the above is done and I have the capacities of T1, T2 and T3 cells, I pick on the ones under 1V to see if they are recoverable and work on the initial overheating ones.
Reply
#4
(05-24-2019, 12:16 AM)Shawndoe Wrote: I have harvested my first 50 cells or so, and I have some questions. 

 
..........I’m curious about the groups experience with this is.
 
Second, I’m curious as to whether there is benefit to disassembling the pack ASAP.  I have noticed people with hundreds of cells, removed from packs and awaiting testing. 
 
Third, when you go to pick up packs from the recycler, and he drops a 55Gal drum of misc. packs in front of you, how do you pick which packs are worth the effort to harvest and test?  My first check is to look at SOC meter if it has one, if it still responds, it goes in my harvest bin.  Next I look for dates that would give me a rough age, but beyond that it is kind of hard to tell what to pick.

Thanks again for your help.

I see you have had your first harvest. Smile

You need to be careful when you are fishing for cell harvesting info and procedure as fish like me will bite the hook to the dismay of some of my peers. Tongue

I will answer your third question first as it is after all where the harvest begins.
Choose your packs wisely, as in quality medical packs if available, and of course laptop packs. Power Tool packs are great if that is all you would be using but are of different chemistries and I would not mix them in with medical and laptop packs.
All your major brand laptop packs are fair game. Stay away from "replacement batteries". Look for the manufacturer's logo on the pack. If no logo let it go. Check the mAh rating of the battery and it will give you an indication of what mAh cells are in it. Look at the size of the pack and you can pretty much guess if it is a 2p, 3p ,4 p etc. Then just take the mAh rating and divide by the p value and you will know what cells are in it.
Example: A 6600mAh cell that looks to be a 3p will have "you guessed it" 2200mAh cells also a 2p rated at 5200mAh will have 2600mAh cells.

Second question:
Taking packs apart as soon as you get them. I really dont think it matters whatever damage the BMS has done if any is done. Just because you take them out of the case isn't going to change anything unless you have a battery that "just" failed and you just got it. When you get 200 to 250 packs at a time it takes some time to liberate all the cells. Also I crack the packs as soon as possible to do my initial testing and then sort the cell within my criteria which I will describe next.

Much to the chagrin of most here I am the "IR" guy.
Internal resistance as far as I am concerned is the most important measurement you can make on a reclaimed 18650 battery, voltage being the second. I usually break apart 75 to 100 packs at a time setting the "guts" of the packs aside and disposing of the plastic carcass. Remember that the plastic is recyclable and to dispose of it properly. You may have some cells with physical damage IE rust, leaks, etc. as soon as I notice that I tape the pack back up and those go into a recycle box. Easier to get rid of as a unit. After I have had enough wrist exercise breaking the packs appart I take the "guts" and remove the BMS and the nickel strips. I personally liberate each individual cell and do not keep them in pairs. Some do, I understand that, but I have found that sometimes ,although very rare, that the IR between to paired cells can be very different. After that I have ~700 cells.
I then sort the cells by manufacturer and model number. After the sorting is done I go and get my IR meter. It is a YR1030+ 4 wire 1kHz AC mΩ meter and I test all cells for IR in accordance with this sheet that I developed after 5000+ cells analyzed.

Anything above the marginal mΩ reading goes into the giveaway box. This process takes very little time and it eliminates all the cells that would not perform well. Sometimes it looks like you are getting rid of a lot of cells but trust me its worth it. They will just disappoint you with poor results.
The cells that have passed the IR test are then sorted by voltage into 2 groups >2V and <2V. They are then finish cleaned and as much of the cell glue and stuff removed as possible. Any Cell >2V goes directly to a charger tester and tested. Any Cell <2V gets put aside and when I have enough (20) they get parallel charged at CC/CV at 4.2V and 50mA per cell for however long it takes for the parallel pack to come up to ~4.18 or so volts. Sometimes it takes 3 days but they will come up. Occasionally checking for stealth heaters with my thermal camera. There are some other things that can happen here with vamp cells, cells that won't let the voltage of the pack come up, but for the most part this works very well in waking hibernating cells up.
If you stick to within the IR chart above you will find that your results will be something like this.


In the final analysis it depends on how you want to do it. If you have resources like Daromer and just throw cells at a charger/tester
(which has worked for him very well) without any preliminary checking so be it.

I do not give heaters a second chance and that may be due to my great supply of packs. But in the long run I don't want a potential heater in my packs anyway no matter if it comes around on the second or third try.

After 7 months of  harvesting cells and refining and maturing my harvesting strategy this is the procedure I have come up with and it works very well. You are welcome to look at my google sheets and look at the Harvested Cell Analysis.xlsx sheet and you will see as you go down the dates 
that after 3/17/2019 when I started to adapt my IR philosophy the results are very good. Mostly ≥80% capacity results. Several tests to push the upper IR limit of some cells but all in all.... well you have to look at it to understand.

Wolf
Shawndoe likes this post
If 18 X 650 = 2200+mAh then we have power! 
May all your Cells have an IR of 75mΩ or less Smile
Last count as of 8/7/2019
Total Number of Cells Recorded and processed                 6149
Total Cells required for PowIRwall                                   2856
Total Cells ≥2200mAh, ≥80%, ≥35mΩ, ≤75mΩ, ≥4.12V   2760
For Info Google Drive
Not your average Wolf       
            Cool
Reply
#5
Quote:Look for the manufacturer's logo on the pack. If no logo let it go.
I found that out with my last run to EH&S, I pulled a couple of generics, one laptop, and one power bank.  The laptop cells were 0.01V and so generic I couldn’t get any info from them, so I binned them on “Trust” issues.  The power bank had cells that tested as borderline self-discharge, and 700mAh tops, so I binned those as well. I did find a few medical packs in my run yesterday.  They look like 3S1P packs for a HeartStart defibrillator unit.  I have yet to get one apart, but I look forward to what I might find inside, the install by dates are all within the last year.

Quote:Taking packs apart as soon as you get them. I really dont think it matters whatever damage the BMS has done if any is done. Just because you take them out of the case isn't going to change anything unless you have a battery that "just" failed and you just got it. When you get 200 to 250 packs at a time it takes some time to liberate all the cells. Also I crack the packs as soon as possible to do my initial testing and then sort the cell within my criteria which I will describe next.

So, it doesn't matter, but you do it anyway.

Quote:Much to the chagrin of most here I am the "IR" guy.

Nooo, not you. Wink

A good IR meter is next on my list of purchases.  I seem to remember you testing various units, do you have any recommendations, or do they all test about the same?  I wouldn't do IR testing on paired packs, only SD and Capacity, of course if you are doing IR testing as your first test, then the pack are coming apart anyway.  What is your IR criteria, do you have a flat line for all cells and manufacturers, or do you go by manufacturer spec to determine condition?

I was reading a paper yesterday about IR testing across the charge curve to diagnose different issues such as SD, Capacity, and heating issues, I'll look and see if I can find it when I get home.  It seemed promising, so the question becomes how would we set up a system to charge pause, test IR, and resume for a fixed period.  I also remember reading that the path of the leads leading back to the meter have a significant effect on the measured IR, due to EMF.

Quote:There are some other things that can happen here with vamp cells,

Vamp cells seem pretty common in packs that show cells < 1V.  I have watched them, and they rise normally, then they begin to peak, usually in the 1-2V range, then they eventually begin loosing voltage from their peak.  If I see this it is an automatic bin.

I live in a small apartment with my wife, so I tend to be conservative in my testing.  If I have any doubt about it, to the bin it goes.  I want to build a powerwall, but until I have a place I can set one up, I have to keep my goals reasonable, so I am building a 1-2kWh power box to act as a power supply and USB distribution when I am in the field with my astro gear, and as backup power and device charging for the occasional power outage at the apartment.  Meanwhile I get practice harvesting cells, and building packs, and just because I can't build a wall now doesn't mean I can't collect the cells, which is more than half the battle.
Reply
#6
When i have some pack's to harvest, i crack, break, grind. saw them open as fast/soon a possible.(incl ebike bat's)
Just like wolf said, you can have leakers, i personally don't like that risk.
You got some packs "fresh" and you allow them for two months to "leak"?
Maybe its me, but the idea that it could ignite a fire and set my shed or worse my house on fire......
I know its me, i really don't like that risk.

We all(i assume) found very weird things in packs, and yes stay away from those replacement packs.

Best
Still learning English. Learning Li ion and solar technology.

4200 cells in packs Exclamation above 2500mah and 90%soh.
1500 waiting for testing.

Saving for 3 times phoenix inverter 48/3000 230v to gain also 380v
3 chargers?

Time is our enemy, must work to, the sun is our friend, must relax to.
With best regards
Reply
#7
(05-24-2019, 08:32 PM)Shawndoe Wrote:
Quote:So, it doesn't matter, but you do it anyway.
Regarding breaking apart packs I break them apart by type. First the medical packs if I get any.
Then I go and try to sort by pack design. I find that once I got a pack design down the harvesting goes a lot faster as you know what part of the pack to attack for the quickest results. I usually try to break apart ~75 packs in one sitting. So yes I break them down as quickly as I can but usually only on the weekends.
Last haul on friday got me a good collection of medical packs and scored ~150 Sanyo UR18650F with an IR of <55mΩ and very few cells above that. The ones above 55mΩ were put into the giveaway box. 14 of them where <2V and were given a CC/CV meal at 4.18V and 50mA per cell which lasted about 3 days till they where fattened up and stopped taking nourishment. The all where <55mΩ and all came back with ~90% capacity.


Quote:A good IR meter is next on my list of purchases.  I seem to remember you testing various units, do you have any recommendations, or do they all test about the same?

The are not all the same but I recommend the YR1030 based units. My favorite has become the YR1030+ https://www.aliexpress.com/item/32817594188.html?spm=a2g0s.9042311.0.0.10c24c4dgRgwez
L
arge display accurate and easy to use.


Quote: What is your IR criteria, do you have a flat line for all cells and manufacturers, or do you go by manufacturer spec to determine condition?
My flat line criteria is anything over 90mΩ goes to the bin. From there it becomes a bit more granular. Yes different manufactures and model numbers have different IR specs.
Here is the chart I am using. It is pretty closely tuned to give you a result of ≥80% capacity for each individual cell manufacturer that I have recorded. It is a work in progress and your mileage may very. But all in all it is a good starting point.



Quote:I was reading a paper yesterday about IR testing across the charge curve to diagnose different issues such as SD, Capacity, and heating issues, I'll look and see if I can find it when I get home.  It seemed promising, so the question becomes how would we set up a system to charge pause, test IR, and resume for a fixed period.  I also remember reading that the path of the leads leading back to the meter have a significant effect on the measured IR, due to EMF.
The leads on the YR1030s are somewhat twisted and looped over each other to combat the EMF interference. I have a 10mΩ reference resistor and the YR1030 based units are dead on. I also have a reference Voltage generator and the Voltage readings on the YR1030 are also very close. the 10th and 100th are right on.

Wolf
If 18 X 650 = 2200+mAh then we have power! 
May all your Cells have an IR of 75mΩ or less Smile
Last count as of 8/7/2019
Total Number of Cells Recorded and processed                 6149
Total Cells required for PowIRwall                                   2856
Total Cells ≥2200mAh, ≥80%, ≥35mΩ, ≤75mΩ, ≥4.12V   2760
For Info Google Drive
Not your average Wolf       
            Cool
Reply
#8
I didn't pay attention to IR in early days as I will going forward but I found this interesting...

I've posted here and there earlier about 5 problem packs that were made from Green G5+G7 cells.  These had voltage swing hi/low significantly more than other packs as my Batrium revealed and I could not easily balance within 70mv of others.   I replaced these cells to match the other packs and solved my overall balance problem.   

When I measured these G5/G7s (400 of them) I thought I'd find destroyed cells 'or something really bad' and instead found about half were in the 180mohm'ish IR level and other half were 180-230mohm IR level per my OPUS BT-C3100(s).   They were all around 3.7v at that point.   They all showed 85% - 95% of manufacturer spec capacity which is why I thought they were good in the first place (before understanding more about IR).

Bottom line is I see you call them out as 'unique' and perhaps that what I ran into when I tried to use them. I've also place them in the 'do not use for battery bank' category based on my experience. AND as you point out - IR seems to be a key factor that is different about them.
Reply
#9
(05-28-2019, 01:32 PM)OffGridInTheCity Wrote: Bottom line is I see you call them out as 'unique' and perhaps that what I ran into when I tried to use them.    I've also place them in the 'do not use for battery bank' category based on my experience.  AND as you point out - IR seems to be a key factor that is different about them.
@OffGridInTheCity

If you see my conclusion in the Sony green cells just suck thread https://secondlifestorage.com/t-Green-So...6#pid48096 it will give you a complete rundown of 66 Sony G8 cells being tested.
As I said there if I had 3000 of just theses cells OK no problem but in a mixed environment no way as you found out.

Its always nice to get verification of a theory I had albeit at your expense of having to replace all those cell in your powerwall.

In the meantime I set all the G* series cells aside till I can figure out what to do with them.

Wolf

(05-25-2019, 09:28 AM)100kwh-hunter Wrote: You got some packs "fresh" and you allow them for two months to "leak"?
Maybe its me, but the idea that it could ignite a fire and set my shed or worse my house on fire......
I know its me, i really don't like that risk.

@100kwh-hunter

I agree with you on the 'leaking cells" but only if they have open sores. As in actually leaking electrolyte.
Now when I go to my recycler and harvest cells they are in a huge bin. All kinds of cells many swollen pouch cells and the smell of lithium in the air.
I rummage through this pit with a well gloved set of hands and pick the packs I am interested in. There is a sea of lithium packs in all kinds of states.
Now I do not want to minimize the safety concerns everyone has as they are well founded and an ounce of prevention is worth a pound of cure.
But the recycler is not worried about 1000 packs catching fire as he dumps more on top of the other ones.
Also after I get done rummaging and displacing all kinds of packs to get to that 12 cell pack at the bottom I have yet to see any sparks or fire.
Use every precaution as is necessary but remember that without any drastic interaction a Li-ion cell will not spontaneously combust.
You need to be doing something to the cell as in overcharging, overheating, puncturing etc. Just look at the firebox testing done by drbacke.
It took a considerable amount of heat to actually get the cells to start to flame. Now once on fire Katy bar the door there is no stopping it.
Our mission is to employ and use all necessary safety measures to insure that does not happen. Holding a Li-ion cell in your hand though should not cause you to feel like you have a bomb in your possession. A potential one but many factors have to come together at the same time for it to go off.
Treat the cell with respect and it will respect you.
Wolf
If 18 X 650 = 2200+mAh then we have power! 
May all your Cells have an IR of 75mΩ or less Smile
Last count as of 8/7/2019
Total Number of Cells Recorded and processed                 6149
Total Cells required for PowIRwall                                   2856
Total Cells ≥2200mAh, ≥80%, ≥35mΩ, ≤75mΩ, ≥4.12V   2760
For Info Google Drive
Not your average Wolf       
            Cool
Reply


Who read this thread?
75 User(s) read this thread:
Kna (09-13-2019, 10:32 AM), friedpenguin (05-28-2019, 10:16 PM), juanjo57 (05-25-2019, 01:19 PM), 100kwh-hunter (06-09-2019, 06:49 PM), Lucan (05-28-2019, 09:03 PM), winny (05-26-2019, 11:37 AM), fleming (05-28-2019, 04:09 PM), Overmind (06-04-2019, 11:30 AM), Richard (05-24-2019, 09:29 PM), bchur83 (05-31-2019, 03:56 AM), zoookii (05-25-2019, 05:38 PM), OffGridInTheCity (05-29-2019, 06:18 PM), mart1nn (05-27-2019, 05:02 AM), kevinjaye (07-17-2019, 01:07 AM), DomingoRP (05-28-2019, 05:20 PM), BmoreThor (05-29-2019, 01:47 AM), ajw22 (05-30-2019, 03:46 AM), schmo (06-06-2019, 05:41 AM), Zaziki (06-03-2019, 06:22 AM), ben teker (05-24-2019, 08:39 PM), hschlech (09-01-2019, 11:44 PM), Magnum164 (08-10-2019, 04:41 AM), Celeritous (07-21-2019, 04:17 PM), krobertson (07-07-2019, 05:04 AM), Androfire (06-30-2019, 01:43 AM), Mdonnelly86 (06-20-2019, 02:46 AM), ilithis (06-18-2019, 06:58 PM), Smackman (06-18-2019, 12:49 PM), Buffalo Board (07-15-2019, 12:14 PM), mackoffgrid (06-09-2019, 11:24 AM), Klutch (06-17-2019, 03:17 AM), Karesi (06-02-2019, 08:35 AM), nero1986 (06-01-2019, 08:56 AM), akpota (05-24-2019, 11:52 AM), jacen17 (06-27-2019, 08:36 PM), Nemo (05-30-2019, 02:37 AM), stevelectric (05-28-2019, 09:48 PM), Generic (06-04-2019, 04:36 AM), Wolf (05-31-2019, 04:52 PM), Chablis_m (05-25-2019, 10:39 AM), djuro (07-11-2019, 12:47 AM), bogptrsn (05-29-2019, 02:22 AM), Oz18650 (05-28-2019, 08:59 PM), ChrisD5710 (05-26-2019, 08:24 AM), Geek (05-28-2019, 11:36 PM), BlueSwordM (07-21-2019, 03:27 AM), Stefaan De Ridder (05-24-2019, 07:13 AM), watts-on (05-28-2019, 03:44 PM), Frnandu Martiński (06-05-2019, 09:26 AM), PAF (05-28-2019, 04:40 PM), Korishan (05-25-2019, 12:49 AM), Majorphill (05-29-2019, 09:25 PM), mike (05-30-2019, 11:16 AM), Sean (05-25-2019, 09:53 AM), daromer (05-24-2019, 05:41 AM), kanchana (06-03-2019, 12:12 PM), Crimp Daddy (05-29-2019, 05:18 PM), jonyjoe505 (05-29-2019, 02:36 AM), Shawndoe (05-30-2019, 05:10 PM), Truth (05-31-2019, 06:59 AM), Beholder (05-25-2019, 06:42 PM), AcE Krystal (07-19-2019, 10:50 PM), HikingGuy42 (05-24-2019, 02:42 PM), 100fuegos (06-02-2019, 01:24 PM), CBlack (06-06-2019, 11:28 PM), Cherry67 (06-02-2019, 11:02 AM), Bubba (05-28-2019, 05:40 PM), cadric (07-01-2019, 08:00 AM), thunderheart (05-31-2019, 09:59 PM), Oliman (05-24-2019, 10:11 PM), davidtrn24 (06-10-2019, 01:31 PM), Onizuka (05-30-2019, 03:41 PM), Ibiza (05-31-2019, 12:26 PM), Redpacket (05-28-2019, 01:48 PM), hbpowerwall (07-23-2019, 09:44 PM)

Forum Jump:


Users browsing this thread: 1 Guest(s)