Cell Level Fusing w/ 24AWG TCW

OffGridHawaii

New member
Joined
Jun 14, 2020
Messages
13
Hello, I have been having issues properly affixing 30AWG tinned copper wire to positive cell terminals. I am using a kweld to spot weld the fuse wire to the cell and it has been incredibly hit or miss in terms of making a proper weld. I am not satisfied and will not be continuing with using Remington 30AWG TCW.

I have decided to use 24AWG tinned copper wire connecting the copper busbars to the cells. These have been welding very securely.

My powerwall is 14s168p with brand new Panasonic cells. I have read the sticky threads in the FAQ on cell level fusing and understand should a cell short, the remaining 167 cells will all force current into the shorted cell.

My question is:

1. Given that if a cell shorts, and the other parallel 167 cells all force current into the shorted cell, this current would easily exceed the ~20A blow point of the 24 AWG wire. Am I wrong in thinking that this wire will be alright to use?

I'd greatly appreciate feedback regarding my use of 24AWG as fuse wire. Thanks!
 

Korishan

Moderator
Staff member
Joined
Jan 7, 2017
Messages
6,418
Increasing the "fuse" blow point is also increasing stresses on the pack. Altho, if the wire pops at 20A, that should still be sufficient.

However, if a cell is partially dead shorting itself, and only pulling 10A from the pack and dumping as heat, the 24awg will not blow.

Remember, not all cells go full on dead short when they die. The can go through a phase of increasing self discharge until they go full short.

Personally, I wouldn't go with 24awg. It's a bit too thick and too high amps needed to pop it, imho
 

OffGridHawaii

New member
Joined
Jun 14, 2020
Messages
13
Korishan said:
Increasing the "fuse" blow point is also increasing stresses on the pack. Altho, if the wire pops at 20A, that should still be sufficient.

However, if a cell is partially dead shorting itself, and only pulling 10A from the pack and dumping as heat, the 24awg will not blow.

Remember, not all cells go full on dead short when they die. The can go through a phase of increasing self discharge until they go full short.

Personally, I wouldn't go with 24awg. It's a bit too thick and too high amps needed to pop it, imho
Thank you. Do you have experience using a kweld spot welding TCW to cells? It was not working properly for me with the 30AWG. Do you think somewhere between 24-30awg would work better and be safe enough?
 

Korishan

Moderator
Staff member
Joined
Jan 7, 2017
Messages
6,418
I haven't done any spot welding, so no experience with that.

When you say it wasn't working properly, what problems are you having? Is it blowing through the wire; not welding to the cap; sparking too much; etc?
 

OffGridInTheCity

Well-known member
Joined
Dec 15, 2018
Messages
1,512
Just a comment for you to take or leave :)

I believe the goal of cell fusingto be a rapid melt (<1sec) as apposed to a 3,4,7sec heat-up/drip/melt.I use 30AWG for my fusesand my (unscientific) tests + youtube lurking (like @AveRageJoe fuse wire tests such as this one on 30AWG -https://youtu.be/eBn3nAnh7fs)indicate more of a15a range before melting 'rapidly'

[size=small]I use 27AWG for my negative side -> bus and its thinner than 24AWG but yet takes noticeably more amps than 30AWG to melt rapidly as I've occasionally shorted it while building a pack.[/size]

Because of the above- I would suggest a double-check that24AWG will not melt 'rapidly' at 20a.

On a positive note - if you do find that 24AWG is not OK, and 30AWG won't spotweld - one solution might be to use those fast-blow axial fuses - that you can get in specific amps. The leads on those will likely spotweld OK.
 

OffGridHawaii

New member
Joined
Jun 14, 2020
Messages
13
Thank you for your feedback. In regard to problems spot welding 30AWG, it's incredibly cumbersome, it sparks 75% of the time, and it blows through the wire making a very weak connection at all power settings. I don't feel safe going forward spot welding that wire. However spot welding is the only method I am going to use to avoid overheating the cells.

I actually started off using 2amp glass axial fuses and needed to stop that quickly as well. Using those is extremely cumbersome. They weld okay but the angles involved don't always match up properly and I feel do not make a structurally safe pack. I have also read that these have quite a lot of resistance.
______________

Considering I have brand new Panasonic cells, I am hoping this will minimize the risk of a cell failure. Would monitoring the packs with an IR gun periodically be a good way to monitor the packs for slowly self discharging cells?

Correct me if I'm wrong, the risk of using 24awg fuse wire would be if a cell started slowly discharging, all of the other cells in parallel would work harder to equalize, and this stresses the other cells.

*Would this take place for a while where I could monitor and see the cell, or is it a rapid escalation?

*Would 24awg prevent a runaway event?

*Would you mind explaining/pointing me in the direction of understanding a dangerous runaway event? This is what I want to avoid happening. I apologize if this has been hashed out and would appreciate links/reiteration (the search feature on this site can be difficult to navigate sometimes). I'd appreciate understanding the chain of events for catastrophic failures.
 

OffGridHawaii

New member
Joined
Jun 14, 2020
Messages
13
I apologize for a redundant post but I'm hoping bumping this thread might give me some closure on this topic. Would someone mind chiming in on:

Considering I have brand new Panasonic cells, I am hoping this will minimize the risk of a cell failure. Would monitoring the packs with an IR gun periodically be a good way to monitor the packs for slowly self discharging cells?

Correct me if I'm wrong, the risk of using 24awg fuse wire would be if a cell started slowly discharging, all of the other cells in parallel would work harder to equalize, and this stresses the other cells.

*Would this take place for a while where I could monitor and see the cell, or is it a rapid escalation?

*Would 24awg prevent a runaway event?

*Would you mind explaining/pointing me in the direction of understanding a dangerous runaway event? This is what I want to avoid happening. I apologize if this has been hashed out and would appreciate links/reiteration (the search feature on this site can be difficult to navigate sometimes). I'd appreciate understanding the chain of events for catastrophic failures.
 

OffGridInTheCity

Well-known member
Joined
Dec 15, 2018
Messages
1,512
One of the main ways to initiate a 'runaway event' (in my mind) is toovercharge cells. See 3:15'ish of this youtube and the aftremath - it can be horrific..https://youtu.be/WdDi1haA71Q Overcharge is not likely with BMS.

The next way (besides overcharge)is due tocell damage internally for various reasons - and dendrites form / a short forms. I've not actually seen a video or anything showing this happenbut I understand that this can lead to fire... which can lead to the video above.

TOPIC1:protect the pack from a dead-short - this is the goal...
The problem with 24AWG is it not typically used that I've read... so I don't see people saying 'yea it works'. But its you're powerwall and your decision. If I was dead set on 24AWG then I woulddo some experiments to verify it burns thru quickly. You cansimulate a dead short with the pack. If it does, without bad side-affects (like meltng metal running down a cell) then you may be satisfied.

You dismissed axial glass fuses..

Another alternative for spot-weldare nickle strips (with fusing)like this-https://batteryhookup.com/collections/cell-level-nickel-fuse BatteryHookup says they'll be back in stock soon. @HBPowerwall just did a fantastic youtube on using nickle strips to build a powerwall -https://youtu.be/PenPYwa00CA

TOPIC2: Self-Discharging Cells/Packs...
You don't need to mix self discharging cells with the topic above.This is not catastrophic with proper BMS to alert you that a pack is bad. The BMS voltage monitoring will make packs with self-discharge visible because they will lag behind the other packs / cause balancing requirements. These should be fixed as a matter of regular monitor/maintenance. I don't imagine a healthy pack will degenerate into a 'self-discharge fire' unless you let it go for weeks/months and balance failed and BMS let it go below minimum voltage. Longbefore that you should fix the pack :)


RECAP:
Fire risk is from overcharge + using damaged cells. Cells get damaged physically or by draining too low or charging too highor by temperature.... and perhaps in very small % - bad manufacture. Fusing is a strongly suggested idea for DIY powerwalls (e.g. large packs) which is shy Tesla create/uses this in their EV batteries.

P.S. You mention your 'cells are new'... but or course as soon as you charge/discharge them 50 times they are no longer new. That's the problem with cells - they do not remain new :)

OK - that's my 2 cents from what I've read, seen and how I think about it with my own 80kwh battery bank. But please, of course, make your decisions :)
 

OffGridHawaii

New member
Joined
Jun 14, 2020
Messages
13
OffGridInTheCity said:
One of the main ways to initiate a 'runaway event' (in my mind) is toovercharge cells. See 3:15'ish of this youtube and the aftremath - it can be horrific..https://youtu.be/WdDi1haA71Q Overcharge is not likely with BMS.

Amazing video. Definitely want to avoid that! I am charging to a total of 4.0 volts per pack (56) and managing the battery with a Batrium BMS, hopefully I can avoid this with those precautions.


The next way (besides overcharge)is due tocell damage internally for various reasons - and dendrites form / a short forms. I've not actually seen a video or anything showing this happenbut I understand that this can lead to fire... which can lead to the video above.

TOPIC1:protect the pack from a dead-short - this is the goal...
The problem with 24AWG is it not typically used that I've read... so I don't see people saying 'yea it works'. But its you're powerwall and your decision. If I was dead set on 24AWG then I woulddo some experiments to verify it burns thru quickly. You cansimulate a dead short with the pack. If it does, without bad side-affects (like meltng metal running down a cell) then you may be satisfied.

You dismissed axial glass fuses..

Another alternative for spot-weldare nickle strips (with fusing)like this-https://batteryhookup.com/collections/cell-level-nickel-fuse BatteryHookup says they'll be back in stock soon. @HBPowerwall just did a fantastic youtube on using nickle strips to build a powerwall -https://youtu.be/PenPYwa00CA

I've strongly considered these, I'm a bit concerned though because they have an iron core and I live in a humid tropical environment near the ocean. I'm very concerned they would rust out quickly. I'd really like to give them a try at some point.


TOPIC2: Self-Discharging Cells/Packs...
You don't need to mix self discharging cells with the topic above.This is not catastrophic with proper BMS to alert you that a pack is bad. The BMS voltage monitoring will make packs with self-discharge visible because they will lag behind the other packs / cause balancing requirements. These should be fixed as a matter of regular monitor/maintenance. I don't imagine a healthy pack will degenerate into a 'self-discharge fire' unless you let it go for weeks/months and balance failed and BMS let it go below minimum voltage. Longbefore that you should fix the pack :)

Excellent, batrium should be very helpful.


RECAP:
Fire risk is from overcharge + using damaged cells. Cells get damaged physically or by draining too low or charging too highor by temperature.... and perhaps in very small % - bad manufacture. Fusing is a strongly suggested idea for DIY powerwalls (e.g. large packs) which is shy Tesla create/uses this in their EV batteries.

P.S. You mention your 'cells are new'... but or course as soon as you charge/discharge them 50 times they are no longer new. That's the problem with cells - they do not remain new :)

OK - that's my 2 cents from what I've read, seen and how I think about it with my own 80kwh battery bank. But please, of course, make your decisions :)

Thank you so much, incredibly informative. I decided to order some 28AWG tinned copper wire to compromise a bit and see if I can get it to bond better than the 30. Aloha.
 

OffGridInTheCity

Well-known member
Joined
Dec 15, 2018
Messages
1,512
FYI - @HBPowerwall just did a fantastic youtube on spot-welding 32AWG fuse wire -https://youtu.be/THVMnyE4yAw He shows greatdetail using K-Weld.
 

OffGridHawaii

New member
Joined
Jun 14, 2020
Messages
13
OffGridInTheCity said:
FYI - @HBPowerwall just did a fantastic youtube on spot-welding 32AWG fuse wire -https://youtu.be/THVMnyE4yAw He shows greatdetail using K-Weld.
Thanks, great video. It seems to me that spot welding TCW thinner than 24 AWG has extremely variable results when welding directly onto cell terminals. The 28AWG proved to be ineffective as well... This nickel strip method looks promising.
 

Redpacket

Well-known member
Joined
Feb 28, 2018
Messages
1,265
You might look at getting the battery packs into an air tight enclosure to reduce corrosion issues?

Wondering if anyone has thought of (or tried) doing eg 30AWG wire soldered to an eyelet, eyelet then spot welded to cell?
Eg these: https://www.jaycar.com.au/solder-lugs-pk-50/p/HP1350

Agree it's be extra work but....?
 

not2bme

Active member
Joined
Oct 16, 2017
Messages
481
OffGridHawaii said:
OffGridInTheCity said:
FYI - @HBPowerwall just did a fantastic youtube on spot-welding 32AWG fuse wire -https://youtu.be/THVMnyE4yAw He shows greatdetail using K-Weld.
Thanks, great video. It seems to me that spot welding TCW thinner than 24 AWG has extremely variable results when welding directly onto cell terminals. The 28AWG proved to be ineffective as well... This nickel strip method looks promising.

Just fyi I use 30 and 32awg TCW with a Sunko 790 with zero issues. I've probably done this for over 1500 cells. Once you find the right setting and tip then it's good. The tip that comes in my probe looks like a 10ga wire. I noticed the ones on hbpowerwall is a fat hunk of copper with a very round radius.
 

Redpacket

Well-known member
Joined
Feb 28, 2018
Messages
1,265
One of the problems with TCW for cell fusing seems to be it conducts too well?
Ie the resistance is so low it doesn't start to heat early enough.
The nickel strip system mentioned seems better because the material is steel/nickel plating so it heats more readily.
Proper fuse wire usually isn't made of tinned copper wire right?
https://en.wikipedia.org/wiki/Fuse_(electrical)
 

Generic

Member
Joined
Oct 23, 2018
Messages
210
It's been a few months since I did this, but I too was initially getting the sparks like you mentioned. There's two things you can try and let me know if they work, as they did for me: 1) be very careful to make sure that your electrode tip is only touching the wire, and absolutely not touching the top of the cell (with 30 ga. wire, it can be hard to tell, but you get a knack for feeling it); and 2) try separating the electrode tips as far as possible on the cell, with one tip on the wire, and one tip directly on the cell. I use the Malectrics spot welder, and I ended up getting good strong welds at the minimum setting (1 ms), I don't know what the equivalent on the kWeld is.
 
Top