Cell level fusing - current failure requirements

[No_Shorty]

Just wanted to check with the community knowledge if there is any reason for cell level fusing other than internal shorts?

My thinking is that in this event the 'fuse' will deal with a huge current. Lets say in a 9p pack you are easily talking hundreds of amps (If I am correct to assume that in the event of a shot the entire pack dumps its shorting current into that one battery).

Therefore when choosing a fuse should the main requirement to choose one that blows/fails as close as possible to the upper limit you intend to draw from each cell, or simply one that blows well below the potential short current?

So, in a given example:

I wish to have an upper limit of 15A from each individual cell. As it's readily available, I choose 0.2x4mm nickel tabs as my fuse connector, with a rough fusing current of 20A (useful table, purely on the maths, not tested:https://docs.google.com/spreadsheets/d/1TKznS77--MO6bC9H0BMu434ZRykqB8wwDxQFHpXDhHM/edit#gid=0).

Lets say the 20A fusing current is way off, and the nickel carries upward of 50A before blowing.

Is this now a problem? By my thinking - the failure it is protecting against should blow the fuse even if calculations are way off.

I think I see the obvious answer: that the momentary 50A short into that cell stands a higher chance of causing thermal runaway that 15-20A would. But if this cell has shorted would this not be a likely event anyway?

This is all with the question in mind: Why bugger about with fuse wire (which on 15A wire has a 15-30 or so amp blow range anyway) when you can just use easily weldable nickel strip.

 

[spinningmagnets]

There are several examples of a "fuse" being part of a nickel strip, by the laser-cutting shaping a narrow section onto the nickel ribbon.

For home backup power-banks, it's an easy calculation. They often use "near free" (*often $1 per cell) 18650's that usually have at least 2000-mAh of life left in them.

Used cells are risky, so individual cell fuses are the common standard. I frequently see these using 80P to 100P, so even if the max draw is 2A per cell, an internal short in one cell would connect the pos and neg of that entire P-group, meaning 200A in a dead short.

Fuse wire size choice is a balancing act. If it is too small, then you are adding un-neccesary resistance to the current path, and any time any conductor in the current path gets warm, it indicates watts being converted to waste-heat. A little warmth may be a necessary evil, but too much heat is poor design.

In the example above, you could use solid copper wire to reduce resistance, something like the thin wires in free salvaged Cat-5 cable. 24-guage might be the thickest you'd want, and thinner wire would blow faster.

If you acquired a LOT of free wire that's a little too thin, then you could double-up on the free wire. (*un-stranding 16-ga stranded wire from the power cable of a free trash-day vacuum cleaner).

Now, on to ebike packs. If you want a small pack with a low P-count (52V / 14S in 3P is only 42 cells) then fuse-wire selection is critical. If we compare a 30Q cell at 15A peaks, then the fuse wire must carry 15A without getting more than barely warm, to prevent voltage sag, leading to weak acceleration.

But...since the pack is only 3P, the peak "fuse-blow" amps are not as high as the previous example. A 30Q may be rated for 15A, but in a dead short, the 3P group will provide over 100A (33A each).

This is a narrower window to aim for. You want low resistance at 15A per cell, but you need the fuse to blow in less than one second if it suddenly flows 33A or more.

Since you dont want your test 30Q cell to blow right away, start with fuses that are likely to be too small (*fuse blows at 15A) and work your way up to a thicker fuse that will blow at the high target (*fuse blows at 33A).

Industrial fuse-wire has a very similar composition to "lead-free" solder, mostly zinc. Once the fuse wire diameter spec increases to a point where there is a significant volume of conductive splatter, the common standard is to switch over to small solid copper wire as the fuse.

I suggest all testing like this be done outdoors on concrete, over a bucket of sand.

 

[No_Shorty]

Thanks for your post, very helpful and more or less in line with my thoughts.

My focus on Nickel was its weldability. You can solder nickel to a copper bus bar, but with the available spot welders copper doesn't play nice.

My use case is also not a bike, I'm building my pack for portable power system, but there are a lot of overlaps to bike pack building which make the information and knowledge on the forum very useful. But the voltage sag of sudden power demands wont be something I will be contending with to the same extent.

 

[electric_nz]

Also note that most individual cells have a built in pressure/temperature safety cutout; so each cell has its own fuse effectively. It’s probably not necessary to over engineer it; but if you’re soldering packs then using a moderate fusewire to connect cells e.g 10A for a nominal 5A peak will give you an extra layer of safety.