The reason for the following very detailed calculation is to recognize things that are not obvious at first. Of course, after 27 pages of speculation, it's just another one. I apologize in advance for the jumble of numbers.
Assumption: one of the 116 cells has a dendrite breakthrough - nearly shorted
condensed summary mOhm:
cell resistance 0.317
48 V bank: 5.8
six 48V banks: 0.966
45V bank: 5.8 but more
Maximum possible current:
443A
Detailed calculating of R_Ohmic foreach Cell with its bus bar:
- Cell resistance 4 years old -condition unknown estimated: 0.3 milliohms.
- Each cell terminal compression connection will have 0.05 milliohms of compression contact resistance of approximately 0.05+0.05 milliohms.
- Busbar: 2mm x 20mm x 70mm Nickel plated cooper core is about 0.07milliohms.
(Nickel plating increases the busbar resistance by about 20%)
see left in picture:
https://diysolarforum.com/attachments/good-battery-connecitons-png.135622/
https://diysolarforum.com/threads/raw-cell-busbar-ampacity.80414/#post-1032751
1. Result for one cell : 0.3+0.05+0.05+0.07 =
0.317 milliohms.
16 cells: 16 x 0.317 milliohms = 5,072 milliohms
2 x 1,2m 70mm2 wire with cabel lug : 0.348 milliohms
Megafuse350 A Littelfuse resistance 0,13 milliohms
https://www.littelfuse.com/media?re...8&filename=littelfuse-datasheet-mega+32v-r2.5
Fuse holder compression contact resistance and steel : ca 0.25 milliohms
2. Result one 48V bank with 16 cells and 2 x 1,2 m wire and fuse with holder: 5,072+0,348+0.13+0.25=
5.8 milliohms
3. Result of 6 parallel 48V banks 5,8 milliohms : 6 =
0.966 milliohms
4. Result one 45 V bank with 16 cells (one cell 0V assumed nearly 0 ohms shorted) and 2 x 1,2 m wire and fuse with holder: 5,072+0,348+0.13+0.25 =
5.8 milliohms (but more)
Now we should see that the six currents of 6 parallel 48V banks are feeding one 45V bank. The difference is only 3 V and nothing else.
Under these conditions, an absolut maximum current of (48V-45V)
3V !!! /(0.966+5,8 milliohms) =
443A through this one bank and its fuse is possible.
73,88A in each of the 6 other banks.
But the real current could be easely in the
264A....349A Range. E.g. the dendrite resistance in the shorted cell are much more higher than 0 Ohm. 349A x (0.13 + 0.25 milliohm) of the Megafuse and holder = 0.132V. The power consumption and heat dissipation of the red glowing fuse in its holder are now 0.132V x 349A=
46.5 Watt for a very very long time. (continous current)
Conclusion:
See Datasheet above: At a room temperatur of 20C the the maximum permissible
continuous current of this 350A fuse is
263A. The range 263A ....349A for this is not allowed, as it is dangerous due to its continous heat dissipation. (greater 350A is allowed because the fuse will break after hours.)
It is clear to me that other conceivable errors can lead to significantly higher currents that would absolutely overwhelm the fuse. E.g.: the internal cell short circuit, which cannot be prevented by any external measure, can also lead to a very undesirable causal chain; enormous heat,
hydrogen !!!, gas development, heating of neighboring cells, etc...
Thanks - great forum