You typically need a 3 to 5 volt difference for a boost converter to do what you describe.
I believe you're thinking about lead acid world, where higher internal resistance needs a larger voltage span to encourage energy flow between them. But why mention this, even if true? The boost converter creates that voltage difference, doesn't it? It's very possible that I have a basic conceptual misunderstanding about something, so please correct me if I'm wrong.
It seems a 1200W DC boost converter
can be obtained for circa $10. Even if there were two needed (one for each direction), it still would work out cheaper than a class T fuse. Downside is that it would lack automation (possibly solvable with an arduino and a few ICs?).
In the lead acid world they use a battery isolator. In the lithium world as long as everything is done correctly the load current and charge current are split evenly between parallel batteries.
The battery isolator actually looks like the right thing, but would need two of those, and operation at 48V. The whole point would be to avoid sharing the load and charge current, in order to eliminate the danger of every possible amp dumping into one bank from all the parallel banks in the event of a shorted cell.
Sunny Boy Storage, an HV AC coupled battery inverter, does that with 3x battery inputs.
Actually, the battery I have for it (LG RESU-10H) does that internally: 150V native battery, bidirectional DC/DC to 400V.
That's interesting... I haven't looked at the SBS too closely. Those kind of voltages (especially DC) start to scare me with DIY stuff.
This could also be accomplished with grid-forming inverter having high current battery, and multiple AC coupled batteries.
That's true... But I was trying to figure out a solution that might be
cheaper than a class T fuse, lol!
But probably not too bad with BMS and disconnect per battery. Protection by disconnecting, just can't connect and bring on line until voltage is very similar so no inrush.
I've been connecting and disconnecting my Ryobi zero turn mower, which has 4S 12V LFP batteries, same as the house. I've been careful to make sure they're fairly close, and monitoring the inrush--it's pretty incredible how quickly and easily power transfers between parallel strings of LFP that have dissimilar SoC. Like a different planet compared to lead acid world! This is actually the "problem" I'd like to possibly solve with DC boost converter, if it doesn't prove too challenging to make it work bidirectionally.
Couple of us plan to do that with one JK BMS per 16s battery, not sure yet how they aggregate current requests for the inverter.
There are a few different "virtual BMS" projects based on ESP32/arduino that use a Victron SmartShunt to speak pylontech over canbus to inverters. I'm trying to get one working for Sunny Islands right now. It's actually working, just needs more testing. Anyway, could be a possible avenue to do the aggregation you were talking about.
I don't wish to derail this thread.
I do want OP
@Jejochen to know that I'm very sorry for your loss, but really thankful that you shared your experience... and wish you and your family well.
I have no clue, I bought 32 of the 400 aH versions of this and that was expensive enough. I can only guess at what one of these costs. That being said, you have Australians that have been using the 200, 400 and 700 aH versions for as long as 20 years with no issues. I contemplated the 700 and the 1000 ah versions but wanted two banks in case one cell had an issue i could disconnect one bank and still have power.
