Jem

Hi, all the best with yours , but quite honestly even if pylontech did fix mine then i would sell them on as i have lost confidence in them, i have  ordered a seplos mason 280 kit including  cells etc,  Pylontech and victron have worked together to come up with a solution, if that fails as in my case then ill go for a diy kit .

N.B Can't tell you the shutdown sequence the electrician followed as it escaped me, but he seemed quite adamant to do things in a very specific order.
Sure, this picture was taken when we started disconnecting the MPPTs and the batteries after the pop:

thanks @BritishRacingGreen, looking into that post now. Cheers!

... continued from previous post .
Ah , beautiful , Pylontech ! I am supposed to be an experienced engineer that can at least change a fuse in an appliance , but no , apparently not !
Can you imagine what the owner thinks of me now !!! 
So the thermal fuse  has a square of solder alloy which acts as the fusing element . So when the current approaches the critical  region  , it get hot and melts away into the cavity. In addition there is a mechanism for the BMS MCU to also trigger a meltdown. This is actually very powerful and adds an additional level of defense when things go wrong. So the MCU will switch on the heater resistor , which sits below the alloy , and concentrates about 10 watts to heat it.
So the jackpot question is whether the fuse blew because of over current or did it blow via the heater resistor activated by the MCU. My opinion is the latter. Firstly, the first fuse array did not blow . Secondly , if there was an overcurrent then the first defense is the MCU turns of both DFET and CFET. Now we know the fets are still working so I assume that it would have cut off the surge current by this means. My take on this is that the battery had been confronted with an over voltage condition from outside. Under this circumstance it is not enough to turn off the fets as the over voltage would still sit on the outboard fets and there could have been a risk that it would kill that fet array . So the MCU blew the thermal fuse . This is the only logical explanation I have and would invite any other explanation from you.
Pylontech fuses are NOT  serviceable for various reasons . Firstly its an inseparable part of a very large BMS control board . Secondly its SMD devices directly soladered onto the cb pads , and there are no less than 8 of them in parallel to make up for the high current rating.
And to get the SMD code of the devices is very difficult. On the Teardown thread of  @PurePower there is a nice view of the US2000 fuse array and at least when can clearly see the 'Z' marking , which is 20A Fast Blow at 65Volts.

Then the 4 pin fuses are probably very hard to get by , and I think if you don't match them correctly , you doing your BMS no good.
So i am basically stepping down , the owner will attempt to get it fixed via warranty or void repair. I will be a watchdog for him to ensure that he does not  pay exorbitant price for a battery that is otherwise working.
Thank you so much for all your valuable inputs.
NOTE : PylonTech looks quality . The only downside for me is the fact that all BMS subsystems are on one large pcb. I would have liked to see a controller pxb , a power train pcb (fets)  and a fuse pcb with surge protection. 
 

Ah, vultures circling very early already😁😁

@Jemyes you are far away from us here on the Southern tip of Africa. Please keep us updated, we have received numerous issues with Pylontech batteries & BMSs doing almost exactly what yours has just done. After the repair of the batteries that was done by @BritishRacingGreenit also appears that they have placed fuses that cannot be replaced onto the BMS board. They can be replaced obviously but not by a normal person without Surface mount Board experience. 
Now I'm not an engineer neither am I a salesman, but sending a battery to world markets that has fuses that need specialist workmanship to replace is not a good way to go in this market. As you can see how far & wide these products go. Now you have a guy in South Africa whom has these skills pretty sure others around you in Belgium but it's just a show stopper. 
I hope they warranty you properly & with knowledge of the EU laws pretty sure you will be in good stead. Here in SA it's a whole other story with regards to what we have to stand on with the Warranty red tape.
Please keep us updated. 

Hi and thanks @Steve87 thanks for your helpful follow-up.
Well, this isn't quite reassuring. And we have no idea of what caused this issue.
In the meantime, a first follow-up from Pylontech asks me to pull the logs from the batteries, just as you suggested; they even provided... wiring instructions. I'm trying to find a cable to extract the logs; I'm a bit surprised they give a wiring diagram to a consumer (RS232 to RJ45), this sounds like DIY & warranty void territory to me.
N.B. For @BritishRacingGreen, this is certainly a nice backup plan, however, being located in Belgium, I see he's a bit far (Germiston, South Africa, right?) ... but I'll certainly reach out for guidance if it comes to that point.