Main fuse blows before MCB.

[Bobble23]

I wonder if anyone can suggest a scenario that could have made this happen.

I did a full rewire on a two bed purpose-built flat a few weeks ago. TNCS, cut-out in cupboard at the stair-well, I ran the tails to a switched fuse inside the flat and on from there to the consumer unit.

All tested up fine and I left it energised for more work to be done (heating, kitchen, bathroom). Was told by the customer that over the weekend the kitchen sockets kept tripping the MCB and that, eventually, the boiler installer tried to connect up and the main fuse in the stairwell blew. The weird thing is that the MCB, the RCD and the switch-fuse didn't trip. I wasn't there myself so I'm not sure what 'boiler installer tried to connect up' specifically means, but I'm assuming he tried to wire into the FCU left for the boiler on the kitchen socket circuit.

I went there today and the problem with the sockets was that a cable had been damaged by the tiler, so I've replaced that leg and all is working fine now.

I'm just curious as to how the main fuse could have gone before any of the other devices. Any ideas?

 

[littlespark]

what did the supply company say when they came to replace the fuse?

If they've been constantly resetting the MCB, then its maybe welded itself shut, thus not tripping on overcurrent.... The RCD wouldn't trip on overcurrent anyway, but will possibly be knackered now as well.

The next step would be the main fuse.

 

[Lucien Nunes]

What was the type, rating and age of the fuse that blew, the one in the switchfuse and the MCB?

Under fault conditions, if the short-circuit current is high enough, all the devices 'start' to blow or trip regardless of their rating, but the process stops when the first one actually breaks the circuit. It's a race, and the only way to ensure that the preferred (usually smaller) device wins is to achieve full co-ordination between them. E.g. the total let-through I²t of the MCB must be less than the pre-arc of the upstream fuse. That involves comparing lollipop diagrams of the devices at the applicable current and voltage. The existing fuse might also have been fatigued or oxidised and therefore had a lower pre-arc than designed.

So when protective devices are relatively close in rating, e.g differing by less than a factor of two (60A upstream, 32A downstream) these effects can occur within certain ranges of fault current. Obviously the results would be different with a longer, lower current. At 75A there would be little doubt that the 32A device would trip first, not so at 2kA.

 

[Bobble23]

Thanks littlespark.

I was wondering who'd pick up on who replaced the fuse. The customer says the boiler installer went off to get a replacement. To be honest, I'm not sure I want to know, but it wasn't me and it wasn't the DNO. Of course that now puts them in a tricky position if they want the DNO to come and check if there is/was a problem with the fuse.

Take your point about the MCB welding shut, but the same one is working fine today with no signs of thermal damage.

 

[Bobble23]

Thanks Lucien, that makes a lot of sense.

Main fuse was a 63A BS88-3. No idea how old it was, but no reason not to think it was the one that went in when the flats were built. MCB was a brand new BS 60898/B 32A.

 

[littlespark]

Plumber replacing a main fuse!!!

Oh hell!


What’s the bets he’s put in a bigger size.

 

[pc1966]

If you check the selectivity tables for the Hager MCBs then for their NBN 32A model and a 63A BS88 fuse the limit is 1.3kA fault current. So if the PFC is below that, the MCB trips, if it is above that then the fuse (probably) also goes.

Basically MCB contact can only open so fast, and while there is a slight reduction in opening time as the fault current goes up, it does not change that much over the range that the 'instant' trip goes. So as the PFC goes up, so the let-through energy also increases and once it gets beyond the pre-arcing value for the fuse it is basically a goner as the wire has melted.

1.3kA is high for an end of circuit value, but not usually high if the supply has a low Ze and the cable short.

 

[pc1966]

Example of MCB let-through energy (grey line at 10kA is limit for that model):

 

[stevels4]

Surprised a plumber managed to get hold of a cut out fuse, maybe a piece of 15mm copper pipe fitted with 2 end caps for an instant red link.

 

[Zlodej]

I think its obvious, Reason was the plumber. Combination of the fact that he tried to go around and the fact that he went to get replacement fuse without finding an issue, is more then enough to call him a cowboy.
Any sensible building trader whether it is plumber, joiner, plasterrer or less mentaly disabled skip filler, would have told the customer to stop switching MCB on, until electrician fixes the issue.

As stated above if you keep switching MCB on it might have eventually failed. Or spikes from switching it on might gradually damage the fuse.
I don't think it was fuses fault, as MCB was tripping before failure.
As stated RCD does not trip for over current, In this stage it looks like fault would probably be between Live and neutral (assuming we are on a single phase) .

 

[Deleted account]

Are you sure that it was the main fuse. Chinese whisper.

plumber tells customer - customer tells you-

did the plumber tell customer correctly
did the customer understand correctly
did the customer tell you correctly.
did you….you get the jist…



My thinking it was the switched fuse that blew not the main incoming, but terminology got mistranslated down the line.

 

[Bobble23]

Thanks all. To be honest Paignton Pete you might be right, though he did say they were rummaging around in the cupboard with the main cut-out in it. Either way interested to read pc1966 and how it's not quite as cut and dried as MCB always trips first, main fuse second. Always something to learn in this job. Mainly to keep away from plumbers.

Thanks again.

 

[GLENNSPARK]

Mcb's are resettable....fuses aren't...
Every time you subject a fuse to fault, it's going to stress it...

Assuming it was a KMF you fitted...
These typically come with an 80A BS88 which won't necessarily provide you with discrimination between 80A & 100A...
(assuming it was a 100A BS88/BS1361 in the DNO's cutout anyway.....could've been as low as 60A)