Fire rating of fuseboard

[Calebp43]

I recently came across a video on YouTube and the guy fit a fuseboard. Very tidy job but I was confused as he used couplers to have the board not flush with the wall and there was massive holes in the back where the cable comes through.
1) doesn't it need to be ip rated so nobody can get there hands in?
2) doesn't it need to be fire rated?

Just posting this as I want to know if there is a flaw in my thinking somewhere. Here is the link to the video -

 

[timhoward]

Is this something you have experienced? I don't generally fit DP rcbos. I like that they make ir testing easier but I don't feel it's worth the extra expense.

If you consider how long the CU is going to sit there, the extra cost is pretty negligible really. And they don't twist as much (thinking Fusebox)

Consider this scenario:
Lighting circuit has N-E fault as outside light filled up with water.
Cooker circuit with no faults.
Both RCBO's on.

Turn on Cooker, the return current has an extra path via the fault in the lighting. If the N-E resistance is low enough one or both of the RCBO's will trip. So far that is exactly the same whether SPSN or SP RCBOs.

The benefit to the customer of the SPSN is that over the phone I can now diagnose it and as long as the SPSN lighting RBCO is off they can now turn the cooker back on and make their kippers on toast. If we skimp and fit SP RCBO's the faulty N path remains and they have to learn to like spam sandwiches.

One could argue that (specifically) for tolerance of N-E faults a fully loaded SP RCBO board is even worse than a dual RCD board.

 

[mainline]

If you consider how long the CU is going to sit there, the extra cost is pretty negligible really. And they don't twist as much (thinking Fusebox)

Consider this scenario:
Lighting circuit has N-E fault as outside light filled up with water.
Cooker circuit with no faults.
Both RCBO's on.

Turn on Cooker, the return current has an extra path via the fault in the lighting. If the N-E resistance is low enough one or both of the RCBO's will trip. So far that is exactly the same whether SPSN or SP RCBOs.

The benefit to the customer of the SPSN is that over the phone I can now diagnose it and as long as the SPSN lighting RBCO is off they can now turn the cooker back on and make their kippers on toast. If we skimp and fit SP RCBO's the faulty N path remains and they have to learn to like spam sandwiches.

Once could argue that (specifically) for tolerance of N-E faults a fully loaded SP RCBO board is even worse than a dual RCD board.

You could always fit DP mcbs. into a split load CU.

 

[pc1966]

Turn on Cooker, the return current has an extra path via the fault in the lighting. If the N-E resistance is low enough one or both of the RCBO's will trip. So far that is exactly the same whether SPSN or SP RCBOs.

With single/dual RCD yes, but not RCBO unless it has an up-front RCD (as sometimes done for TT systems). For RCBOs to interact in a non-obvious manner it would need to be a N-N sort of fault.

The N-E fault tripping out other circuits happens because of the common RCD to several MCBs in older board. Basically as far as that RCD is concerned they are all the same circuit.

I would always go with DP (SP + N) RCBOs if I could as they make testing easier. But where it really matters is systems with RCDs higher in the supply chain, such as TT with incomer RCD, or a farm with 300mA RCD for fire safety reasons, or sub-main where you can't disconnect in time simply on Zs, etc. Without switching N there you don't have proper selectivity and any N-E fault will trip out the upstream protection.

 

[timhoward]

With single/dual RCD yes, but not RCBO unless it has an up-front RCD (as sometimes done for TT systems). For RCBOs to interact in a non-obvious manner it would need to be a N-N sort of fault.

I agree with you that I was wrong about which device will trip, I wasn't thinking straight. The "large load" RCBO (not shown) will not trip as everything "coming in" is also "going out".
But I think my wider point is correct - this kind of thing can still happen (supply at top, large load with own RCBO on left, another RCBO with faulty circuit on right).
And with a SP+N RCBO the customer can isolate the fault until someone can come and prod it.

 

[happysteve]

Consider this scenario:
Lighting circuit has N-E fault as outside light filled up with water.
Cooker circuit with no faults.
Both RCBO's on.

Turn on Cooker, the return current has an extra path via the fault in the lighting. If the N-E resistance is low enough one or both of the RCBO's will trip. So far that is exactly the same whether SPSN or SP RCBOs.

Eh? I don't understand how this would work...

If you have an upstream (e.g. time-delayed) RCD then sure, you need the RCBO to disconnect the N so that it clears as well as detects the fault. But if there is no upstream RCD, you're fine... surely?

EDIT: just seen your later post, fair enough!

 

[Aaron b]

With single/dual RCD yes, but not RCBO unless it has an up-front RCD (as sometimes done for TT systems). For RCBOs to interact in a non-obvious manner it would need to be a N-N sort of fault.

That was my thought. With the neutral being connected via a fly lead all of the current measured should be equal.

I wonder if it could be the switching effecting the electronics in the rcbo. With lighting I would expect a shared neutral and a botched cu change.

 

[mainline]

With single/dual RCD yes, but not RCBO unless it has an up-front RCD (as sometimes done for TT systems). For RCBOs to interact in a non-obvious manner it would need to be a N-N sort of fault.

The N-E fault tripping out other circuits happens because of the common RCD to several MCBs in older board. Basically as far as that RCD is concerned they are all the same circuit.

I would always go with DP (SP + N) RCBOs if I could as they make testing easier. But where it really matters is systems with RCDs higher in the supply chain, such as TT with incomer RCD, or a farm with 300mA RCD for fire safety reasons, or sub-main where you can't disconnect in time simply on Zs, etc. Without switching N there you don't have proper selectivity and any N-E fault will trip out the upstream protection.

I've had it twice where a recently wired building, TNC-S full sp rcbo board re energise the tripped rcbo and it would trip and also take out a random one with it, fault found to be N To Cpc fault in outside lighting in both cases.

The energising of the faulty circuit caused the trip in another.

 

[Pretty Mouth]

I've had it twice where a recently wired building, TNC-S full sp rcbo board re energise the tripped rcbo and it would trip and also take out a random one with it, fault found to be N To Cpc fault in outside lighting in both cases.

The energising of the faulty circuit caused the trip in another.

Very strange. Out of interest, why do you think a 1PSN RCBO protecting the faulty circuit would have prevented the tripping of random others, seeing as the faulty circuit would be complete on energising anyway?

 

[mainline]

Very strange. Out of interest, why do you think a 1PSN RCBO protecting the faulty circuit would have prevented the tripping of random others, seeing as the faulty circuit would be complete on energising anyway?

It wouldn't, but it isolates the N to E faulty circuit.

 

[pc1966]

I've had it twice where a recently wired building, TNC-S full sp rcbo board re energise the tripped rcbo and it would trip and also take out a random one with it, fault found to be N To Cpc fault in outside lighting in both cases.

The energising of the faulty circuit caused the trip in another.

Do you remember what brand of RCBO?

If you get enough of a kick on the supply, either fault current induced close to the RCBO, or voltage spike with cable capacitance, that can be enough trip to residual current detection. But those would not change with N-switching as such.

Still, I personally think all RCBO should be N-switching due to the easier testing for faults (e.g. open each in turn and IR the outgoing cables) as well as the slight protection it could give for reversed polatiry (not tripping on over-current, but on current imbalance, and probably "once only" if you are lucky on a major fault as usually the "N switching" side lacks the arc chute, etc, needed to live another day).

 

[mainline]

Do you remember what brand of RCBO?

If you get enough of a kick on the supply, either fault current induced close to the RCBO, or voltage spike with cable capacitance, that can be enough trip to residual current detection. But those would not change with N-switching as such.

Still, I personally think all RCBO should be N-switching due to the easier testing for faults (e.g. open each in turn and IR the outgoing cables) as well as the slight protection it could give for reversed polatiry (not tripping on over-current, but on current imbalance, and probably "once only" if you are lucky on a major fault as usually the "N switching" side lacks the arc chute, etc, needed to live another day).

Both CUs were Fusebox.
I only ever use miniature ones now.

 

[Aaron b]

Both CUs were Fusebox.
I only ever use miniature ones now.

If I have two I might try simulating it. I've been installing elucian recently, the rcbos are only single pole but they are a better design than fusebox. I've had a couple of iffy fusebox rcbos. One didn't trip with the test button and one was on in the off position.