What fault can turn a TT into a TNCS and disable the RCD?

[apprentice87]

Awwww, i would have had a newbie stab at this!!!! [And i am crap at this sort of thing] but here goes;

Right, this is a TT system. We will assume then that the RCD is fitted at the origin.

We know Ra is 60 odd ohms or whatever it is, but Zs at a socket is 1.2 ohms [or whatever it was]

So, how does the meter arrive at its findings?? To measure Zs it applies a load between phase and earth and measures the volt drop and arrives at a figure for earth loop impedance.

We have a N/E fault after the RCD presumably....

So, when we do our Zs test, we are really just measuring L/N inpedance up to that point in the installation and we came up with 1.2 ohms or whatever.

The RCD does not trip because, [remembering that there is a L/N fault] even though the meter is connected between phase and earth, the RCD just sees [in the main] a load across phase and neutral, so it will not think anything is amiss...

The test button will still work though because it creates an imbalance be switching a resistor between phase on one side of the RCD and neutral on the other side, and thereby creating an imbalance across the RCD.

Now, as to the danger caused...

If we assume that the cpc's in the installation were sized by calculation, that calculation would have been based on the PEFC that would flow with the Ra of 60 odd ohms. Now we have a fault though, the PEFC further down the line from the fault will have massively increased, as now it is really a phase/neutral fault. This means that in the event of an earth fault of low impedance, that our now VERY undersized for the fault current CPC's will be cooked. Once this happens we have no earthing...

Am i miles out???

john....

 

[OnOff]

On the right track.but just think of the route MOST OF the fault current will take NOW with the existing N-E short at a WALL LIGHT fitting wired in 1mm T&E.

Cooked N terminal at the wall light? (aka burnt chocolate block....)

 

[Andy™]

Cooked N terminal at the wall light? (aka burnt chocolate block....)

only if its a dodgy connection. overload would effect the entire length, so more then just the terminal that will be cooked

 

[sparkytim]

Awwww, i would have had a newbie stab at this!!!! [And i am crap at this sort of thing] but here goes;Right, this is a TT system. We will assume then that the RCD is fitted at the origin.

We know Ra is 60 odd ohms or whatever it is, but Zs at a socket is 1.2 ohms [or whatever it was]

So, how does the meter arrive at its findings?? To measure Zs it applies a load between phase and earth and measures the volt drop and arrives at a figure for earth loop impedance.

We have a N/E fault after the RCD presumably....

So, when we do our Zs test, we are really just measuring L/N inpedance up to that point in the installation and we came up with 1.2 ohms or whatever.

The RCD does not trip because, [remembering that there is a L/N fault] even though the meter is connected between phase and earth, the RCD just sees [in the main] a load across phase and neutral, so it will not think anything is amiss...

The test button will still work though because it creates an imbalance be switching a resistor between phase on one side of the RCD and neutral on the other side, and thereby creating an imbalance across the RCD.

Now, as to the danger caused...

If we assume that the cpc's in the installation were sized by calculation, that calculation would have been based on the PEFC that would flow with the Ra of 60 odd ohms. Now we have a fault though, the PEFC further down the line from the fault will have massively increased, as now it is really a phase/neutral fault. This means that in the event of an earth fault of low impedance, that our now VERY undersized for the fault current CPC's will be cooked. Once this happens we have no earthing...

Am i miles out???

john....

On the button there John

For those who like pictures a VERY rough sketch of how things would go;

 

[sparkytim]

Cooked N terminal at the wall light? (aka burnt chocolate block....)

As Andy has stated the WHOLE 1mm T&E could be fryed as the majority of the Fault current flows down the 1mm CPC to the N-E fault and BACK UP the 1mm N to the C Unit and off home thru the RCD, (which does not see any imbalance) and back to earth via the supply N and centre point of the supply Transformer.

 

[apprentice87]

Yeehh! Thanks Sparkytim!!!

I am only a self taught amateur, not even passed the 2382. Not brave enough to have a go at it yet.

I do try though, i want to be happy that i have a fairly good understanding of electrical things before i make a prat of myself by trying to pass exams!!!!

john....

 

[sparkytim]

2382 a doddle John, just need to be able to use the index

.

For any who would like the full story connected with the above see Post 6 in the link below

http://www.talk.electricianforum.co.uk/electrical-inspection-testing-forum/20057-testing-week-multiple-fault-finding.html

 

[OnOff]

2382 a doddle John, just need to be able to use the index.

For any who would like the full story connected with the above see Post 6 in the link below

http://www.talk.electricianforum.co.uk/electrical-inspection-testing-forum/20057-testing-week-multiple-fault-finding.html

I concur about the 2382. I did the 60 questions in 59 minutes and think I got 98.3%. Means I can read a book quick, that's all! OK, there were a few cable calcs, v.drop etc but none of it's rocket science in there. As this thread proves there's no substitute for experience. Many thanks to the OP - I've learnt something once again! Always amazes me how LITTLE I know!

 

[PC Electrics]

This is why I don't really understand Steps enthusiasm for TT over TN-C-S. Had this install under scrutiny been a TN, then we would have had a low resistance path to earth, so we would have got a decent sized earth fault current and the RCD would have tripped to alert the owner to a fault.

Empirically, I would say that N-E faults are quite common, certainly more common than dropped Ns on a TN.

 

[sparkytim]

This is why I don't really understand Steps enthusiasm for TT over TN-C-S. Had this install under scrutiny been a TN, then we would have had a low resistance path to earth, so we would have got a decent sized earth fault current and the RCD would have tripped to alert the owner to a fault.Empirically, I would say that N-E faults are quite common, certainly more common than dropped Ns on a TN.

Not wanting to upset ANYONE, but I agree, NO system ia 100% safe, and this situation goes to show that includes TT.

In defence the Zs would have been low enough to Trip the MCB, but not before a sizable fault current would have travelled its route down the 1mm and back, (don't think it would pass an Adiabatic calculation)

I have found like You and many others I suspect, that N-E faults are common, they can sit for years on 3036 protected circuits with no effect,

So the Moral of the story is ; Test, Test and Test again, no way can you bang in a new CU and hope all is Ok, RCDs can be outwitted!.

 

[steptoe]

This is why I don't really understand Steps enthusiasm for TT over TN-C-S. Had this install under scrutiny been a TN, then we would have had a low resistance path to earth, so we would have got a decent sized earth fault current and the RCD would have tripped to alert the owner to a fault.Empirically, I would say that N-E faults are quite common, certainly more common than dropped Ns on a TN.

like I said right at the very beginning, something else would/should have been done/would need to occur to prevent this ever being a danger,

situations like this normally only ever occur due to a limited understanding of the systems in place.

notwithstanding the additional fault, this situation could well be easily avoided,

BTW, I think you will find that all the RCBO fanboys would have the whole installation tripped out with a fault like that under TN, where-as the dual RCD club would still have half a house energised.

 

[sparkytim]

This was a duel RCD installation and the other RCD tested fine. The interesting thing is that the N-E fault had sat there undetected untill a Zs Test taken during other work set alarm bells ringing.

The owner was un-aware that he had caused the problem installing the wall light and the RCD had not tripped on this TT install for the reasons as explained. Had it been TNS or TNC-S then it would have tripped.

So as already mentioned NO system is Totally imune to un-safe situations.

 

[steptoe]

Why would it?

Path of least resistance is still the neutral, would depend on the load really.

BTW, when Im testing Zs I manufacture a fault like that on TN so I anyway at high current and not trip the RCD, I find low current testing inaccurate.

So, a very similar fault on TN dont necessarily trip the RCD.

 

[OnOff]

I'm rural and TT as the place is quite old. Some newer, local properties are TNC-S. Speaking to the EDF guy about whether it was worth changing to TNC-S when he came to do a disconnect for me and he said if the TT is working OK then he'd stick with it as the local pikeys keep stealing the intermediate (true) earth cables off the poles - apparently it's rife down this way. So for the TNC-S neighbours - say the earth cables get nicked and there's a broken neutral........wouldn't they be better off with TT? I know its IF upon IF!

Planning for the future and in considering a new CU WAS considering RCBO all the way. I've found 2-pole RCBO's for like a tenner - Garo make I think - any good or cheap rubbish?

 

[batty]

Don't think they are spankers any more. Not sure how you spell sponcerers but thats what spell check came up with.