How a simple job can go wrong quickly....

I am not convinced by the the reversed polarity theory, I think the electrician carried out the safe isolation including testing for voltage being present.


So I do not think there was a big bang and the electrician was aware that there was a problem until being told that there was by the shop keeper.


 Andy Betteridge

I also think the electrician spoke to Bod to ask what could have gone wrong, if there had been reversed polarity and he had touched a neutral conductor with 240 volts on it down to earth he would have been telling Bod what had happened, because he would have known about it.


 Andy Betteridge

If the neutral had broken first, the lines would not have been far behind. Would the duration have been long enough to have done the damage?


There must be plenty of 4-pole switches out there (RCD or otherwise) which could present exactly the same risk, but I doubt that this sort of occurrence is commonplace. Moreover, wouldn't this have happened previously during quarterly, now six-monthly exercising of the test button? ?

Don’t bite off cable insulation, I have a chipped and dead front tooth because I did that when I was about twelve years old.


 Andy Betteridge

Just to confirm; in the original post, the RCD tripped when the Neutral and the CPC touched. This I would expect. There was NO mention of any such touching of a Live to Neutral or Earth.

My thoughts are that the 4-pole RCD did not Close the Neutral first or Open it last, as it should have done. Googling 4-pole RCD and looking at the pictures, each I found had the Neutral pole marked with an N. So without seeing the wiring, my guess is that the RCD was not wired correctly and that one of the three Lives was connected to the N Pole (with the corresponding Live out being on the outgoing N) and the Neutral conductor was taken to the unused Live in, with the outgoing Neutral connected to the associated L out.

So when the RCD tripped, the Neutral lifted first and due to three phase loads being unbalanced, the single phase supply with the now 'duff' equipment on it was now effectively connected phase to phase and reacted accordingly.


Would guess that if the RCD had ever been tested with the button, then all loads had been switched off first.


Clive

Well, interesting replies as I expected. 

The problem?

When the RCD was reset, the neutral didn't but the three phases did causing as some have thought, a loss of neutral leaving the star point floating and subsequent voltage issues. 

As for how long the neutral was floating, no one knows but long enough to damage some sensitive electronic kit.

A replacement RCD was installed and installation OK but a lot of equipment suffered,,,,,,,,,,,,,,

I might be involved with the inevitable "blame allocation" but it does raise some interesting past history installation methods, I&T and T questions.

As to how anyone could anticipate this in their method of work is also interesting....

Regards

BOD

Thanks BOD!


None of us thought about the neutral reconnecting slowly, but it does seem more likely than disconnecting too fast. Oxidised contacts which had not been exercised quarterly! ?


How do you know that the damage was done on reconnection? You don't seem to be saying that it was found that the lines had connected, but the neutral hadn't; just that it was slow.


Incidentally, is it TT or TN?

I think that this is a design issue. Fitting a 3ph RCD to other than 3 phase loads is asking for trouble and shouldn't be done, as I have come across similar problems. My own solution would be to have a separate single phase DB for single phase loads, with its own RCD as required. Chris is right that perhaps RCBOs should be the correct choice, without a 3 phase RCD also, but there could also be some 3 phase sockets! This area probably needs looking at in BS7671, as it seems that more problems may be made by this supposed "safety" situation of RCD protection. It could be that some hefty single phase surge suppressors might have saved things but I am doubtful.