RCD problems..

[ash_s3]

Hi all

installed some downlights for a customer the other day, nice easy job (as they all start out to be)
did the usual tests, Earth loop was a high reading, checked supply and it was TT even though a lead sheathed underground supply cable.
Ze was high 290ohms, bonding to gas and water all good 10mm which was bringing the zs down a fair bit.
suggested we try and get dno to change to tn-cs or if not possible we will get a better reading on a new rod.

Heres the headscratcher..

RCD as main switch, non split load DB, although fairly new Crabtree star breaker setup.
With all loads switched off the RCD will not test, will not trip at 1x or 5x or ramp test.
Tried brand new RCD, exactly the same. Disconnected the neutral fly lead from RCD and it trips perfectly within the permitted times.
I then went through each neutral one at a time disconnecting them and testing the RCD, found that with one ring finals neutrals take out it tests fine, turns out it's not a complete ring, line is ok but high resistance on neutral end to end.

Further investigation required but couldn't figure out why this would stop the RCD from operating.

The guy that changed the DB obviously had no tester as its a right cock-up but that's another story..

 

[R G]

IMO , all RCD testing should be carried out at the DB with all outgoing MCBs in the off position anyway , as regards the faulty circuits these are another issue and should be sorted separately , another point is the Earthing system should be checked and Ze recorded before any work starts , hence the need for testing before any work is specified

Only Friday I went to a potential job to do the pre testing and I found an Up front stand alone 30Ma Rcd that was faulty with a very high Ze , supply is TT , TNCs is available because of the new ABC upgrade outside the property , but regardless the lady said someone else has quoted the work and has told her that everything at the DB was all ok, without doing any testing , anyway she has paid me for the pre testing which in the end found many faults on the existing circuits as well .....

 

[ash_s3]

RCD test was carried out at the DB with all outgoing MCBs in the off position, however the fault would not disappear until the faulty N conductors were removed from the N terminal strip.

I get that the circuit is faulty and needs investigating, just don't understand why that stopped the RCD operating under test.

Test button works all the time though.

 

[R G]

Sounds very similar to the problem I had too , faulty circuits would not allow rcd to trip , and when I tested with circuits isolated at the DB , found I couldn't get any readings and would not trip , Test button works o/k though , I came to the conclusion that the RCD has become faulty over a period of time with the faulty circuits connected and continuously reset against the fault ,the whole lot was quite old anyway....

 

[kenny7askew]

It monitors the imbalance current between the live condutors. Nothing to do with the CPC.

agree but current needs to flow in cpc to create imbalance,with no load connected and cpc and N short rcd wont trip but will with button because that is internal through resistor

 

[davesparks]

RCD tests should be carried out with the outgoing conductors disconnected. Just turning the MCBs off is not enough, unless the MCBs are double pole then this would be acceptable.

With the N-E fault present the test current which normally flows to earth will partially flow through the N-E fault and return via the neutral conductor. With that ridiculously high Ra very little of the test current will actually flow to earth.

 

[davesparks]

Sounds very similar to the problem I had too , faulty circuits would not allow rcd to trip , and when I tested with circuits isolated at the DB , found I couldn't get any readings and would not trip , Test button works o/k though , I came to the conclusion that the RCD has become faulty over a period of time with the faulty circuits connected and continuously reset against the fault ,the whole lot was quite old anyway....

Did you test with all outgoing cables disconnected from the RCD, or did you just switch off MCBs/remove fuses?

 

[Piratepete]

i don't know why, either, but i've come across it a couple of times. once was on a dual RCD CU, where neither RCD would trip, not on the test button or x1 and x5. dissed the radial with a N fault and RCDs then tested perfectly.

What bothers me about this, is that you have a situation here where the RCD is unlikely to trip in case of a real fault e.g Electric iron goes live and client dies.
OP says the test button works (see page 2), so the client will have no idea that it may not trip in practice.

I've just been doing an EICR on a domestic where a ring had a low N/E IR (not low enough to trip the RCD) but the RCD (Eaton) still tested OK. I find the reliance on RCDs worrying. Perhaps it was a bit premature to do away with supplimentary bonding!

 

[Leesparkykent]

RCD tests should be carried out with the outgoing conductors disconnected. Just turning the MCBs off is not enough, unless the MCBs are double pole then this would be acceptable.

With the N-E fault present the test current which normally flows to earth will partially flow through the N-E fault and return via the neutral conductor. With that ridiculously high Ra very little of the test current will actually flow to earth.

I totally agree with this. One of the issues contributing to this is the high RA with the neutral offering less resistance than the earth.

 

[Plonker 3]

agree but current needs to flow in cpc to create imbalance,with no load connected and cpc and N short rcd wont trip but will with button because that is internal through resistor

What if someone grabs a live cable, will that not cause the RCD to trip too?

 

[kenny7askew]

What if someone grabs a live cable, will that not cause the RCD to trip too?

course it will,the person provides the earth route

 

[telectrix]

it should, dilb, but i'm only likely to try that if i feel suicidal.

 

[Leesparkykent]

it should, dilb, but i'm only likely to try that if i feel suicidal.

Like if you owned that metrel from the pic last night? Lol

 

[telectrix]

yeah. it's even uglier than the yellow peril fluke.

 

[HappyHippyDad]

course it will,the person provides the earth route

I think Dillb probably knows that Kenny... and that was the point.

 

[telectrix]

if this forum was south park, we could kill kenny.