Earth Leakage problem

[Octopus]

Yes I do

what is the "back ground" leakage like across the installation?

 

[J.Spark]

How does one test for back ground leakage?

 

[Lucien Nunes]

Murdoch dunno if you spotted the bit about the ELCB still tripping with the installation off at the main switch? I think that is what was meant (J.Spark, can you confirm?) in which case I don't think it's leakage from this installation but from elsewhere circulating via the MET. This is one of the rare cases where it's actually helpful to clamp either of the earth leads going through the ELCB, normally clamping a CPC can give misleading results but here it's relevant because that's the current in the trip coil. If it's more than the current measured with the two tails clamped together (this system's leakage) then the problem is external.

 

[Marvo]

In the OP he states he replaced the 'earth leakage unit' with a new one, after I read that I wrote off any possibility it was a VOELCB. Photos would be handy if JSpark can attach a couple..

 

[J.Spark]

Murdoch dunno if you spotted the bit about the ELCB still tripping with the installation off at the main switch? I think that is what was meant (J.Spark, can you confirm?) in which case I don't think it's leakage from this installation but from elsewhere circulating via the MET. This is one of the rare cases where it's actually helpful to clamp either of the earth leads going through the ELCB, normally clamping a CPC can give misleading results but here it's relevant because that's the current in the trip coil. If it's more than the current measured with the two tails clamped together (this system's leakage) then the problem is external.

Yes....still trips with the 3 phase main switch shut off. Tested no voltage to the top of the ELCB....which I find very odd. Cannot wrap my head around that.

 

[J.Spark]

In the OP he states he replaced the 'earth leakage unit' with a new one, after I read that I wrote off any possibility it was a VOELCB. Photos would be handy if JSpark can attach a couple..

Will do so when I am on site again

 

[Lucien Nunes]

An N-E fault can trip a TP+N RCD with the three lines isolated (not all 3-phase main switches isolate the neutral, unlike single phase DP). Current e.g. from a TN-S neutral above earth potential can pass to earth at the fault via the RCD N which the RCD will sense as a residual. Hence, the importance of the IR tests even though the fault doesn't require the system energised to trip.

But, what is this thing we're discussing? RCD? VOELCB? I was taking it to be a VOELCB because you said it was, but now I see Marvo has a point.

 

[J.Spark]

OK....let me rather attach some pics of the ELCB and take it from there. Thanks for the help guys. Appreciate this. Will be in contact during the course of the day tomorrow.

 

[J.Spark]

I am from South Africa so the wording may be a little different lol!

 

[Marvo]

I am from South Africa so the wording may be a little different lol!

Ahh, I've got a feeling your terminology is causing kak boet.

 

[Lucien Nunes]

'n buitelander! Marvo, let vertaal dit vir ons...

 

[J.Spark]

Hahaha haha yeah I am sure lol!

 

[Marvo]

Is it the usual black CBI samite mount earth leakage breaker?

 

[J.Spark]

The din rail type....schneider electric make

 

[Marvo]

Then it's a standard 30mA RCD. Is it being used as the main breaker or is it just supplying the sockets?

The only way it can trip with the installation isolated is if there's power still on it from another source which is unlikely as you said you checked for supply voltage after isolating so something's not adding up.

Note for the UK guys, South African DB's (CU's) generally have all RCD's and MCB's supply at the top and final ccts connected at the bottom and it's common practice for the entire installation to have a single upfront 30mA RCD.

 

[J.Spark]

Yes only for sockets....but I find it strange that it trips like this....even trips with putting on some circuits which are not running through the earth leakage as well as when there is 0 voltage on the top.

 

[Marvo]

Yeah, it's not making sense, an RCD can't trip unless there's power coming from somewhere. Is there an inverter back-up system, solar pv system or UPS in the premises perhaps?

 

[J.Spark]

Nope.....nothing like that....I double checked voltage on top all the time to make sure because this confused me quite a bit. Never came across something like this. Very strange to say the least.

 

[Marvo]

I'd check for voltage on the load side as well. Do all the usual tests like megger test the circuits, test for crossed circuits/borrowed neutrals (continuity between circuits) and a leakage test with an earth leakage clamp meter around the L+N at the RCD.

With the RCD tripping when the power is isolated I'd suggest extreme caution, treat the DB and all circuits as live until you find the cause of that.

 

[J.Spark]

I did check on load side as well. No voltage. But will do all the tests again and double check on all again.....hopefully it has fixed itself overnight lol! One can only wish lol!