Interesting problem

[PEG]

Is the RCBO single pole(SP) OR single pole with switched neutral(SPSN)? To find out out which, isolate the RCBO and measure continuity between Nin and Nout when RCBO open and then closed. (There is a circuit capacitance effect which can be troublesome to residual current detectors when a switched neutral conductor is first energised (at a similar(but not absolutely identical) moment to the switched line conductor becoming energised).

Have you done standard FRC checks ? What where the result ? Of particular interest are the figure of 8 readings at each socket taken in step from the first socket on the 'outward leg' and ending on the first on the 'home leg'. As well as the L-N and L-E figure of 8 also do an N-E figure of 8 set of tests. (This is to explore whether the ring is symmetrical to currents flowing clockwise and anti-clockwise or if there is in fact a direction around the ring with a lower conductance path for the line currents which is opposite to the direction of lower conductance path for the neutral currents). Or in other words, the currents at the beginning of a leg should equal eg: 10A/10A and 20A/20A not 10A/15A and 20A/15A.

That occurrence,may need further explaining...

 

[marconi]

I was called out to look at an RCBO tripping that powers a ring final circuit in a youth cafe.
Checked the IR values at the DB - and all >299. I removed the conductors from the RCBO and checked just that circuit.
Now weirdly, when I go to one of the sockets and test either way, the value for L-E is 0.12 when the conductors are back in at the DB end. To prove a point I then removed the conductors at the DB and stuck in some wago's, i.e. L-L, N-N and E-E. When testing at the socket again, >299.

Also checked the RCBO on its own, that seemed to function just fine.

Anybody got any ideas?

1. As a check did you/can you disconnect the FRC conductors from the RCBO and connect to the load terminals a 100Watt filament lamp and check correct operation?

2. Also, in my research om MEMSHIELD2 RCBOs I turned up this document - http://www.hevacomphelp.com/elec/1pdf/mcp/0047.pdf in which there is this note: SPSN RCBOs must NOT be fitted in TPN type B distributionboards. SPSN RCBOs are limited for use in DP versions of type A andtype B distribution boards.Is this the case or not in the DB you are working on?

3. With the FRC disconnected could you also measure the potential difference between N and E at this DB with the RCBO open and then closed and then do the same with the FRC connected to the RCBO?

4. Is there another RCBO you can connect the FRC in to temporarily to establish whether the RCBO which trips on closure is faulty?

5. Are all the other final circuits supplied via their own RCBO or a shared RCD? Have you tried turning off all the other Final circuits and then turning on the Ring FRC alone? If the FRC's RCBO holds closed now turn on the other Final circuits one by one.

6. Does the FRC supply any equipment which has bonded and/or earthed metalwork eg: A gas boiler.

7. With FRC completely disconnected could you check the IR between the DB's N-E and L-E. And then between each of the FRC's conductors and DB's L, N and E.

7. What is the FRC wired in? Flat T/E; MICC, SWA...?

8. Is equipotential bonding in place and proved satisfactory by conductance measurements?

 

[marconi]

Nothing plugged in, that was my suggestion on the phone when the client first rang me. Oh and yes neutral fly-lead + earth lead removed whilst testing.
With either side of the ring connected to the RCBO (in turn) will operate fine. When both all are connected, RCBO trips.
I had all the sockets off and removed the conductors from terminals and IR tested all conductors individually, all >299.[/QUOTE

I have been pondering this all day and pursuing different theories. If you do not find any defects in wiring then the statement highlighted makes me think that the distributed LCR characteristics of each of :- the supply itself - the RCBO - and the FRC wiring are amalgamating to create out of balance transient currents when the RCBO is closed. As far as the supply-RCBO combination is concerned,the transmission line characteristics of a one end only fed ring is not the same as a those of a both end connected ring. I have an explanation being crafted but for now I have the following practical suggestions:

1. Turn the FRC into one radial on a 20A RCBO or two radials each on an RCBO.
2. Replace the MEMSHIELD RCBO by a Hager type which have better transient performance.

 

[sponge23]

Hi Marconi, thanks for your comments. I happened to have another MEMshield RCBO in my van that day so I swapped it over and still same effect.

All the other circuits are on RCBO's, all wired in T+E. Its a small circuit only 2 x 2g sockets and 2 x FCU's, and yes i removed the load conductors from the FCU's.

Must admit I didn't do any ring checks, but will do when I go back.

 

[marconi]

Hi Marconi, thanks for your comments. I happened to have another MEMshield RCBO in my van that day so I swapped it over and still same effect.

All the other circuits are on RCBO's, all wired in T+E. Its a small circuit only 2 x 2g sockets and 2 x FCU's, and yes i removed the load conductors from the FCU's.

Must admit I didn't do any ring checks, but will do when I go back.

Thank you for the update. I will await your next report. In it could you tell me whether the RCBO is SP or SP and switched neutral and, the outcome of the other checks I suggested please. Do check the IR of the other circuits too, one by one. Lastly, what do the FCUs supply?

 

[Octopus]

What are the end to end results for the ring?

Bump!!

 

[marconi]

If the testing suggested by Murdoch and me does not discover any faults then also do the following testing:

Steps 1 - 3 This is to discover (perhaps) damage to the ring wiring which is putting conductors in touch with the fabric of the building.

First isolate/switch off the DB to de-energise all circuits and check presence and conductive paths of equipotential bonds.

1. Completely disconnect the FRC (Ls, Ns and Es) at the DB. Then measure resistance using your Megger between Main Earth Terminal and each of the FRC's conductors in turn. Note down readings and look for anomalies.

2. Now, do these tests between the 'outward' leg conductors only and the MET, after removing and separating completely the conductors at the sockets and FCUs in turn. So, if the outward leg connects to S1, S2, FCU1 and FCU2, in that order moving away from the DB, remove the connections from FCU2 first and make the measurements. Then, leaving FCU2 disconnected, disconnect FCU1, make measurements and so on until all sockets and FCUs are disconnected. Note down readings and look for anomalies.

3. Finally, measure the resistance between MET and each of the conductors of the 'home' leg.

As you remove lengths of the FRC you may see a marked change in resistance measurements between the FRC and the MET which is indicative of the last section removed having some contact with the fabric of the building.

4. Now that the two sockets and two FCUs have been removed, join up the Ls, Ns and Es at each of the box positions and then reconnect the outward leg end of the ring to the RCBO. Close the RCBO and see if it holds. Then connect both legs of the ring to the RCBO and close it. (This is to investigate whether one of the sockets or FCUs is faulty).

 

[Octopus]

I wonder if the OP has an earth leakage clamp metre? It would be interesting to check the installation leakage!

 

[sponge23]

Hi Marconi, sorry but what is the use of checking the IR of the other circuits?
One FCU supplies a wall heater, the other has nothing on it.

 

[marconi]

Hi Marconi, sorry but what is the use of checking the IR of the other circuits?
One FCU supplies a wall heater, the other has nothing on it.

To investigate conductive paths for earth leakage currents.