Trying to figure cause of tripping

[Gavin John Hyde]

Done a CU change today, brief testing before hand I knew there was a problem on one circuit, believed to be the element in the immersion tank so put the very low IR readings down to this.
Fitted new CU with plumber arranged to look at the immersion, changed it and thought that would be that. House previously had no main rcd protection at CU.
When I did the IR testing today before powering up i got >199 on all circuits accept the immersion heater radial.
L-E 0.9 Mohms N-E 3Mohms L-N 2.3Mohms.
So knew something was a miss. Bit of investigation and tucked behind the water tank was a rcd socket with 2 shower pumps plugged in. so the immersion radial does both the tank and a double socket.
Removed rcd socket and got >199 ohms on all tests
Put Rcd socket back in and found that the 2 shower pumps (Both 3 amp fused plug in sorts) would trip the rcd in the consumer unit but not the rcd socket. It had not tripped previously when the old re-wireable fuses were in use. Customer didn't even know it was there!
The pumps when plugged into a normal socket did not trip the CU rcd. The RCD in the Cu would trip with the pumps plugged in but not even working with both showers off.
Ramp tested the rcd socket and it trips at 24mA, The 2 pumps when plugged into a normal socket do not trip the CU rcd, it only trips if plugged into the rcd socket.
Taken it out to achieve discrimination or come 2019 selectivity, but its baffled me as it ramp tests as okay at 24Ma, I put the rcd socket on the upstairs ring and put the vacuum on it, no tripping. so seems as if the 2 pumps are the cause but as said when on a normal 2 gang socket the Cu rcd doesnt trip.
Could it be leakage across both the rcd socket and the 2 pumps combined? Only other thing on consumer unit rcd is the cooker circuit which at 500v IR got >999 on all tests.
its good to know for future reference as I like to know why things tripped.
rcd socket shown below:

 

[Octopus]

did you use your earth leakage clamp meter?

 

[telectrix]

i'd suspect a fault in the RCD socket. with that out of circuit, i understand that nothing trips, even with the loads plugged in.

 

[Wilko]

Hi - if you are going back, perhaps test the pumps too ?

 

[Gavin John Hyde]

did you use your earth leakage clamp meter?

No as there was a lack of space and it took me enough time trying to even get to and remove the said socket. Did consider it though, but once i got to the tank which at first i believed to be the only load on circuit, it was when i went to disconnect the heater from the radial that i saw the the rcd socket

 

[telectrix]

Hi - if you are going back, perhaps test the pumps too ?

suck, slurp. suck, slurp,. tested OK..

 

[Gavin John Hyde]

i'd suspect a fault in the RCD socket. with that out of circuit, i understand that nothing trips, even with the loads plugged in.

Thats my thinking but it never tripped before i changed the old plug in fuse board to a modern one, it no longer trips with a normal double socket in and all tests okay, so clearly the rcd socket but not sure what exactly

 

[Octopus]

Thing is, removing the RCD, which IMHO isn't necessary, fixes the problem .....

next

 

[freddo]

Those Power Breaker RCD sockets get damaged if insulation tested, and end up with all sorts of faults. I've managed to wreck 3 so far.

If you have RCD protection at the board, then get rid of the RCD socket.

 

[Gavin John Hyde]

Hi - if you are going back, perhaps test the pumps too ?

I didnt test the pumps themselves, I did plug them into the upstairs ring via an extension lead and they didnt trip the main rcd for the ring. After these were fitted the heating engineer had come and filled the space with pipes making it a damn awkward space to work in, i should get a gold medal for contortionist skills for this job

 

[Deleted member 9648]

Still not clear what the IR reading was on the pumps though?
Edit as post above crossed over......It'd be useful to eliminate a pump fault

 

[Gavin John Hyde]

Still not clear what the IR reading was on the pumps though?
Edit as post above crossed over......It'd be useful to eliminate a pump fault

It could be good to know for future reference and peace of mind, I am going back in few weeks to fit some extra sockets, change some plastic accessories for brushed steel so may if time permits have a look at the pumps a bit more.

 

[littlespark]

Faulty RCD socket. Not needed now so remove it and replace with normal socket.
You cant test the RCD socket now because of the CU RCD, (unless you bypass) but IMO the socket is stuffed

 

[Richard Burns]

I would say that with the socket in place the house RCD trips, with the socket removed the house RCD does not trip. Ergo the RCD socket is faulty.
Why it is faulty is difficult to say but you could try installing a new RCD socket as a test and see if the house RCD then trips, however with the access limitations this might not be worthwhile (or chargeable to the customer) but since effectively removing the redundant socket solves the problem stick with that.
You could take the socket home and try running things on it and see if it trips your RCD!

 

[Gavin John Hyde]

You could take the socket home and try running things on it and see if it trips your RCD!

Done just that and put on my kithen ring, it doesn't trip my Hager RCBO, board i fitted was BG. So can't explain that.
Customer hasn't rung me to say its gone bang this evening so all seems good!

 

[telectrix]

Faulty RCD socket. Not needed now so remove it and replace with normal socket.
You cant test the RCD socket now because of the CU RCD, (unless you bypass) but IMO the socket is stuffed

post #3.

 

[Toneyz]

Sorry, Gavin bit late on your post agree what others have said faulty RCD socket. If you're going back to do further work suggest incorporating new double socket on RFC that is accessible.

 

[Leesparkykent]

Done just that and put on my kithen ring, it doesn't trip my Hager RCBO, board i fitted was BG. So can't explain that.
Customer hasn't rung me to say its gone bang this evening so all seems good!

 

[Gavin John Hyde]

Sometimes the BG boards if you have only a few circuits are okay, sometimes they are a damn pain and do not align properly!! It is very much a lottery. If I have quite a few circuits then i have been fitting Schneider boards of late as they are good value and actually very good quality.

 

[Lucien Nunes]

Those initial IR readings, in the 1-3MΩ range, what was causing those - the pumps, socket or immersion? Clearly even 0.9MΩ isn't going to affect an RCD much with 0.26mA of leakage, but they are low enough to suggest a problem worth finding.

 

[coolsparx]

RCD sockets & spurs, surge protected leads & USB sockets all play havoc on test results if not found & removed prior to testing.

 

[Gavin John Hyde]

Those initial IR readings, in the 1-3MΩ range, what was causing those - the pumps, socket or immersion? Clearly even 0.9MΩ isn't going to affect an RCD much with 0.26mA of leakage, but they are low enough to suggest a problem worth finding.

When i took the rcd socket out of the circuit the readings all went upto the 199 range. So clearly the rcd socket was causing issues, for some reason the socket tripped the main cu rcd. as the pumps on a newly installed standard socket did not trip the cu rcd. which indicates problem being the rcd socket.