RCD and replacement fails to operate

[PEG]

A suspected faulty RCD,should not be condemned,until it is tested in isolation,that is,removed,from any and all other possible,additional faults. That way,you are only testing one component,and can diagnose accurately (EM issues aside)

 

[Des 56]

You should test the earth loop impedance of the rod rather than a ZS
It may be that it does not exist or is exceptionally high
Couple that with a neutral-earth fault and the Rcd may yet see no imbalance

 

[Ian1981]

Turning off the MCBs will not remove the neutral connection! You need to remove the neutral link from the RCD as well.

It won't isolate the neutral but I think he's just saying remove any loads on the outgoing circuits first and test the rcd without anything drawing current.
With a Ra so low apparently during his earth electrode test the rcd should trip on a neutral earth fault on a energised circuit.
An insulation resistance test is possibly the next step after rcd is tested again with loads disconnected.
Wonder if the earthing conductor was disconnected from the MET during the electrode test because that Ra is very low

 

[mhar]

There is no such test as a 5 x trip current on a 100mA rcd (only applies to 30mA) so it may be possible your meter 'knows' this and is not doing anything.
What meter are you using? A couple of months ago I went to test 6 mem 30mA rcbos all sitting side by side in a board, none would trip even when I set my meter to ramp test on a 300mA rcd. I was testing at the terminals. However when I tested from the circuits, they all behaved perfectly, contrary to what I was expecting. This was using a Metrel Mi3101

 

[Ian1981]

There is no such test as a 5 x trip current on a 100mA rcd (only applies to 30mA) so it may be possible your meter 'knows' this and is not doing anything.
What meter are you using? A couple of months ago I went to test 6 mem 30mA rcbos all sitting side by side in a board, none would trip even when I set my meter to ramp test on a 300mA rcd. I was testing at the terminals. However when I tested from the circuits, they all behaved perfectly, contrary to what I was expecting. This was using a Metrel Mi3101

I've had this on mem 2000 when using the pods to covert the mcb to rcbo.
Won't work with the wire connected into the terminal, needed to get my long thin prob into the terminal with circuit removed

 

[saspm]

It will be interesting to see the outcome! I've had similar things happen with suppression capacitors being the culprit.

 

[telectrix]

There is no such test as a 5 x trip current on a 100mA rcd (only applies to 30mA) so it may be possible your meter 'knows' this and is not doing anything.
What meter are you using? A couple of months ago I went to test 6 mem 30mA rcbos all sitting side by side in a board, none would trip even when I set my meter to ramp test on a 300mA rcd. I was testing at the terminals. However when I tested from the circuits, they all behaved perfectly, contrary to what I was expecting. This was using a Metrel Mi3101

this is common with some RCBOs; a short length of 2 core flex in series with your test leads will overcome it.

 

[Octopus]

Wonder if the op has done any IR tests.....

 

[Risteard]

If there was a N-E fault as soon as load turned on the RCD would trip if I am not wrong

Not always.

 

[Midwest]

I've experienced a non tripping RCD (including test button) on a TN-C-S install, with an N-E fault on one circuit, even had similar when doing monthly calibration on my MFT, having left the washing machine connected.

I know someone explained the reasoning before, but would the same issue be replicated on TT supply?

 

[westward10]

If you have a neutral/earth fault on a TT system the impedance of the earth path maybe such that the current will choose the neutral route and hence no imbalance is detected.

 

[Adam Lench]

I'd definitely recommend disconnection of the load. Would be interested to know how you get on. Good luck

 

[weevilward]

Yeah I learnt that the hard way too, in a hurry to get out of a pub (!) before they opened - did the live RCD testing without fully unplugging everything - found the trip times were really slow - way past the limits, so I thought I should change it. Came next day with a new RCBO. When I powered everything down and removed the outgoing conductors I thought "I'll just do one more test before I remove it..." and it tested spot on. That was a Merlin RCBO, and the cause was a fridge. I've seen the same effect a few times and it only seems to be things with motors that cause it ???

 

[jibjob]

I had a similar problem on a TT a few years back due to a N-E fault. I can't remember the details but I'm sure it was a single split load board with a main switch & a standalone 100mA RCD. The fault was on the non-RCD side of the board which added to the confusion. Once I'd located the faulty circuit I left the neutral disconnected & then retested the RCD & hey presto.......!
This may not be the issue here but it sounds very familiar to the one I had, hope this helps if you haven't already sorted it.