Rcd fault not tripping

[Pretty Mouth]

no continuity between any circuits with mcb's off, no continuity between neutral bars with rcd's off and no neutral earth / live earth with main switch off until boiler is connected.

Just checking: this was a continuity (low ohms) test? Have you tried an IR test at 250V between the two RCDs? (both RCDs open, all MCBs closed, boiler spur closed, boiler plug connected to boiler. IR test from N and busbar of one RCD to N and busbar of other RCD.)

IR all clear apart from boiler and that reading is 0.00

Have you tried testing this as a continuity test?

I have not done insulation resistance between live and neutral on lights as I haven’t removed all the lamps.

I'd say this is probably unnecessary, it's not something I would bother doing if it was me.

 

[UKMeterman]

Suspect someone has done a dodgy perm live for the heating and is crossing over the two RCDs.

 

[Tom256]

Wanting to explain this yet do not know how you are testing so basics:

One reason why a RCD won't trip is a N - E fault that is allowing a small part of the current to flow back along the Neutral. Because of this the RCD does not see the full imbalance that the test is creating.

A small enough resistance that your IR tester won't have the resolution/range to indicate it :
Q. you are showing 0.00 M ohms - Is that correct?
I have assumed you are reading this at the boiler plug.
Q. Same test using the circuit continuity setting on your MFT and a resistance reading that will be greater than the range on your continuity test - perhaps >200 ohms - Is that correct? .

I hinted at this a few posts back. I gave an example figure of 1500 ohms and asked what would this resistance read if measured using your insulation tester.

Q. The boiler once cleared from the system (unplugged) reverts all RCD's to operational. When the boiler is plugged in the RCD will not trip - Is that correct?

A N - E will not be lost isolating the circuit from the rcd by the single pole MCB
A N - E fault in circuit at certain values of current flow can stop an RCD from tripping.

It has been already mentioned to forget the lights for now.

Getting late and delayed coming back to you. Consider what I have said.
I have asked the questions here in this post because I am not clear how you are testing and where your results are being obtained. Hence the N - E situation that could be the issue that I have layed out.