MCB Change on Socket Circuit - No RCD

I am not surprised that the customer does not want to pay for RCDs - 3-phase ones cost a lot more than 3 single-phase ones.

411.3.3 requires RCD protection for socket-outlets not exceeding 32 A. However, other than in a dwelling, they may be omitted where, "a documented risk assessment determines that RCD protection is not necessary."

So if the customer does not want to pay, get him to get his pen out pretty pronto.

In a sense, that shifts liability from you to the customer. The Regs say nothing about the quality of the documented risk assessment.

What Code did you apply for lack of RCD protection when you prepared the EICR?

The lack of RCD was coded as C3 and the failure for max zs was recorded as C2

Yes that was my thinking as well, might be a suggestion to him. I just wanted to check on the requirements to install if only making a small alteration due to a potentially dangerous situation. 

some welding sockets have failed, some due to the size of the MCB in relation to the cables current carrying capacity and some due to Zs. The loading is not an issue so we are looking to downgrade the size of the breakers which will sort out both issues.

You've probably already ahead of me on this, but just in case... some welding gear can have quite significant inrush/switch-on surge currents - so sizing MCBs just on nominal loading and/or reducing the type of MCB might risk nuisence tripping,

   - Andy.

Yes I believe that's why they were changed for a higher rating at some point, the current over rated breakers are not what was initially installed according to the customer, we have now failed them but they no longer have the larger welders and have said they are happy for them to be downgraded again. It is a machining workshop so they will only be used for certain purposes, but I also don't want it to come back and bite me

When you say the welding sockets "have failed" do you mean they have been burnt out,  damaged by overload or just that you have not given them an EICR pass ?

If the former then really they should be changed for the next size up, or the overload removed.

If you mean you have found a 16A socket on a 20A D curve breaker and really you would prefer to see a 16A C type, then it may well be that no instant action is really required, if the load on the sockets is in practice known and not likely to be changed.

Also, the problem of the high Zs may be made acceptable in some cases by lowering R2 relative to R1 - i.e. supplementary earthing, so that during an L_N fault at the far end, the R1/R2 voltage division gives a touch voltage that stays below 50v - then we do not need to worry about disconnection times in quite the same way.

It is not often done, and it may not be the normal way to think about things, but so long as the highest touch voltage under fault conditions is less than 50, or whatever the safe shock limit is for the situation,  the  disconnection time can safely be extended to "never" . Of course in reality it would still go off in a few seconds, and you would really  want it to, if not from a shock perspective, then  from considerations of hot wires and a large electricity bill !!

Mike

New sockets (and possibly the cables feeding them) need RCD protection, but you're not installing new sockets or cables. So I don't think you need to add RCD protection for the new work to be regs compliant.

I know what you mean, but I don't think that it would be compliant without the documented risk assessment.

However, I don't think that remedial work must eliminate all non-compliances, just the dangerous or potentially dangerous ones, i.e. C1 and C2.

644.1.2 basically says that the safety of the new work mustn't be compromised by defects in the existing installation. Which is why if you add a socket to an existing circuit, it must be RCD protected, but there is no requirement to add RCD protection to all the other existing sockets on that circuit at the same time.