Modifying existing power

my client only seems to want to test the new section

Up to a point, that might be right - in that technically you're only responsible for your own work rather any existing parts of the installation.

However in this case you're making an alteration as well as an addition, so you need to ensure that the alteration doesn't adversely affect the existing installation - which it is quite likely to do if you're effectively lengthening the cables. For instance you need to be sure that Zs and voltage drop at the far end of the circuit are still within acceptable limits (or no worse than they were before if they were already beyond current limits).

In any event, some tests - e.g. insulation tests cannot be done without involving the existing installation - if you make the test prior to connection to the existing then you're not testing part of your new work. Ditto for c.p.c. continuity.

The distinction between old and new can be useful with some tests however - say the existing installation insulation tested to something awful (<< 0.5Meg say) - provided you tested both before and after you should be able to verify that your work had a perfectly adequate insulation resistance even if the rest of the circuit didn't. (Hence you can still fill out a cert (if with something rude in bit letters written in the comments on the existing installation box) finish the job and get paid even if the customer refuses to have the existing improved.)

   - Andy.

I am not at all sure that I understand the concept. 2 new ends implies that there will be a new segment from DB to old cable; the old cable; and a further new piece to the loads. It may be possible to test CPC continuity of the separate parts, but I am at a loss as to how Zs could be measured without a complete circuit.

IR testing is mandatory - 643.1. It is to be conducted with final circuits connected - 643.3.2.

So I think that James is right and the client is wrong.

And yes, this is initial verification, so test to BS 7671:2018+1. It might be argued that these are not new circuits and a MEIWC is all that is required. That still has places for IR and Zs.

I would fully test the circuit before starting to identify any existing issues. 

Then testing of the entire circuit will be required upon completion to ensure that any new issues have not been created. 

If it were just an extension of an existing circuit I might limit the insulation testing to the new work, but it’s argumentative if that’s okay, read the minor works certificate to see what you are actually signing for.

Andy, just the kind of insightful response I required. So, I'm not going mad then.

Thank you Chris, the circuit is from DB, old cable, new cable, old cable, load. I take your point on the mandatory testing. Thanks. 

Thank you

James Edward Duncan: 
Thank you Chris, the circuit is from DB, old cable, new cable, old cable, load. I take your point on the mandatory testing. Thanks. 

So just splicing in a new bit of cable - I had imagined it the other way.

I suppose that there is an argument that it doesn't need to be tested at all on the basis that it is repair work and not initial verification.

Back in 1970 something we had an argument with a Building Inspector, he wanted the extension to a sewer installed in the 1930’s around forty years earlier tested and the only way to test the new part was to include the old part in the test.

We said there was no point testing it as it would fail because of the old part and we weren’t under any obligation to replace the old part if it leaked, so he just needed to accept we had made a good job of the new part and let it go without testing.

In the end I went to the Council Office and had a word with his boss, he laughed and said he would tell him to sign it off without testing, because it was obvious the old part would leak.

Unfortunately though you cannot  extend a faulty circuit if it comprises the safety of the new part, so the existing part needs to be in an acceptable condition.

Hm! I think that I can see the point. So you do a Zs at the furthest appliance and it fails. Then you go back to the downstream junction and it fails; and then the upstream junction and it also fails. Shouldn't you have tested it before you started?

My ideal way to handle the job would be to test the existing circuits first, determine that they are suitable and safe for continued use, (an EICR would be ideal) and then do the new works, and then re-test the whole lot. 

There would seem to be little point in only testing the new bit - if its in the middle of the circuit especially. 

An addition on the end of a circuit - well perhaps just checking the new bit has some merit, but - not if the addition is at the start or in the middle of the circuit……..