Testing when modifying a circuit

[lozarus]

Afternoon all. Bear with me on this one...

Despite having been involved in the electrical industry for many years, I'm only just about to embark on 2391 - its more of a personal endeavour than anything, having been very interested in inspection & test for a long time (I know, not everybody's idea of a good time). I've obviously read the GN back to front and am enjoying the Chris Kitcher literature.

I'm preemptively trying to put this in the context of my employment (I'm a Head of Estates with electrical bias). I've asked a few of the different electrical contractors (who are all really supportive with my learning) we use for large projects and got mixed answers - I'll share these later. My question is as thus:

Like for like replacement of accessories and fittings (I include changing a flourescent for an LED light fitting in this for example). I know in general this is viewed by the industry as "maintenance" and no testing isn't strictly required (I like to do a Zs however to prove a CPC to the point I'm working at and check ring continuity with my MFT).

However when it comes to modifying a circuit - extending an RFC to add additional sockets, providing an FCU as a spur off an RFC, adding lighting to a radial and so on - I'm a bit unsure as to what's expected / what people actually do in the real world regards testing.

The MEIWC suggests: R1 + R2 / R2, RFC Continuity, IR, Polarity and Zs

My understanding is that to complete R1+R2 and your IR would require isolating at the board, disconnection and testing from the source.

My feedback from various people is mixed as to whether all these tests are done on every such job. I note the MEIWC form states "where relevant and practicable". So would it for example be reasonable to omit an R1+R2 and IR if its unreasonable to isolate the board to facilitate this and all other tests prove acceptable (and ideally, results from a previous EICR are available).

I get a feeling the answer is going to be "its up to the policies, procedures and risk assesment of the organisation". Well guess who writes them for my department? Hence wanting to seek some thoughts on this.

Interestingly I spoke to the insurers about this a few months back. After some consulting with the underwriters and technical experts their response was "everything's generally OK as long as you have evidence of your 5 year EICR".

Thoughts appreciated.

 

[loz2754]

I know in general this is viewed by the industry as "maintenance" and no testing isn't strictly required

Testing is always required, even if no certification is produced. An in-house maintenance form could be used to record the results. (Zs)

So would it for example be reasonable to omit an R1+R2 and IR if its unreasonable to isolate the board to facilitate this and all other tests prove acceptable

I would say that, even if the existing circuit cannot be fully tested due to operational reasons, the extended part of the circuit should have the full range of tests done and results recorded on the MEIWC. If fed from an RCD, then RCD results will need to be provided.

 

[westward10]

I think you are on the right track, test first or repent later we have seen this many times on here.

 

[cloverman]

Have always had to carry out testing and documented on inhouse recording system

 

[Eddy Currents]

My understanding is that to complete R1+R2 and your IR would require isolating at the board, disconnection and testing from the source.

You obviously isolate the circuit to extend the circuit so once finished then you do your IR and R1+R2 on the isolated circuit

You can either calculate the ZS or test the ZS on the newly extended circuit once the circuit is back on

There should be no reason for you to work on a live circuit and so no reason you cannot test that circuit once you finished

 

[westward10]

You obviously isolate the circuit to extend the circuit so once finished then you do your IR and R1+R2 on the isolated circuit

You can either calculate the ZS or test the ZS on the newly extended circuit once the circuit is back on

There should be no reason for you to work on a live circuit and so no reason you cannot test that circuit once you finished

No you test before you start then confirm your new alteration after.

 

[timhoward]

You obviously isolate the circuit to extend the circuit so once finished then you do your IR and R1+R2 on the isolated circuit

You can either calculate the ZS or test the ZS on the newly extended circuit once the circuit is back on

The thing is, the OP is right in as much as none of us should be doing dead-testing within a live enclosure. A lot of us are comfortable doing so after lots of practise. (That isn't an encouragement to anyone to do this though)

There should be no reason for you to work on a live circuit and so no reason you cannot test that circuit once you finished

You could then energise a dead short. The point is initial verification is to make sure it's safe to energise.

 

[Eddy Currents]

No you test before you start then confirm your new alteration after.

Ideally yes, but not really required to test before as long as it is tested afterwards and results acceptable before circuit is put back on

The point I was trying to make is the tests can be done on the dead circuit even if the rest of the board is live

 

[westward10]

Never alter a circuit until you have confirmed insulation resistance and earth continuity as a minimum. There are several threads on here where people have fallen short on this then have to discover is it me or was it already there.

 

[lozarus]

You obviously isolate the circuit to extend the circuit so once finished then you do your IR and R1+R2 on the isolated circuit

You can either calculate the ZS or test the ZS on the newly extended circuit once the circuit is back on

There should be no reason for you to work on a live circuit and so no reason you cannot test that circuit once you finished

Isolate and lock off the circuit yes....but (and I know everyone does it) - the works wouldn't necessarily entail opening a live 3ph board supplying power to other live services to facilitate the test of one circuit with a minor addition/alteration

Of course I know it's done, but I'd almost consider that a slightly greater risk than omitting an R1+R2 - especially if there's evidence of recent periodic testing. Notwithstanding the alteration itself being tested and the original having all other reasonable tests done

Or am I wrong? I'm not challenging anyone here. Just ----ing ideas about and fact finding.

 

[timhoward]

Of course I know it's done, but I'd almost consider thay a slightly greater risk than omitting an R1+R2 - especially if there's evidence of recent periodic testing.

It's particularly important to ensure that ADS (automatic disconnection of supply) will occur under fault conditions. The only way to be sure is to know that the CPC path has a sufficiently low impedance to allow enough current to flow to trip the protective device.

Worst case - no IR testing so you don't know you have a N-E fault in your new section, no R1+R2 testing so you don't know you have no earth path back, and you could end up with metalwork that's gives someone a shock if touched.

If I had an EICR showing a circuit was recently in fine health, and a board that was particularly problematic to work on for any of a number of reasons, at an absolute minimum I might do a live Zs before I start on the unmodified circuit, IR the new section, and do a live Zs when I finish. This at least tells me that there is earthing and no cable faults in my addition/alteration.
I'm not encouraging this sloppy practise but it's better than no testing at all.

 

[westward10]

Evidence of an EICR means little unless you have access to the Report. The cìrcuit you are working on may have Code 2 issues.

 

[lozarus]

It's particularly important to ensure that ADS (automatic disconnection of supply) will occur under fault conditions. The only way to be sure is to know that the CPC path has a sufficiently low impedance to allow enough current to flow to trip the protective device.

Worst case - no IR testing so you don't know you have a N-E fault in your new section, no R1+R2 testing so you don't know you have no earth path back, and you could end up with metalwork that's gives someone a shock if touched.

If I had an EICR showing a circuit was recently in fine health, and a board that was particularly problematic to work on for any of a number of reasons, at an absolute minimum I might do a live Zs before I start on the unmodified circuit, IR the new section, and do a live Zs when I finish. This at least tells me that there is earthing and no cable faults in my addition/alteration.
I'm not encouraging this sloppy practise but it's better than no testing at all.

Genuine question here - and I'm probably missing something obvious. You say no R2+R2 means no proof of earth back - surely a Zs measurement within the specification of the device proves there is?

Again I may be missing something here...

 

[lozarus]

Evidence of an EICR means little unless you have access to the Report. The cìrcuit you are working on may have Code 2 issues.

I commission the EICRs here - and know the results like the back of my hand. Appreciate that's only relevant in my scenario but still.

 

[timhoward]

Genuine question here - and I'm probably missing something obvious. You say no R2+R2 means no proof of earth back - surely a Zs measurement within the specification of the device proves there is?

A Zs certainly gives a strong indication all is well which is why I suggested doing this if a board truly can't be worked on. That said...

An R1+R2 is proving good connections and path along a specific CPC.
A live Zs test can prove there is a path back, but it could well be via metal trunking, bonding and a copper water pipe.
But as above, it gives some reassurance and is a lot better than nothing.