Insulation Resistance Test on Ring Circuit

[JohnW2]

If there is a 0 reading for a continuity test ...

What do you mean by a "0 reading"? If you mean "0 Ω", then that would be a 'pass' in a 'continuity' test - but I presume you are talking about a 'fail' (broken/disconnected conductor), aren't you?

and there is no obvious visual clues, when would the ring halving method not be a plausible approach to proceed?

Plausible, yes - but, as I said, not necessarily the best, or necessarily even the most sensible, approach.

Also, once you have tracked down the fault and subsequently proved with a temporary cable that the tests are now successful...

I'm not sure that is really necessary. If you have properly identified the nature and location of the fault, there's probably no need to prove it further by 'bypassing the fault.

, is there any issues with the 2 radial approach, if replacing the cable is deemed impractical?

Provided you are certain that the only fault is in the one segment you've taken out of service, then I can see no problem with that approach Indeed, a good few people would probably have preferred it as two radials (rather than a ring) in the first place!

I’m sure an expert understanding of electrical concepts will provide one with more options but I’m trying to understand if there are any obvious flaws with my suggested approach.

That understanding (of whether there were any 'obvious flaws) would also result from having "an expert understanding of electrical concepts'.

Kind Regards, John

 

[Harry Bloomfield]

Understood. Presumably you will only separate when the IR reading is too low when using that combined earthing terminal? That’s when we want to test with each individual CPC to identify where the issue is.

The technique is one of halving the ring, to see which half has the issue, then quartering it. So it progresses until you find the problem section itself.

 

[Jupiter01]

For one thing I would regard a '0 result' as representing continuity (or 0 ohms).

Sorry, I meant the meter maximum which indicates the open circuit.
I was thinking of the insulation test with a zero reading. Would the halving method be a plausible response in that situation too?

You asked me to provide some more detail around the intended approach. Upon a failed test, I would find the midpoint of the ring circuit (or approximate) and remove a socket here. I would then re-test each leg of the ring (now a radial), I would expect one of these to now pass the test and one leg to fail (unless there are multiple faults but let’s assume that is unlikely…). At this point, I will add some more sockets to the good ring (if a small ring, may be one socket at a time). The failing leg of the ring is getting smaller with each additional socket being added to the good ring. I would repeat the test after each socket(s) being added to the previously good ring and at some point, I am expecting this to fail. At this point, I would test the previously failing ring and expect that to pass. I now know where the fault resides. As a further validation, I would take a length of 2.5mm T&E and use it (along the surface) instead of the defective section of the ring. I would then retest the complete ring and expect it to pass.

Clearly, the surface cable is a temporary measure and I now need to consider a permanent solution.
This is where I raised the question on whether I can split the two leg of the rings as a radial with 20amp breakers.

I would welcome your feedback on this. Thanks again.

 

[DetlefSchmitz]

I think I was concerned that you do not state between which two points you are 'testing this leg'. It may be obvious or it may be that you had something else in mind. But it's hard to tell if you are doing it right unless you are specific, as there is more than one way to approach this.

 

[Jupiter01]

I think I was concerned that you do not state between which two points you are 'testing this leg'. It may be obvious or it may be that you had something else in mind. But it's hard to tell if you are doing it right unless you are specific, as there is more than one way to approach this.

Sorry, I had P+N to E (whilst connected to the main earth terminal) in mind for this. Is that correct?

 

[Harry Bloomfield]

Sorry, I meant the meter maximum which indicates the open circuit.
I was thinking of the insulation test with a zero reading. Would the halving method be a plausible response in that situation too?

Simple answer - Yes! Halving, quartering and so on, is the standard approach to all similar faults, where there is no obvious visible issue and of course the quickest way to a result.

 

[Jupiter01]

Simple answer - Yes! Halving, quartering and so on, is the standard approach to all similar faults, where there is no obvious visible issue and of course the quickest way to a result.

I thought so too but @JohnW2 comment left me in doubt.

 

[JohnW2]

I thought so too but @JohnW2 comment left me in doubt.

I think some of the confusion results from the fact that we seem to be jumping around a bit between 'continuity' and IR testing. For example, what does this refer to? ...

Sorry, I had P+N to E (whilst connected to the main earth terminal) in mind for this. Is that correct?

As for ...

.... At this point, I will add some more sockets to the good ring (if a small ring, may be one socket at a time).

As Harry has said, the normal (and most efficient) process would be one of 'successive halving', rather than successively adding some (maybe one) sockets to the 'good leg' each time.

Kind Regards, John

 

[Jupiter01]

As Harry has said, the normal (and most efficient) process would be one of 'successive halving', rather than successively adding some (maybe one) sockets to the 'good leg' each time.

Apart from the adding-one-socket-at-a-time to the good leg taking longer than successive halving, are there any issues? I would have thought that the time aspect is negligible too, as the domestic ring circuit is likely to be relatively small.

 

[Harry Bloomfield]

Apart from the adding-one-socket-at-a-time to the good leg taking longer than successive halving, are there any issues? I would have thought that the time aspect is negligible too, as the domestic ring circuit is likely to be relatively small.

It can take a surprising amount of time, in a domestic environment to trace faults, if you think about it. All that guessing at how the ring has been run, moving furniture and lifting floorboards, assuming a cable fault. Sometime a signal type cable tracer system can help with the routes cables take. Imagine the difficulty of tracing a single floorboard nail, which has been put through such a cable, somewhere on a ring and that happens with fair regularity.

Usually all much easier in an industrial environment, where cable routes are on the surface, much more accessible and damage is much more visible.

 

[JohnW2]

Apart from the adding-one-socket-at-a-time to the good leg taking longer than successive halving, are there any issues? I would have thought that the time aspect is negligible too, as the domestic ring circuit is likely to be relatively small.

There are no issues other than time, and I agree that if the ring is small it won't make a lot of difference - but nor is there a reason for not using 'successive halving'. If there are more than half a dozen sockets, successive halving will, 'on average', usually save time.

You seem to have bought into that idea to some extent by doing your initials 'halving'. You could, if you had wanted, have started at one end of the ring and then moved along one-socket-at-a-time if you wanted, but you obviously realised that it was more efficient to start by 'halving' - so why not continue halving?

One warning ... the exercise can become difficult and confusing if (as sometimes happens) the way in which the sockets have been wired does not correspond to what you would expect from the physical location of the sockets - since what you thought was 'the next socket along the ring' might not be!

Kind Regards, John

 

[Jupiter01]

One warning ... the exercise can become difficult and confusing if (as sometimes happens) the way in which the sockets have been wired does not correspond to what you would expect from the physical location of the sockets - since what you thought was 'the next socket along the ring' might not be!

Is unwiring subsequent sockets and running a continuity test between wires (L,N and E as one of them may have a break) from this socket to what you suspect is the next one, the surest way to proceed and ensure you are moving through the circuit in the right order?

 

[JohnW2]

Is unwiring subsequent sockets and running a continuity test between wires (L,N and E as one of them may have a break) from this socket to what you suspect is the next one, the surest way to proceed and ensure you are moving through the circuit in the right order?

Although not unheard of, I would imagine that the situation of the order of sockets (on the ring) not being 'as expected' is probably pretty rare, so I don't think it makes sense to anticipate the possibility in every case.

If you get to the end of your systematic testing process ('progressive halving' or whatever) without locating the fault, with certainty, then you can start scratching your head and contemplating procedures such as you suggest.

Bear in mind that the more sockets you 'disturb', disconnect and then reconnect in the name of testing, the more likely are you to introduce new faults into the circuit [this is another reason for favouring 'successive halving', rather than 'one-socket-at-a-time']

Kind Regards, John

 

[EFLImpudence]

Has it been said that to make the ring circuit two legs for testing, it must also be split at a second location; usually the CU?

 

[JohnW2]

Has it been said that to make the ring circuit two legs for testing, it must also be split at a second location; usually the CU?

I think that it depends upon what testing one is talking about, and how it is being done. Testing 'continuity' of conductors (of one leg) back to the CU does not necessarily require the leg being tested to be disconnected from the other one at the CU - although, in practice, one probably would do that.

In terms of IR testing, I don't think it has been said explicitly, but would hope it is obvious that two legs cannot be tested 'separately' if they are joined together (at the CU or anywhere else).

Kind Regards, John