Verifying polarity on radial ccts

[Hudson]

Hi
I've got my 2391 assessment at the end of the week and I've managed to confuse myself (and not for the first time)
I get how performing R1+R2 and R1+Rn on a ring also confirms polarity.
The thing that is confusing me is how does performing only a R1+R2 on a radial socket cct (and operating switches to make sure it breaks) confirm polarity.
I get it confirms you are not switching the neutral but it doesn't confirm that the line and cpc are not swapped over. I can't see what dead test would confirm this isn't the case and you don't want to be energising a cct if the line is connected to the CPC terminal!
Thanks

 

[pc1966]

R1+R2 as you say only confirms L and E are somehow connected to L and E of the socket.

R1+Rn confirms that L and N are somehow connected to L and N of the socket.

Together you know that L has to have been at the L socket. (And thus E and N are also right)
[automerge]1598194159[/automerge]
In this sense it is no different to the ring case, there you also depend on the pair of tests to know unambiguously that you have them wired up right.

What the ring test adds is a very good check of resistance as you know they should all be the same in the figure of 8. A poor joint or dodgy switch, etc, with even 0.05 ohm extra is easier to spot than the radial case where at different lengths to multiple sockets you have different cable resistances to allow for.

 

[radiohead]

Hi
I've got my 2391 assessment at the end of the week and I've managed to confuse myself (and not for the first time)
I get how performing R1+R2 and R1+Rn on a ring also confirms polarity.
The thing that is confusing me is how does performing only a R1+R2 on a radial socket cct (and operating switches to make sure it breaks) confirm polarity.
I get it confirms you are not switching the neutral but it doesn't confirm that the line and cpc are not swapped over. I can't see what dead test would confirm this isn't the case and you don't want to be energising a cct if the line is connected to the CPC terminal!
Thanks

R1+R2 as you say only confirms L and E are somehow connected to L and E of the socket.

R1+Rn confirms that L and N are somehow connected to L and N of the socket.

Together you know that L has to have been at the L socket. (And thus E and N are also right)
[automerge]1598194159[/automerge]
In this sense it is no different to the ring case, there you also depend on the pair of tests to know unambiguously that you have them wired up right.

What the ring test adds is a very good check of resistance as you know they should all be the same in the figure of 8. A poor joint or dodgy switch, etc, with even 0.05 ohm extra is easier to spot than the radial case where at different lengths to multiple sockets you have different cable resistances to allow for.

I get where the OP is coming from. An R1+R2 test on a radial is carried out with the line and cpc joined either at the DB or at the farthest point on a circuit. A low range resistance test is then carried out at the opposite end between line and cpc. The test will confirm that the line and cpc are continuous but will not confirm that cpc and line connections at a socket outlet have been reversed. The test will produce the same result at a socket outlet whichever way line and cpc are connected. One way to verify line and cpc are not reversed at a socket outlet would be an R2 test to the earth pin at each outlet, but this is never carried out on socket outlets during initial verification. Or as stated an R1+Rn....I am not aware of anyone carrying out that test either.
But it is a requirement that testing is carried out alongside visual inspection, which should be sufficient to spot a wrong connection before a circuit is energised.

 

[pc1966]

So you would not do a R1+Rn test on a radial?
I'm confused here!

 

[Ian1981]

The cpc should be tested at each point of the circuit for initial verification

 

[radiohead]

I don't carry out an R1+Rn.

 

[Ian1981]

I don't carry out an R1+Rn.

Except on ring final circuits?

 

[radiohead]

So you would not do a R1+Rn test on a radial?
I'm confused here!

The cpc should be tested at each point of the circuit for initial verification

It should, but the point is an R1+R2 test at each socket outlet on a radial will not verify L and cpc are in the right terminals.

 

[Ian1981]

It should, but the point is an R1+R2 test at each socket outlet on a radial will not verify L and cpc are in the right terminals.

It will if aided by a visual inspection but obviously you do not want to be removing every thing on the circuit

 

[radiohead]

Except on ring final circuits?

We are discussing radials
[automerge]1598197830[/automerge]

It will if aided by a visual inspection but obviously you do not want to be removing every thing on the circuit

I stated that in post #3

 

[pc1966]

Sockets with the obvious bare CPC terminal are very unlikely to be wired L/E swapped for sure, but those with 3 similar looking terminals do have that risk.

Going round with a wander lead and poking the fixing screws is certainly an easy way to verify CPC connectivity and then you know the R1+R2 verifies L/N polarity.

 

[James18288]

On initial verification R1+R2 needs to be checked at every point on the circuit not just the furthest. This will then prove CPC continuity and polarity. if you can’t see the connections, for example sockets on a radial, what’s to stop you doing a quick R1+Rn or a R2 at the earth pin if you can’t verify polarity visually?

On a periodic inspection, R1+R2 could be checked at just the furthest point or it could be checked at every point, it would be up to you as the inspector and whoever has ordered the test to agree limitations. Also the test sequence doesn’t have to be followed to the letter as the system is already energised so polarity can be confirmed via Zs. Although granted I wouldn’t have done that in my 2395 exam and doing the tests in order and taking a couple of sockets off to confirm polarity is no bother for the exam.

 

[Pretty Mouth]

So you would not do a R1+Rn test on a radial?
I'm confused here!

I was taught to test R1+Rn on a radial as well as R1+R2. It makes sense to me as it confirms polarity and shows any poor connections. But I note that there's no mention of it in either the OSG or guidance note 3

 

[James18288]

I agree with you and think I was taught the same. I know I was always told that there is nothing stopping you recording the values on an EIC even if they aren’t required (R1+Rn or resistance of MPBC). Even though there is no mention specifically of an R1+Rn, I think it does say you have to prove visually or with a low resistance ohm meter and this is probably the best way to do that. Even on a large 32a radial it’s only going to add another 10 minutes to your testing

 

[Hudson]

Thanks for the replies .
I guess a visual inspection would probably do the trick. However, in the exam I understand there are some accessories which you are not allowed to take off. I guess on a test rig the easiest thing to do is an R1+R2 and an R2 which would tell you all 3 connections are ok. Seems strange GN3 doesn't cover this eventuality in it's test sequence.

 

[Lucien Nunes]

I always test R1+Rn during commissioning, not least because it seems daft not to verify each conductor that you have just installed. And as per the above explanation by PC1966, the combination of testing R1+R2 and R1+Rn proves the line conductor is right so therefore the other two must be right.

 

[pc1966]

Exactly. You would not IR test only one of the L/N conductors, so why only check loop resistance of one?

 

[radiohead]

There is no requirement or even suggestion in GN3 or Bs7671 of carrying out a R1+Rn test on radials (as far as I am aware), which is why I don't, and I would suspect I am not alone given the deafening silence from many of the forum stalwarts on this thread. I am confident that a combination of 'required' testing and visual inspection is sufficient when verifying my own work, but taking on board the comments on this thread I will consider carrying out the additional test in future, particularly when verifying radials in which others have been involved in the 2nd fix.

 

[pc1966]

To be honest I never clocked that GN3, etc, did not included that for radials as I just assumed it would be the same idea (check all 3 wires) but without the figure-of-8 loop or the end-end r1/r2/rn checks.

I see the point that if you know the L & E are not swapped from inspection then R1+R2 is your key point for over-current ADS to be measured and compared to the MCB Zs, but having had the odd bad socket in the past always though it wise to check them all.

Without wanting to upset anyone over testing strategies, it might go some way to explaining the common argument that radials are much faster to test than RFC!

 

[Vortigern]

Good points above. For myself, I do not check continuity of L-N. There are two scenarios where such checking may take place a new installation and an existing installation.

A new installation means I connect the various cables to the termination points. I know by visual sighting that the Earth is connected to the earth point at the socket, likewise the Neutral and Live. Therefore I just do R1/R2 on all sockets. On commissioning the circuit loop test confirms polarity and correct connection. The R1/R2 test also shows that I have not connected the N to the L or E terminal. Which I already knew when I wired the socket.

Existing installation; What am I doing there? EICR or fault finding. The first test I usually do by way of investigation as it is easiest is EFLI and L-N loop test. This tells me that All cables are teminated (or not) correctly, it confirms (or not) polarity and shows up any sockets that are defective. Any anomalies are picked up and investigated from there.
Anything wrong with that?

Bottom line; As you may all know there are occasions when you may be fitting up to fifty or more sockets in dado trunking, are you really telling me you will do R1/R2 R1/Rn when you know you just wired all those sockets when you could just pick up any problems with EFLI test, your meter will complain if you get it wrong!