Insulation Resistance Test

[leascliffe]

Hi All,

I'm doing a c/u change for my first assessment which is next week and I'm getting myself confused over the Insulation Testing, hope you can help.

The setup is a TT supply, coming from the meter into a new 100A switch isolator (where the old 100mA RCD was) then 25mm tails across to the new Crabtree Starbreaker dual rccb board.

I have read that the insulation resistance test should be carried out on the whole installation. When I did my training, we just tested between L-N, L-E, N-E with L & N being the incoming side of the main switch.

Please could you let me know how to test the whole of my installation in one go for my assessment? I just don't want to risk damaging the rccb's in the board...

PS. Also there is no visible busbar in the Starbreaker to clip onto.

Thanks.

 

[Engineer54]

How are you going to perform your overall IR tests through 2 RCD devices??

 

[Deleted member 9648]

Test on main isolator load side,and isolator off, with all upstream devices in the CU on....Test L-E,and N-E....If you intend to test L-N you will have to disconnect all loads,or you will get a false reading and cause possible damage......
Sounds like GN3 would be a good investment as this explains all.

 

[SPARTYKUS]

Hi leascliffe

you might refer to the installation-wide N-E test. Clearly there should be complete separation & infinite insulation resistance between these, and you can deduce quickly if there is a problem with an old board say, before upgrading to an dual RCD board as you are doing. So if you connect your meter across the Main earth terminal and the neutral bar, and your result for the entire installation comes back as >299M Ohms (or whatever your posh meter goes up to) then you can safely assume that ALL circuits have an IR reading of >299M Ohms between N-E.

I may have missed a few steps out there but assume you will be following the Onsite guide / or regs / or GN3 to the letter if you are not sure how to perform this test.

 

[Deleted member 26818]

Are the tails from the 100A switch isolator double insulated?

 

[Richard Burns]

When you were training you probably had an isolator before the main switch (as you have in this case) but this is not always the case.
Normally as wirepuller says Main switch off, and test on the load side of the switch not the incoming side.

There is some suggestion that doing an IR test through an RCCB may damage it, so you would then not do an entire installation test in one go but split it up for each RCCB protected section.
All loads removed, disconnected or bypassed.
Switch off each RCCB and switch on each MCB covered by it.
Test at the outgoing side of the RCCB L+E, L+N, N+E
Then test in the same way at the other RCCB.
Whole system then tested RCCBs not damaged.
As previously said GN3.

 

[Des 56]

Leave your Rcds in the off position,switch all mcbs to the on position,then get the 2 pieces of cable with crocodile clips that you keep in your test box, link out an mcb from each section and then 2 outgoing neutral bars and away to go

 

[tony mc]

When doing IR testing between Line & Neutral IMO always do a soft test on 250V first just in case you have missed something still connected better safe than paying out for a expensive piece of equipment !

 

[spark 68]

On a Domestic CU replacement (as opposed to a new installation), is it neccessary to do the L to N test at all ?, as when I was at last years Elex show, there was a seminar on this very subject, and it was suggested to just do a L+N to E test at the new board, to save disconnecting loads of accessories neons, flossies, boiler controls and alarms etc.

Also on a Domestic board replacement, again as opposed to a new install, I usually prefer to test the circuits individually anyway, as although the regs say min of 1Meg, I was always told to investigate anything less than 10Meg, which then usually means splitting the circuits up anyway, besides you need the individual results for the SoTR, having said that I usually do a quick full IR test on the board with the RCD's/main Isolator open before the circuits are connected, just in case of anything naughty.

 

[topquark]

I think the suggestion of L+N->E is just for pretest (before changing the CU, in an attempt to remove any possible tripping equipment from the installation first). I would suggest that you always test all circuits fully after replacing the CU.

 

[Octopus]

Have you installed the new board? If so did you test the IR of each circuit as part of the test?

 

[leascliffe]

Hi,

Thanks for all your replies, really appreciate it:

Are the tails from the 100A switch isolator double insulated?

Yes.

Have you installed the new board? If so did you test the IR of each circuit as part of the test?

I tested each circuit individually on the old rewireable fuse board and all was ok. Have since fitted the new Dual RCD Crabtree Starbreaker and once again tested each circuit individually and all is ok but just wanted to clarify how I test the installation as a whole as noted in GN3.

When you were training you probably had an isolator before the main switch (as you have in this case) but this is not always the case.
Normally as wirepuller says Main switch off, and test on the load side of the switch not the incoming side.

There is some suggestion that doing an IR test through an RCCB may damage it, so you would then not do an entire installation test in one go but split it up for each RCCB protected section.
All loads removed, disconnected or bypassed.
Switch off each RCCB and switch on each MCB covered by it.
Test at the outgoing side of the RCCB L+E, L+N, N+E
Then test in the same way at the other RCCB.
Whole system then tested RCCBs not damaged.
As previously said GN3.

Thanks RB, I'll give that a try and let you know!

Thanks again.

L.

 

[Risteard]

besides you need the individual results for the SoTR

Nothing to stop you testing the installation as a whole. In fact, this is recommended by GN3. The separate entries in the Schedule of Test Results were, I believe, just to make it look neater!

 

[spark 68]

Hi Risteard,

Nothing to stop you testing the installation as a whole. In fact, this is recommended by GN3. The separate entries in the Schedule of Test Results were, I believe, just to make it look neater!

I didn't say there was anything wrong with this, and I know what GN3 recommends, but the Op was having difficulty in doing the test, I was merely suggesting that he could use the individual readings in Lieu, and test in seperate parts, in fact I said he would have to do the individual tests anyway for the SoTR ,as this method is also valid I believe, as his long as his readings were within bounds, and this was likely a plastic (Domestic) CU and a replacement on an existing install, I also said I was taught to investigate any readings under 10 Meg.,

The only other way would be to jumper the RCD's and measure at the load side of the main Isolator, and then run the risk of damaging vunerable equipment etc..etc.., recommendations are one thing, practicalities are another, so as long as he tested the individual circuits I don't see a problem with what I suggested, as I was not compromising safety.

Also on a Domestic board replacement, again as opposed to a new install, I usually prefer to test the circuits individually anyway, as although the regs say min of 1Meg, I was always told to investigate anything less than 10Meg, which then usually means splitting the circuits up anyway, besides you need the individual results for the SoTR, having said that I usually do a quick full IR test on the board with the RCD's/main Isolator open before the circuits are connected, just in case of anything naughty.

Iam not trying to argue the ---- here.

Cheers,
Spark 68

 

[leascliffe]

Hi,

I spoke to Crabtree Tech Department today, they confirmed that a 500V IR Test will not cause any damage to the RCCB's...Hooray!

They went onto say the only risk is to RCBO's, which should be disconnected before the IR test; the test will not cause any damage them, it's just that the RCBO will give inconsistent (fluctuating) readings between tests.

Thanks for all your replies,

L.

 

[Richard Burns]

Interesting to know about fluctuating readings for RCBOs did not know that and certainly useful to remember, if one forgets to disconnect.
I have heard anecdotal evidence that suggests that sometimes RCDs do not work after an IR test so I have always not tested through them, but I may relax a bit now (especially for crabtree ones).