Insulation Test Results

[alancoulson]

A caravan site is having problems with tripping on one of the mains incomer switches.

I have isolated the cable that feeds the distribution board that appears to be where the problem is and have carried out insulation tests with the following results.

N - L1 = >500meg ohm, N - L2 = 86m, N - L3 = 220m

L1 - L2 = >500m, L1 - L3 = >500m L2 - L3 = 220m

E - L1 = >500m, E - L2 = 84m, E - L3 = 180m

L2 is quite a bit lower and L3 is a little low compared to L1 but even at 84meg I wouldn't expect this to be a problem, this dis board then feeds another 2 dis boards with several hook up points fed from the second dis boards. Non of the dis boards have their own ELCB protection so the only ELCB protection is on the mains feed to the first dis board.

the mains feed protection is 320A with 300ma, 1 SEC delay ELCB.

Any thoughts would be appreciated.     

 

[ProDave]

an incommer SWITCH won't trip.

So what is it, an MCB, and RCD or what?

 

[alancoulson]

OK its a BSEN 60947 type 2 rated @ 300A 36Ka breaking, RCD.

 

[Andy™]

so you have tested a cable feeding the board that 'appears' to be the probem, but not tested anything else to actually see what is causing the problem?

 

[alancoulson]

all I want is a second opinion on the test results, if as I think they are acceptable the problem gets much bigger as there are a lot of small dis points and hook up points.

 

[Sidewinder]

On the face of that the results "look" fine.

However, you would need to calculate the cumulative leakage to see if there is a real issue.

Can you clamp the cores to check the actual leakage?

 

[alancoulson]

yes I think I can, thanks I hadn't thought about doing this I should be able to detect which phase has the fault or if more than one phase has leakage. Thanks for the input.

 

[binky]

if you've got an RCd as the incomer for the whole site, you'll be a long time looking for the fault :facepalm:

is there no 30mA RCDs for the individual pitchs? Think I would switch of 1 dis board at a time, and see which one clears the fault, then work on that board.

 

[NozSpark]

Also,, the current regs say that every individual socket outlet has its own RCD protection

 

[ProDave]

Yes a very unsatisfactory install for a caravan site.

Get rid of that rcd incomer, and do the job properly with an rcd for each bollard at least.

It only takes one caravan with something dodgy and the whole site is off.  Does the warden spend all his time resetting it? and how does the warden determine which caravan has the fault?

 

[NozSpark]

Could be TT though

 

[binky]

Trying to think how old that type of RCD is? The fact it was initially desribed as an ELCB probably answers that!! Might just be a case of it knackered after many trips over many years....

 

[Sidewinder]

So the incoming "main switch" is a BS EN 60947-2, 320 (or 300?) Amp, 300mA, with a 36kA breaking capacity, and an adjustable time delay, set to 1s?

Is this also an ACB?

Do you know the make & model?

Is it a 3 or 4 pole device?

Is the supply, TT, TN-S, TN-C-S, or other?

 

[Blue Duck]

So the incoming "main switch" is a BS EN 60947-2, 320 (or 300?) Amp, 300mA, with a 36kA breaking capacity, and an adjustable time delay, set to 1s?

Is this also an ACB?

Do you know the make & model?

Is it a 3 or 4 pole device?

Is the supply, TT, TN-S, TN-C-S, or other?

Just take a picture. ...

 

[alancoulson]

its a TT supply with Ra @ 67ohms, the switch is 4 pole, i'm  now unsure exactly what, but I will check tomorrow the earth leakage trip is adjustable from a 1000ma to 300ma, from 1 sec to 300ms.

fed off this switch are 2 dis boards with MCB's feeding various hook ups, all with rcd or rccbo protection for each socket that's just basic. also there are 2 toilet blocks which have RCD protection in their separate CSU.

So if the feeder cable is OK I am thinking it might be a cumaltive build up of earth faults from each dis board going back to the main feed control switch.  

 

[binky]

it could be a cumaltive build up, or an actual problem with the feeds to the pitchs. Is the tripping is fairly random, which would suggest cumalative, or at fairly set times, which would suggest a particular cct when it's in use. There's only really 2 ways to approach this, either test for mA leakage, or take a walk and eyeball the system and hope to spot some obvious fault, like a damaged IP junction box sat in a puddle. Methodical testing is the professional approach to ensure you clear fault(s).