Confusing earthing issue

[J_T]

Hi
I am having a few issues with the earthing at my house. This is a bit long winded but hopefully you can understand.

I am not a time served electrician but I have my 17 edition.

The supply to the house from NIE is a TT system. There is 11 KVA over head going to the transformer pole which is about 50 metres from my house. On the transformer pole, from what I can make out there is a 3 phase transformer that is going to a neighbouring farm approx 200 meters away and 2 single phase transformers, 1 of which I am assuming is feeding my house.
Again from what I can see there is an earth spike at the transformer and there are 3 overheads coming away from the transformer pole. 2 of which are the live and neutral that supply my house and 1 which is an earth that travels 1 pole away from my house and runs down that pole to a 2nd earth spike.
It is a new build house that was wired 4 years ago. The live and neutral from the NIE pole runs through a 30 meter duct underground to my meter box in a small brick house 30 meters from my distribution box inside the house. The meter box and distribution board are connected using a 16mm 3 core SWA cable.

I have a 100mA RCD at the meter box which supplies a live an neutral ISCO and an earth spike here that is connected to a 3rd ISCO. My house connections are taken from here.
At my distribution board the house circuits are split into 2 groups, each fed off a separate 30mA RCD.

At the time of installation we tested the earth loop impedance and got a result of 133 ohms.
4 months ago after finishing the wiring upstairs I thought it would be a good idea to have the wiring tested and signed off as acceptable.

The guy carrying out the testing was happy with the internal circuits but when it cam to testing the speed of cut out on the RCDs, the results were intermittent. At 1 times the fault current it wouldnt trip every time and at 5 times the fault it wouldn't carry out the test and displayed <50V (I think it was < and not >).
After checking the onsite guide it does mention in section 11.5 about a 50V rise in the protective conductor potential not sure if this is what it refers to.
The earth loop impedance had risen to >400ohms.

I put in an additional earth closer to the transformer, less than 10 meters from it. I think the ground is quite stony. The earth loop reduced to 300 ohms and we seen the same results for the RCD tests.

As a test I have daisy changed this closer earth spike to the fore mentioned second earth that I am pretty sure is connected to the transformer and actually clamped it to the transformer earth.

Using a different meter to test the earth loop (Fluke) before calling the testing guy back it is throwing up an error 4 message and won't carry out the test.
On further testing there is approx 20VAC between the neutral and earth. No matter what I do this voltage is there. I have disconnected all but my original earth spike. I have tested with all load disconnected so it is just the NIE supply neutral to earth.
This is also tripping one of my 30mA RCDs at the dist board even with all supplied load circuits disconnected so I have temporarily replaced with a 2 pole isolator. I still have the 100mA protection at the meter.

Last night the 100mA at the meter tripped and would not reset. The normal 20VAC neutral to earth was still there, I disconnected the earth to see if that made a difference and isolated the house. The voltage difference had reduced to 0VAC. I though maybe it was something capacitive that was bouncing around between earth spike (grasping at straws).
Today the voltage was still 0VAC I thought I would chance putting the 30mA RCD back in. The voltage earth to neutral was 0VAC, as soon as I tried to connect the RCD it tripped, the voltage was 20VAC again!!

I am really stumped here, I have asked NIE to take a look as I am half suspecting the transformer. I have see it in the past having sparking issues. Maybe the earth I am connected to is from 1 of 2 transformers on the pole and my neutral is from the other..?
Could I have too many earths in? Or generating a potential difference between rods?? I know there is a minimum distance and I am sure I am exceeding this.

They have said they can supply me a dedicated earth but I would have to move my meter into the house which is a lot of work. I am not sure if this is only for TNS as I would be happy with TNCS if they could do that without relocation.
Besides I think I have generated a dedicated earth at present by connecting directly to the transformer spike.

I would really appreciate any feed back on this as I feel we have tried and tested everything we can.

Sorry for the long windedness of this post.

Thanks

JT

 

[J_T]

Ok. I got a lend of a fluke tester.
I plugged in at a plug socket and carried out the loop impedance test (no trip) option. This gave a result of 1.5 ohms which doesnt sound right for tt. It also line up with the assumption that thete is a neutral earth fault somewhere.
I discinnected the outside lights. No difference then continued to disconnect circuits and perform the test untill i seen a different reading for a lighting radial... which was Err5.
I put my normal fluke multi meter across neutral and earth and measured 0.9ohms. I then disconnected the earths at different celing roses until i identified the fault (in the most awkward and out of reach fitting possible).
When i took the light fitting off the resistance went high (>1M).
I tightened up the wiring in the fitting (which is metal and earthed) and reconnected.
Everything seems ok now with the rcd tripping when tested and staying on when reset.
I think the Err5 is when you try and test with a load connected...
I didnt today but will do an insutalion test tomorrow to make sure there are no earth faults but hopefully that was the issue.
Does what im saying sound sensible?

 

[J_T]

Ok. I got a lend of a fluke tester.
I plugged in at a plug socket and carried out the loop impedance test (no trip) option. This gave a result of 1.5 ohms which doesnt sound right for tt. It also line up with the assumption that thete is a neutral earth fault somewhere.
I discinnected the outside lights. No difference then continued to disconnect circuits and perform the test untill i seen a different reading for a lighting radial... which was Err5.
I put my normal fluke multi meter across neutral and earth and measured 0.9ohms. I then disconnected the earths at different celing roses until i identified the fault (in the most awkward and out of reach fitting possible).
When i took the light fitting off the resistance went high (>1M).
I tightened up the wiring in the fitting (which is metal and earthed) and reconnected.
Everything seems ok now with the rcd tripping when tested and staying on when reset.
I think the Err5 is when you try and test with a load connected...
I didnt today but will do an insutalion test tomorrow to make sure there are no earth faults but hopefully that was the issue.
Does what im saying sound sensible?

 

[Octopus]

Seeing as this has been going on for nearly 1 year, would it be worth getting a decent spark involved who knows what they are doing ?

 

[J_T]

I appreciate it has been going on for a year but there have been 2 different faults here. Neither of which are run of the mill and 1 which required the suppliers intervention which is not always quick.
I was under the impression this was a forum for discussion/advice but if the advice is get an electrician i could have came to that conclusion myself rather than trying to understand the issue myself.

 

[Lucien Nunes]

Yes it does sound like the explanation, as per my previous post. Either your previous IR test was in error or quite possibly the fault was intermittent. E.g. when you took the IR test it was not quite in contact but by the time you were delving deeper into the Zs it had made contact. These things can be a royal pain to find, which serves as a reminder not to trust anything, anywhere. If I get conflicting or confusing results I always go back and recheck all measurements with extra care, avoiding the assumption that what seemed to be decent IR readings at the time remove all suspicion from the insulation as the culprit.

 

[Midwest]

Friggin MFT's. Whilst checking for extraneous pipework in bathroom today, I connect mine to a piece of copper pipe in a plasterboard wall, devoid of anything electrical and not physically connected to anything, just for the hell of it. Meter told me to pee off 'cos there was < 50v there

 

[UKMeterman]

Less than 50 volts? Check your symbols. But if you do find stray volts an insulation test might be a good idea.

 

[Midwest]

Less than 50 volts? Check your symbols. But if you do find stray volts an insulation test might be a good idea.

Apologies >50V