TT Earthing problem?

[MrDuck]

My (new to me) property has a TT supply with a 30mA RCD protecting all the MCBs (not ideal but not causing inconvenience so far). There is a substantial Earthing Rod robustly connected to the earth bar of a modern Consumer Unit. PME is not available so far as I am aware.

The RCD trips via its Test button or if I simulate a 30mA leak from Live to Earth. However, if I short Neutral to Earth at a ring main socket the RCD does NOT trip, which I understand to be incorrect (and certainly not as per my previous property).

If I measure the current which flows from N to E it's about 18mA (not enough to trip) and this reduces (!) to about 16mA if I increase the house load e.g. by turning on a cooker. There is a constant approx. 5 volts ac between Neutral and Earth. I have tried measuring the earth rod resistance by installing a second earth rod some distance away but bizarrely even with the RCD off (so no house load) there is about 5 volts ac between this new rod and the old one which messes up any attempt to measure resistance with a multi-meter.

I'm guessing that my earth rod has a problem but is it also possible that the neutral from the supply company has a poor earth? I live in a very rural area. Any suggestions much appreciated before I call in a professional with the right test gear.

 

[EFLImpudence]

How have you measured the 18mA?

The fact that shorting N to E does not trip the RCD is not "incorrect".
The fact that it does in TN systems is a result of unavoidable current flow; it is not a design feature.

 

[MrDuck]

I measured the 18mA ac by using a suitable multimeter onto a cable plugged into a ring main (with the live well insulated!). My previous property had PME, this one doesn't.

 

[Taylortwocities]

IIRC The standard spec for RCDs is that it must not trip at 50% of its rated current but must trip within 40mS at 100% of its rated current. These tests are done both at 0 and 180 degrees of a sine wave.
What you describe is fairly standard stuff for a TT. An earth to neutral test is not going to give you consistent results and , as said, has a lot of variables. Its something that is sometimes observed, but not a standard test in any way.

I think you are ok, if you do anything, I would get the RCD checked using a pukker RCD tester.

 

[Spark123]

To test the resistance of a rod to ground you need a special type of meter with two more rods.
It isn't really necessary when a Ze (or Ra) test can be carried out with an earth fault loop impedance tester. A multimeter is neither of these.

 

[MrDuck]

Thanks everyone for your feedback and I will reduce my worry level!. I plan to get a professional with the right kit to do some earthing tests 'just to make sure'.

Would a conversion to PME (if available) be worthwhile?

 

[oceansoul]

The RCD trips {snip} if I simulate a 30mA leak from Live to Earth.

How did you test that

 

[slippyr4]

With an 8k resistor, at a guess. Bit dodgy because it'll get 7W in it. But for a few milliseconds it'd probably be ok.

 

[JohnW2]

The RCD trips via its Test button or if I simulate a 30mA leak from Live to Earth. However, if I short Neutral to Earth at a ring main socket the RCD does NOT trip, which I understand to be incorrect (and certainly not as per my previous property).

As has been said, the failure to trip in response to an N-E short does not, in itself, give rise to any concern (as has been said, it is dependent upon many factors) but the fact that it does trip with 30mA L-E leak confirms that the RCD is functioning correctly, and that your earth is at least 'not too bad', but it doesn't tell us much more than that.

If I measure the current which flows from N to E it's about 18mA (not enough to trip) ... I'm guessing that my earth rod has a problem ....

Whilst none of us can say that it doesn't have a problem, nothing you have told us really gives any indication that there is any problem. The question you were asked about how you generated the 30mA L-E leakage current is relevant in this respect. A TT electrode should really have a resistance/impedance less than about 100Ω (certainly less than 200Ω). In that situation, as has been suggested, (and assuming 230V) a resistor of around 7.6kΩ should result in a 30mA current, hence an RCD trip. However, if you had to use a lower value resistor than that (say, less than 7kΩ) in order to get 30mA, then that would suggest that the resistance of your earth rod was probably far to high.

.... but is it also possible that the neutral from the supply company has a poor earth?

Again, nothing you have told us really gives any indication that such is the case, although that proves nothing. The supply neutral will be connected to 'true earth' at the transformer/substation, but there will be an appreciable voltage drop in that conductor (due to the current flowing through it), so it would be expected that there would be a potential difference between 'your end' of that supply cable and earth - increasingly so as your distance from the transformer/substation increases. Indeed, if there were no appreciable potential difference between N and E in your house, then an N-E short certainly would not cause your RCD to trip, since no appreciable current would flow through your 'short'.

Having said that, if you have concerns then the obvious course is, indeed, to get things checked out by someone who has the appropriate test gear.

Kind Regards, John

 

[ericmark]

To test an earth rod has some danger, using an earth rod tester means you have to disconnect the earth rod so no earth during the test, to use a loop impedance meter again you should disconnect and if not it can produce a mains voltage on earthed items during the test.

So too options one is assume nothing wrong which is likely the case and two is get some one with the correct equipment. The EICR done when a new occupent takes over a home should have included testing

 

[ericmark]

To test an earth rod has some danger, using an earth rod tester means you have to disconnect the earth rod so no earth during the test, to use a loop impedance meter again you should disconnect and if not it can produce a mains voltage on earthed items during the test.

So too options one is assume nothing wrong which is likely the case and two is get some one with the correct equipment. The EICR done when a new occupent takes over a home should have included testing

 

[Taylortwocities]

The EICR done when a new occupent takes over a home should have included testing

Oh, is that a new requirement then?

 

[MrDuck]

Thanks for all this valuable feedback. Regarding my 'RCD tester' I examined this today and it contains five 39k 1/2W resistors in parallel (7.8k) so is close to the nominal 30mA. I have used this at a previous property which had a battery of RCBOs and it was very helpful as a go/no-go tester after repairs etc. At my current property it trips the RCD 'instantly'.

 

[Lectrician]

IIRC The standard spec for RCDs is that it must not trip at 50% of its rated current but must trip within 40mS at 100% of its rated current. These tests are done both at 0 and 180 degrees of a sine wave.
What you describe is fairly standard stuff for a TT. An earth to neutral test is not going to give you consistent results and , as said, has a lot of variables. Its something that is sometimes observed, but not a standard test in any way.

I think you are ok, if you do anything, I would get the RCD checked using a pukker RCD tester.

Within 40ms at 5x rated current, so set your meter to 150mA. Depending on RCD, 200 or 300ms at rated current.

An RCD may trip with a short from N to E on a TN/PME system as neutral current diverts down the low impedance earth. On a TT, the earth resistance is unlikely to be low enough for the current to decide to flow that way, rather than the low impedance neutral.

 

[ericmark]

At 15 volt between neutral and earth 500 ohms resistor would give 30 mA but as that voltage drops to 3 volt then the resistor would be 100 ohms, the problem is the current at 230 volt would be really high current to get it to test at the neutral to earth voltages so it is just not something we can do.

In real terms we want it to trip if the neutral to earth voltage exceeds 50 volt, but for that to happen we have a much more serious situation and the simple RCD is not there to do that.

Standard is 6 tests, in three pairs pos and neg part of wave form with 15 mA, 30 mA and 150 mA the last test can be missed if the time is less than 40 mS with 30 mA test. You simply can't measure 40 mS with a stop watch so you have to use a proper RCD tester.

I know many don't but you should have an EICR or as it was called PIR every ten years or with change of occupant which ever is shorter, in Scotland for rented it is law, rest of UK only recommended. The point is you have not a clue what the last occupant has done, and that may not have been the owner, so any questions to seller don't really count. Can't find all faults, but you can find a lot, so it really does make sense.

What is odd, is except for Scotland, there is no set qualification to do the testing, most consider C&G 2391 as showing you have the skill, but one can do inspection and testing without being a scheme member so there is nothing to say it's do correctly. We see loads of threads on the forum where it would seem the guy testing has not followed the laid down system of coding, and has decided to make up his own, so I am a little suspect of EICR's which makes them worthless unless you know the electrician and can trust him.