Generator earthing

[ricicle]

GN3 and GN8 cover it I think. It's not that hard.

 

[JohnW2]

...In other words, under what circumstances could a measurement obtained by loop testing be significantly 'wrong'?

Where the electrode resistance areas can clash, loop testing is only for use where you are testing an electrode on a TT system which has RCD protection. An electrode used as the means of earthing of a system needs to be done the old fashioned way.

I still don't really understand. Clashing of the resistance areas of what electrodes with which? Furthermore, in the final analysis isn't it loop resistance (which can't possibly be less than Re) that one is really most interested in?

The method is completely alien to a majority of sparks bar the likes of Westies men and lightning conductor men who will probably do this a lot. I would have to consult my books on this type of testing as I have never needed to do it before.

It sounds as if many of you "haven't lived"! Before I had a loop tester, I had to measure electrode resistances with a transformer, ammeter, voltmeter and a couple of earth rods

Kind Regards, John

 

[Spark123]

...In other words, under what circumstances could a measurement obtained by loop testing be significantly 'wrong'?

Where the electrode resistance areas can clash, loop testing is only for use where you are testing an electrode on a TT system which has RCD protection. An electrode used as the means of earthing of a system needs to be done the old fashioned way.

I still don't really understand. Clashing of the resistance areas of what electrodes with which? Furthermore, in the final analysis isn't it loop resistance (which can't possibly be less than Re) that one is really most interested in?

The area around the rod can overlap with other electrodes or extraneous conductive parts which too have a resistance area which can give an artificially low reading. As the rod Rob is sinking in, when in use, is nothing to do with the fault path back to the distribution transformer then you need to ask what meaning measuring the rod as part of a TT system is worth. I'll use capital letters to signify Rob's new rod, a loop measurement will give you a reading for tT. It is going to be used as a Tn-s for the house etc and a Tt for the milking area.
Ie the rod is providing the means of earthing of the generator supply, it is the primary, not the secondary rod in the system.
You need to prove the rod is in good contact with ground without any other rods/ecps interfering with this measurement.

 

[JohnW2]

The area around the rod can overlap with other electrodes or extraneous conductive parts which too have a resistance area which can give an artificially low reading.

I obviously don't know exactly what Rob did, and it's possible that he did simply measure the EFLI of the installation (with extraneous-conductive-parts connected), in which case what you say above would be correct. However, if one wants to estimate the Re of an earth electrode by loop testing, one tests the electrode in isolation - i.e. although one uses the supply L to undertake the test, one disconnects the rod from the installation and connects it only to the meter. Most loop testers measure both L-E loop impedance and L-N loop impedance. They display the former (usually calling it Zs), but also subtract half of the L-N loopimpedance from it and display the result as 'Re'.

That displayed 'Re' consists of two components - 'T' plus 't' per the terminolgy you used (i.e. the impedances of the electrode being tested, T, plus the impedance to earth from the DNO's transformer's star point, t). That latter impedance is a complex affair, because it consists of lots of things in parallel, starting with the DNOs rod at the transformer but also (assuming it is a TN supply, as is the one Rob is dealing with) all the other routes to earth from there - e.g. via the extraneous-c-ps of all consumers connected to that transformer and, in the case of PME, all the PME electrodes. Because there are so many parallel paths to earth, the effective resistance ('t') is likely to be very low - let's face it, what we are talking about is essentially the Re component of the EFLI we are used to seeing when we measure from any TN installation connected to the transformer in question - in other words, well under 1Ω .

So, assuming it is a TN supply, the loop test (if done on the electrode in isolation) will display T+t, and 't' is going to be extremely small. The displayed 'Re' will therefore be very close to 'T' (the resistance of the electrode under test) and can never be higher than 'T'.

If the transformer is not providing a TN supply, there will be far fewer parallel paths to earth, so 't' could potentially be as high as 20Ω/21Ω (which appears to be the maximum the DNO allow for the rod at the transformer) - so that the Re displayed by a loop tester would be ~(T+20). For many purposes, that would be adequate (as a maximum figure for the rod's resistance), but I agree that it would probably not be adequate if one was looking for a very low electrode resistance (as Rob is) - but we know that Rob is dealing with a TN supply.

Kind Regards, John

 

[Spark123]

Rob is dealing with the means of earthing for his own TN system (and a TT system as an aside), the DNOs earthing system has nothing to do with it hence must be left out of any testing.

 

[JohnW2]

Rob is dealing with the means of earthing for his own TN system (and a TT system as an aside), the DNOs earthing system has nothing to do with it hence must be left out of any testing.

That may be the 'party line' but, as I've explained, electrically-speaking one can (if the DNO's supply is TN) get a very good measurement (and crucially, a measurement which can't be higher than the truth) in the manner I described. In fact, in electrical terms, the only real difference between using a loop tester and a 'proper earth resistance tester' is that, in the latter case, one is 'in control' of the reference electrodes one uses, whereas in the former case one is using the DNO TN earth for that purpose.

Kind Regards, John

 

[EFLImpudence]

although one uses the supply L to undertake the test, one disconnects the rod from the installation and connects it only to the meter. Most loop testers measure both L-E loop impedance and L-N loop impedance. They display the former (usually calling it Zs), but also subtract half of the L-N loopimpedance from it and display the result as 'Re'.

Not if you use the high current two-lead test, which you would.

This will not display the Re measurement but the line value (separately determined) can be deducted from the total.

 

[JohnW2]

Not if you use the high current two-lead test, which you would. This will not display the Re measurement but the line value (separately determined) can be deducted from the total.

OK, if that's how your meter works - but whether the meter subtracts the line impedance automatically or one subtracts it manually, what I wrote about the method still applies. In practice, of course, the line impedance is likely to be almost negligible compared with the resistance of an 'average' earth electrode, but it becomes more significant when dealing with very low earth resistances such as RF is looking for. 'Ignoring' (not subtracting) the line impedance obviously would simply make the measurement more 'conservative' (higher than the truth).

Kind Regards, John

 

[EFLImpudence]

All true but my main point in answering your previous post is that -

you would not use the low-current 'no-trip' method to test/measure a rod.

 

[JohnW2]

All true but my main point in answering your previous post is that - you would not use the low-current 'no-trip' method to test/measure a rod.

Agreed (although, of course, some loop testers/MFTs, particularly older ones, only offer that method) - I wasn't thinking totally straight when I talked about the automatic determination/display of Re. However, as you've said, one can determine it separately and subtract it manually.

As for the use of loop testing to measure electrode resistance, having thought a bit more, I can think of just one situation in which it could possibly give 'low' answers, and that's if the electrode under test is in very close proximity to something (buried) connected to the DNO earth - either a bonded extraneous-c-p or a PME electrode (the latter not relevant in RF's TN-S case). In all other situations, the loop impedance (even after subtraction of line impedance) has got to be at least as as high as the electrode resistance. However, maybe that one possible exception is regarded as a reason for 'measuring it properly'.

Kind Regards, John