Initial Re measurement on TT rod

[JohnW2]

Is this perhaps a recognised phenomenon? I initially thought that it was just a fluke (of my Fluke!), but it is so consistent that I now find that hard to believe.

Whenever I measure (estimate) the Re of one of my earth rods by measuring the loop impedance of the rod (in isolation), the first measurement I get is always appreciably higher (usually to the tune of ~50&#937 than all subsequent measurents over the next few hours.

For example, this morning (with my Fluke 1652) I got an initial result of about 120Ω. An immediate repeat (within seconds) gave a result of about 70Ω and several repeats over the following 3-4 hours gave similar results, not varying from the second reading (i.e. ~70&#937 by more than an ohm or three.

As I asked, is this a 'known phenomenon' and, if so, what is the explanation? Is it perhaps some sort of electrolytic phenomenon - although the test waveform is described as 'sinusoidal', it's duration is only 10mS (i.e. one half-cycle at 50Hz {assuming it does use 50Hz!}) - so, if it starts at zero crossing, it will actully be effectively DC.

... or is there some other possible explanation? Given the consistency of measurements (excluding the initial one), I find it hard to believe that the meter is faulty - but I suppose anything is possible.

Kind Regards, John

 

[bernardgreen]

It is most probably a skin effect on the rod that gives the higher reading. The resistive skin is created by chemical action between rod and soil when there is no current passing from rod to soil. The current or voltage from the initial test seems to break the skin thus improving the connection between rod and soil.

 

[JohnW2]

It is most probably a skin effect on the rod that gives the higher reading. The resistive skin is created by chemical action between rod and soil when there is no current passing from rod to soil. The current or voltage from the initial test seems to break the skin thus improving the connection between rod and soil.

That makes sense. Are you implying that this is something you have observed?

If this is, indeed, a general phenomenon, I'm surprised that I've never seed anything advising that one should 'discard' the very first loop impedance measurement made on a TT installation.

Kind Regards, John

 

[bernardgreen]

I have not observed it personally during measurement of my earth rods. Next time I measure them I will pay more attention. I test by forcing a known current into the rod and measuring the voltage between rod and ground using a remote earth rod or the incoming neutral as a ground reference. Possibly the skin effect has been broken by the time I measure the voltage which is always a few seconds after I apply the voltage

In case any one is wondering R = V / I is the equation used to calculate the resistance at current I .

 

[JohnW2]

I have not observed it personally during measurement of my earth rods. Next time I measure them I will pay more attention. I test by forcing a known current into the rod and measuring the voltage between rod and ground using a remote earth rod or the incoming neutral as a ground reference.

I assume that my machine applies a known voiltage and measures the current.

Possibly the skin effect has been broken by the time I measure the voltage which is always a few seconds after I apply the voltage.

Possibly. As I said, my meter's manual appears to be saying that the testing current only flows for 10 msec - and that seems to be enough to produce whatever effect I'm seeing. However, I'm a bit confused by that, because it takes a long time (probably about 20 secs) to actually complete the measurement with Re values as high as one gets with a rod, although it is almost instantaneous with TN earths. In fact, what the manual actually says is "Maximum test current at 230V: 12A sinusoidal for 10ms". Maybe it lasts for much more than 10ms if the current is a lot less than 12A (as it would be with my rod).

Kind Regards, John

 

[ericmark]

I at one time spent 6 months installing and testing earth rods but never noticed a variation between first test and later tests. I used the standard two test probe method there was no power on the site to use a earth loop impedance meter so all done the hard way.

The readings were not done quickly it takes some time to set up the two probes and there was always three tests moving the centre probe to ensure tested over enough distance.

I would always wait until the reading was steady which I suppose is not an option with a loop impedance meter which will likely measure in less than 40ms so as not to trip any RCD.

So it would seem there could be a problem with skin effect.

 

[JohnW2]

The readings were not done quickly it takes some time to set up the two probes and there was always three tests moving the centre probe to ensure tested over enough distance. ... I would always wait until the reading was steady which I suppose is not an option with a loop impedance meter which will likely measure in less than 40ms so as not to trip any RCD.

Indeed, since it certainly does not trip the RCD (which was in circuit, although I obvioulsy could have bypassed it if I so wished) - so that would be consistent with the 10ms the manual mentions. However, as I said, I don't understand why it takes so long to complete the measurement when Re is relatively 'high (as with a rod) - I've just timed it, and it's about 25 seconds for the rod, but only 1-2 seconds for a TN earth. I can't imagine what it can be 'doing' for all of that 25 seconds! I presume that it does a L-N loop measurement as well as a L-E one, since it displays both loop impedance and 'Re', but even that doesn't explain the time taken.

So it would seem there could be a problem with skin effect.

Indeed, that does sound like a credible explanation. BTW, whilst timing the meter I was able to confirm that I'm still getting about 70Ω, now some 8 hours after that intitial 120Ω reading (with all subsequent ones being ~70&#937. So, if it is a skin effect issue, it goes away for a long time after a bit of current is passed through it.

Of course, this probably doesn't only affect Re measurements. It ought to also apply to the very first Zs measurement one did on a TT installation.

Kind Regards, John