UK Higher than expected Zs

[Trotm1]

Hi,

I did my first EICR on a TT system today and the Zs was a lot higher than Ze+R1+R2 on everything.

Zs was around 3.5 ohms on all of the ring circuits and the oven circuit and lighting circuits were between 3-4.5 ohms.

The Ze was 1.1 ohms and R1+R2 was between 0.9-0.23 ohms at the various points. I bypassed the type AC RCD but still got the same results.

The previous EICR report from Napit also had higher than expected Zs but they only tested for R2 and passed the installation.

I’m not that familiar with TT systems, is this expected / acceptable or was it not property investigated previously, what could be the issue?

 

[pc1966]

For a TT system having Ze (presumably rod Ra ?) of 1.1 is exceptionally low (good). Usually tens of ohms from a rod is considered good.

Did you test at the CU Zdb with bonding attached, rather than isolating the earth rod (with installation off) and testing it alone ? Having metallic service pipes or foundation metalwork bonded to the MET would explain that value.

But you are quite right that the supply impedance and R1+R2 is not really adding up to the circuit Zs values, even though they are all reasonably consistent (3.5 ohm RFC, 3-4.5 ohm others).

Can you try measuring the "Zs" at the MCBs to see if they are within the ohm of the end of circuit values? Something odd with RCD "uplift" seems the most probably explanation.

 

[Trotm1]

For a TT system having Ze (presumably rod Ra ?) of 1.1 is exceptionally low (good). Usually tens of ohms from a rod is considered good.

Did you test at the CU Zdb with bonding attached, rather than isolating the earth rod (with installation off) and testing it alone ? Having metallic service pipes or foundation metalwork bonded to the MET would explain that value.

But you are quite right that the supply impedance and R1+R2 is not really adding up to the circuit Zs values, even though they are all reasonably consistent (3.5 ohm RFC, 3-4.5 ohm others).

Can you try measuring the "Zs" at the MCBs to see if they are within the ohm of the end of circuit values? Something odd with RCD "uplift" seems the most probably explanation.

Thank you for the response, yes I measured the earth rod Ze with the parallel paths disconnected.

The CU was split load, RCD for the rings and oven and no RCD for the lighting and smoke alarm circuits but these had been upgraded to RCBO’s.

I put one of the rings on a spare MCB by passing the RCD altogether but had the same result.

What I think I’ll do is mark it on the report as FI and go back and investigate further. I just wanted to ask in case there was something to do with being a TT system I’m not aware of.

 

[pc1966]

Thank you for the response, yes I measured the earth rod Ze with the parallel paths disconnected.

OK that is unusual low then, probably it has a "near miss" with some buried metal such as pipe work, etc.

The CU was split load, RCD for the rings and oven and no RCD for the lighting and smoke alarm circuits but these had been upgraded to RCBO’s.

I put one of the rings on a spare MCB by passing the RCD altogether but had the same result.

OK.

What I think I’ll do is mark it on the report as FI and go back and investigate further. I just wanted to ask in case there was something to do with being a TT system I’m not aware of.

It is unlikely here, but one aspect of note is the earth rod impedance can change a lot if it goes from dry soil to wet under heavy rain following a drought. Not instant, but measurements an hour apart, etc, might not be as consistent as you would see with TN earthing.

Again, if you can do the CU Zdb and then the circuit supply point Zs (probe at terminals) to see if they are consistent first, then check something like the RFC Zs to see if it is as close as your R1+R2 values would point to.

Something odd with your MFT calibration might be a factor, but I would try the above to see at what point you get them diverging.

 

[timhoward]

The easy thing to check - A decent Ze, good R1+R2 but a high Zs on more than one circuit would lead me to check the main incoming earth cable, it's connection at the consumer unit and it's connection at the MET (if any).

I'm also slightly suspicious of an exceptionally (almost unbelievably!) good Ze and then high Zs values, so it might be worth being open to the idea that your MFT is playing up or it's leads is doing something strange.

(I wouldn't have FI'd it - I know the result is a problem so I'd give it a C code. Finding out why is remedial work!)

 

[Trotm1]

The easy thing to check - A decent Ze, good R1+R2 but a high Zs on more than one circuit would lead me to check the main incoming earth cable, it's connection at the consumer unit and it's connection at the MET (if any).

I'm also slightly suspicious of an exceptionally (almost unbelievably!) good Ze and then high Zs values, so it might be worth being open to the idea that your MFT is playing up or it's leads is doing something strange.

(I wouldn't have FI'd it - I know the result is a problem so I'd give it a C code. Finding out why is remedial work!)

I was surprised by the low Ze as well.

I’ve ruled out the leads as I used my probes for the spurs and I have a separate plug socket lead both giving the same high readings.

The MFT is also not the issue. It came back from its annual calibration a few weeks ago and I've just did a Zs on my ring at home which gave expected results.

I haven’t submitted the report yet but I think you’re right with a C code as it’s not right.

 

[HappyHippyDad]

I know you've said you've bypassed the RCD's but did you then do the Zs test on the high setting rather than the low setting (used when you don't want the RCD to trip)? I have found that I sometimes get higher ZS readings when tested on a low current setting.

Also, I (think) I sometimes get higher readings when I do a 2 lead test (available on some MFT's) rather than utilising the neutral as well.

It does sound (70%) likely that it's a testing issue rather than a possible fault.

 

[Risteard]

Bypassing RCDs for loop testing, particularly on Direct Earthing (TT) is not advisable!

 

[westward10]

TT systems can be a law unto themselves you are better off doing continuity tests (R1 + R2) and adding them to Ze. Same with any ring finals.

 

[timhoward]

TT systems can be a law unto themselves you are better off doing continuity tests (R1 + R2) and adding them to Ze. Same with any ring finals.

Agreed and I normally do this.
The thing irking me with this one is that if the same results were obtained with the RCD (unwisely) bypassed, presumably using a high current test, it's possible that there's a piece of the earth path that hasn't been tested - maybe there's an external MET to the CPC bar with loose connection one end.

I'd like to see a high current loop test from the incomer to the CPC bar, assuming there is no upstream RCD and compare that with the Ze to see if it's plausible. I'm sure @pc1966 suggested this many posts ago.

 

[OLDBOY]

As previously posted, almost certainly due to differences in readings obtsined with high and low current measurements on the meter, low current readings are notoriously unreliable.