R1,R2,Rn versus r1,r2,rn continuity readings on EICR.

[Ray Pooley]

Looking at the EICR form. GENERIC SCHEDULE OF TEST RESULTS

Looking for clarification on R1,R2,Rn versus r1,r2,rn continuity values.

I have reached the some conclusions.

Do r1,r2 and rn only apply to ring circuits and R1,R2 and Rn (not on form) to radial?

It just looks like that might be the case when you look at columns 10 to 14.

But having said that I have seen R1,R2 being reference to ring circuits too in a document on the subject.

That document seems to suggest that the difference between r1 and R1 is that r1 is the
maximum impedance of the final circuit live cable (source to end) whereas R1 is measured at each
outlet on the circuit (source to outlet). Meaning that R1 applies to both ring and radial circuits.
On a radial, for example, R1 at the outlet furthest from the source would be equal to r1.

Same for r2/R2 and rn/Rn.

Am I interpreting that correctly?

Cheers

 

[telectrix]

not quite. r1 etc. is end to end and only applies to ring circuits. ( these measurements are to confirm that the ring is unbroken).
R1 etc. applies to all circuits and is from MCB to end of radial or mid-point of a ring. useful to know that R1 should be 0.25x r1. as being a ring, conductor size is doubled, and you measure R1 to mid point. hence divide by 2 twice.

 

[loz2754]

r1, r2 ,rn only applies to ring final circuits. r1 applies to the measured end to end resistance (not impedance) of the line conductor. Similarly with r2 for the ring CPC conductor, and rn for the ring neutral conductor.

R1, R2 applies to ring final circuits and also radial circuits.

 

[pc1966]

As above, the lower case r1,r2,rn are specific to the ring circuit (though you could argue that they are measured like a radial that just happened to end up back at the CU).

The upper case R1,R2,Rn can be applied anywhere, but usually it is for the last socket (or only load) in a radial system. As @telectrix says you could measure R1+R2 on a RFC as well, but you generally don't need to as the end-end tells you all that is needed as your worst-case R1+R2 = (r1+r2)/4

However, the RFC end-end test is not checking any spurs off a RFC (just as the R1 & R2 on the end socket is not testing branches off a radial socket set) so really for a new installation, or if 100% coverage is needed, you should check every socket. In the radial case that means measuring R1+R2 each time (with L & E linked at CU) where as for the RFC you do the figure-of-eight test by linking L1 & E2 and also linking L2 & E1 so you can then measure the same worst-case R1+R2 at every socket, with any spurs being obvious by a value higher than the (r1+r2)/4 values expected (and of course socket polarity being verified).

Now many folk seem to thing that testing the RFC is more complicated than testing the radial, and it is. But what the above sorts of tests are doing is giving you far greater fault coverage and making even a small bad connection more obvious during the test.

A simple example is you are expected to measure rn for the RFC every time, but no one is expected to test Rn for a radial (though some sparkys might) due to the rather basic assumption that if it is open then some sockets won't work so it gets noticed and fixed. Not so a poor overheating joint...

Also it is easier to measure RFC end-end at the CU during an inspection than to identify which socket is the last one for a radial's test. But again if the RFC has spurs, or the radial has branches, you are not testing everything by simply checking the end point.

 

[Ray Pooley]

Thanks for the responses.

I am using the word impedance because impedance extends the concept of resistance to AC circuits, and possesses both magnitude and phase, unlike resistance, which has only magnitude.

So I am assuming AC supply which is normally the case.

When I use the word "source" I am referring to the MCB.

But what about the reference to "outlet" in the above snippet?

The impedance of the circuit from the source to a particular outlet is going to vary slightly at each outlet as you go along the circuit.

In the case of radials it will increase along the circuit at each outlet.

In the case of ring circuits there will be two values. I presume you would use the smallest value in any calculation of leakage to ground current.

Is it necessary to test each outlet or is that only necessary if the source to end impedance at any point leads to the conclusion that the leakage to ground current produced is insuffient to trip any RCD present?

Do you ever, in practice, measure and record at each outlet?

Thanks

 

[pc1966]

Thanks for the responses.

I am using the word impedance because impedance extends the concept of resistance to AC circuits, and possesses both magnitude and phase, unlike resistance, which has only magnitude.

So I am assuming AC supply which is normally the case.

Yes, but in reality the impact of inductance (and skin effect on resistive part) at 50Hz for cables is small and only noticeable for big cables. E.g. if you look at tabulated "volt drop" impedance for cables such as 4D4B in the regs you only see a 1% difference in r & z by time you reach 50mm.

For things like motors, transformers, relay coil, etc, then L is usually be dominant.

When I use the word "source" I am referring to the MCB.

But what about the reference to "outlet" in the above snippet?

Typically a socket, could be a FCU. For a fixed appliance it would be the connection point so maybe a cooker switch, etc.

The impedance of the circuit from the source to a particular outlet is going to vary slightly at each outlet as you go along the circuit.

In the case of radials it will increase along the circuit at each outlet.

In the case of ring circuits there will be two values. I presume you would use the smallest value in any calculation of leakage to ground current.

Leakage in wiring is usually negligible, the insulation resistance values are in the M ohm range if OK and will not be different to any measurable degree where you test it.

The fault loop impedance R1+R2 will increase as you move away from the supply/source, and the maximum values is the one to note as that give the worst-case fault disconnection time. Normally for a radial that is at the furthest point, for a RFC it is in the mid-point (i.e. furthest by both "legs") but given by (r1+r2)/4.

Basically the goal of a CPC is to keep any metalwork close to Earth potential and cause a lot of current to flow if something tries to alter that. Enough current so the supply MCB/fuse will open fast so limiting the fault energy (i^2 * t) as well as the potential shock duration of anyone in contact with the CPC/metalwork and the true Earth (or other CPC).

Relying on a RCD to disconnect is usually reserved for TT systems (separate earth rod) as insufficient current would flow for most MCBs, and for cases when you can't reasonably keep the loop impedance down. In most TN final circuit cases you would expect Zs = Ze + R1+R2 to be low enough to clear the MCB's Zs limit, otherwise you are likely to be reaching the point where voltage drop under normal circumstances is too great.

Is it necessary to test each outlet or is that only necessary if the source to end impedance at any point leads to the conclusion that the leakage to ground current produced is insuffient to trip any RCD present?

Do you ever, in practice, measure and record at each outlet?

On an inspection you would normally check only an agreed percentage of sockets, perhaps targeting any that look particularly old or otherwise dodgy. If any fail to meet the required Zs then it is usually a fault that needs correction (C2).

On initial installation, of if a client wants 100% testing and agrees to pay for the time, then you should check every socket.

If you have found a fault that needs rectification then I would say that you ought to check all of them on that circuit (or even the installation) as part of the costed repair as clearly the system was not 100% good, and you won't really know if that was a one-off problem of if the installation was poorly done/hacked by DIY and needs more than just the observed fault fixed.

 

[Ray Pooley]

Yes, but in reality the impact of inductance (and skin effect on resistive part) at 50Hz for cables is small and only noticeable for big cables. E.g. if you look at tabulated "volt drop" impedance for cables such as 4D4B in the regs you only see a 1% difference in r & z by time you reach 50mm.

For things like motors, transformers, relay coil, etc, then L is usually be dominant.


Typically a socket, could be a FCU. For a fixed appliance it would be the connection point so maybe a cooker switch, etc.


Leakage in wiring is usually negligible, the insulation resistance values are in the M ohm range if OK and will not be different to any measurable degree where you test it.

The fault loop impedance R1+R2 will increase as you move away from the supply/source, and the maximum values is the one to note as that give the worst-case fault disconnection time. Normally for a radial that is at the furthest point, for a RFC it is in the mid-point (i.e. furthest by both "legs") but given by (r1+r2)/4.

Basically the goal of a CPC is to keep any metalwork close to Earth potential and cause a lot of current to flow if something tries to alter that. Enough current so the supply MCB/fuse will open fast so limiting the fault energy (i^2 * t) as well as the potential shock duration of anyone in contact with the CPC/metalwork and the true Earth (or other CPC).

Relying on a RCD to disconnect is usually reserved for TT systems (separate earth rod) as insufficient current would flow for most MCBs, and for cases when you can't reasonably keep the loop impedance down. In most TN final circuit cases you would expect Zs = Ze + R1+R2 to be low enough to clear the MCB's Zs limit, otherwise you are likely to be reaching the point where voltage drop under normal circumstances is too great.


On an inspection you would normally check only an agreed percentage of sockets, perhaps targeting any that look particularly old or otherwise dodgy. If any fail to meet the required Zs then it is usually a fault that needs correction (C2).

On initial installation, of if a client wants 100% testing and agrees to pay for the time, then you should check every socket.

If you have found a fault that needs rectification then I would say that you ought to check all of them on that circuit (or even the installation) as part of the costed repair as clearly the system was not 100% good, and you won't really know if that was a one-off problem of if the installation was poorly done/hacked by DIY and needs more than just the observed fault fixed.

Thanks for the in depth account. Much appreciated. Will paste it into a document and preserve it.

 

[loz2754]

@pc1966 has it pretty much covered. As usual.

I'll just throw this in:
Guidance Note 3 Inspection and Testing stipulates that, even on a periodic inspection, an earth fault loop impedance measurement should be taken at every socket outlet, with no allowance for samples. With good reason, as portable hand held equipment could be plugged into any socket.

Also, when I earlier commented that the R1 + R2 measurement is the resistance of the cables, rather than the impedance which we normally talk about in AC circuits, it's because we are measuring the literal resistance of the copper in this test, as there is no AC present during such a dead test. As opposed to the measuring of Zs, or earth fault loop impedance, which is a live test, wherein we're measuring the actual impedance of the circuit. The clue being in the abbreviation letters used: R for resistance, Z for impedance.

 

[Ray Pooley]

@pc1966 has it pretty much covered. As usual.

I'll just throw this in:
Guidance Note 3 Inspection and Testing stipulates that, even on a periodic inspection, an earth fault loop impedance measurement should be taken at every socket outlet, with no allowance for samples. With good reason, as portable hand held equipment could be plugged into any socket.

Also, when I earlier commented that the R1 + R2 measurement is the resistance of the cables, rather than the impedance which we normally talk about in AC circuits, it's because we are measuring the literal resistance of the copper in this test, as there is no AC present during such a dead test. As opposed to the measuring of Zs, or earth fault loop impedance, which is a live test, wherein we're measuring the actual impedance of the circuit. The clue being in the abbreviation letters used: R for resistance, Z for impedance.

Got it. Thanks. I shall grab that for future reference.

 

[pc1966]

Guidance Note 3 Inspection and Testing stipulates that, even on a periodic inspection, an earth fault loop impedance measurement should be taken at every socket outlet, with no allowance for samples. With good reason, as portable hand held equipment could be plugged into any socket.

Thanks, that is useful clarification.

Also, when I earlier commented that the R1 + R2 measurement is the resistance of the cables, rather than the impedance which we normally talk about in AC circuits, it's because we are measuring the literal resistance of the copper in this test, as there is no AC present during such a dead test. As opposed to the measuring of Zs, or earth fault loop impedance, which is a live test, wherein we're measuring the actual impedance of the circuit. The clue being in the abbreviation letters used: R for resistance, Z for impedance.

Yes, strictly speaking all AC measurements for protection, etc, are |Z| = sqrt(r^2 + x^2) as it is only the magnitude of the current that matters, and not any phase shift / power factor.

But in most cases, e.g. domestic or light industrial/commercial supplies of 100A or so, the DC resistance measured for R1+R2 and the AC-measures Ze can be combined to estimate Zs with sufficient accuracy that the phase angle/inductive part 'x' can be largely ignored.