Part of the outcome of this test, is to:
"ensure that the earthing system has been installed in a manner that will cause circuit protective devices to operate if there is a fault between live parts, other than the neutral, and the mass of earth"
Can this be satisfied by measuring Rphe and confirming that it complies with table 8.2?
What methods do people use for the "continuity of the earthing system" testing, as it relates to final subcircuits?
Continuity of the earthing system
Re: Continuity of the earthing system
This is a (part) post from someone on another forum.
--------------
Remove the active and the earth from their respective terminals and link with a wago.
At each point of connection measure the resistance of active and earth with my 1825 mtf with the plug adapter.
I should get a gradually increasing resistance the further from the active - earth link.
If at any stage I get an OL I can swap the connections to see why I have a crossed polarity or no connection at all.
(For example neutral - earth crossed polarity I would get OL as the neutral is not connected to the earth as it has been removed, and I can confirm this by swapping the neutral and earth plugs on my adapter and I should then see a resistance, if not it is then no connection to the outlet or a different circuit).
At the end of line I can measure the resistance, this would be my internal Rphe.
If this value is lower than specified in table 8.2 I meet the requirements of clause 8.3.5 for protective conductors (proving disconnection times).
Testing to the standards of AS/NZS3017.
---------------
--------------
Remove the active and the earth from their respective terminals and link with a wago.
At each point of connection measure the resistance of active and earth with my 1825 mtf with the plug adapter.
I should get a gradually increasing resistance the further from the active - earth link.
If at any stage I get an OL I can swap the connections to see why I have a crossed polarity or no connection at all.
(For example neutral - earth crossed polarity I would get OL as the neutral is not connected to the earth as it has been removed, and I can confirm this by swapping the neutral and earth plugs on my adapter and I should then see a resistance, if not it is then no connection to the outlet or a different circuit).
At the end of line I can measure the resistance, this would be my internal Rphe.
If this value is lower than specified in table 8.2 I meet the requirements of clause 8.3.5 for protective conductors (proving disconnection times).
Testing to the standards of AS/NZS3017.
---------------
Re: Continuity of the earthing system
The requirement specified is actually only relevant for circuits using automatic disconnection of supply as the means of providing fault protection.
As such, the test should be mandatory only for those circuits. However very few circuits use any other form pf fault protection, so this error doesn't cause many problems. For example, a circuit of sockets supplied from isolating transformer ; can't comply with the requirements of this test and still remain isolated. Just another instance of Wiring Rules having a somewhat blinkered approach.
There's a reason "3017" doesn't suggest any other method than trailing lead;
and why 8.3.5.2. Note 2 points us to the Re values of the table, rather than the R phe values.
Yes, measuring the A-E loop is a valid method for meeting this particular requirement.
Arguably slightly better than simply measuring PEC resistance, because it removes the uncertainty around contribution of active(s); but still based on several assumptions about the rest of the earth fault loop.
And because it involves a lot of disconnecting; increases the chance of something not being re-connected.
So in fact measuring just the PEC, using a trailing lead, is a better option for most earth testing.
The A-E loop method is suggested for EFLI testing, but again it's a second-rate option compared with the live test.
And to avoid problems, should be done early in the test sequence, always being followed by confirming PEC continuity after re-connecting all PECs.
This is one of several reasons that EFLI tests are required only for circuits supplying sockets, and then only if they don't have additional (fault) protection by RCD.
As such, the test should be mandatory only for those circuits. However very few circuits use any other form pf fault protection, so this error doesn't cause many problems. For example, a circuit of sockets supplied from isolating transformer ; can't comply with the requirements of this test and still remain isolated. Just another instance of Wiring Rules having a somewhat blinkered approach.
There's a reason "3017" doesn't suggest any other method than trailing lead;
and why 8.3.5.2. Note 2 points us to the Re values of the table, rather than the R phe values.
Yes, measuring the A-E loop is a valid method for meeting this particular requirement.
Arguably slightly better than simply measuring PEC resistance, because it removes the uncertainty around contribution of active(s); but still based on several assumptions about the rest of the earth fault loop.
And because it involves a lot of disconnecting; increases the chance of something not being re-connected.
So in fact measuring just the PEC, using a trailing lead, is a better option for most earth testing.
The A-E loop method is suggested for EFLI testing, but again it's a second-rate option compared with the live test.
And to avoid problems, should be done early in the test sequence, always being followed by confirming PEC continuity after re-connecting all PECs.
This is one of several reasons that EFLI tests are required only for circuits supplying sockets, and then only if they don't have additional (fault) protection by RCD.
Re: Continuity of the earthing system
I use a trailing lead and make sure MEN link or incoming N is disconnected so I eliminate any N-E transpositions
Re: Continuity of the earthing system
@Alec - yes the situation in question was non-residential and didn't have RCDs. So a loop test was required for socket outlets, at the appropriate stage in the testing.
Similarly, in the other thread, people were referring to the EFLI testing, and A-E loop, and concentrating their responses to socket outlets, but -
My contention was that the required outcome of 8.3.5 was an accurate value for Re, as shown in 8.3.5.2 (a) and (b), and needed to apply to many different types of circuits. There was a mix of socket outlet, water heaters, A/C units and lighting that needed to be tested.
In this specific case, the installation had mostly 4mm² on the socket outlet circuits, and most of that was the old type with 4mm² earth. But some circuits had been extended with 2.5 in some areas. So there was a mixture of conductors and calculating a proportion or Rphe as Re, would have been impractical.
@Jamie - yes that's what I did. Removed the MEN link, test that N-E is a suitably high resistance (had to repair a N-E short on one circuit), then do the trailing lead test for Re.
Having the neutral disconnected when doing this test, means that the testing also confirms that it is the earth which is connected to the earth pin on socket outlets, or appliances. This takes care of the hardest part of polarity/correct circuit connections - ensuring there aren't any N-E transposition as you said.
Similarly, in the other thread, people were referring to the EFLI testing, and A-E loop, and concentrating their responses to socket outlets, but -
My contention was that the required outcome of 8.3.5 was an accurate value for Re, as shown in 8.3.5.2 (a) and (b), and needed to apply to many different types of circuits. There was a mix of socket outlet, water heaters, A/C units and lighting that needed to be tested.
In this specific case, the installation had mostly 4mm² on the socket outlet circuits, and most of that was the old type with 4mm² earth. But some circuits had been extended with 2.5 in some areas. So there was a mixture of conductors and calculating a proportion or Rphe as Re, would have been impractical.
@Jamie - yes that's what I did. Removed the MEN link, test that N-E is a suitably high resistance (had to repair a N-E short on one circuit), then do the trailing lead test for Re.
Having the neutral disconnected when doing this test, means that the testing also confirms that it is the earth which is connected to the earth pin on socket outlets, or appliances. This takes care of the hardest part of polarity/correct circuit connections - ensuring there aren't any N-E transposition as you said.
Re: Continuity of the earthing system
if you look at Table 8.2 closely; you'll find that the values for Rph, Re, & Rphe all vary depending on the rating of the protective device, and not depending on the conductor sizes. If the two columns for 'conductor size' were completely removed, the Table would still work.
EG the max permitted values for a 16 A protective device are the same regardless of whether the circuit conductors are 1.5 mm2 or 2.5 mm2 - or in fact any other size.
This is not really surprising, because - as you've pointed out - the key thing is operating the protective device. That's directly related to resistance; but only indirectly related to conductor size.
Which makes using the Re values by them selves the simplest method of showing compliance. no need to worry about what specific size, or sizes, of conductor have been used, or in what proportions. Yes calculating the permitted Re of a mixed-cable circuit based on active conductor size(s) would be difficult. But what Table 8.2 tells us is that it really doesn't matter.
Just look at the type & rating of the protection, and you'll be given an Re values that works for any conductor size or combination of sizes.
If the Re isn't low enough, and you've eliminated any bad connections in the earthing; de-rate the protection.
There are those who want to remove table 8.2 completely, and force us to make a lot of dreary detailed calculations for both EFLI and earth continuity.
I like the fact that it gives us a 'deemed to comply' that we can easily use - and that's near enough for all practical purposes.
But it would be even better without those 'conductor size" columns; which don't add anything useful, but instead mislead people into thinking conductor size is directly relevant when it isn't.
------------
This does highlight that in most cases (ie for all circuits using automatic disconnection of supply as the means of fault protection); earth continuity is actually part of EFLI. The key to why measuring just the PEC is deemed adequate testing is that for most circuits the required trip time is 5 seconds. That's why the re values for 'fuses' (ie HRC fuses) is so much higher than the values given for mcbs - all the 'mcb' values are calculated for 0.4 seconds, as required for circuits supplying sockets.
For a non-socket circuit, if the PEC is good enough to be OK for a 0.4 sec trip time, then it's going to be better than OK for a 5 sec trip time. There's no point going to the trouble of measuring Rph as well; because if there's a n impedance in the active that would prevent operation within 5 sec, it's going to make itself known as soon as supply is connected to load.
Every time we disconnect something for testing purposes, we have to re-connect it - and we should then prove that the re-connection was properly done. Therefore the idea is to do only the minimum amount of testing that will show compliance. Doing a disconnected part-EFLI test when a simple continuity test would provide a sufficient level of proof isn't a good idea; because it increases the risk of the testing process causing a problem.
So we don't generally do EFLI test except for sockets. And even for sockets, we only require EFLI if there's no RCD protection. That's because an RCD has a 0.3 sec operating time (25 % faster than required); almost regardless of the impedance of the PEC (because with an RCD we're not relying on passing a high fault current).
So for any group of circuits, including circuits with different cable sizes and different types of load; the only relevant factors are:
(a) the rating of the overcurrent protection; and
b) whether the circuit supplies any sockets (including eg sockets for luminaires) that don't have RCD protection.
- If so an EFLI test is required and the Rphe values of Table 8.2 apply (alternatively, use live test and table 8.1).
- If not, just a continuity test using the re values
EG the max permitted values for a 16 A protective device are the same regardless of whether the circuit conductors are 1.5 mm2 or 2.5 mm2 - or in fact any other size.
This is not really surprising, because - as you've pointed out - the key thing is operating the protective device. That's directly related to resistance; but only indirectly related to conductor size.
Which makes using the Re values by them selves the simplest method of showing compliance. no need to worry about what specific size, or sizes, of conductor have been used, or in what proportions. Yes calculating the permitted Re of a mixed-cable circuit based on active conductor size(s) would be difficult. But what Table 8.2 tells us is that it really doesn't matter.
Just look at the type & rating of the protection, and you'll be given an Re values that works for any conductor size or combination of sizes.
If the Re isn't low enough, and you've eliminated any bad connections in the earthing; de-rate the protection.
There are those who want to remove table 8.2 completely, and force us to make a lot of dreary detailed calculations for both EFLI and earth continuity.
I like the fact that it gives us a 'deemed to comply' that we can easily use - and that's near enough for all practical purposes.
But it would be even better without those 'conductor size" columns; which don't add anything useful, but instead mislead people into thinking conductor size is directly relevant when it isn't.
------------
This does highlight that in most cases (ie for all circuits using automatic disconnection of supply as the means of fault protection); earth continuity is actually part of EFLI. The key to why measuring just the PEC is deemed adequate testing is that for most circuits the required trip time is 5 seconds. That's why the re values for 'fuses' (ie HRC fuses) is so much higher than the values given for mcbs - all the 'mcb' values are calculated for 0.4 seconds, as required for circuits supplying sockets.
For a non-socket circuit, if the PEC is good enough to be OK for a 0.4 sec trip time, then it's going to be better than OK for a 5 sec trip time. There's no point going to the trouble of measuring Rph as well; because if there's a n impedance in the active that would prevent operation within 5 sec, it's going to make itself known as soon as supply is connected to load.
Every time we disconnect something for testing purposes, we have to re-connect it - and we should then prove that the re-connection was properly done. Therefore the idea is to do only the minimum amount of testing that will show compliance. Doing a disconnected part-EFLI test when a simple continuity test would provide a sufficient level of proof isn't a good idea; because it increases the risk of the testing process causing a problem.
So we don't generally do EFLI test except for sockets. And even for sockets, we only require EFLI if there's no RCD protection. That's because an RCD has a 0.3 sec operating time (25 % faster than required); almost regardless of the impedance of the PEC (because with an RCD we're not relying on passing a high fault current).
So for any group of circuits, including circuits with different cable sizes and different types of load; the only relevant factors are:
(a) the rating of the overcurrent protection; and
b) whether the circuit supplies any sockets (including eg sockets for luminaires) that don't have RCD protection.
- If so an EFLI test is required and the Rphe values of Table 8.2 apply (alternatively, use live test and table 8.1).
- If not, just a continuity test using the re values
- Rating: 16.67%
Re: Continuity of the earthing system
Yes I understand and agree with all those points.
The argument quite simply, was about whether Re needed to be determined in its own right, or getting the correct Rphe reading was sufficient. My point was, that because of the mixture of cable sizes, it would be impossible to calculate Re from the Rphe measurement (and especially impractical when testing 40 old circuits, with no idea where they start or finish).
The other persons point, was that if Rphe complies, then the circuit protective device will operate as required.
While that may be the case, the requirement of 8.3.5 is to prove the resistance of the PECs, and that hasn't been achieved.
That's before even getting into all the impractical steps that needed to be taken, like disconnecting and reconnecting conductors, and doing calculations. All because someone seemed to have a aversion to running around with a trailing lead - being too "old school".
The argument quite simply, was about whether Re needed to be determined in its own right, or getting the correct Rphe reading was sufficient. My point was, that because of the mixture of cable sizes, it would be impossible to calculate Re from the Rphe measurement (and especially impractical when testing 40 old circuits, with no idea where they start or finish).
The other persons point, was that if Rphe complies, then the circuit protective device will operate as required.
While that may be the case, the requirement of 8.3.5 is to prove the resistance of the PECs, and that hasn't been achieved.
That's before even getting into all the impractical steps that needed to be taken, like disconnecting and reconnecting conductors, and doing calculations. All because someone seemed to have a aversion to running around with a trailing lead - being too "old school".
Re: Continuity of the earthing system
8.3.5 has 2 required results;
so need to see whether measuring resistance A+E , and getting a result less than the relevant Rphe value from table 8.2 satisfies both.
For (a), low enough for operation of protective device, clearly "Yes";
because that's exactly what an EFLI test is about; albeit the method makes several assumptions about the rest of the fault loop.
To that extent doing the EFLI is, in theory, better than measuring just the Re alone; but that "advantage' is mire than offset by the disadvantage of having to disconnect & reconnect conductors.
For (b), consistent with length, CSA & type of conductor, also "Yes".
We're not required to show that the Re values are compliant with the table; that's not as a requirement. It's given to us as guidance (in a Note); and basically it provides a "deemed to comply" value..> But other methods / measurements that achieve the stated required results(a) & (b) are equally acceptable.
You seem to be hung up on a perceived need to calculate an re value for a mixed run of conductor sizes.
But that would be a complete waste of time; because the stated max values aren't based on conductor size; they only change with rating & type of protection. They apply regardless of Cu, Al, or Fe being the PEC material, and regardless of whether it's a PEC in a cable, of a length of conduit or trunking.
Clearly if the PEC is not part of the ciruit cabling, only a trailing-lead type measurement can provide an answer. But where we do have a PEC that's part of a cable, we can choose to measure A + E and we'll still get a result that allows us to conform that the circuit meets both requirements.
True the Table is based on the circuit having all same cable size throughout the circuit route length, and all Cu. But any differences can only be very minor over a typical subcircuit. The only value that matters for 8.3.5 is the Pe, so the question becomes how a Rphe measurement can demonstrate compliance of Re.
For requirement (a); we are given Rphe as a deemed-to-comply for 8.3.9 EFLI, and we can also apply this as a deemed-to-comply for "will allow sufficient current to pass" - because the underlying requirement is identical.
For requirement (b); we need to do some maths to justify the assumption that Rphe can be used to show "consistent with length / type / CSA" when we're testing all-Cu circuits with integrated PEC.
For a cable with equal-sized conductors, we can expect a 50 / 50 split of Rphe. For unequal conductors, from the figures in the Table for 4mm2 with 2.5 PEC; we can expect it will be approx 38 % Rph & 62 % Re, of the max permitted Rphe. The question becomes whether we can just assume this to be true; or whether we need to show that we haven't got an ab-normally low Rph combined with an unacceptably high Re.
The best Rph we can have is with zero-resistance connections, ie the resistance of that length of cable with no joints. Resistance of 2.5 mm2 Cu is 8.14 ohms / km, and 4 mm2 is 5.06 ohms/km ('3008.1.2', Table 35; conductor temp 45 degC).
Compare 0.8 given in 8.3.5.2 (b) for 100 m of 2.5 mm2; and clearly we're only looking to be not excessive, rather than being expected to have no allowance for all the connections along an installed circuit.
'3000' Table B1 gives us guidance on max circuit route length in C-curve mcb of 68 m for 2.5 mm2 (2 0 A) and 67 m for 4 mm2 (25 A).
But worst case will be a short run, so use a 20 m run as example (ie 0.02 km). Rph can't be much less than .02 of the value (we'll be measuring at < 45 deg conductor temp, but the difference is tiny. 0.02 x 5.06 = 0.1 ohm
Assume - again worst case - the Rphe reading is at the max permitted. On 20 A Type C; we're allowed 0.49 for Re; or 0.98 for Rphe. So 0.98 - 0.1 gives worst-case Re as 0.88 ohms. True that's nearly double the "max" Re value we'd use for measuring just the PEC; but that's not a problem
We know the combined A=E loop resistance is compliant; so we're only looking to show "not excessive" for conductor type / length / CSA.
If that PEC was all 2.5 mm2, we'd be starting with 0.02 km x 8.14 ohm/km = 0.163 ohm for a straight length of cable, plus something for all the connections along the route. Compare Note 1 suggesting 0.8 ohm for 100m, so around 0.16 for 20 m.
We know that for this worst-case scenario the Re can't be > 0.88, and in practice it will be much less; probably closer to the 38 / 62 % split at around 0.61 .
Is the difference - not more than 0.7 ohm max ,but probably closer to 0.12 - enough to say Re is not consistent (ie excessive)? That's for the person signing the CoC to decide.
Bearing in mind that even a high-quality DMM has a several-digit error; so that even if we're using a meter resolving to 2 decimal places we can only rely on the 1st-place value.
And remembering we're not looking for absolute match with a required value, just for "consistent with".
-----------------------
Personally I wouldn't normally make that big an assumption about which connections were AOK and which were maybe sub-standard
I prefer to measure PEC alone; it's not hard to do and I can be more confident that the result satisfies both requirements.
I generally leave any EFLI testing for after livening, and use the live method.
If I ever do measure the A+E loop, I only disconnect the A, not the E. There's absolutely no need to disconnect the E, and doing so risks leaving it disconnected. Just use a jumper to link A&E (measurement to be done at load end(s) of circuit; so every branch gets tested)
Alternatively, do the EFLI (dead) test before connecting a new subcircuit; and then do the PEC trailing lead test after.
so need to see whether measuring resistance A+E , and getting a result less than the relevant Rphe value from table 8.2 satisfies both.
For (a), low enough for operation of protective device, clearly "Yes";
because that's exactly what an EFLI test is about; albeit the method makes several assumptions about the rest of the fault loop.
To that extent doing the EFLI is, in theory, better than measuring just the Re alone; but that "advantage' is mire than offset by the disadvantage of having to disconnect & reconnect conductors.
For (b), consistent with length, CSA & type of conductor, also "Yes".
We're not required to show that the Re values are compliant with the table; that's not as a requirement. It's given to us as guidance (in a Note); and basically it provides a "deemed to comply" value..> But other methods / measurements that achieve the stated required results(a) & (b) are equally acceptable.
You seem to be hung up on a perceived need to calculate an re value for a mixed run of conductor sizes.
But that would be a complete waste of time; because the stated max values aren't based on conductor size; they only change with rating & type of protection. They apply regardless of Cu, Al, or Fe being the PEC material, and regardless of whether it's a PEC in a cable, of a length of conduit or trunking.
Clearly if the PEC is not part of the ciruit cabling, only a trailing-lead type measurement can provide an answer. But where we do have a PEC that's part of a cable, we can choose to measure A + E and we'll still get a result that allows us to conform that the circuit meets both requirements.
True the Table is based on the circuit having all same cable size throughout the circuit route length, and all Cu. But any differences can only be very minor over a typical subcircuit. The only value that matters for 8.3.5 is the Pe, so the question becomes how a Rphe measurement can demonstrate compliance of Re.
For requirement (a); we are given Rphe as a deemed-to-comply for 8.3.9 EFLI, and we can also apply this as a deemed-to-comply for "will allow sufficient current to pass" - because the underlying requirement is identical.
For requirement (b); we need to do some maths to justify the assumption that Rphe can be used to show "consistent with length / type / CSA" when we're testing all-Cu circuits with integrated PEC.
For a cable with equal-sized conductors, we can expect a 50 / 50 split of Rphe. For unequal conductors, from the figures in the Table for 4mm2 with 2.5 PEC; we can expect it will be approx 38 % Rph & 62 % Re, of the max permitted Rphe. The question becomes whether we can just assume this to be true; or whether we need to show that we haven't got an ab-normally low Rph combined with an unacceptably high Re.
The best Rph we can have is with zero-resistance connections, ie the resistance of that length of cable with no joints. Resistance of 2.5 mm2 Cu is 8.14 ohms / km, and 4 mm2 is 5.06 ohms/km ('3008.1.2', Table 35; conductor temp 45 degC).
Compare 0.8 given in 8.3.5.2 (b) for 100 m of 2.5 mm2; and clearly we're only looking to be not excessive, rather than being expected to have no allowance for all the connections along an installed circuit.
'3000' Table B1 gives us guidance on max circuit route length in C-curve mcb of 68 m for 2.5 mm2 (2 0 A) and 67 m for 4 mm2 (25 A).
But worst case will be a short run, so use a 20 m run as example (ie 0.02 km). Rph can't be much less than .02 of the value (we'll be measuring at < 45 deg conductor temp, but the difference is tiny. 0.02 x 5.06 = 0.1 ohm
Assume - again worst case - the Rphe reading is at the max permitted. On 20 A Type C; we're allowed 0.49 for Re; or 0.98 for Rphe. So 0.98 - 0.1 gives worst-case Re as 0.88 ohms. True that's nearly double the "max" Re value we'd use for measuring just the PEC; but that's not a problem
We know the combined A=E loop resistance is compliant; so we're only looking to show "not excessive" for conductor type / length / CSA.
If that PEC was all 2.5 mm2, we'd be starting with 0.02 km x 8.14 ohm/km = 0.163 ohm for a straight length of cable, plus something for all the connections along the route. Compare Note 1 suggesting 0.8 ohm for 100m, so around 0.16 for 20 m.
We know that for this worst-case scenario the Re can't be > 0.88, and in practice it will be much less; probably closer to the 38 / 62 % split at around 0.61 .
Is the difference - not more than 0.7 ohm max ,but probably closer to 0.12 - enough to say Re is not consistent (ie excessive)? That's for the person signing the CoC to decide.
Bearing in mind that even a high-quality DMM has a several-digit error; so that even if we're using a meter resolving to 2 decimal places we can only rely on the 1st-place value.
And remembering we're not looking for absolute match with a required value, just for "consistent with".
-----------------------
Personally I wouldn't normally make that big an assumption about which connections were AOK and which were maybe sub-standard
I prefer to measure PEC alone; it's not hard to do and I can be more confident that the result satisfies both requirements.
I generally leave any EFLI testing for after livening, and use the live method.
If I ever do measure the A+E loop, I only disconnect the A, not the E. There's absolutely no need to disconnect the E, and doing so risks leaving it disconnected. Just use a jumper to link A&E (measurement to be done at load end(s) of circuit; so every branch gets tested)
Alternatively, do the EFLI (dead) test before connecting a new subcircuit; and then do the PEC trailing lead test after.
Re: Continuity of the earthing system
Thank you for the question.
It's been very helpful to me; as "30i7 is being currently being revised, and this has forced me to think through the issues behind the requirements.
Which is something we should all understand, given we have freedom to choose "other" methods, but ONLY as long as they provide equally valid results.
Which depends not only on methods, but also what order we do them; but mostly requires that we understand what is, and isn't a 'valid' result.
I don't think this stuff gets taught to apprentices; certainly wasn't in my day
It's been very helpful to me; as "30i7 is being currently being revised, and this has forced me to think through the issues behind the requirements.
Which is something we should all understand, given we have freedom to choose "other" methods, but ONLY as long as they provide equally valid results.
Which depends not only on methods, but also what order we do them; but mostly requires that we understand what is, and isn't a 'valid' result.
I don't think this stuff gets taught to apprentices; certainly wasn't in my day
- Rating: 16.67%
Re: Continuity of the earthing system
Doesn't get taught to apprentices at all really or it's incorrect or vague
I believe testing is one of the biggest flaws in our industry, it's because the people testing don't really understand it themselves, I'm still trying to get my head around certain aspects of testing but most of what I know is self taught and from the great information I've received on here
Thing is it's been hard to get my head around things even actively trying to learn it let alone people who just want to forget about work once they get home
When the your supervisor doesn't really know how can you expect an apprentice to learn, a lot of people don't even know 3017 exists but I'm glad it's getting a modern reworking and really look forward to getting my hands on it
I believe testing is one of the biggest flaws in our industry, it's because the people testing don't really understand it themselves, I'm still trying to get my head around certain aspects of testing but most of what I know is self taught and from the great information I've received on here
Thing is it's been hard to get my head around things even actively trying to learn it let alone people who just want to forget about work once they get home
When the your supervisor doesn't really know how can you expect an apprentice to learn, a lot of people don't even know 3017 exists but I'm glad it's getting a modern reworking and really look forward to getting my hands on it
Re: Continuity of the earthing system
Great discussion guys!
I was an early 70's apprentice an cant remember getting taught any of this, i hope the apprentices of today appreciate the information available and knowledge out there, that they can learn from, from what I have observed though a lot of them dont.
I was an early 70's apprentice an cant remember getting taught any of this, i hope the apprentices of today appreciate the information available and knowledge out there, that they can learn from, from what I have observed though a lot of them dont.
Re: Continuity of the earthing system
I hope the question is helpful in re-fashioning some additional clarity into the clause.. because in my view, 8.3.5.2 (a) and (b) are quite clearly only talking about the PEC.
The first sentence "The resistance of protective earthing conductors shall be:" says to me, that the result must be only the PEC. Not a total of Rphe being less than the table value. If that was all that was necessary, I would expect that the clause will be clear that Rphe is acceptable.
Rphe is a valid answer for a dead EFLI test, as stated in the EFLI test section.
I have no issue with whatever method people want to use, resistors with instruments that calculate the value, or testing the value of A & N loop resistance, then halving that to use one conductor as your "trailing lead". One situation raised in the other post was a 300m run to a pump (for example), where a trailing lead would be impractical. Sure, in that case, where it's one known cable, then using Rphe would be fine to do the calculation of Re.
Then you get to practical situations like the one I had - wanting to test a complete existing (old) installation. Disconnecting 40+ circuits to test individually would be madness.
The way I tested it, was to remove the main neutral (it's a DB, not MSB) - then test N-E at the switchboard to ensure there wasn't any faults - then locate and repair the fault (N-E short), then do the trailing lead test to all earthed items.
Testing this way ensures that the earth is connected correctly, and also found a N-E reversal which wouldn't have been picked up otherwise as it is non-residential and doesn't have any RCDs.
After the PEC testing (and repairing the faults), carry on with the IR, then live polarity and loop testing.
Photos of faults:
The first sentence "The resistance of protective earthing conductors shall be:" says to me, that the result must be only the PEC. Not a total of Rphe being less than the table value. If that was all that was necessary, I would expect that the clause will be clear that Rphe is acceptable.
Rphe is a valid answer for a dead EFLI test, as stated in the EFLI test section.
I have no issue with whatever method people want to use, resistors with instruments that calculate the value, or testing the value of A & N loop resistance, then halving that to use one conductor as your "trailing lead". One situation raised in the other post was a 300m run to a pump (for example), where a trailing lead would be impractical. Sure, in that case, where it's one known cable, then using Rphe would be fine to do the calculation of Re.
Then you get to practical situations like the one I had - wanting to test a complete existing (old) installation. Disconnecting 40+ circuits to test individually would be madness.
The way I tested it, was to remove the main neutral (it's a DB, not MSB) - then test N-E at the switchboard to ensure there wasn't any faults - then locate and repair the fault (N-E short), then do the trailing lead test to all earthed items.
Testing this way ensures that the earth is connected correctly, and also found a N-E reversal which wouldn't have been picked up otherwise as it is non-residential and doesn't have any RCDs.
After the PEC testing (and repairing the faults), carry on with the IR, then live polarity and loop testing.
Photos of faults:
- Attachments
-
-
Re: Continuity of the earthing system
Agree that for a very long run it's probably OK to use one of the other cores as the equivalent of a trailing lead.
Also that existing installations may need a different approach than new work.
Fault-finding is different again.
Your example highlights that testing has seldom been done as thoroughly as it should be.
That N-E reversal should have been picked up with initial testing - but many of us were not trained to do all that "time-wasting" testing.
When I first got a loop tester, I used it to test all the suspended fluo lights on the job we were doing;
and out of 60 fittings, 3 weren't earthed -all errors of connection at either ceiling rise or luminaire.
That's 5% fail rate; and taught me that none of us are as good as we think we are.
Yet we all tend to walk into any instalation and assume that, when new, it was all compliant.
Also that existing installations may need a different approach than new work.
Fault-finding is different again.
Your example highlights that testing has seldom been done as thoroughly as it should be.
That N-E reversal should have been picked up with initial testing - but many of us were not trained to do all that "time-wasting" testing.
When I first got a loop tester, I used it to test all the suspended fluo lights on the job we were doing;
and out of 60 fittings, 3 weren't earthed -all errors of connection at either ceiling rise or luminaire.
That's 5% fail rate; and taught me that none of us are as good as we think we are.
Yet we all tend to walk into any instalation and assume that, when new, it was all compliant.