Imported: Earth Fault Loop Impedance Testing - Residential

Post Reply
User avatar
DougP
Posts: 203
Joined: Sun Apr 12, 2020 10:11 pm
Answers: 3
Has thanked: 11 times
Been thanked: 27 times

Imported: Earth Fault Loop Impedance Testing - Residential

Post by DougP »

JonnoM
Nov 14 2017 23:25


This topic has been covered a few times but I still couldn\'t get a definitive answer. Sorry for the novel but there is no simple way to explain all this. I am a 3rd year apprentice and work for my father who I have also confused on this matter. My confusion is around the requirements for testing the earth fault loop in particular a new build residential installation with supply connected (commonly when the test is done). ASNZS:3000 8.3.9.1 states where an installation includes socket outlets NOT protected by an RCD, the impedance required for automatic disconnection of supply shall be verified on each circuit...And 8.3.9.2 Note 3 states if an RCD operates during the test the result is considered satisfactory.

Q1. Does that mean in this scenario with ALL socket outlets protected by RCDs you don\'t have to do the earth loop test at all and you can just test the operation of the RCD with the test button and thats enough? Or do you still need to do an Earth loop test at the furtherest socket of each circuit so to prove the RCD tripped?

Q2. Is testing earth loop impedance only required for circuits containing socket outlets or are there other areas that need to have this test done, e.g I have been told you need to test the mains for earth loop impedance at the main switchboard?

Q3. Is the method of testing with the supply connected the best/satisfactory method when RCDs are in place or is it better to test the resistance of individual circuits with the supply not connected to actually verify the ELZ and calculate the PSSC, or thirdly is it better practice to use an ELZ tester with a no trip function?

Q4. Lastly in terms of any new installation of circuits with socket outlets in a residential scenario, e.g new power point circuit installed in an existing home, the requirement would be an RCD needs to be installed at the board, therefore as previously stated I wouldn\'t need to test earth loop? Basically the Earth Loop Impedance test is obsolete in any new residential installation...

AlecK
Nov 15 2017 09:13


Direct answers first, then some explanation.

Q1.
a) \"Does that mean in this scenario with ALL socket outlets protected by RCDs you don\'t have to do the earth loop test at all and you can just test the operation of the RCD with the test button and thats enough?\"

Yes, that\'s exactly what it means (for EFLI test).

b) \"Or do you still need to do an Earth loop test at the furtherest socket of each circuit so to prove the RCD tripped?\"

No.

Q2.a) \"Is testing earth loop impedance only required for circuits containing socket outlets\"

Yes.

b) \"or are there other areas that need to have this test done, e.g I have been told you need to test the mains for earth loop impedance at the main switchboard?\"

There is NO requirement to do EFLI tests for mains or for non-socket circuits. However there is a requirement for all isolation and protection devices to be rated for the PSSC at the location they are installed; and since there\'s a direct relationship between EFLI and PSSC a test at main switch can be useful to confirm you got the design right.



Q3. \"Is the method of testing with the supply connected the best/satisfactory method when RCDs are in place or is it better to test the resistance of individual circuits with the supply not connected to actually verify the ELZ and calculate the PSSC, or thirdly is it better practice to use an ELZ tester with a no trip function?\"

Since no EFLI tests are required on RCD-protected circuits, all three procedures suggested are additional / not required. however of the three, the most useful is usikg the \"no-trip\" function of an EFLI tester that has one.


Q4. \"Lastly in terms of any new installation of circuits with socket outlets in a residential scenario, e.g new power point circuit installed in an existing home, the requirement would be an RCD needs to be installed at the board, therefore as previously stated I wouldn\'t need to test earth loop? Basically the Earth Loop Impedance test is obsolete in any new residential installation...\"

Correct - though I wouldn\'t say \"obsolete\", just \"not required\".
---------------
This clause has been considerably revised for the next edition; with more explanation. Which should reduce confusion by increasing understanding.

First thing that has to be understood is what we are checking for: fault protection[2.4]. There are several acceptable methods, but the most common is automatic disconnection of supply [2.4.1 (a), & 2.4.2]; and for that we use a system of protective earthing of exposed conductive parts, plus either a circuit overcurrent device or an RCD. Using overcurrent devices requires low EFLI; using RCDs doesn\'t.

Why test only socket circuits?
Because only socket circuits have a required operation time of 0.4 sec - all other circuits are 5 sec. Clearly a longer time allows a lower EFLI - you can get an idea as to what sort of EFLI will be effective from last 2 columns of Table 8.1, which give figures for HRC fuses. Remember you\'ve already tested the PEC part of the EF loop, and ensured it\'s good enough for 0.4 sec. Any impedance in the active part bad enough to prevent operation of fault protection is going to show up as volt drop and / or performance failure under load.

Why not test RCD-protected circuits?
because the operating time of the RCD is no more than 0.3 sec.

Why is the live test preferred?
Because it tests the actual EFLI, rather than testing only the subcircuit\'s contribution and making (well-founded) assumptions about the rest.
For (additional) testing when an RCD is in circuit, using a \"no-trip\" function gives an actual reading; though typically less accurate than the standard test used for a non-RCD circuit. Whereas tripping the RCD just tells you the RCD works; and you\'ll be doing that later anyway as part of RCD testing. So while that\'s enough to comply (for EFLI) you must be looking for more or you wouldn\'t be doing extra EFLI tests.


pluto
Nov 15 2017 13:38


The above theory lesson was OK until this paragrph which contains a major error.
I quote

Why test only socket circuits?
Because only socket circuits have a required operation time of 0.4 sec - all other circuits are 5 sec.

Clearly a longer time allows a [\"lower\" The error; it should read \"Greater\"] EFLI - you can get an idea as to what sort of EFLI will be effective from last 2 columns of Table 8.1, which give figures for HRC fuses.

end of quote with my correction in [ ]

Unless my as/nzs 3000 has a misprint in table 8.1 (2 right hand columns) the 5 secs EHLI values are greater than the 0.4 sec values.



AlecK
Nov 15 2017 14:00


Yes, I should have said the longer time allows \"higher\" EFLI; due to not needing as much current to flow.

SparkyJoe
Nov 15 2017 15:38


So are you dead testing (r1+r2) (Live earth) at your end point before powering up?

Or are you just doing a insulation resistance test and then pressing the trip button when its powered up

AlecK
Nov 15 2017 16:44


Neither.
For an RCD-protected circuit supplying sockets, the \"dead\" EFLI test is just as unnecessary as the live one; so I don\'t do it.
But neither do I \"just\" do I.R. test followed by the first part of an RCD-operation test.

The sequence I generally follow for a single new subcircuit is:
First, earth continuity, at each point on the circuit.
Then I.R. testing
Then correct circuit connections.
Then liven for purposes of testing.
Then test that the RCD actually disconnects all live conductors of the circuit when it operates.
Finally polarity, again at every point on the circuit.
All of which is required; so if not done then the CoC (saying \"work was done lawfully and safely\") would be false.





benben
Nov 15 2017 19:56


interesting topic, on my current job i am required to EFLI test every circuit, lights,power,hwc.... the list goes on.

i am told we have to test the earth is connected and an efli covers this.

i understand that you only test power points with no rcd. i also would do main anyway.

testing lights are a pain in the ass.

what are you guys recomendations.

AlecK
Nov 16 2017 09:07


There\'s another aspect to the \"live\" EFLI test: it can be hazardous on non-socket circuits because it requires access to live terminals of equipment. And with motors that brings mechanical risks as well as electrical.

The benefit of doing any test has to be weighed against the risks of doing it. That\'s why most recommended test methods are done on de-energised circuits.

An EFLI test done at last point of circuit can help towards confirming the PEC is properly connected at that point - depending what other testing you do. But it does nothing for all the other points on the circuit - every earthed point needs to be tested.

Important to remember that while alternative test methods are allowed, they have to give \"equally valid results\". That\'s much harder to achieve than it sounds, as every possible fault has to be taken into account. The specified testing isn\'t a bunch of individual tests, but a system where results of each type of test work together with results of other tests. A \"wrong\" test result can put earlier \"pass\" results of other types of test into doubt. And some test methods risk creating faults of types already tested for & \"passed\".

Neither of the accepted EFLI test methods is enough by itself to establish for certain that the PEC
a) is connected correctly; and
b) has acceptable impedance.
So if EFLI is being used as a means of avoiding proper earth continuity tests, then the testing system is incomplete and therefore non-compliant.


EdTPole
Nov 16 2017 10:17


Aleck you said

\"Neither of the accepted EFLI test methods is enough by itself to establish for certain that the PEC
a) is connected correctly; and
b) has acceptable impedance.
So if EFLI is being used as a means of avoiding proper earth continuity tests, then the testing system is incomplete and therefore non-compliant.\"

Has acceptable impedance? Is that not what an earth fault loop impedance test does?

My tester has both a trip and non trip function. The non trip operation must pass less than 30mA to the PEC from the active and it also loads the neutral as it requires a neutral connection. The high current test requires just the PEC and active conductors and gives me two results. Earth impedance and PSCC.

When EFLI is tested at every socket with an assistant at the board to reset the RCD does this not prove that the PEC is connected and have acceptable impedance if it meets the requirements of AS/NZS 3000?

AlecK
Nov 16 2017 11:31


The obligation to test earth continuity is separate from and additional to tghe obligation to test EFLI> It\'s also universal; ie every point on every circuit, compared with just the end point on circuits supplying non-RCD sockets.

The generally accepted way of testing EC is using a trailing lead; so if you want to use a different method it has to produce results that are equally valid. And an EFLI test (either method) can\'t do that in every case and for all possible faults.

For the \"live\" test, firstly the connections; as you say most low-current (\"no trip\" EFLI testers require a \"neutral\" as well as an \"earth\" connection. It\'s not loading the N in any significant way, just using it to carry test current that would otherwise trip the RCD. But the tester can\'t tell whether the N & E connections have been reversed. So you could be actually testing a loop that includes subcircuit N instead of PEC.

Secondly the impedance value.
The reading you get is for the entire actual EFL; from the distribution tx away down the street, via lines, mains, submains and final subcircuit. The submain PEC is a tiny part of it; so the reading will always be larger than the actual impedance of the subcircuit PEC - and that\'s the impedance that 8.3.5 requires us to confirm.

True, if the reading is below the limit set by Table 8.2 for Re; you can safely assume that the \"PEC\" part of it is OK. For readings above the limit in the Table; you\'re guessing / making assumptions. The further above the limit value, the less reliable those assumptions are.

And all of that is only provided that you know that it\'s the PEC that you did the test on - which you can\'t unless you\'ve done other tests to confirm that.
Basically if you haven\'t proved the PEC is correctly connected at all points (including at switchboard), then your IR tests, your polarity tests, your correct circuit connections tests, and your EFLI tests are all unreliable. That\'s because the standard test methods for these other tests rely on the earthing system being correctly set up.

Also with cables above 2.5, the N is often larger than the PEC; so for a N-E reversal, the value will be artificially low.

Yes, I know that when load is added, the RCD will trip for most types of incorrect connection; but at this stage (testing prior to placing in service) there is no actual load, and on a \"no-trip\" setting you don\'t add enough - while on the full EFLI setting you\'ll just trip the RCD and either not get a reading at all or get one you can\'t rely on.

With the \"dead\" test; again the reading (Rphe) is not the one required, but a higher value; and while assumptions may be made about how much of that reading is Re and how much is Rph they are assumptions; they can\'t provide certainty unless the Rphe reading is lower than the limit for Re.


Testing can\'t be based on assuming everyone has done everything right; the whole point is to prove, beyond doubt, that no mistakes have been made. So for any test method, we have to consider how possible faults will affect the results, not only for that test but also for other types of test. Same applies to the order we do the testing in.

EdTPole
Nov 16 2017 11:51


I don\'t use the EFLI test that does not trip an RCD as like someone mentioned earlier it does not give an accurate reading.
As for PEC and neutral transpositions under a high current test if the RCD does not trip then this can indicate a broken PEC or PEC/N transposition.

AlecK
Nov 16 2017 13:05


Another thing to consider is safety while testing. The new edition will specifically cover this point, along with selection of appropriate test equipment.

Going straight to a live EFLI test without first checking that PEC is correctly connected risks livening exposed conductive parts of the circuit.

Yes using the high-current EFLI test should trip the RCD, and if it doesn\'t you know something is wrong. But it can\'t give a reading for continuity of PEC; and that\'s what is needed to satisfy both parts of 8.3.5.2.
So there\'s no way it gives results \"equally valid\" for earth continuity as the \"normal\" test.




EdTPole
Nov 16 2017 13:24


Exposed metal parts on a power circuit?

So an earth loop impedance reading does not give a measure of continuity?

An impedance test stresses the PEC in the manner that it was designed for.



EdTPole
Nov 16 2017 14:47


Let me be clearer in my question Aleck.

In your initial response to the OP, you stated,

\"Since no EFLI tests are required on RCD-protected circuits, all three procedures suggested are additional / not required.\"

It was my understanding that EFLI tests are required on all circuits, RCD protected or not. Also owhere can I find any reference to where it states only on circuits connected to socket outlets?

OP \"Q2.a) \"Is testing earth loop impedance only required for circuits containing socket outlets\"

Aleck \"Yes.\"

AS/NZS 3000:2007 8.3.3 Lays out the mandatory tests, no exceptions for RCD protected circuits.

8.3.3 e) Verification of impedance required for automatic disconnection of supply (earth fault-loop impedance), in accordance with clause 8.3.9.

8.3.9.2 b) 3) If an RCD operates during the test, the test result is considered satisfactory.

No mention of any exclusion for RCD protected circuits or circuits that don\'t supply socket outlets. Please correct me if I\'m wrong.



AlecK
Nov 16 2017 15:03


while it\'s unlikely any plug-in appliances will be connected at time of testing a new subcircuit, circuits are not restricted to only supplying only one type of outlet 9eg sockets. So it\'s very possible that there can be exposed conductive parts that would be livened during testing.

Yes a no-trip EFLI test give a reading of impedance - but not of the impedance Re that 8.3.5 requires us to confirm. The full-current test typically doesn\'t give any reading on an RCd-protected circuit, just trips the RCD.

True a normal EFLI test \"stresses\" the PEC (and the active); though not \"in the manner it was designed for\". The test current is typically around 23A, whereas the current the PEC is designed for is much greater. The test current is designed not to stress the circuit conductors & terminations, but simply to allow the impedance to be calculated with reasonable accuracy.

And it does not confirm Re.

So it\'s a possibly useful extra, but it\'s no substitute for an EC test. The first and most important part of which is to ensure that you are actually measuring a PEC and not some other conductor that may have been connected to a E terminal.

AlecK
Nov 16 2017 15:53


Yes 8.3.3(e) is verification of impedance required for automatic disconnection. The first result of that is that if the circuit doesn\'t use automatic disconnection for fault protection, then confirmation of impedance to achieve it is not required.
Where the test is required, it must be done \"in accordance with 8.3.9\".
Each of the tests \'required\" by 8.3.3 is similar, in that they all refer to a later clause for the details.

8.3.9.1 states where / when EFLI testing is required. The only testing required by the clause is for \"circuits containing socket outlets not protected by RCD\". What is not required need not be done, and clearly not required on all circuits; though the current wording has been improved for next edition eg will reinforce the \"only for automatic disconnection\" aspect.

As so often words not there are just as important as words included. If it doesn\'t say you \"must\" or \"shall\", you don\'t have to.
So not required, for example, if each socket has a built-in RCD, not required for lighting, not required for submains, etc.

Where it is required; high fault current and therefore relatively low impedance is required to operate the over-current device.

For RCD-protected sockets, as outlined previously, automatic disconnection by RCD does not require a low-impedance fault current path. Only a few mA needs to flow for the fault protection to operate within time limit. And it\'s that time limit that\'s crucial, because pretty much any fault current via PEC (or any other return path) will cause RCD to trip; and when it trips it will be within time limit regardless of the fault current level. So the impedance of PEC can be relatively high. It needs to be pretty much open circuit before the fault protection stops working, and we know it isn\'t that high, because we tested earth continuity as required by 8.3.3(a) and 8.3.5. And therefore we already know that the fault protection will work.

For all other circuits, eg mains, submains, lighting, fixed appliances,
1 the time limit is much greater;
2 while 8.03.3(e) calls for testing, it does so \"i.a.w. 8.3.9\" and that clause does not require a test; and
3 we tested EC.

Put another way, the reason 8.3.9 doesn\'t require circuits without non-RCD sockets to be EFLI tested is because it would amount to a double-up, the impedance of PEC has already been verified sufficiently under EC testimg. Especially when the risks arising from the test are considered (eg measuring EFLI on a motor while it\'s running).

What you appear to have been suggesting, and what I don\'t accept, is doing the EFLI test instead of an EC test. Section 8 doesn\'t accept it either.

For non-RCD sockets, because of the higher risk that a person might be on the receiving end of a fault, EC testing by itself is not considered enough, and the added risk of an EFLI test is not great. The \"live\" test gives absolute confirmation, and the \"dead\" test is considered an adequate alternative because any impedance in the rest of the EF loop will show up as an operational failure long before there\'s any earth fault.





EdTPole
Nov 16 2017 16:57


\"What you appear to have been suggesting, and what I don\'t accept, is doing the EFLI test instead of an EC test. Section 8 doesn\'t accept it either.\"

Not at all. The way I read the standard is that an acceptable result must be obtained for socket outlets not protected by RCD and that if there is an RCD then it tripping is an acceptable results.

What you seem to be implying is that I think continuity tests do not need to be done. That I do not accept.

I will continue carrying out EFLI test on any socket I have worked on be it protected by RCD or not. This means I will test power sockets. Range sockets and I always check EFLI when having worked on hot water cylinders. Why wouldn\'t you?

A partially cut PEC will still pass a continuity test but it might not pass an EFLI test if the RCD is faulty.

At the end of the day I suppose it is the case of doing the minimum required vs a professional job.

Funnily enough just the other day I had an inspection after installing an earth stake, MEC, earth, and neutral bar with MEN link. I tested the MEC but the inspector didn\'t. He disconnected the main earth and did an EFLI test on it. I quizzed him on it. He said it could be bond wire instead and I should have a label on it. My argument was it does not require a label and if it was a bond it could be to pipework and he would still be in the dark about the MEC and earth stake.

AlecK
Nov 16 2017 19:42


Ok, seems I picked up a wrong message.

So in terms of just EFLI testing of installation work, it\'s only required for sockets that don\'t have RCD protection.

And yes, the requirements of Wiring rules are the minimum. You can do it on other circuits as well, but you don\'t have to.

For your MEC example, one requirement is that the continuity of the MEC be not more than 0.5 ohm. The \"EFLI\" test gives an idea of the overall effectiveness of the electrode, but isn\'t required; largely because there is absolutely no requirement for effectiveness. And it can\'t prove the resistance of the MEC - or any of the other requirements of a main earthing system.

The inspector is right that it can be difficult to ID the MEC, and a label certainly helps. A label used to be required at swbd end but you\'re right that it isn\'t any more.
And that just doing the \"EFLI\" (even on a labelled wire) doesn\'t prove that the other end of that wire is what\'s connected to the electrode.

However the testing required of inspectors is not the same as the testing required for verification of work. All the inspector has to do is \"whatever testing is required to satisfy\" that the work complies and will be safe. Some may be more easily satisfied than others, and there\'s no specific required tests. Personally I wouldn\'t be satisfied with just an \"EFLI\" test, and I don\'t usually do one.

I do like to disconnect the MEC when testing its continuity, but even then you can\'t be absolutely certain there are no parallel paths


DougP
Nov 16 2017 20:07


EdTPole, do you do a EC test on every earthed item and every socket outlet or not?

When you said \"When EFLI is tested at every socket with an assistant at the board to reset the RCD \" - it appears that you use this instead of EC testing? How does that work with PC items like towel rails, wall ovens or fan heaters for example?

It seems ridiculously tedious to EFLI test every outlet, and have to reset the RCD, rather than do the normal trailing lead EC test as required by 8.3.5 and in the sequence of tests. The trailing lead test is simple to complete by one person. And it give the required result as per 8.3.5.2.

Plus, if you do the EC test with RCDs off (therefore N disconnected), you have taken care of half of your polarity test for RCD protected circuits as well - because you know you can only be testing for the earth conductor on the earth terminal. Probably more than once I have found cables which were \"lost\" from a flush box, just from the initial EC test.

EdTPole
Nov 16 2017 20:30


Doug Yes I do as per 8.3.3 and the following flow chart shows the best order to test in. I use an extension lead with the plug cut off one end wired into the de-energised SWB observing correct polarity and removing the MEN link for obvious reasons. Don\'t you?

EdTPole
Nov 16 2017 20:37


No worries Aleck. I actually have an AVO Earth tester that has a couple of stakes that you drive at 10m and 20m from the stake for testing earth continuity. Have used it on a couple of large commercial builds just for reassurance.

The Inspector I mentioned puzzled me with the things he complained about but was happy to take my word for it that the MEC was connected to the stake and had a resistance of 0.1 ohms. Every other inspector I have come into contact with brings a trailing lead to test it.

DougP
Nov 16 2017 21:53


I know how to carry out the required tests Ed.

Maybe you could just take this opportunity to absorb some other peoples knowledge on the matter.

EdTPole
Nov 16 2017 22:09


Good I\'m glad we both know how to test Doug. No need to try and prove any superiority you think you might have over anybody else. Unless you care to give your name registration details?

Sarmajor
Nov 17 2017 10:02


No worries Aleck. I actually have an AVO Earth tester that has a couple of stakes that you drive at 10m and 20m from the stake for testing earth continuity. Have used it on a couple of large commercial builds just for reassurance.

Reassurance of what exactly.
Sure you prove that you have a connection from your earth electrode to the mass of earth but it is not a required test and there is no maximum value in the regs or AS/NZS3000:2007.

Still does not prove the MEC or effectivelness of the EFLI circuit, given that almost all of the fault current in a phase to earth fault only flows in the installation PEC\'s until it reaches the MEN link. From there the path of least resistance is via the PEN back to the transformer.

EdTPole
Nov 17 2017 14:49


Sarmajor that the earth stake was conductive to the general mass of earth in some very dry environments.

It was in response to Alecks comment,

\"The \"EFLI\" test gives an idea of the overall effectiveness of the electrode, but isn\'t required; largely because there is absolutely no requirement for effectiveness.\"

Just because it\'s not required does not mean I can not do it.



Sarmajor
Nov 17 2017 17:20


Given that the EFLI test is carried out with the MEN link in place it will tell you bugger all about the effectiveness of the MEC.
The earth fault current in an installation will take the path of least resistance which is via the PEN back to the star point of the Transformer.

Even if your earth test gives a low reading for the connection of the electrode to the mass of earth, the resistance of the mass of earth back to the Transformer earthing system (and other installation earthing systems) will still prevent any signifigant current from flowing through the dirt path.

EdTPole
Nov 17 2017 17:39


Sarmajor

\"Given that the EFLI test is carried out with the MEN link in place it will tell you bugger all about the effectiveness of the MEC.
The earth fault current in an installation will take the path of least resistance which is via the PEN back to the star point of the Transformer.\"

Please tell me where I stated that this was or was not the case.

Stating the obvious brings what to the table?

Maybe you should read the thread again.

I\'m not here for an argument sorry.
Post Reply