Earth bonding question.

lucernarian

A quick question on earth bonding, people talking about a mininum 6mm2 being used for any earth bonding work, but what about light fittings and metal switches etc?! Surely you can't be shoving down 6 or 10mm2 wires and hoping they'll fit into the screw terminals of a back box or a metal faceplate?! If the rule is at least half the size of the conductor, there wouldn't be a problem there.

But anyhow, this conversation seems to be just about overall earth bonding for the house so I'll be off:p

Bruthal

To_be_confirmed wrote: »

A quick question on earth bonding, people talking about a mininum 6mm2 being used for any earth bonding work, but what about light fittings and metal switches etc?! Surely you can't be shoving down 6 or 10mm2 wires and hoping they'll fit into the screw terminals of a back box or a metal faceplate?! If the rule is at least half the size of the conductor, there wouldn't be a problem there.

But anyhow, this conversation seems to be just about overall earth bonding for the house so I'll be off:p

Bonding is the earthing of extraneous metal such as pipes and sinks which form no part of the electrical installation, so that in the event of a neutral failure on a neutralised installation, all exposed and extraneuos metal will be at the same voltage, even if thats 230v, which would reduce the chance of a fatal shock, which would be more dangerous if one pipe was at 230v and another was at ground potential.

Earthing of light fittings and metal back boxes is an earth to protect the box from becoming live due to a fault from contact with the circuit supplying the light or switch box, and so the earth to them only needs to be half the size of the circuit conductor. These items would not be considered extraneous or exposed metal like a sink or pipes etc would be.

lucernarian

I wonder if I should be concerned when I record 15-30V differences between copper water pipes and say a nearby socket's earth...:eek: This was in the family's house, in the attic which has a socket installed there on a ring main.

Bruthal

To_be_confirmed wrote: »

I wonder if I should be concerned when I record 15-30V differences between copper water pipes and say a nearby socket's earth...:eek: This was in the family's house, in the attic which has a socket installed there on a ring main.

Test between the socket neutral and its earth, and the socket neutral and the pipe. This should tell you which earth is above ground potential, assuming the socket neutral is ok. If its a neutralised installation the pipes should be at the same potential as all earths in the house.

lucernarian

It's definitely the pipes which are above ground potential. From what I've seen of the house, it appears to use TN-C. I've found nothing that suggests there's an earth connected to anything but the neutral bar near the meter.

Bruthal

Its not really possible when testing between 2 items with a voltmeter to tell which is the higher potential without using some known reference. The socket earth is known, but it may well not be a good socket earth, and a non connected (to earth bar) socket earth wire can show a few volts through induction as its in a live cable. But it could be the pipes above ground potential alright.

But anyway, all pipes should be bonded in a TN-CS setup.

M cebee

robbie7730 wrote: »

Bonding is the earthing of extraneous metal such as pipes and sinks which form no part of the electrical installation, so that in the event of a neutral failure on a neutralised installation, all exposed and extraneuos metal will be at the same voltage, even if thats 230v, which would reduce the chance of a fatal shock, which would be more dangerous if one pipe was at 230v and another was at ground potential.

Earthing of light fittings and metal back boxes is an earth to protect the box from becoming live due to a fault from contact with the circuit supplying the light or switch box, and so the earth to them only needs to be half the size of the circuit conductor. These items would not be considered extraneous or exposed metal like a sink or pipes etc would be.

- the sink is extraneous but the metal light fitting is exposed


- the 'protective conductor' has to be the same size as the phase/line(up to 16) except when using T&E

Bruthal

M cebee wrote: »

- the sink is extraneous but the metal light fitting is exposed

True, obviously enough:o

- the 'protective conductor' has to be the same size as the phase/line(up to 16) except when using T&E

This more or less happens automatically while wiring in a domestic situation anyway.

M cebee

ya:)

i find it's a good way to learn -checking the rules when a question comes up


if you were using pvc conduit and singles you'd have to use same size 'live conductors' and CPC


can you use a smaller size earth on final circuits in steel conduit as they're paralleled with the steel ?

(i know in theory you don't need them but industry practise is separate earth)
i usually do them same size-but copper is expensive now

Bruthal

M cebee wrote: »

ya:)

i find it's a good way to learn -checking the rules when a question comes up

Well the exposed metal one is sort of common sense, something which dissapears after a few early morning cans:D

if you were using pvc conduit and singles you'd have to use same size 'live conductors' and CPC


can you use a smaller size earth on final circuits in steel conduit as they're paralleled with the steel ?

(i know in theory you don't need them but industry practise is separate earth)
i usually do them same size-but copper is expensive now

In industrial wiring we used to use 4 square and 2.5 earth as far as i can remember in socket circuits alright, and 1.5 earth in the 2.5 lighting circuits. Its been a while since i did any final circuit wiring with individual singles now though, so the memory might be a bit dodgy. More switchroom terminating stuff on any industrial jobs i was at, even thats been a while now.

2011

Duplicate post deleted

2011

In industrial wiring we used to use 4 square and 2.5 earth as far as i can remember in socket circuits alright, and 1.5 earth in the 2.5 lighting circuits. Its been a while since i did any final circuit wiring with individual singles now though, so the memory might be a bit dodgy. More switchroom terminating stuff on any industrial jobs i was at, even thats been a while now.

+1

That was my experience up to the year 2000 on large industrial projects. Since then the CPC and phase have been sized the same on the jobs with steel conduit that I have worked on. This means that the prospective short circuit current will be higher and consequently disconnection time reduces.

M cebee

2011 wrote: »

+1

That was my experience up to the year 2000 on large industrial projects. Since then the CPC and phase have been sized the same on the jobs with steel conduit that I have worked on. This means that the prospective short circuit current will be higher and consequently disconnection time reduces.

larger cpc's will make little difference to PSCC or disconnections times i would say-as they're paralleled with the steel

steel takes 80-90% of fault current in a fault scenario afaik


they're sized same as live conductors where i work-i assume the reasoning is that if you're not depending on steel as a protective conductor then they should be regulation size evehn though they're paralleled with steel

Bruthal

M cebee wrote: »

larger cpc's will make little difference to PSCC or disconnections times i would say-as they're paralleled with the steel

steel takes 80-90% of fault current in a fault scenario afaik


they're sized same as live conductors where i work-i assume the reasoning is that if you're not depending on steel as a protective conductor then they should be regulation size evehn though they're paralleled with steel

It should not make very much difference between the half size earth and full size even without the conduit in parallel, if you take a loop impedence for example, at 2.5 ohms, thats nearly 100 amps, which shouldnt make much difference on a 20 amp circuit.

With the steel conduit there certainly would be no difference once its a continous path.

I think myself the main improvement about full sized earths is they have the better mechanical strength/strong connections, especially at connection terminals, for a given size circuit conductor.

2011

M cebee wrote: »

larger cpc's will make little difference to PSCC or disconnections times i would say-as they're paralleled with the steel

steel takes 80-90% of fault current in a fault scenario afaik

That is a very sweeping statement
Although I see your point and agree that in many cases you are correct. But how much fault current a steel conduit takes depends on the impedance of the conduit and the CPC. The impedance of the CPC is easy to calculate with a high degree of accuracy, but the impedance of the conduit will depend on many factors. These include:

1) How many joins there are in the conduit (couplers etc.)
2) How tight these joins are. A loose joint will present far more resistance.
3) If the joins corrode or loosen over time.
4) What diameter the conduit is (20, 25, 32mm) and how thick the walls of the conduit are (class 1,2,3,4,)
5) If trunking makes up part of the earth fault loop, what size is the trunking

etc.....

In fact there are so many variables that it is very hard to calculate form a design point of view. There is also a concern that in the future a modification will be made such as flexible conduit (like Kopex) is installed in a section and continuity will be lost.


Therefore it is often ignored and the calculation is done on the CPC instead.

In reality I agree that in many cases increasing the CPC from say a 2.5 to a 4 sq. will make little difference if it is in steel conduit, but not in all! Design is generally done on a worst case sinario i.e steel conduit does not provide any connection to earth.

they're sized same as live conductors where i work-i assume the reasoning is that if you're not depending on steel as a protective conductor then they should be regulation size evehn though they're paralleled with steel

Regulation size is not always equal to the size of the phase conductor. If it can be shown that the CPC selected satisfies the "adiabatic equation" (section 543.1.2 in ET101:2008) the size can be reduced. For example standard 2.5 T & E has a 1.5 CPC.

I have worked on many large projects in Ireland where the CPC is 1/2 the size of the phase conductor for most circuits (offices, colleges, hospital, government buildings etc..).

Frequently on large projects the electrical designer will specify that the CPC must be the same size as the phase conductor (particularly in ATEX zoned areas) but this does not mean it is a regulation.

Bruthal

2011 wrote: »

That is a very sweeping statement
Although I see your point and agree that in many cases you are correct. But how much fault current a steel conduit takes depends on the impedance of the conduit and the CPC. The impedance of the CPC is easy to calculate with a high degree of accuracy, but the impedance of the conduit will depend on many factors. These include:

1) How many joins there are in the conduit (couplers etc.)
2) How tight these joins are. A loose joint will present far more resistance.
3) If the joins corrode or loosen over time.
4) What diameter the conduit is (20, 25, 32mm) and how thick the walls of the conduit are (class 1,2,3,4,)
5) If trunking makes up part of the earth fault loop, what size is the trunking

etc.....

In fact there are so many variables that it is very hard to calculate form a design point of view. There is also a concern that in the future a modification will be made such as flexible conduit (like Kopex) is installed in a section and continuity will be lost.


Therefore it is often ignored and the calculation is done on the CPC instead.

In reality I agree that in many cases increasing the CPC from say a 2.5 to a 4 sq. will make little difference if it is in steel conduit, but not in all! Design is generally done on a worst case sinario i.e steel conduit does not provide any connection to earth.




Regulation size is not always equal to the size of the phase conductor. If it can be shown that the CPC selected satisfies the "adiabatic equation" (section 543.1.2 in ET101:2008) the size can be reduced. For example standard 2.5 T & E has a 1.5 CPC.

I have worked on many large projects in Ireland where the CPC is 1/2 the size of the phase conductor for most circuits (offices, colleges, hospital, government buildings etc..).

Frequently on large projects the electrical designer will specify that the CPC must be the same size as the phase conductor (particularly in ATEX zoned areas) but this does not mean it is a regulation.

On a well installed counduit and trunking run, it will make little if any difference.

I would think the thickness of the steel conduit wall would be almost irrelevant compared to the effect of how well the conduits and trunking is installed. 1mm thick walled 20mm steel conduit has a CSA of about 60mm square, and its likely to be a little thicker than that, so that really is not much of a factor compared to the likely size of the earth conductor in such a conduit.

I think myself the steel conduit should be discounted anyway when deciding on the circuit size. Thats just my opinion though. As said here, a few corroded or loose couplers on the conduit, or loose bushings into trunking would not be unheard of, especially when flanged couplers are not used for the conduit into the trunking.

2011

robbie7730 wrote: »

On a well installed counduit and trunking run, it will make little if any difference.

Agreed. But well installed now may be in a very different condition in a year or two.
Also, conduit is not always well installed.

I would think the thickness of the steel conduit wall would be almost irrelevant compared to the effect of how well the conduits and trunking is installed.

It all matters. The thickness of the conduit is comparable to the CSA of a cable.

1mm thick walled 20mm steel conduit has a CSA of about 60mm square

60 sq. ??? Do you mean 60 sq. copper?? What size conduit?

I think myself the steel conduit should be discounted anyway when deciding on the circuit size. Thats just my opinion though.

In general it is. I agree with you.

As said here, a few corroded or loose couplers on the conduit, or loose bushings into trunking would not be unheard of, especially when flanged couplers are not used for the conduit into the trunking.

+ 1

Bruthal

2011 wrote: »

Agreed. But well installed now may be in a very different condition in a year or two.
Also, conduit is not always well installed.

True alright, that is why i said it should be discounted. I think the day where the conduit could actually be the earth is long gone now, or should be.

It all matters. The thickness of the conduit is comparable to the CSA of a cable.

Not really. A 20mm outer diameter hollow pipe with a wall thickness of 1mm will have a conducting CSA of 60 square mm, i.e. ten times what a 6 square copper wire has. So a 20mm steel conduit is the same. So it is a far bigger conductor than any earth thats likely to be in it, thats why i would think a slight difference in thickness would play very little part in the circuit earth path.

pipe CSA = 0.785x(outer diameter(20)^2-inside diameter(18)^2)
0.785(400-324) = 59.66

And for the same reason, once the conduit is in parallel with the earth, the earth size being changed will make little difference either, once the conduit/trunking path is sound, which as you say, is not highly reliable.

60 sq. ??? Do you mean 60 sq. copper?? What size conduit?

The conduit is 20mm outer diameter, and i am giving it as 18mm inside, but its usually about 17mm. 20mm outside minus the 18mm inside diameter gives a CSA of 60mm sq to conduct any current. So its quite a big conductor, so a slight thickness variation, while it will change the conduit resistance, will be insignificant in the scheme of things with a 2.5, 4, or 6sq earth in parallel with it.

2011

A 20mm outer diameter hollow pipe with a wall thickness of 1mm will have a conducting CSA of 60 square mm, i.e. ten times what a 6 square copper wire has

Apples and oranges. You have to compare like with like:

The resistivity of copper is very different to that of steel. That is why it is generally not used as a conductor.

ρ copper = 1.68×10^−8 [Ω/m]

ρ steel = 10 to 100 ×10^−8 [Ω/m]

Bruthal

2011 wrote: »

Apples and oranges. You have to compare like with like:

Your missing my point though. We are not comparing the conducting properties of the two.

Apples and oranges have no bearing, as i was talking about how the thickness of the wall of the conduit affects the parallel path of the conduit in parallel with the CPC. Varying the thickness of a conduit will only very marginally affect its resistance, and as its now in parallel with a CPC, the change is even smaller in the circuit.

We are not talking about how good a conductor the conduit is compared to copper, and that wont affect the percentage change in its own resistance for a thickness change.

Of course its thickness has to affect it, but im suggesting a slight variation in conduit thickness is insignificant in the overall scheme of things. After all, the proper CPC will be adequate on its own, so a conduit lowers this impedence of course, but a slight variation in thickness of conduit will have no major effect after that. Also, on just about any circuit, the conduit run will be much shorter than the circuit run, so its actual wall thickness again has little bearing.

My overall point was, the thickness will have much less bearing than the quality of fitting of the conduit, tight non corroded couplers etc.

The resistivity of copper is very different to that of steel. That is why it is generally not used as a conductor.

ρ copper = 1.68×10^−8 [Ω/m]

ρ steel = 10 to 100 ×10^−8 [Ω/m]

If we take a CPC of 2.5 Sq on a 50 meter circuit, its resistance is about 0.37 ohms, and 2 lengths of steel conduit coming from the trunking, and their impedence is also 0.37 ohms, even though it likely will be far lower as they only form a fraction of the circuit run, and are 24 times the size of the CPC, then the earth path is now about 0.185 ohms.

If the conduit thickness is increased by 10 percent, then the conduit impedence will now be 0.33 ohms, so the parallel path will now be 0.17 ohms, not a major difference(0.015). It wont matter what resistance you give for the conduit above, the variation in thickness will still have the same effect on its own % resistance, nothing to do with the fact its not a copper conductor. And in fact, the higher the resistance of the conduit, the less its variation will affect the parallel resistance with the CPC.

Its like saying put a 120 sq in parallel with a 2.5 CPC, then try a 50 Sq. For the overall 20 amp circuit, the difference between the 2 will be negligible. But either will be a lower reading than the 2.5 alone.

Dont forget, i am not saying the use of steel conduit will have little effect on the path, im suggesting a variation in thickness will have little bearing overall. Thats why the Ω of steel comared to the Ω of copper is not in question here. Any electrician would realise steel is not the conductor copper is.

Overall im not disagreeing with you, im just saying the variation in conduit thickness is of little bearing, and the major factors are its proper installation. But since the CPC will be adequate on its own, or should be, the conduit path variation will be of less importance anyway. Not that its an excuse for a sloppy installation though.

On a well installed installation, the conduit path back will not be exclusively the cinduit and trunking, other conduits from the trunking will be earthed at socket boxes, light switches etc, as well as click fittings on the trunking possibly, so the conduit path back does tend to be a low resistance path, once the conduits are well connected to the trunking, and the trunking joins are done properly.

2011

Overall I am agree, except for this statement:

A 20mm outer diameter hollow pipe with a wall thickness of 1mm will have a conducting CSA of 60 square mm, i.e. ten times what a 6 square copper wire has

This is because a steel conductor will have a far higher resistance than a copper conductor of identical CSA and length. Your statement above suggests that this is not the case.

Bruthal

2011 wrote: »

Overall I am agreeing except for this statement:

A 20mm outer diameter hollow pipe with a wall thickness of 1mm will have a conducting CSA of 60 square mm, i.e. ten times what a 6 square copper wire has

This is because a steel conductor will have a far higher resistance than a copper conductor of identical CSA and length. Your statement above suggests that this is not the case.

My statement does not say this, are you seriously suggesting i believe steel conducts like copper? Come on, give me some credit now:D, i think i was 10 when i realised that one, it was a long time before the apprentice days anyway.

But that quote merely shows the CSA is not small like you suggest it is here, and my quote was a direct reply to this. Maybe i read what you meant wrong, not uncommon for me to do that, and getting worse:eek:.

The thickness of the conduit is comparable to the CSA of a cable

I mentioned 6 Sq simply because it was unlikely to have anything bigger in a 20mm conduit. And also ye wouldnt think a conduit has that CSA. And so the conduit was likely to be roughly comparable in resistance to any CPC in it, as it was 10 times bigger than a 6 Sq.

Absolutely nothing to do with copper v steel as conductors. Its actually hard to believe a conduit would have that CSA, and in actual fact, thats very conservative. The inner diameter is usually about 17mm (i just measured now), so CSA is almost 90 mm Sq. Nearly a 95 Sq. So i think that would compare to the resistance of a 6 square copper earth probably.

But even if it does not, it still changes nothing in the fact a slight change in the conduits thickness will account for little in the overall earth path, especially as the steel is a far poorer conductor per Sq mm than the copper its in parallel with.

And the reason i mention the CSA of the conduit? To show that a slight variation in its thickness wont have a major affect on the earth path which has an independently sized CPC in parallel with it. How well steel conducts does not matter to this fact.

A earth loop impedence test can be done at a socket with its CPC disconnected, and the socket earth just connected to the steel box, then do it with the socket wiring just in connectors so the CPC is isolated from the conduit.

Can also measure the earth - live conductor loop resistance bact to the DB with power off, then conduit - live conductor loop, and compare. We did similar to these before, but i cant remember much about the results.

2011

To me the statement below appeared to suggest that a steel conductor CSA 60 would be ten times better than a 6 sq copper cable.

......conducting CSA of 60 square mm, i.e. ten times what a 6 square copper wire has

But if you say I misread that, then fair enough. But I think you will appreciate that it is a little open to interpretation.

Bruthal

2011 wrote: »

To me the statement below appeared to suggest that a steel conductor CSA 60 would be ten times better than a 6 sq copper cable.

conducting CSA of 60 square mm, i.e. ten times what a 6 square copper wire has

But if you say I misread that, then fair enough. But I think you will appreciate that it is a little open to interpretation.

yes but a different interpetation is helped by a shortened version of my sentence, creating a different context version of my sentence above. Here is the full one.

A 20mm outer diameter hollow pipe with a wall thickness of 1mm will have a conducting CSA of 60 square mm, i.e. ten times what a 6 square copper wire has

It says a 20mm pipe has a cross sectional area of 60mm Sq, ten times what a 6mm sq wire has. Now what would you say its comparing in that sentence in fairness? The word has right at the end of the sentence would indicate whats being compared there. Especially as it was a direct reply to you saying the pipe has the same thickness as a wires CSA. No reference is made to how good a conductor anything is, Just what CSA they have.

Im not the best at writing on boards for sure, but after all the posts we had in this forum, the fact you would think i believe steel is the conductor that copper is, is surprising. I dont think any sparks i seen posting in it would think that. Its electrical common sense really.

2011

I speed read most posts (especially the long ones) so sometimes I miss stuff, I'm not perfect.

To me it looked like a comparison, no big deal.
So, as I said:

But if you say I misread that, then fair enough

Bruthal

2011 wrote: »

I speed read most posts (especially the long ones) so sometimes I miss stuff, I'm not perfect.

To me it looked like a comparison, no big deal.
So, as I said:

I speed write them, so thats a good combination with a speed reader coming along afterwards:D:D