Earth wire on gas main pipe - advice on regs sought

[holmslaw]

..

 

[Wontdothatagain]

Sorry it has taken me so long to reply to this but i have been working away.

I have spoken to the Gas Safe Inspector, we were having a meeting on a large build along with the site manager and a senior installation engineer. I printed this thread off and showed him and this was his reply,

There is no Gas Safe regulation that states the Bonding cannot be attached to the gas pipe work. It is quite common on large builds to have the bonding cable tied to the pipework for several meters until it can go onto tray or basket and back to the MET. The whole purpose of the cable is to bring the installation to the same potential as all other conductive and extraneous conductive parts of the installation.

As for Benards question, there are lots of complaints from meter readers about getting "shocks" from gas meters but this is mainly found out to be from un-bonded pipes and he went on to tell me about a very large domestic property where the PME electric service entered the property from the West side from high ground and the gas from the East side on low ground. The pipe work was not bonded and the customer complained about getting shocks. They got the pipes bonded and all was fine.

 

[bernardgreen]

The pipe work was not bonded and the customer complained about getting shocks. They got the pipes bonded and all was fine.

I am a bit sceptical and think there was more involved in removing the cause of the shocks from the gas meter. The "earth" from the PME supply would not be at true ground and even without the bonding the metallic gas pipe would almost certainly be in contact with electrical "earth" via the boiler and its electrical earth. I suspect the shocks at the gas meter discovered a fault with the quality of the PME provided "earth" and that was rectified at the same time as the bonding was added.

 

[Wontdothatagain]

The way he explained it was that because there was a potential difference between the 2 supplys and with no bonding to the gas work close to where it enters the property that was the cause of the shock.

 

[derrickbowman]

I have just been having this debate about strapping the earth to the gas pipe and was given the following reason for not doing it. Reg 543.3.1 says something about the earth protection/bonding wire not being mechanically supported by the service it is protecting. This reg is to ensure that a plumber/gas man in re-routing the gas (or water) service pipe does not cut the wire attached to it. By having an independent route for the earth wire from source (<600 mm inside boundary) to the main earth terminal one disassociates the wire from the pipe; so any pipe work changes should not cause a change to the continuity of the earth wire. This regulation is about protecting the earth wire (not protecting the gas pipe). Of course there is nothing to stop a builder chopping and independently routed earth wire but that is another issue.

However from a gas pipe protection point of view other regulations say that all wires that carry a current should be not less the 25 mm apart from a <=28 mm gas pipe (35 mm for larger pipe) without a second insulating barrier between the pipe and wire. This is to prevent any risk of a spark or shock hazard between wire and pipe.

But one might say that there will be no current flow in the earth wire as it is essentially earth potential running alongside the pipe making it safer then a separated wire would be. But the logic behind a wire that is making extraneous points at equal potential is that they do it by pulling any voltage difference down to the same value i.e. if a water pipe earth, gas pipe entry and supply earth were at different potentials then they are shorted together by the bonding wire to make the same potential. This implies that if there was an earth potential difference then some current must flow in the bonding wires to equalise that potential between all extraneous earth points and the MET.

Normally this is unlikely but I recently received a report of an instance of quite high earth bond wire current. It was explained thus:

The supply was TN-C-S so the MET was a neutral potential. The sub station was a pole mounted transformer and that is where the source earth existed and was connected to the neutral there. The feed ran across farmland and fed a farm complex and a hamlet of houses. When demand was high the neutral lines resistance could let the neutral voltage rise quite few volts relative to earth. So the MET would show this voltage rise. But the house water pipe (being lead) had a very good earth back to the substation so current would flow from the MET to the water pipe.

This meant that current flowed along the bond wire which was a single continuous wire from the MET along the gas pipe and then daisy chained onto the water pipe in the same cupboard as the gas pipe.

The electrician said he put a clamp meter on the bond wire and was quite surprised how high the current was. Certainly a disconnection would cause a spark at the terminal (and let a reasonably high voltage difference to occur at the earth points).

Interesting I thought. Maybe further justification for keeping the earth wire wire away from the gas pipe and a reason to have separate wires for MET to each extraneous entry point.

 

[Risteard]

For Electric cables Separation of at least 25mm to be provided for domestic gas pipework up to 35mm....For pipework over 35mm then 50mm of separation is needed.

OOI, does the same apply if you use DI cable, or put it in conduit?

No such thing as double insulated cable.

 

[Risteard]

Apologies. Have just noticed the antique nature of this revived thread...

 

[EFLImpudence]

I have just been having this debate about strapping the earth to the gas pipe

It's not an earth; it is a bond because the pipe is earthed.

and was given the following reason for not doing it. Reg 543.3.1 says something about the earth protection/bonding wire not being mechanically supported by the service it is protecting.

543.3.1 does not say that. Did you quote the correct number?

This reg is to ensure that a plumber/gas man in re-routing the gas (or water) service pipe does not cut the wire attached to it.

We cannot legislate for idiots.

By having an independent route for the earth wire from source (<600 mm inside boundary) to the main earth terminal one disassociates the wire from the pipe;

Why would that be desirable?

so any pipe work changes should not cause a change to the continuity of the earth wire.

Unless at point of connection.

This regulation is about protecting the earth wire (not protecting the gas pipe). Of course there is nothing to stop a builder chopping and independently routed earth wire but that is another issue.

Quite.

However from a gas pipe protection point of view other regulations say that all wires that carry a current should be not less the 25 mm apart from a <=28 mm gas pipe (35 mm for larger pipe) without a second insulating barrier between the pipe and wire. This is to prevent any risk of a spark or shock hazard between wire and pipe.

Except that this wire is connected to the pipe.

But one might say that there will be no current flow in the earth wire as it is essentially earth potential running alongside the pipe making it safer then a separated wire would be.

There can in a fault be current in the bond (don't keep calling it earth).

But the logic behind a wire that is making extraneous points at equal potential is that they do it by pulling any voltage difference down to the same value i.e. if a water pipe earth, gas pipe entry and supply earth were at different potentials then they are shorted together by the bonding wire to make the same potential.

Correct.

This implies that if there was an earth potential difference then some current must flow in the bonding wires to equalise that potential between all extraneous earth points and the MET.

No, there would be potential difference and current could flow through you if you touched two of them.
That is why they are bonded to remove the difference.

Normally this is unlikely but I recently received a report of an instance of quite high earth bond wire current.

Then there is a fault somewhere.

Interesting I thought. Maybe further justification for keeping the earth wire wire away from the gas pipe and a reason to have separate wires for MET to each extraneous entry point.

But they are all connected together for that very purpose.
The wire is insulated.

I don't know if the gas industry has a regulation for some reason.

You seem to worry a lot about sparks; do you think the gas in the pipe is going to explode?

 

[derrickbowman]

EFL Impudence

thanks for your thoughts. I could not reply with your words quoted (funnies with my browser) so I put your words between [ ] brackets.


[It's not an earth; it is a bond because the pipe is earthed.]

Okay, fair point, bond or protective wire then. (The green and yellow 10 mm wire used for bonding together extraneous earth points to the MET to clamp all extraneous earth points and the MET to the same potential to protect the system against potential differences). Whatever the preferred term used that is the purpose I mean.


[543.3.1 does not say that. Did you quote the correct number?]

Okay I was just doing a cut and paste from the report so it is their words not necessarily the regs literatum. The words were:

'Main bonding conductors should not be supported by the service pipes they are connected to,Reg 543.3.1'.

That might be a Guidance Notes 8 or similar interpretation from a gas board document. But it is about protecting the 'protective' wire.

[We cannot legislate for idiots.]

[Why would that be desirable?]

I agree but it is reasonable to assume that if a protective wire is detached from a pipe then anyone working on the pipe won't affect the wire. So it could be a good bit of advice to give if experience has shown that some tradesmen, when working on gas or water pipes, may break the wire and not rejoin it properly or replace it. Also regulations being revised and/or modified by authorities for building regs, gas regs etc. need to be done on the basis that there is little or no dependency on another service. By advising that the pipe and wire take separate routes using disciplines peculiar to their nature then there will be reduced risk of one affecting the other at rework times.

[Unless at point of connection.]

True. But as there has eventually to be a point of termination it is reasonable to put that as close to the service entry to the building to minimise any affect during pipe run rework. Also at <600 mm it is more reasonable to assume that only a fully qualified a professional worker will be doing the work and that they will understand the need for the protective wire. Plus inspectors are more likely to see a missing termination but not notice a break in the run or the protective wire.


[Except that this wire is connected to the pipe.]

I agree, this is just one of those interpretations of the reg 528.3.3 & .4 - 'proximity to non-electrical services'. The On Site Guide says that distribution cables (i.e. cables carrying a working current) should be separated by the quoted distances. If the protective bond wire is carrying current should this also be separated?


[There can in a fault be current in the bond (don't keep calling it earth).]

Is that only the case for a fault condition? If the logic of the bond wire is that it is making extraneous earth points be at equal potential by bonding (or shorting) them together with a low resistance wire then they must do this by pulling any voltage difference down to the same value. So if there were differences in potential there must be current. If there are no differences ever then there is no need for a protective bond wire.

The question is can we have a valid reason for a potential difference without there being a fault condition?


[No, there would be potential difference and current could flow through you if you touched two of them. That is why they are bonded to remove the difference.]

So ipso facto current must flow through the protective bonding wire which is what the engineer reported to me.

[Then there is a fault somewhere.]

But is it a fault. It is reasonable for the neutral line voltage to rise up from ground earth under load (and the live line to drop a bit). So if the neutral was the house MET (as in TN-C-S) then one could have a neutral and supply earth with a higher potential then a ground earth (e.g. water pipe). This is not a fault condition provided the earth loop resistance and supply voltage drop are within limits.


[You seem to worry a lot about sparks; do you think the gas in the pipe is going to explode?]

I am trying to compare the thoughts that distribution wires are kept away from a gas pipe but a protective bond wire carrying current legitimately need not be. Certainly if something severed a gas pipe and, at the same time, any clipped protective bond wire that carried a current due to earth potential difference then gas could be released at the same place as the severed wire may spark.

That was my thinking as to why it may be advisable to run the protective bond wire elsewhere to the gas pipe run.

 

[EFLImpudence]

[543.3.1 does not say that. Did you quote the correct number?]
Okay I was just doing a cut and paste from the report so it is their words not necessarily the regs literatum. The words were:
'Main bonding conductors should not be supported by the service pipes they are connected to,Reg 543.3.1'.
That might be a Guidance Notes 8 or similar interpretation from a gas board document. But it is about protecting the 'protective' wire.

Ah, 5.2.3 of GN8 does say that and refer to 543.3.1 which does not.

I agree but it is reasonable to assume that if a protective wire is detached from a pipe then anyone working on the pipe won't affect the wire. So it could be a good bit of advice...

Well possibly but what about where the wire did run?

[Unless at point of connection.]
True. But as there has eventually to be a point of termination it is reasonable to put that as close to the service entry to the building to minimise any affect during pipe run rework.

We do not put it there to avoid pipe rework .

Also at <600 mm

The 600mm. only applies to indoor meters.
The rule otherwise is 'at point of entry where practicable'.

it is more reasonable to assume that only a fully qualified a professional worker will be doing the work and that they will understand the need for the protective wire. Plus inspectors are more likely to see a missing termination but not notice a break in the run or the protective wire.

You have more faith than I.

[Except that this wire is connected to the pipe.]
I agree, this is just one of those interpretations of the reg 528.3.3 & .4 - 'proximity to non-electrical services'. The On Site Guide says that distribution cables (i.e. cables carrying a working current) should be separated by the quoted distances.

Mmmm. Yes - to avoid damage but people can avoid damaging them and, as above, what about avoiding damage where they are subsequently placed?

If the protective bond wire is carrying current should this also be separated?

I presume these are your words and not a continuation of the above.
The Main Protective Bonding will only be carrying current in fault conditions (apart from leakage).

[There can in a fault be current in the bond (don't keep calling it earth).]
Is that only the case for a fault condition? If the logic of the bond wire is that it is making extraneous earth points be at equal potential by bonding (or shorting) them together with a low resistance wire then they must do this by pulling any voltage difference down to the same value. So if there were differences in potential there must be current.

There can be no potential difference on a negligible impedance connection.

If there are no differences ever then there is no need for a protective bond wire.

There could be without the bonding. That is why it is applied - to remove it.

The question is can we have a valid reason for a potential difference without there being a fault condition?

As above, apart from very small earth leakage current - no.

[No, there would be potential difference and current could flow through you if you touched two of them. That is why they are bonded to remove the difference.]
So ipso facto current must flow through the protective bonding wire which is what the engineer reported to me.

No, it (the potential difference) has been removed

[Then there is a fault somewhere.]
But is it a fault. It is reasonable for the neutral line voltage to rise up from ground earth under load (and the live line to drop a bit). So if the neutral was the house MET (as in TN-C-S) then one could have a neutral and supply earth with a higher potential then a ground earth (e.g. water pipe). This is not a fault condition provided the earth loop resistance and supply voltage drop are within limits.

Not if bonded. That is the point.

[You seem to worry a lot about sparks; do you think the gas in the pipe is going to explode?]
I am trying to compare the thoughts that distribution wires are kept away from a gas pipe but a protective bond wire carrying current legitimately need not be. Certainly if something severed a gas pipe and, at the same time, any clipped protective bond wire that carried a current due to earth potential difference then gas could be released at the same place as the severed wire may spark.

I don't think that is the reasoning behind the spacing.
I may be wrong.

That was my thinking as to why it may be advisable to run the protective bond wire elsewhere to the gas pipe run.

The advised spacing with live wires is so little anyway I think the bonding wire is unimportant in this respect.

 

[bernardgreen]

[No, there would be potential difference and current could flow through you if you touched two of them. That is why they are bonded to remove the difference.]
So ipso facto current must flow through the protective bonding wire which is what the engineer reported to me.

No, it (the potential difference) has been removed

If there was potential between pipe and MET before the bond was added then there will still be a potential after the bond is added but it will be reduced to a very low, near zero difference. There will be current flowing in the bond. The potential and current will depend on the impedance between the pipe and the source of the extraneous potential and it ratio to the impedance of the bond cable.

In most cases the impedance of the extraneous source is several tens of ohms or more and the bond is less than a ohm so the potential is to all practical purposes zero,

But were the extraneous source is very low impedance, as in a metallic water main, the ratio allows for a significant voltage and hence a significant current.

In cases where the MET voltage rises due to an elevated neutral potential and the bond is to a water pipe connected to a metallic mains then the current in the bond can be several amps.

 

[derrickbowman]

[No, there would be potential difference and current could flow through you if you touched two of them. That is why they are bonded to remove the difference.]
So ipso facto current must flow through the protective bonding wire which is what the engineer reported to me.

No, it (the potential difference) has been removed

If there was potential between pipe and MET before the bond was added then there will still be a potential after the bond is added but it will be reduced to a very low, near zero difference. There will be current flowing in the bond. The potential and current will depend on the impedance between the pipe and the source of the extraneous potential and it ratio to the impedance of the bond cable.

In most cases the impedance of the extraneous source is several tens of ohms or more and the bond is less than a ohm so the potential is to all practical purposes zero,

But were the extraneous source is very low impedance, as in a metallic water main, the ratio allows for a significant voltage and hence a significant current.

In cases where the MET voltage rises due to an elevated neutral potential and the bond is to a water pipe connected to a metallic mains then the current in the bond can be several amps.

This is exactly how the report came in to me. The neutral line was of the order of 0.08 ohm to earth at the substation (and presumably the live the same). The water pipe was an earth ground of about 0.48 ohm to the neutral line (presuming 0.4 for ground earth and the 0.08 neutral back to complete the loop). So when the farm and hamlet were drawing say 100 amp on the feed the supply dropped 16 volts (+8 on the neutral and -8 on the live). As the neutral earth and ground earth had a 5 to 1 resistance ratio that 100 amp flowed the same ratio i.e 84ish in the neutral and 16ish in the ground. That meant that equi-potential protective bond wire carried 16 amps and ran adjacent to the gas pipe clipped to it. Any drop between MET and water pipe was trivial as that is the purpose of the 10 mm bond wire, low resistance so effectively a short between MET and water (via gas pipe).

 

[EFLImpudence]

Yes. Thank you Bernard.

I think I was not appreciating the actual problem and concentrating on the safety aspects of the bonding.

Apologies.

 

[securespark]

I don't believe you should strap a cable to pipework. Cable supports should be independent.

Sure there's something in the 14th about that....

As for BG engineers strapping cables to pipes, I can't believe they could do that and get away with it.

I'd be in a disciplinary if I did that.

 

[bernardgreen]

it gets more interesting.

If there is some current flowing in the bond to the gas pipe ( assuming plastic supply to the meter ) then it is flowing along the internal gas pipe to where the gas pipe meets a water pipe. ( probably at the boiler gland plate ).

Assume the gas fitter shuts of the gas supply at the gas meter and then disconnects the gas pipe from the boiler. When the pipe is disconnected the current flowing along the pipe will be interrupted. That could create a spark. There will be gas still in the pipe mixing with air as the pipe is disconnected.

Is it possible that prolonged and heavy current along the gas pipe could result in over heating and possible melting of s soldered joint leading to gas leaks and sparking ?