Main earth bonding

[Cabbyfan]

Hi,

Hopefully a fairly straight forward question ...

Fitted new CU, system is TN-C-S. it's a 17th dual rcd board split load etc.

10mm main bond to gas is easy but my water main is plastic and at rear of house. It's not difficult to run in as all the floors and carpet is up. However I'm pretty sure the water main pipe is plastic so can't bond that but to me it makes sense to bond the water pipe on the copper side.

Also, 4mm to heating pipes ?

Thanks.

 

[JohnW2]

Hopefully a fairly straight forward question ...

Indeed - and. frankly, one which I wouldn't have expected someone who felt competent to change a CU to have to ask.

10mm main bond to gas is easy but my water main is plastic and at rear of house. It's not difficult to run in as all the floors and carpet is up. However I'm pretty sure the water main pipe is plastic so can't bond that but to me it makes sense to bond the water pipe on the copper side.

If the water service enters the house in plastic, none of the water pipework needs main bonding.

Also, 4mm to heating pipes ?

Only metal pipes which enter the building from outside (and not necessarily even all of them) require main bonding.

Kind Regards, John

 

[PrenticeBoyofDerry]

Hi,
Hopefully a fairly straight forward question ...

Fitted new CU, system is TN-C-S. it's a 17th dual rcd board split load etc.

10mm main bond to gas is easy but my water main is plastic and at rear of house. It's not difficult to run in as all the floors and carpet is up. However I'm pretty sure the water main pipe is plastic so can't bond that but to me it makes sense to bond the water pipe on the copper side.

Also, 4mm to heating pipes ?

I think this question deserves a straight forward answer

Find a competent electrician to prove what you have done thus far is safe!

 

[bernardgreen]

Are plastic water pipes incapable of introducing an extraneous potential ?

Water is said to be non conductive when inside a water pipe but is considered to be conductive in a bath room.

I suppose the water in the supply pipe is pure and that in a bathroom could be full of dissolved soap and other chemicals which could increase it's conductivity.

 

[davelx]

Are plastic water pipes incapable of introducing an extraneous potential ?.

Yes. As it's not a conductor it cannot introduce a potential!

Water is said to be non conductive when inside a water pipe but is considered to be conductive in a bath room.

I suppose the water in the supply pipe is pure and that in a bathroom could be full of dissolved soap and other chemicals which could increase it's conductivity.

Even if the water is conductive, and therefore capable of introducing a potential, bonding the plastic pipe wouldn't stop that, would it? You'd have to bond the water itself - so you'd need an electrode inside the plastic pipe, that could be bonded to the MET

 

[bernardgreen]

You'd have to bond the water itself - so you'd need an electrode inside the plastic pipe, that could be bonded to the MET

Which is probably best achieved by a short length of copper pipe ( or the metallic stop cock ) in the water supply line being bonded to the MET.

 

[Cabbyfan]

Thanks guys.

Its one of those odd questions, the regs clearly state you need a min of 10mm main bonding to gas / water etc but doesn't clarify if its a plastic pipe. When I was an apprentice spark I stopped once I did my 15th Edition and there was a time when if it was metal you bunged an earth on it.

Personally I would much prefer having my hot/cold water pipework bonded but that's just me.

 

[SimonH2]

Indeed - and. frankly, one which I wouldn't have expected someone who felt competent to change a CU to have to ask.

I've seen plenty of situations where someone is competent to do a job, but feels the need for clarification when faced with conflicting advice

If the water service enters the house in plastic, none of the water pipework needs main bonding.

It doesn't seem to be that clear cut, even our very own Wiki says otherwise :
//www.diynot.com/wiki/Electrics:main_equipotential_bonding

If the water or gas supplies are plastic but the installation pipework is metal then main protective bonding conductors are required to connect the metal pipework to the MET. If both the supply and the installation pipework are plastic then no main protective bond is required.

That would seem to me to fit in with the purpose of the MEB - to ensure that all exposed metalwork is bonded to earth so as to ensure it can't become live (by creating a low impedance circuit to earth and ensure protective devices feeding the faulty circuit(s) operate in a timely manner.

Taking a practical example. Gas and water both enter in plastic, so the argument goes that they don't need bonding. The only earth for the water and heating systems is via the boiler mains connection. A fault develops between the live of (say) the 32A circuit feeding the cooker or an RFC and a water or heating pipe. The only route to earth is therefore via the boiler mains supply terminal block, a length of possibly only 0.75mm^2 flex, and then whatever circuit the boiler feeds off back to the CU. The impedance of this circuit could quite feasibly be sufficient to allow a "significant' fault current to flow, raising the water & heating pipework to a dangerous potential, and also creating a risk of fire as the thin flex burns out.
As I read the graphs, a B curve breaker needs at least 5x it's rated current to guarantee to trip in 0.4s - that's 160A for a 32A breaker. That means (if my mental arithmetic is correct) a total impedance for the fault, supply, earth, and everything else of no more than 1.5ohms

 

[davelx]

You'd have to bond the water itself - so you'd need an electrode inside the plastic pipe, that could be bonded to the MET

Which is probably best achieved by a short length of copper pipe ( or the metallic stop cock ) in the water supply line being bonded to the MET.

Which would probably accumulate an insulating layer of lime-scale over time and become ineffective. As would an electrode unless it was of a material on which lime-scale did not deposit.

A glass of cold water from the tank in my loft measures 2KOhms at 500V.
A glass fresh from the mains measures 3KOhms.

Hardly non-conducting!

 

[JohnW2]

Indeed - and. frankly, one which I wouldn't have expected someone who felt competent to change a CU to have to ask.

I've seen plenty of situations where someone is competent to do a job, but feels the need for clarification when faced with conflicting advice

Of course. I was merely using much more gentle language in anticipation of what was inevitably going to follow my post

If the water service enters the house in plastic, none of the water pipework needs main bonding.

It doesn't seem to be that clear cut, even our very own Wiki says otherwise : //www.diynot.com/wiki/Electrics:main_equipotential_bonding

Our wiki is wrong. The only things that require main bonding are extraneous-conductive-parts. Only metal pipes which enter the equipotential zone (and not necessarily even all of them) qualify as that, and require main bonding. If one bonds internal metal pipework when the service enters the premises in plastic, that is not main equipotential bonding, since it cannot fulfil the purpose of MEB.

If the water or gas supplies are plastic but the installation pipework is metal then main protective bonding conductors are required to connect the metal pipework to the MET. If both the supply and the installation pipework are plastic then no main protective bond is required.

That would seem to me to fit in with the purpose of the MEB - to ensure that all exposed metalwork is bonded to earth so as to ensure it can't become live (by creating a low impedance circuit to earth and ensure protective devices feeding the faulty circuit(s) operate in a timely manner.

You are talking about EARTHING, not bonding of any sort, and certainly not MEB. If one decides to bond otherwise 'floating' internal metal pipework, one has to balance the opposing risks - the one you mention (should the pipework somehow become live) against the increased risk of electric shock which arises by 'unnecessarily' increasing the amount of earthed metal around the place.

In practice, of course, in an all-metal-pipework house, the pipework will almost inevitably already be earthed (via CPCs of boilers, immersions, CH pumps/valves etc.), so that 'unnecessary' additional earthing of it makes no difference.

As for the example you go on to give, it's nothing to do with bonding, let alone MEB. The whole point of the design of the various circuits in the house (and the adequacy of their live conductors, CPCs and EFLI) is to ensure that protective devices will operate rapidly (before damage to any cables/CPCs occurs) whenever the potential of exposed conductive parts (which, in effect, includes metal pipework) rises to potentially dangerous levels due to a fault.

Kind Regards, John

 

[EFLImpudence]

Its one of those odd questions, the regs clearly state you need a min of 10mm main bonding to gas / water etc

Only IF an extraneous-conductive-part , a point nearly always overlooked.

but doesn't clarify if its a plastic pipe.

What do you mean by clarify?
Again, IF an extraneous-conductive-part.
A material which does not conduct electricity cannot be an e-c-p.

When I was an apprentice spark I stopped once I did my 15th Edition and there was a time when if it was metal you bunged an earth on it.

Yes - stupid in the extreme.
Not sure if you meant to but you are correct to say earth; this would not be bonding.
Never assume those in charge know what they are doing.
(Present new regulations suggest the situation is no better now)

Personally I would much prefer having my hot/cold water pipework bonded but that's just me.

Let us hope you do not get a shock from something else while you are touching one of these pipes.
Plus if these pipes did not require bonding (not an e-c-p) then you would not be bonding them, it would be earthing - much the same as a window frame or the proverbial spoon.

 

[EFLImpudence]

Further:

People seem to think that earthing is a good thing in its own right.

That is not the case. It is necessary to ensure disconnection of the supply should a fault occur.
It is connected to parts which may become live because of the fault and may be touched.
It would be better to ensure that they did not become live and so remove the need for earthing.

Therefore to apply earthing to parts which cannot become live because they are not part of an electrical appliance or accessory is not beneficial.
Some seem to think that live conductors can suddenly jump out and attack random pipes - we do not have to cater for this.

 

[EFLImpudence]

A glass of cold water from the tank in my loft measures 2KOhms at 500V.
A glass fresh from the mains measures 3KOhms.

Hardly non-conducting!

How wide is your glass?

What would be the figure for a metre of water? 25 x 2kΩ ?

 

[JohnW2]

People seem to think that earthing is a good thing in its own right. ... That is not the case. It is necessary to ensure disconnection of the supply should a fault occur. ... It is connected to parts which may become live because of the fault and may be touched.
It would be better to ensure that they did not become live and so remove the need for earthing. ... Therefore to apply earthing to parts which cannot become live because they are not part of an electrical appliance or accessory is not beneficial. ... Some seem to think that live conductors can suddenly jump out and attack random pipes - we do not have to cater for this.

All very true - as I just wrote, if one is going to go around earthing things which do not necessarily need to be earthed, one has to 'risk-assess' the relative risks of earthing and not earthing the parts.

Of course, as far as metal pipework is concerned, it almost inevitably will be earthed by virtue of CPCs (of immersions, boilers or CH components etc.), so additional ('unnecessary') earthing will usually not make things any worse.

Kind Regards, John

 

[SimonH2]

As for the example you go on to give, it's nothing to do with bonding, let alone MEB. The whole point of the design of the various circuits in the house (and the adequacy of their live conductors, CPCs and EFLI) is to ensure that protective devices will operate rapidly (before damage to any cables/CPCs occurs) whenever the potential of exposed conductive parts (which, in effect, includes metal pipework) rises to potentially dangerous levels due to a fault.

So, taking the example I gave - and leave out any bonding/earthing of the pipework.
Which circuit is not designed properly ?
The RFC is designed properly - if there's a L-E fault in the circuit then the breaker will trip.
The boiler supply is designed properly - if there's a L-E fault in it or the boiler then the breaker will trip and/or the fuse in the FCU/plug will blow.
Connect the live in the larger circuit to the CPC in the other and it's not guaranteed.

Therefore to apply earthing to parts which cannot become live because they are not part of an electrical appliance or accessory is not beneficial.
Some seem to think that live conductors can suddenly jump out and attack random pipes - we do not have to cater for this.

Now we take a hypothetical installation where the sparky has been round and done his bit. And the sparky determined that the water pipes didn't need any earth or bonding connection. What's more, none of his cables are on/close to any pipes.
Then the plumber comes round and puts the plumbing in, and like many plumbers, doesn't seem to care about cables. So somewhere there's a cable sat on a hot pipe. Over time, with the heat and movement, the insulation slowly wears through until fate decided that the live conductor which happened to be on (the wrong side) contacts the pipe.

Because someone didn't cater for it, we now have live pipework with only the boiler wiring to "earth" it - hence an inadequate fault current to make the breaker trip in a timely manner.
But it's alright if anyone dies in the ensuing fire, because it wasn't the sparkies responsibility to account for that - or that's how I read your statement.

Of course, these days the circuit will almost certainly have RCD protection. That's not the case in older properties that haven't yet been upgraded.

But I believe the regs disagree with you. Bonding of conductive parts in a bathroom IS required unless certain conditions are met. That sounds very very much to me like those who wrote the regs consider that the electrician MUST cater for the potential of a live conductor randomly attacking a pipe.