Earthing Steel Bath

[JohnW2]

I'm happy enough with 10mA - which will be 'safe' for the vast majority of people (give or take consequential injuries due to 'jumps'/falls etc.).

23kΩ is the cut-off point, then.

Indeed - provided one is confident that a measurement not much over 23kΩ, if one ever got it (we agree, very unlikley), would still be over 23k if one measured it a few days, weeks or months later.

You don't need to measure a metal waste in the ground - it should be main bonded.

Is not the whole reason for the measurements you advocate in order to identify a hitherto unexpected/unsuspected low impedance path to earth, whatever the reason for that path?

However, and more to the point, I think that some may read your statement as implying that a measured resistance <23k&#937; means that bonding is required

It is.

- e.g. (per your measurement to MET) that bonding of the bath would be required if, for example, it were plumbed in copper (almost inevitably with a low resistance to MET).

But, if extraneous, copper tube will be bonded so the value will be very low.

Yes, but you appear to be saying, without qualification,that a metal bath which has a low resistance to MET (by implication, for any reason) 'needs to be bonded'- which will make people think of G/Y cables. Are you considering the path through copper pipe (to main bonding, hence MET) to represent that required 'bonding'?

I'm still confused about what you are trying to say. Is it that a bath which is extraneous and bonded by the pipework will read as if it needs bonding? I.e. all bonded parts will give a reading which requires them to be bonded.

See above. What you are saying makes total electrical sense. However, as I said, 'bonding' (of a bath) makes people think of G/Y cables (connected to bath), and I don't think you are really suggesting that such cabling is required just because a metal bath has a low impedance path to MET via copper pipework and main bonding (and/or incidental paths via CPCs etc.), are you?

Kind Regards, John

 

[EFLImpudence]

Is not the whole reason for the measurements you advocate in order to identify a hitherto unexpected/unsuspected low impedance path to earth, whatever the reason for that path?

I suppose so, but in this thread that would be a totally unexpected path - not visible metal pipework as there is none.
The purpose here (in this thread) is to verify the reading as (probably) infinite.

Yes, but you appear to be saying, without qualification,that a metal bath which has a low resistance to MET (by implication, for any reason) 'needs to be bonded'- which will make people think of G/Y cables. Are you considering the path through copper pipe (to main bonding, hence MET) to represent that required 'bonding'?

Of course, but here it is all plastic.

Were there copper pipes then they would be tested and bonded as required.

I'm still mystified by your logic.
You seem to be saying that if the bath has a low impedance because it is connected to bonded pipework does it need to be bonded.
One reason that bonding is not required would be the fact that it is already bonded satisfactorily.

What you are saying makes total electrical sense.

Good.

However, as I said, 'bonding' (of a bath) makes people think of G/Y cables (connected to bath), and I don't think you are really suggesting that such cabling is required just because a metal bath has a low impedance path to MET via copper pipework and main bonding (and/or incidental paths via CPCs etc.), are you?

No, but it would only have a low impedance path because it is (already) bonded.

If the bath in this thread were installed by contractors who bolted it to an extraneous metal structure of the building without the owner realising then after testing it may, indeed, be necessary to connect some green and yellow to it if the readings were low enough.

 

[JohnW2]

Is not the whole reason for the measurements you advocate in order to identify a hitherto unexpected/unsuspected low impedance path to earth, whatever the reason for that path?

I suppose so, but in this thread that would be a totally unexpected path - not visible metal pipework as there is none.

Indeed, but you did write, in this thread:

Strictly speaking the only way to tell definitively is to do an insulation resistance test between the bath and the MET after installation of all equipment.

...so assumed that you must have been thinking of the possibility of 'a totally unexpected path', otherwise your comment would have been redundant.

I'm still mystified by your logic.
You seem to be saying that if the bath has a low impedance because it is connected to bonded pipework does it need to be bonded. One reason that bonding is not required would be the fact that it is already bonded satisfactorily.

Maybe it's just me, but the context was a discussion about (possible) 'bonding' of a bath, and I don't think many people would take 'bonding', in such a context, to include that provided by (already bonded) copper pipework. As I said, to most people,'bonding' implies G/Y-insulated cables, not copper pipes.

Electrically speaking, there really is no difference between our views on this matter. The only issue was by concern that your wording might have confused some people - but maybe they are not as easily confused as I was fearing!

Kind Regards, John

 

[EFLImpudence]

Strictly speaking the only way to tell definitively is to do an insulation resistance test between the bath and the MET after installation of all equipment.

...so assumed that you must have been thinking of the possibility of 'a totally unexpected path', otherwise your comment would have been redundant.

Isn't that the purpose of testing?
If I were in the room and could see all round the bath then perhaps it would be pointless but I am sure you would have been 'after me' had I, from here, given a blanket "NO".

I'm still mystified by your logic.
You seem to be saying that if the bath has a low impedance because it is connected to bonded pipework does it need to be bonded. One reason that bonding is not required would be the fact that it is already bonded satisfactorily.

Maybe it's just me, but the context was a discussion about (possible) 'bonding' of a bath, and I don't think many people would take 'bonding', in such a context, to include that provided by (already bonded) copper pipework.

You have mixed this thread (with all plastic attachments) by introducing copper pipe.

As I said, to most people,'bonding' implies G/Y-insulated cables, not copper pipes.

That may be but it is not so.

Electrically speaking, there really is no difference between our views on this matter. The only issue was my concern that your wording might have confused some people - but maybe they are not as easily confused as I was fearing!

I don't doubt that they are.

What I don't understand is why you think connecting g/y to the bath (or not) would be confusing if that were the only way to do it.

That is probably why the manufacturers fit a 'tag' for the purpose - not because it is mandatory but just in case.

 

[JohnW2]

Strictly speaking the only way to tell definitively is to do an insulation resistance test between the bath and the MET after installation of all equipment.

...so assumed that you must have been thinking of the possibility of 'a totally unexpected path', otherwise your comment would have been redundant.

Isn't that the purpose of testing?

Exactly. I'm not criticising the concept of testing at all. I agree that one should always entertain 'the possibility of the totally unexpected'. Were that not the case, we would, for example, probably omit most 'testing for dead' - since we don't usually 'test for dead' unless we are pretty certain that the circuit is dead!

You have mixed this thread (with all plastic attachments) by introducing copper pipe.

Not really. You implied that 'bonding' was required if there was less than 23k&#937; resistance between the bath and the MET - and I merely pointed out that having the bath plumbed with copper pipework would be one situation in which that resistance would be low.

As I said, to most people,'bonding' implies G/Y-insulated cables, not copper pipes.

That may be but it is not so.

I think that's the crux of our 'talking at cross-purposes'. It is because I believe most people would think that way that makes me feel that they might be confused by your comment about the need for bonding. I'm I wrong in that belief, then I'm talking nonsense.

What I don't understand is why you think connecting g/y to the bath (or not) would be confusing if that were the only way to do it.

I've never suggested that it would be (confusing). I'm suggesting that the bath with a very low resistance path to the MET, because of copper piepwork, does not need a G/Y bonding conductor because of that low resistance path to the MET via the pipework.

Kind Regards, John

 

[mfarrow]

Just to add to what's been said, I completely agree about designer needing to take into consideration changes in the ohmic value of the earthing both on the bath side and at the MET. But...

When you measure the resistance with your IR meter, you are only measuring the direct resistance from the metallic item to the MET. In reality, there will be a human being in between which would allow a slightly lower value of measured resistance.

GN8 gives the value of a dry human with 230V applied across it as being 1000 ohms, which you can either take off the calculated maximum allowed value (giving you 22k) or by adding a 1M resistor in series when you test. Either way, 22k vice 23k isn't a great deal, and therefore sticking to 23k would be a pragmatic approach in ensuring the credibility of the measurement and as a way of justifying any further (in)action.

 

[bernardgreen]

I'm suggesting that the bath with a very low resistance path to the MET, because of copper piepwork, does not need a G/Y bonding conductor because of that low resistance path to the MET via the pipework

It probably needs a bond to ensure the original "earthing" via pipe work is replaced by an earth via bonding if the connecion from bath the MET via pipe work is lost.

It just seems illogical that if the bath by chance has a connection to the MET then it must be bonded but if it has no existing connection to the MET then it must not be bonded. Connection being less than 23 K ohms at the time the electrician measures it. The conductivity may increase when the humidity in the bath room is raised.

Maybe the logic is that pipe work may be altered and the connection to MET via pipe work is lost so bonds are needed. But if an electrician was to be asked bond or not bond after the pipe work was altered he would find no connection between bath and MET and would therefore decide bonding was not needed.

As regards 23 K ohm. That at 230 volts is 10 mA which is enough current to create a hazard to the person whose body is part of the circuit.

To bond the bath or not seems to be dependent on tests made in ideal conditions ( dry room ) when there are no faults in the installation. Surely it is better to make the decision based on the conditions that will occur when the bath room is being used and should also take into account faults that may occur.

Personally if I am in a bath of water I want to be assured that the potential of that water will be the same as the potential on any exposed metal I can touch. The bath taps being the obvious metal that I will be touching.

 

[viewer]

This is a topic on which I've taken advice from the posters above and am happy with that , so please forgive my interest here. If (if) I've read it correctly, the hypothetical scenario has two copper pipes to metal taps on a metal bath. There may also be an electrical shower. The pipes are properly bonded. The question is whether the bath should be bonded to the pipes.
The argument for is that the taps and bath will always be at the same potential and someone in the bath touching the taps would be safe. For this to be done, as generally taps are mounted on plastic inserts there will, when the installation is new be no automatic bonding and the G/Y from tag to pipes will be needed.
The argument against is that by bonding the bath there is a pathway to earth which did not previously exist and therefore produces a risk with the introduction of a potential (faulty appliance, fan heater on extension etc).
The spanner in the works seems to be the plumbing. Copper pipes may be tested by the electrician to ensure conductivity/connections but ten minutes or ten years later the plumber fits a plastic section to repair a leak. The plastic inserts which isolated the tap will fill with water.. and so on.
My suggestion is that the introduction of a length of plastic pipe between any copperwork (supply and drain)and the bath would solve the problem and future-proof the installation. Probably not realistic for many situations, but most questions about bonding and electricity in bathrooms arise during refits or new work and in those cases easy to achieve and may be a suggestion which could be incorporated into how-to-it guides.
I'm ready to be shot down, but would appreciate views on my (?) logic.

 

[EFLImpudence]

GN8 gives the value of a dry human with 230V applied across it as being 1000 ohms, which you can either take off the calculated maximum allowed value (giving you 22k) or by adding a 1M resistor in series when you test. Either way, 22k vice 23k isn't a great deal, and therefore sticking to 23k would be a pragmatic approach in ensuring the credibility of the measurement and as a way of justifying any further (in)action.

Whilst that is true, I cannot see that it is worth bothering about.

It is extremely unlikely that the reading will be within the range of 22 - 23k&#937;.
It will be relatively low or very high.

The figure of 23k&#937; is only used because that is the 10mA value at th nominal voltage of 230V.
A variation in voltage and actual voltage will result in the figure rising to 24.5k&#937; anyway.

 

[EFLImpudence]

This is a topic on which I've taken advice from the posters above and am happy with that , so please forgive my interest here. If (if) I've read it correctly, the hypothetical scenario has two copper pipes to metal taps on a metal bath. There may also be an electrical shower. The pipes are properly bonded. The question is whether the bath should be bonded to the pipes.

No. because the bath, itself, has no potential. It would in a fault scenario if earthed.
Any potential that it may have - because of the connections with the pipes - will, with a perfect join (electrically) be the same as the pipes and with a less than perfect join, less and so less hazardous.

The argument for is that the taps and bath will always be at the same potential and someone in the bath touching the taps would be safe. For this to be done, as generally taps are mounted on plastic inserts there will, when the installation is new be no automatic bonding and the G/Y from tag to pipes will be needed.

Don't forget that the purpose of supplementary bonding is to equalise the potential between (simultaneously accessible) parts which have become live because of a fault.
The bath itself cannot be one of these parts unless it is connected to metal parts which are and these should be correctly bonded.

The argument against is that by bonding the bath there is a pathway to earth which did not previously exist and therefore produces a risk with the introduction of a potential (faulty appliance, fan heater on extension etc).

Yes.

The spanner in the works seems to be the plumbing. Copper pipes may be tested by the electrician to ensure conductivity/connections but ten minutes or ten years later the plumber fits a plastic section to repair a leak. The plastic inserts which isolated the tap will fill with water.. and so on.

If you mean because of the conductivity of the water, I believe that can be discounted if the plastic section is more than 200mm.
Other than that, it would be better if the pipes were insulated and the bath isolated but it is not done by plumbers.
You could fit long insulating flexible tap connectors in your own property.

My suggestion is that the introduction of a length of plastic pipe between any copperwork (supply and drain)and the bath would solve the problem and future-proof the installation.

It would.

Probably not realistic for many situations, but most questions about bonding and electricity in bathrooms arise during refits or new work and in those cases easy to achieve and may be a suggestion which could be incorporated into how-to-it guides.
I'm ready to be shot down, but would appreciate views on my (?) logic.

Yes and ,as I believe, in a previous thread you discovered that bonding and earthing do not have yes or no answers - it all depends on the situation.


I think this thread has been muddled after my suggestion that a metal bath supplied with plastic attachments should be tested to verify (for certain) that it was isolated and replies enquiring about it being supplied with copper pipes which have lead to confusing and contradictory replies.

 

[JohnW2]

I'm suggesting that the bath with a very low resistance path to the MET, because of copper piepwork, does not need a G/Y bonding conductor because of that low resistance path to the MET via the pipework

It probably needs a bond to ensure the original "earthing" via pipe work is replaced by an earth via bonding if the connecion from bath the MET via pipe work is lost.

I must be writing unclearly, because people don't seem to be getting my point ... If the only reason for a low resistance path from bath to MET (hence requiring bonding, per EFLI) were the copper pipework, if that path from bath to MET were 'lost' (e.g. by inserting plastic interuptions into the pipework), then the resistance from bath to MET would rise to a very high (near 'infinite') value, and therefore the need for bonding (per EFLI) would disappear. As you go on to say:

Maybe the logic is that pipe work may be altered and the connection to MET via pipe work is lost so bonds are needed. But if an electrician was to be asked bond or not bond after the pipe work was altered he would find no connection between bath and MET and would therefore decide bonding was not needed.

... which is exactly what I've written above - interupt the path to MET via pipework, and the need for bonding disappears, so there's no need to guard against that possibility by installing 'G/Y'bonding.

Surely it is better to make the decision based on the conditions that will occur when the bath room is being used and should also take into account faults that may occur.

Precisely. That's the point I'm always making, and the reason why I'm nervous when I see things being written which could be taken by some to mean that the decision can/should be based on a measurement (at one point in time).

Kind Regards, John

 

[EFLImpudence]

I must be writing unclearly, because people don't seem to be getting my point ... If the only reason for a low resistance path from bath to MET (hence requiring bonding, per EFLI) were the copper pipework, if that path from bath to MET were 'lost' (e.g. by inserting plastic interuptions into the pipework), then the resistance from bath to MET would rise to a very high (near 'infinite') value, and therefore the need for bonding (per EFLI) would disappear.

You're not writing unclearly but confusing two situations.

In this thread was an (apparently) isolated bath which I suggested testing to the MET (to verify isolation) in case, hypothetically, it were connected to a metal structure which made it extraneous.
This would apply whether or not the bath was connected to plastic or copper pipes and would involve bonding the bath itself rather than just the pipes.

If it were supplied by copper pipe then the pipes would be tested and bonded appropriately but it is likely that the connection to the metal structure would go unnoticed by an electrician after the installation.
I suppose, therefore, that the bath should be tested before connection to the copper pipes but a plumber fitting the bath would know how he had fitted it.

Maybe the logic is that pipe work may be altered and the connection to MET via pipe work is lost so bonds are needed.

Not unless connected to metal structure.

But if an electrician was to be asked bond or not bond after the pipe work was altered he would find no connection between bath and MET and would therefore decide bonding was not needed.

Yes, unless connected to metal structure.

which is exactly what I've written above - interupt the path to MET via pipework, and the need for bonding disappears, so there's no need to guard against that possibility by installing 'G/Y'bonding.

It depends on the situation.
There are no definitive answers when talking about two different installations in one thread.

Surely it is better to make the decision based on the conditions that will occur when the bath room is being used and should also take into account faults that may occur.

I presume you mean used for bathing but what about being used for cleaning, using electric shaver, repairs etc.
You cannot have variable bonding and so a compromise has to be made and so we work to the times when people are naked and wet.
After all, if it were not for the bath supplementary bonding would not be required at all.

Precisely. That's the point I'm always making, and the reason why I'm nervous when I see things being written which could be taken by some to mean that the decision can/should be based on a measurement (at one point in time).

It depends on which time that is, surely, and how do we compensate for any changes.


RCDs save a lot of uncertainty.

 

[JohnW2]

You're not writing unclearly but confusing two situations. In this thread was an (apparently) isolated bath which I suggested testing to the MET (to verify isolation) in case, hypothetically, it were connected to a metal structure which made it extraneous.

With respect, I don't think I'm confusing two situations. Maybe my mistake was to use the example of a connection from bath to met via metal pipework. As you indicate, I could just as easily have used the example of a connection from bath to MET due to the bath being bolted to (bonded) structural metalwork - the discovery of which, as you say, is one of the (albeit extremely unlikely) reasons for your test.

Precisely. That's the point I'm always making, and the reason why I'm nervous when I see things being written which could be taken by some to mean that the decision can/should be based on a measurement (at one point in time).

It depends on which time that is, surely, and how do we compensate for any changes.

Indeed, but statements about testing to decide whether bonding is necessary usually/often don't mention those "it depends" qualifications. As for how one compensates for any changes, my personal view is that if the resistance (to MET) is less than at least a few hundred k&#937;, I would want to investigate the reason for that 'sub-infinity' measurement, and then, once I had discovered the reason, I would make a judgement as to whether it was a resistance which could perhaps reduce to much lower levels under certain environmental conditions.

RCDs save a lot of uncertainty.

... and also create senses of (sometimes) false confidence, sometimes even 'complacency'. I need not tell you that RCDs do fail (sometimes dramatically, as I've just discovered!!) and, even if they work as intended, are not a guarantee against death or serious injury. I'm sure you would agree that one should not allow the presence of an RCD to allow one to design an installation any less 'safely' than one would in the absence of an RCD. The RCD is obviously there as 'additional protection' (over and above everything else one can do to achieve 'safeness'), not as a means of compensating for less-than-ideal design.

Kind Regards, John

 

[bernardgreen]

Surely it is better to make the decision based on the conditions that will occur when the bath room is being used and should also take into account faults that may occur.

I presume you mean used for bathing but what about being used for cleaning, using electric shaver, repairs etc.

Any use which creates humidity in the bathroom will affect the impedance of the electrical path from bath to earthed items ( or to ground directly ). The effect could be to significantly reduce the impedance to below this arbitary fidue of 23 K ohms.

You cannot have variable bonding and so a compromise has to be made and so we work to the times when people are naked and wet.

Precisely..... and when is the measurement of the 23 K ohms made ? Very unlikely that it will be made when someone is in the bath or efven shortly after the person gets out of the bath so the decision to bond or not to bond is not made on data from the main use of the bathroom.

Metal bath feet on damp floor boards can provide a conductive path to items that are earthed in some way. ( either direct or by bonding connections )

After all, if it were not for the bath supplementary bonding would not be required at all.

In some old buildings the walls are sufficiently conductive to introduce true ground potential into the bath room. With a small area of contact to the wall the impedance will be high but iif the bath is in contact with the wall then the bath is going to act as a contact electrode and the larger area of contact with the wall will effectively reduce the impedancde between bath and true ground. Also earthed item in the wall such as pipes need to be considered.

It depends on which time that is, surely, and how do we compensate for any changes.

We cannot compensate over time. it has to be correct from the beginning.

RCDs save a lot of uncertainty.

Only if the total impedance of the circuit that includes the human body is low enough to allow 30 mA of earth current to flow.

 

[JohnW2]

It depends on which time that is, surely, and how do we compensate for any changes.

JohnW2 did not write that, although he did quote and comment on it.

Kind Regards, John