Plumbers. Bless them.

[RF Lighting]

Ok the - so the furthest point from the CU is my WiFi stat now with its unused cpc. so - If I have you right - Insulation test, polarity test. ELI and RCD (if Applicable) and we are safe?

The proper sequence is

Continuity of CPCs. You can do either R1+R2 or just R2 measurement for this
Ring final circuits end to end measurements (not applicable here)
Insulation Resistance
Polarity (dead test)
Polarity (live test)
Loop Earth fault loop impedence test
Earth electrode resistace (only required on a TT supply)
Devices. Functional testing of RCDs etc.

Assuming no cpc wire in the WiFi stat flexi for the minute, - then the same detail but the earth from the earth tab on the boiler?

You should take the reading at the furthest point on the installation. In this case if there is no CPC at the furthest point, then the next furthest point. If there are branches on the circuit such as individual wires to a cylinder stat and a room stat then you really ought to confirm there is a satisfactory earth at each position.

If doing the Insulation test ( bearing in mind its usually a spur on a ring main) would u need to unplug sensitive equipment to 500vdc? or would 250 v dc sufficed or Join L&N & just do the one 500v option?

For a new circuit I'd expect each condutor to each other conductor at 500V DC. You could do this test once all the wiring is in place but before making the actual connection to the boiler for example so there is no risk of frying the components on the PCB.

 

[viewer]

This is a quenuine and serious question ...

The bath (because it is in the concrete floor) should be tested for continuity with the met (your connection removed, obviously). ... If less than 23kΩ it should be bonded as above. If >23kΩ it should NOT be bonded.

Would you be happy to make a decision not to bond (should conditions for omission of bonding not be satisfied) on the basis of a 'acceptably high' resistance obtained during a long dry spell of weather and when there was no recent history of the floor having been frequently drenched with water from above?

Kind Regards, John

Thanks both; I see your logical approaches. The work was done a few years ago when I moved in. The system was TT (now PME) with the rods just outside the bathroom and the whole house very very wet. I guess (no memory) that this was why I joined all together as well as fitting an RCBO.
As the house has been repaired it is now generally dry, although John is spot on with his assessment of my unruly bathing habits. As work progresses I am constantly checking and questioning; hence my post. I want it right. Shower is not in use at the moment so a good time to get it right. Thanks again.

 

[bernardgreen]

If >23kΩ it should NOT be bonded.

My opinion is that if you can stand in a bath and at the same time touch metal items then I would want to be certain for ever that bath and the metal items were at the same potential and the only way to ensure that is by bonding them together.

23 kOhm at 230 volt allows 10 mA to flow. That current can be lethal depending on points of entry and exit. It is too low to trip an RCD so the shock will be prolonged.

Even if the shock is not lethal there is a un-acceptably high risk of subsequent injury from a fall or other shock induced muscular contractions.

John makes a very valid point that testing on a dry day will give a different result than the same test on a damp day or when the room is humid due the bath having been used.

 

[ban-all-sheds]

My opinion is that if you can stand in a bath and at the same time touch metal items then I would want to be certain for ever that bath and the metal items were at the same potential and the only way to ensure that is by bonding them together.

So your opinion is that one should take metal items which are not in any way part of an electrical installation and deliberately connect them to one.

My opinion is that you are David Cockburn, or a disciple of his.

 

[JohnW2]

My opinion is that if you can stand in a bath and at the same time touch metal items then I would want to be certain for ever that bath and the metal items were at the same potential and the only way to ensure that is by bonding them together. 23 kOhm at 230 volt allows 10 mA to flow. That current can be lethal depending on points of entry and exit. It is too low to trip an RCD so the shock will be prolonged.

I agree that the arbitrary 23kΩ/10mA appears to be a bit marginal but, having said that, even with 23kΩ, a current of 10mA would only flow if the 'other metal item' was at a potential of 230V - something which would be unlikely to arise (or persist for a significant time) in a satisfactory installation (certainly if a TN supply) and which would almost certainly result in an RCD operating very rapidly.

Having said that, if I felt that there was any prospect that (due to wetness) the resistance from bath to earth could ever fall to anything like as low as 23kΩ (even if it were very high when measured under 'dry' conditions), and if any other metal could be touched whilst in contact with the bath, I probably would want it to be bonded. This is only really likely to (possibly) arise if the bath is in contact with a thin concrete floor which is laid directly on soil.

Kind Regards, John

 

[bernardgreen]

So your opinion is that one should take metal items which are not in any way part of an electrical installation and deliberately connect them to one.

If there is a reasonable chance of the metal item coming into contact with a live conductor then it is my opinion that common sense requires consideration be given to each specific case.

My opinion is that you are David Cockburn, or a disciple of his.

My opinion is you are a person with a closed mind that finds it easier to do as the book requires instead of making the effort to evaluate the situation that actually exists or may exist.

The rules assume no one would put a double insulated appliance into a sink full of water. So the rules do not consider it possible that a steel sink will become live. What would you consider is better protection should that accident happen ? The rules or a bond from sink to CPC or MET to ensure the RCD can trip BEFORE someone is electrocuted.

 

[JohnW2]

The rules assume no one would put a double insulated appliance into a sink full of water. So the rules do not consider it possible that a steel sink will become live. What would you consider is better protection should that accident happen ? The rules or a bond from sink to CPC or MET to ensure the RCD can trip BEFORE someone is electrocuted.

As you will understand, that is one side of an argument to which there are two sides - i.e. a risk-balance assessment has to be made in each case. The other side of the argument is that, should someone be simultaneously in contact with a 'floating' metal sink and some other bit of metal that had become live, nothing would happen, but if the sink were bonded then the person would receive a shock, and would be reliant on an RCD to reduce the risk of that shock being fatal.

In that case, the risk assessement would depend upon which scenario one considered more likely - (a) a potential being introduced into the sink (e.g. by dropping a Class II appliance into the sink when full of water) or (b) some other metal within the room becoming live, due to a fault. Opinions will obviously differ, but I think I would personally regard the latter as the more probable, hence tending to shift the balance in the direction of 'do not bond'.

Kind Regards, John

 

[EFLImpudence]

Of course 23kΩ allows 10mA to flow. That's how the figure 23kΩ is determined - from 10mA, a figure deemed not dangerous.
You may use a higher figure if you wish.

The measurement is not going to be 22999 or 23001Ω.
It will either be relatively low or very high, hopefully nearly infinite.

The bath with plastic supplies definitely should not be earthed - this would not be bonding - but because it is in concrete we do not know if the bolts go down into the earth so I suggested it should be tested.

I do not consider the concrete, in itself, to be a problem as concrete is, in any case, waterproof (unless, I suppose, very poorly made).

The likelihood of the bath becoming live by being touched by a live conductor is almost nil - especially if there are no conductors nearby.
Even with a cable underneath the chance of this happening is remotely small compared to a person touching a faulty appliance or cable whilst leaning on the bath or sitting on it using a faulty hair drier were it wrongly earthed - this would not be bonding.

Bonding is a compromise and while it would be possible to think of a situation where had it been done differently a hazardous situation could have been avoided we cannot make an installation perfectly safe for all occurrences.


An example of this is the MP's daughter who was tragically killed, which, I believe, led to the introduction of part P and the accompanying regulations, even though, apart from perhaps cable routing, I do not think there was anything wrong with the installation.

She was unfortunately killed when she hung a metal spoon on a metal hook which was fixed with a screw which had penetrated a live wire while her leg was touching a washing machine.
I am assuming there was no RCD protection.

In this case, it would have been preferable had the hook been earthed and the washing machine carcass not earthed.
There are several permutations which would have been 'safer' for her but amongst other scenarios, to have the washing machine carcass unearthed would, normally, be more likely to cause a shock and for the hook to be earthed as a matter of course would be silly and could lead to a shock the other way around also being fatal.

So, the hook should have been tested. People may do such a thing now but, to be honest, was that to be expected? Would it have been noticed at the time or did it become live by some later movement?
Either way, it was just a tragic accident but earthing parts that should not be bonded could just as easily lead to similar tragic consequences.

We can only do what is considered sufficient to avoid the more likely of many different hazards.

 

[viewer]

Hi, sorry, didn't mean to start an argument; but, (not withstanding the advice which I've already taken onboard) if I've read it correctly, both John and EFLI agree that, it is a matter of risk (hazard x likelihood). The hazard doesn't change but the likelihood does. Texts I've read tended to be dogmatic and gave no explanation of how a decision was made.

I can now understand, as both said, that the likelihood of the bath becoming live is much less than the likelihood of something else becoming live and the bath acting to complete the circuit through the person. Using a vacuum cleaner in a bathroom must be normal behaviour. Likewise, I now understand the reason for not bonding radiators, etc served with plastic plumbing.
Thanks again

 

[JohnW2]

Of course 23kΩ allows 10mA to flow. That's how the figure 23kΩ is determined - from 10mA, a figure deemed not dangerous.

As I said, 23kΩ would result in a current of 10mA flowing (through a person) IF the 'other metal' being touched were at 230V (and if one ignores the impedance of the person). However, as I also said, if the electrical installation as a whole is properly designed and constructed, it's very unlikely (theoretically 'impossible'?) that the potential of any exposed-c-ps or other metalwork could rise to anywhere near 230V for longer than it takes an RCD (ifpresent) to operate. Hence, as I said, in practice the current though a person may well be less than 10mA, and should persist for a significant time - hence probably pretty unlikely to result in a fatal shock.

The measurement is not going to be 22999 or 23001Ω. It will either be relatively low or very high, hopefully nearly infinite.

In most situations, that's probably true - but it's not inevitable. I have a particular 'issue' in some parts of my cellar, where the floor consists of (in many case cracked!) unglazed quarry tiles sitting on sand over soil (which can vary from 'bone dry' to 'waterlogged'). The resistance between a metal object sitting on that can vary from 'nearly infinite' (>500M&#937 after prolonged dry spells (e.g. summer 2011) to very low after prolonged wet spells (e.g. summer 2012!), when the water table is often higher than the cellar floor, and only the active drainage system prevents the whole cellar being submerged! ... and, of course, at various times, any resistance between those extremes.

I do not consider the concrete, in itself, to be a problem as concrete is, in any case, waterproof (unless, I suppose, very poorly made).

Yes, that's probably true - it's really when bolts etc. (or structural metalwork) pass through (or may pass through) or, as is quite common, if the concrete is cracked, that an issue may arise.

The likelihood of the bath becoming live by being touched by a live conductor is almost nil - especially if there are no conductors nearby.
Even with a cable underneath the chance of this happening is remotely small compared to a person touching a faulty appliance or cable whilst leaning on the bath or sitting on it using a faulty hair drier were it wrongly earthed - this would not be bonding. Bonding is a compromise and while it would be possible to think of a situation where had it been done differently a hazardous situation could have been avoided we cannot make an installation perfectly safe for all occurrences.

Exactly - and, as I wrote to Bernard, my personal view that, in this case, the 'compromise' (balancing of risks) is probably in favour of not bonding the bath (or metal sink). ... as you go on to conclude ...

We can only do what is considered sufficient to avoid the more likely of many different hazards.

So, even if you perhaps thought otherwise, I don't think we're actually disagreeing about anything.

Kind Regards, John

 

[EFLImpudence]

Well done.

Remember - only parts which are, themselves, earthed need bonding.

 

[EFLImpudence]

I don't think we're actually disagreeing about anything.

No, I wasn't arguing with you.

It was Bernard who queried the 23kΩ because it would result in 10mA even though that's how the number was reached.

Plus explaining my other points for you and viewer.

 

[JohnW2]

Hi, sorry, didn't mean to start an argument; but, (not withstanding the advice which I've already taken onboard) if I've read it correctly, both John and EFLI agree that, it is a matter of risk (hazard x likelihood). The hazard doesn't change but the likelihood does. Texts I've read tended to be dogmatic and gave no explanation of how a decision was made.

You've started nothing - it's a very common cause of 'robust debate' around here!

As a matter of detail, even the 'hazard' as you call it can differ. If you consider the hypothetical situation of a bath with a resistance of just over 23kΩ to earth, if unbonded that resistance would limit the current flowing to earth from someone touching the bath as well as something else that were live to a maximum of 10mA - which usually would not prove fatal. On the other hand, if the bath were bonded (hence connected to earth), then then the current through someone touching the bath and also something like the infamous faulty vacuum cleaner which had somehow become 'live' could be very much greater than 10mA (impedance from bath to earth much lower), and hence much more probably fatal. Hence, not only is 'electrification' of a bath less likely than the 'electrification' of something else (hence favouring not bonding the bath), but the current through the person as a result of electrification of something else could be much higher if the bath were bonded.

Kind Regards, John

 

[ban-all-sheds]

If there is a reasonable chance of the metal item coming into contact with a live conductor then it is my opinion that common sense requires consideration be given to each specific case.

But that's not what you said - you mentioned nothing about reasonable chances and nothing about live conductors:

My opinion is that if you can stand in a bath and at the same time touch metal items then I would want to be certain for ever that bath and the metal items were at the same potential and the only way to ensure that is by bonding them together.

My opinion is you are a person with a closed mind that finds it easier to do as the book requires instead of making the effort to evaluate the situation that actually exists or may exist.

On the contrary - it is not I who is not thinking about it - it is you who is following a faulty line of reasoning which goes "ooh - it's metal, ooh - it might become live through some event, ooh - I might be touching it and touching something else metal at the time, ooh - I'd better bond everything metal together".

It is you who is not making any evaluation of the probabilities of the various faults you envisage, it is you who wants to blindly connect things to your electrical installation even if that would make it more dangerous on balance than less.

The rules assume no one would put a double insulated appliance into a sink full of water. So the rules do not consider it possible that a steel sink will become live. What would you consider is better protection should that accident happen ? The rules or a bond from sink to CPC or MET to ensure the RCD can trip BEFORE someone is electrocuted.

The rules, without a shadow of a doubt, because the event you postulate is so remarkably unlikely that I do not consider it anything but stupid to connect my sink to the electrical installation so that it will ALWAYS become live in the event of a fault to earth anywhere in the installation just to "be on the safe side in case" an event of vanishingly small probability happens.

Because it's not "on the safe side" - it, and you, are dangerous.

 

[bernardgreen]

so that it will ALWAYS become live in the event of a fault to earth anywhere in the installation

Surely the fault to earth will trip one or more protective devices or are you saying that a fault to earth ( or more accurately ) the CPC will make all "earthed" metal items. With PME systems a broken neutral can make the CPC live. That cannot be detected by protective devices. Equipotential bonding makes that situation acceptable safe for anyone inside the the equipotential zone bit hazardous for anyone across the boundary ( external water tap etc )

Because it's not "on the safe side" - it, and you, are dangerous.

If I had a live CPC in my cottage the shock hazard from touching a damp wall while holding a metal kettle is significant. Hence a TT install is on balance a better option.