to bond or not to bond

[leep82]

I know its been done to death but in going to ask for your thoughts opinions anyway.

Just turned up on a job that ive been on for a while where i have had no communication with the plumber. Everything has been through the customer. The plumber has fitted a new central heating system with a boiler that is oil fed. The bund for the oil sits on a concrete pad and has been in for weeks. I was told by the customer that the pipe to feed the boiler would be surface fixed and of plastic, hence, no need for any bonding. Standing looking at it now and the pipe has been buried for 2/3 mtrs and although it has a thick plastic coating is quite clearly a copper pipe. Is the plastic coating enough to suggest that bonding may not be necessary?

 

[telectrix]

test it to see if it's extraneous. yo may well find that the pipe is bonded at the tank.

 

[Richard Burns]

It is possible that a metal tank on a concrete base may provide enough of an earth connection (especially once it has been the for a while) to test as extraneous. Though if the tank is plastic (as they mainly are now) obviously not applicable.
Damage to the plastic sheath over time in shifting underground material could also permit an earth reference.
However at the present time Tel's advice is the best.

 

[leep82]

Thanks guys

 

[leep82]

Quick update, following on from the advice i tested to see whether or not the pipe was extraneous and it seems it is. I had to run out a wander lead as the part in question is some way from the DB. I disconnected the main earth conductor and then performed an IR test at 500v between my main earth conductor and wander lead. My reading was 0.00Mohms, and so less than 0.02Mohms which would deem the pipe not extraneous. Is this right?? The boiler is all connected and so could i be getting a parallel path through the already bonded water?

 

[Richard Burns]

The test you are doing is to see if the oil pipe is electrically connected to earth and so could introduce an earth potential into the house.

If there is limited electrical resistance (less than 23kΩ) then the pipe is connected to earth and so it is extraneous and requires bonding back to the MET.

There is some consideration that if the resistance is less than 1667 and the installation is fully 30mA RCD protected then it may not need bonding but this is not reflected in the regulations.

If you get a very low reading such as 0.00MΩ then it is easy to change to low ohms resistance testing and if the value is say 0.01Ω to 6Ω then this is very likely to be a connection via a parallel path via other bonding or circuit cpcs and therefore giving you a false reading.

If the pipe is connected to earth then you can think of it like an earth rod which, if just placed in the ground, is likely to give readings from about 50Ω to 2000Ω though there is wide variation possible.

If the pipe is only tenuously connected with earth then the reading is likely to be in the hundreds of MΩ.

if you tested to the disconnected main earth conductor and the bonding is not connected at the service head then you should not be able to get parallel paths from the water bonding or from circuit cpcs.

 

[davesparks]

if you tested to the disconnected main earth conductor and the bonding is not connected at the service head then you should not be able to get parallel paths from the water bonding or from circuit cpcs.

Actually you could get a parallel path I think. If the oil pipe being tested is connected to the boiler, which the water main is also connected to, and if the water main is itself extraneous, then the oil pipe will test as extraneous via the water pipe.
But of course this is getting in to the realms of ridiculousness as far as practicality goes.

 

[leep82]

if you tested to the disconnected main earth conductor and the bonding is not connected at the service head then you should not be able to get parallel paths from the water bonding or from circuit cpcs.

Actually you could get a parallel path I think. If the oil pipe being tested is connected to the boiler, which the water main is also connected to, and if the water main is itself extraneous, then the oil pipe will test as extraneous via the water pipe.
But of course this is getting in to the realms of ridiculousness as far as practicality goes.

This is what i was thinking, there is no way to check the oil pipe on its own as it is connected to the boiler along the the water

 

[davesparks]

This is what i was thinking, there is no way to check the oil pipe on its own as it is connected to the boiler along the the water

It's a fun game isn't it!

 

[yellowvanman]

If there is limited electrical resistance (less than 23kΩ) then the pipe is connected to earth and so it is extraneous and requires bonding back to the MET.

I notice in the May edition of the Professional Electrician & Installer the NICEIC are saying its NOT extraneous if the resistance is > 7.67kohm

 

[Leesparkykent]

I notice in the May edition of the Professional Electrician & Installer the NICEIC are saying its NOT extraneous if the resistance is > 7.67kohm

yes I noticed that....got a copy in front of me now.

 

[Pete999]

Yep I wondered about that, can anyone enlighten me?

 

[davesparks]

Yes well the NICEIC do like to print a lot of bog-roll

 

[Richard Burns]

If we accept 30mA is safe for RCDs, then 30mA should be safe for bonding as well, I suppose.
Though it always assumes that the protective device operates.

 

[davesparks]

If we accept 30mA is safe for RCDs, then 30mA should be safe for bonding as well, I suppose.
Though it always assumes that the protective device operates.

Eh? I'm sure I'm missing something obvious here, but how is the presence of an rcd going to assist with the bonding?

 

[Richard Burns]

Sorry I am just saying that a 230V pd on 7670Ω would generate 30mA of current.

Since additional protection by an RCD at 30mA is considered generally non life threatening then can we apply this to a conductive part that may present a 30mA current to someone holding it in the case of a fault?

 

[Pete999]

If we accept 30mA is safe for RCDs, then 30mA should be safe for bonding as well, I suppose.
Though it always assumes that the protective device operates.

https://www.youtube.com/watch?v=v38PfIVy2rI

 

[yellowvanman]

A similar article was in Certsure's (NICEIC) quarterly magazine about 18 months ago, and I queried it with my asessor last year. He was unaware of it and felt it was incorrect and did say he would investigate further. When we met this year he didn't get anywhere with it as no one in the NICEIC was prepared to say it was incorrect!

I think one of the GN specifically mentions the 23k value.

 

[Leesparkykent]

Yep I wondered about that, can anyone enlighten me?

Apparently within IEC/TS 60479 it gives a value of 30mA for the value of current through the body that should not be exceeded 230V/0.03=7666 but surly this figure should even have 1000 subtracted due to impedance of the human body? The way I was taught and I think many others the current that's accepted to not be exceeded through the human body is 10mA. 230V/0.01=23000 subtract 1000 due to the impedance of the human body gives us the figure of 22000 ohms. I Know which figure I would rather stick to and its not what the NICEIC state :crazy:

 

[davesparks]

Sorry I am just saying that a 230V pd on 7670Ω would generate 30mA of current.

Since additional protection by an RCD at 30mA is considered generally non life threatening then can we apply this to a conductive part that may present a 30mA current to someone holding it in the case of a fault?

I will honestly have to put pencil to paper, and probably consult the books to clear this one up in my mind I think.
Though moving house at the moment means I haven't a clue where any of the above items are currently located

 

[Richard Burns]

I have just looked in GN8 and it specifies that it is up to the designer to choose the value of current that is acceptable
0.5mA - perception threshold
10mA - let go threshold
30mA - <300ms muscle contraction, <5s breathing difficulty, revesible heart disruption, immobilization, but generally no organic damage.

The resistance of the human body approximating to 1kΩ gives resistance values of:
459kΩ for 0.5mA
22kΩ for 10mA
6.7kΩ for 30mA

since bonding faults could be reasonably long term in some cases I would tend to aim for 10mA as at least you can get away from the fault .

 

[yellowvanman]

The other problem is how do you measure 6.7kΩ as opposed to 22kΩ. Most insulation testers have a 10kΩ resolution at these levels of resistance, my Fluke's lower limit is 0.02M, lower than that it displays 0.00!

 

[davesparks]

The other problem is how do you measure 6.7kΩ as opposed to 22kΩ. !

With a test meter which has a suitable measurement range

 

[leep82]

Wow this has escalated. If i assume it is indeed extraneous, my next problem is where to actually place my clamp. The oil feed enters the building and terminates immediately into the back of the boiler. There is no exposed metal pipe work that is accessible inside the building.

 

[davesparks]

Wow this has escalated. If i assume it is indeed extraneous, my next problem is where to actually place my clamp. The oil feed enters the building and terminates immediately into the back of the boiler. There is no exposed metal pipe work that is accessible inside the building.

At the point of entry to the installation or as near as practicable. The point of entry to the installation is presumably the point at which it passes through the external wall, so bond it either on one side of the wall or the other.

But of course that information is irrelevant if there is no exposed metal inside the building because it does not require bonding if it is not able to be touched!

 

[leep82]

Wow this has escalated. If i assume it is indeed extraneous, my next problem is where to actually place my clamp. The oil feed enters the building and terminates immediately into the back of the boiler. There is no exposed metal pipe work that is accessible inside the building.

At the point of entry to the installation or as near as practicable. The point of entry to the installation is presumably the point at which it passes through the external wall, so bond it either on one side of the wall or the other.

But of course that information is irrelevant if there is no exposed metal inside the building because it does not require bonding if it is not able to be touched!

this is what i have been asking myself. In fact there is very little exposed metal work throughout the entire run of the pipe as it is covered in a plastic coating. There are small sections where an elbow joint is used and again where it terminates into the bund. These sections arent big enough to fit a standard clamp but being as though they are outside and there is no chance that you can physically touch any of the exposed pipe and any other exposed metal part, does it actually need it?

 

[Richard Burns]

If the oil pipe joins the boiler on a non conductive point and the oil pipe inside the boiler is non conductive such that there is no continuity between the oil pipe and the boiler itself then there are no extraneous conductive parts exposed to touch within the equipotential zone and no requirement to bond.

 

[leep82]

I dont think that the pipe does join the boiler on a non conductive point, and am wondering if this is why, when i did my test to check if it was extraneous my reading was 00.0Mohms, as the water feed to the boiler is also at the same point

 

[telectrix]

basically BS7671 allows that a 10mA current through the body is safe (22k ohms. but niceic accepts 30mA. ( 6.7k ohms ). any niceic officials want to test their theory? "get hold of these while i switch on". i know where i'd sooner be if coming into contact with a live cable, considering the state of my furred up arteries.

 

[Risteard]

Apparently within IEC/TS 60479 it gives a value of 30mA for the value of current through the body that should not be exceeded 230V/0.03=7666 but surly this figure should even have 1000 subtracted due to impedance of the human body? The way I was taught and I think many others the current that's accepted to not be exceeded through the human body is 10mA. 230V/0.01=23000 subtract 1000 due to the impedance of the human body gives us the figure of 22000 ohms. I Know which figure I would rather stick to and its not what the NICEIC state :crazy:

Indeed, or you could even go a step further to limit any current to below the perception threshold where a human being would not even be aware of it.

 

[davesparks]

basically BS7671 allows that a 10mA current through the body is safe (22k ohms. but niceic accepts 30mA. ( 6.7k ohms ). any niceic officials want to test their theory? "get hold of these while i switch on". i know where i'd sooner be if coming into contact with a live cable, considering the state of my furred up arteries.

I have been on the receiving end of a shock in which my life was saved by a 30mA RCD. It knocked me out cold for about half an hour, and it hurt like hell, the last thing I remembered when I came round was hearing the click as the RCD operated.

What was I doing to get such a shock? I was tacking a BT cable round an architrave with a metal stapler, I discovered that the switch cable had been tucked in behind the architrave when I fired a staple into it. Because my hand was clenched to fire the staple and I was pushing on the stapler at the time I didn't stand a chance of getting away from it.

 

[yellowvanman]

With a test meter which has a suitable measurement range

Thats a helpful response!

Are you aware of a specific instrument that has the necessary resolution, when using a 500V insulation tester? If so, I would be interested in what it is.

 

[davesparks]

Thats a helpful response!

Are you aware of a specific instrument that has the necessary resolution, when using a 500V insulation tester? If so, I would be interested in what it is.

I made no mention of tester resolution, only measurement range plus How would using a 500V insulation tester affect another test instruments resolution?

I think what you are trying to ask is whether or not I know of a test instrument which tests at 500V and has a suitable measurement range? In which case no I do not, but I'm not entirely sure why you would want one?
This is not a measurement of insulation resistance, it is a measurement of resistance and as such any meter capable of accurately measuring resistance will do the job.

Insulation resistance testers are generally used as they are the most convenient instrument for most electricians and generally have a suitable range when considering a 10mA current. But as far as I know they are not required.

Personally I use a megger AVO410 multimeter.

 

[LeeH]

I think this is a perfect example of a time where it's harder to prove it isn't extraneous then it is to bond it.

Just my 10 pence...

 

[davesparks]

I think this is a perfect example of a time where it's harder to prove it isn't extraneous then it is to bond it.

Just my 10 pence...

I completely agree

 

[madmaz]

I've just had to complete strange bonding thanks to a plumber !!! Main water supply to a detached outbuilding enters property in alcathene then drops to copper after stop tap, this copper drops under the concrete into speed fit which runs in the concrete supplying hand basins, each set of pipes to each basin are copper connected to a speed fit tee under the concrete the bottom 1" of each copper pipe is buried in concrete when I tested each pipe the resistance to met was below the 23000 ohms but different to each other, readings of 2.9k - 14.5k I've had to bond all pipes together thanks to this,

 

[yellowvanman]

I made no mention of tester resolution, only measurement range plus How would using a 500V insulation tester affect another test instruments resolution?

I think what you are trying to ask is whether or not I know of a test instrument which tests at 500V and has a suitable measurement range? In which case no I do not, but I'm not entirely sure why you would want one?
This is not a measurement of insulation resistance, it is a measurement of resistance and as such any meter capable of accurately measuring resistance will do the job.

Insulation resistance testers are generally used as they are the most convenient instrument for most electricians and generally have a suitable range when considering a 10mA current. But as far as I know they are not required.

Personally I use a megger AVO410 multimeter.

Not aware of how GN8 defines how the measurement is made, but from an ECA publication it states:

'Using a 500 volt DC insulation tester,
measure the insulation resistance between
the item and the main earthing terminal.
If the resistance value is 23 000 ohms or
greater, and inspection confirms that the
resistance is unlikely to deteriorate, then the
item can reasonably be considered not to be
an extraneous-conductive-part.'