Does this need to be bonded?

[HappyHippyDad]

I have just put in a sub distribution board in the middle of a field for a new supply for some poly tunnels (soil heating mats and lighting). I have made it a TT.

I have noticed that they have concreted in a number of separate metal hollow poles (see picture) which will be part of the polytunnel. I haven't tested these yet, but they will test as extraneous. If they remain as they are (i.e accessible) do they require Main bonding?

I do get confused with bonding! Every year I re-read up on it, feel I've got it, and then either forget it or a new scenario comes along!

My thoughts are that it will need a 10mm into the nearby DB as it is an 'exposed metallic structural part of the building', which is one of the examples of an extraneous conductive part (411.3.1.2). However, each post is a separate piece of extraneous, so I need to continue this 10mm around the poly tunnel to each post. Is this right? Or, does it not require bonding for whatever reason?

PS.. there will be no livestock in the field at any time.

 

[LastManOnline]

It would still be fun to loop test them, if only
to have a good laugh if they have lower impedance than the official earth electrode….

Yes. Would be interesting to know what the total Fli is with all metal legs tied together

We do have to remember that bonding in this case is joining a metal thing in the ground to another metal thing in the ground.

Bonding in a home assumes there is little or no hand to foot fault current current flow.
This does, nt apply outside.

Bonding in the home assumes connected metalwork has a low resistance to earth.
This is unlikely to apply in OP, s case.

So what will bonding achieve?. If there is a fault in a class 1 accessory all metalwork becomes live providing multiple fault paths to "real earth".
I think I would prefer unbonded metalwork in this case and ideally use class 2 light fittings /accessories that may be attached to the metalwork

 

[timhoward]

I'm in agreement with @LastManOnline. As far as I can see, in this situation, all bonding does is give you more chances to die if the RCD fails.

If the person was reasonably insulated from real earth, e.g. standing on a giant rubber mat, then there are obvious benefits to to ensuring that a faulty appliance/light casing and the poles are held at the same potential, so under fault conditions touching either or both wouldn't cause a shock.
But in this case there is an instant ~1000 ohm path to real earth via the human to soil. Touch one thing live, never mind two, and you have a problem. So in this situation limiting the number of items you can touch at mains potential does seem preferable to me.
The only other aspect would be the poles becoming directly live themselves, e.g. a faulty class 1 heater resting against one. As it's TT then we'd be relying on the RCD anyway as limited current would flow. I like the idea of an additional up-front S type 100ma RCD and 30ma RCBOs so you get some redundancy in fault conditions.

 

[brianmoooore]

I suspect they'll be the hooped plastic pipe type supports... so MDPE water pipe just stuffed into those hollow pipes. The PVC sheeting, when tight, will be part of the structure..

This is the method of construction of the ones I've seen around here.

 

[HappyHippyDad]

Went back there today for 20 mins and tested the posts. They test 101ohms between posts and the same between post and rod.
I didn't do an EFLI test on the posts though, sorry!
I can't quite see how bonding helps in this case.
All posts will have a similar Ra to the rod. If there is a fault to earth the RCD will trip.
If a person came into contact with one of the posts whilst simultaneously touching an exposed conductive part that was somehow live, the RCD would trip as the posts are <1667ohms.

 

[cliffed]

This may help determining an extraneous conductive part

 

[HappyHippyDad]

This may help determining an extraneous conductive part

Hi cliffed,
These posts are definitely extraneous as they measure 101Ω to the MET. However, they are pretty much the same as the rod in the ground (66ohms). If anyone touched them under fault conditions the RCD would trip as they are below 1667ohms.

I've heard conflicting arguments about this.

Its clear to me why an Extraneous Conductive Part (ExtCP) needs bonding IF it measures between 1667Ω and 22KΩ to the MET, as between this range the current into the human (poorly worded) between a fault and the ExtCP will be >10mA.

The bit that is not clear to me is why do we still need to bond IF the ExtCP has a sufficiently low resistance to MET, meaning that the RCD will trip if we come into contact with it under fault conditions.

I have used 1667Ω as this is the known figure to trip a 30mA RCD in order to limit voltage to 50V. However, we use a figure of 200Ω for stability reasons AND we would have to be certain that the ExtCP is not going to be altered in any way (thus perhaps increasing its resistance to between 1667Ω and 22KΩ).

In my scenario, the posts are <200Ω to earth AND are not going to be altered, so why do they need bonding? What risk do they pose?

 

[cliffed]

Hi cliffed,
These posts are definitely extraneous as they measure 101Ω to the MET. However, they are pretty much the same as the rod in the ground (66ohms). If anyone touched them under fault conditions the RCD would trip as they are below 1667ohms.

I've heard conflicting arguments about this.

Its clear to me why an Extraneous Conductive Part (ExtCP) needs bonding IF it measures between 1667Ω and 22KΩ to the MET, as between this range the touch voltage between a fault and the ExtCP will be >10mA.

The bit that is not clear to me is why do we still need to bond IF the ExtCP has a sufficiently low resistance to MET, meaning that the RCD will trip if we come into contact with it under fault conditions.

I have used 1667Ω as this is the known figure to trip a 30mA RCD in order to limit voltage to 50V. However, we use a figure of 200Ω for stability reasons AND we would have to be certain that the ExtCP is not going to be altered in any way (thus perhaps increasing its resistance to between 1667Ω and 22KΩ).

In my scenario, the posts are <200Ω to earth AND are not going to be altered, so why do they need bonding? What risk do they pose?

They may need bonding for a definite earth connection reference point

 

[DPG]

Can't see the point of connecting a metal rod in the ground to another metal rod in the ground.

Or can I?

 

[HappyHippyDad]

Can't see the point of connecting a metal rod in the ground to another metal rod in the ground.

Or can I?

Covering your arse

 

[pc1966]

Can't see the point of connecting a metal rod in the ground to another metal rod in the ground.

Or can I?

Depends on the goal. Linking them lowers Ra and also under high current faults could drop the step potential gradient a bit, but really to goal is to avoid it being needed in the first place!

 

[DPG]

Depends on the goal. Linking them lowers Ra and also under high current faults could drop the step potential gradient a bit, but really to goal is to avoid it being needed in the first place!

Well, a bit. But the Ra of the proper earth rod is not going to be helped much by the resistance of the hoops I wouldn't have thought.

 

[pc1966]

Well, a bit. But the Ra of the proper earth rod is not going to be helped much by the resistance of the hoops I wouldn't have thought.

Probably not.

But ultimately for a TT system the issue of CPC being not being elevated comes down to reliable RCD protection, I would go for a dual RCD box (100mA incomer, few 30mA RCBO with DP switching per circuit), a decently reliable rod, and not bother with any bonding.

 

[LastManOnline]

This may help determining an extraneous conductive part

Am curious as to why the attachment you posted stated 7.67 Kohm, s as the cut off point for what is regarded as extraneous.? I expected to see a figure of 22 Kohm, s?

 

[Pretty Mouth]

Am curious as to why the attachment you posted stated 7.67 Kohm, s as the cut off point for what is regarded as extraneous.? I expected to see a figure of 22 Kohm, s?

If you use 10mA for the calculation, you get 22k ohms, if you use 30mA you get 6.67k ohms (not, i believe, 7.67k ohms)

 

[HappyHippyDad]

If you use 10mA for the calculation, you get 22k ohms, if you use 30mA you get 6.67k ohms (not, i believe, 7.67k ohms)

You would get 7.67k ohms with the calculation given. It's just that the calculation we generally use includes taking off 1k ohms for resistance through the body. Hence 22k rather than 23k, or in this case 6.67k rather than 7.67k.