Tray and basket bonding

[i=p/u]

What if a cable gets damaged on the tray, what if the tray is within persons reach and could be touched every day by people walking by?

 

[MJPD29]

Well, there is a statement. Are you an Electrician?

Are you? Or are you a meter fitter UKPN?

 

[pc1966]

It is one of those "it depends" situations. For data cables alone then you don't need to bond it as no risk of shock.

However, there is always the chance it gets used for some 3-core cable, etc, in the future!

I always bond the tray, partly as it was always done under the EEBADS, and partly for EMC reasons as well (telecoms background). On one case it really mattered as a tree came down in high winds, a 240V cable going outside was wrenched out and shorting to the bonded tray was what tripped the OCPD (not RCD'd)!

You don't need to use brass fixings as the roofing bolts are fairly good conductors in the first place and the majority of the current flows between the crimped lug and the tray, with the bolt simply holding the two together. As mentioned above, in most cases the mechanical assembly is sufficient bonding, and the wires are needed only to link between air-separated section and ultimately to the MET via whatever CPC is available for it.

Being a bit paranoid I usually have two connections to the installation's CPC just in case one gets broken.

 

[pc1966]

Here is a photo of the pulled cable I mentioned:


And close up of the OCPD trip results:

 

[Rockingit]

Interesting that this discussion has just popped up today as I've been mulling it over in my head as I've got my annual inspection tomorrow - commercial with a bunch of tray, mix of inside and outside, some making contact with wet pipe work, some not, some being attached to the same lengths of strut that panels are also mounted on. I've 'belt and braces' bonded the lot back to the MET, wondering if it might become a discussion with the engineer.

 

[pc1966]

My view has always been that if you can't be sure that it is not in contact with anything that could become a hazard, best to bond it.

In fact, even if it is all just screwed to brickwork, etc, if the supply is TN-C-S you might find it is below 22k to true Earth and so might be considered an "extraneous conductive part" as it could give enough of a shock w.r.t. some class I appliance's CPC-bonded metalwork to fall off a ladder, etc.

 

[Rockingit]

My view has always been that if you can't be sure that it is not in contact with anything that could become a hazard, best to bond it.

In fact, even if it is all just screwed to brickwork, etc, if the supply is TN-C-S you might find it is below 22k to true Earth and so might be considered an "extraneous conductive part" as it could give enough of a shock w.r.t. some class I appliance's CPC-bonded metalwork to fall of ladder, etc.

I've actually found (admittedly it was central London) a final circuit with a Zs of 0.17 - with a disconnected PME! So that totally vindicates your point here.

 

[Ian1981]

My view has always been that if you can't be sure that it is not in contact with anything that could become a hazard, best to bond it.

In fact, even if it is all just screwed to brickwork, etc, if the supply is TN-C-S you might find it is below 22k to true Earth and so might be considered an "extraneous conductive part" as it could give enough of a shock w.r.t. some class I appliance's CPC-bonded metalwork to fall off a ladder, etc.

Earth it or bond it?
Bonding something because your not sure is no reason to bond anything.
It’s an exposed conductive part so earth it it , if it’s not then don’t, it’s extraneous then bond it, it’s not extraneous then don’t.
lumping a quite possibly large protective bonding conductor on to metal work because your not sure is not a valid reason to do so.

 

[pc1966]

Earth it or bond it?
Bonding something because your not sure is no reason to bond anything.

It is an exposed conductive part.

But it might not be in any way dangerous unless it also has the ability to introduce a voltage. So that means either:

• It can come in contact with a power source above 50V AC, etc.
• It can connect to the true Earth and you are on a TN-C-S supply

In the original discussion the talk was of data cables, and for Ethernet (even PoE) they do not meet the first point. But that might not continue in the future. That is the uncertainty, and while it is often argued we cannot engineer for future changes, my expression of "not sure" comes down to this aspect.

It’s an exposed conductive part so earth it it , if it’s not then don’t, it’s extraneous then bond it, it’s not extraneous then don’t.
lumping a quite possibly large protective bonding conductor on to metal work because your not sure is not a valid reason to do so.

The argument about bond/Earth could run and run and often does:

• Are you wiring back to the MET? Then its a CPC.
• But it is extraneous and back to the MET? Then it is a main protective bond
• Are you linking it to nearby metalwork such as data racks? Then it is supplementary bonding.

Most likely it will be 4mm cable (from 544.2.1) unless the whole TN-C-S aspect kicks in and whatever cable try structure really is able to divert significant current in the event of an open PEN then its Table 54.8 which probably means the usual 10mm.

Though unless the tray assembly is in contact with major parts (structural steel, service pipes, external cable work) it won't event see such currents, so a quick check of imeadance to MET before bonding could save you going down the big girth route if, say, it is > 100 ohms.