well, I imagine the person who put it in has decided that it is safer to earth the tray, or RSJ or what have you than not.


It may or may not be, depending if it is outdoors, and how the main supply earth is arranged.

On a metal post set a foot or so into cement or metal trays or beams held with a handful of anchor bolts in to masonry, the "electrode resistance" it provides is likely to be pretty minimal ,hundreds or even thousands of ohms between the metal and terra-firna. So even with a full 230V fault, the current flowing to earth will not trouble the link wire.

But, at least if a live wire hits that tray, due to a vandalised light fitting or whatever,  there should now be enough of a connection to operate the ADS.


If this was a metal pipe running underground for many tens of metres the answer would be different, then it maybe a few ohms to terra-firma, and 'self earthing'

The regs are written assuming that all such metalwork is well grounded, and so requires a substantial main earth bond, as large currents may flow into the earth from the substation CPC.

This is not correct of course (quite a few assumptions in the regs are a bit iffy  - another good one is that all faults are zero resistance....), but it does err on the side of safety, by ensuring that most of the time, the bonding is over-sized. The only times problems really arise is where the same chunk of metal is earthed twice, bot not quite to the same voltage - this can happen with service pipes or with larger metal framed structures with more than one incoming supply. I'm going to ignore railways, mines, shipyards and other special cases, as these need, and have, their own rules.


Assuming that SWA is not a very large size, it is quite possible the armour is not really providing a low enough resistance to satisfy the regs in terms of the cross-section needed for a main bonding conductor, but I suspect that it is not really a problem either  - if the box had been metal the same connection would have been made but less obviously, and no-one would even notice.

Equally, a light switch on a metal post rarely really needs a 10mm2 cable, so there would need to be a lot of upgrading to meet the letter of the regulations.

So maybe not compliant, unlikely to be an issue in practice.


Mike

A banjo on the outside of a plastic box probably isn't regarded as best practice either - at least if it's just trapped between the gland and the box (I can't quite see the detail in that one) -  the issue is that plastic is somewhat softer than the steel of traditional enclosures, so can distort over time and result in the contact between the banjo and gland being less that tight. If you have to use a banjo with plastic enclosures the usual advise is to trap the banjo between two lock nuts inside the box so avoiding relying on the plastic to maintain a tight contact - or better still a propriatory "earthing nut" which makes a much better job of making a solid reliable contact with the thread of the gland.


   - Andy.

AJJewsbury:

A banjo on the outside of a plastic box probably isn't regarded as best practice either - at least if it's just trapped between the gland and the box (I can't quite see the detail in that one) -  the issue is that plastic is somewhat softer than the steel of traditional enclosures, so can distort over time and result in the contact between the banjo and gland being less that tight. If you have to use a banjo with plastic enclosures the usual advise is to trap the banjo between two lock nuts inside the box so avoiding relying on the plastic to maintain a tight contact - or better still a propriatory "earthing nut" which makes a much better job of making a solid reliable contact with the thread of the gland.


   - Andy.

Agreed ... these kind of products:

 

• https://www.tlc-direct.co.uk/Products/PRPEN20.html?source=adwords&ad_position=&ad_id=415703895078&placement=&kw=&network=u&matchtype=&ad_type=&product_id=PRPEN20&product_partition_id=934892185560&campaign=shopping_cable_accessories&version=finalurl_v3&gclid=Cj0KCQiA_qD_BRDiARIsANjZ2LBjsyv88ws4kfQXHV5ygjrdbqxh_r_1mbvpK7MzXeBeuqw2d1zASaUaAgtAEALw_wcB

• https://www.tlc-direct.co.uk/Products/WKEC308.html?source=adwords&ad_position=&ad_id=415703895075&placement=&kw=&network=u&matchtype=&ad_type=&product_id=WKEC308&product_partition_id=934892185320&campaign=shopping_accessories&version=finalurl_v3&gclid=Cj0KCQiA_qD_BRDiARIsANjZ2LDqliGURTjg0zd9lhb77XLzR3s7JKMqdiBJ5b_Pky5EeHmXRgQFg6caAoE8EALw_wcB

Yeah that’s it just a plastic box and switch. The so reason for main bonding being such a large conductor is due to the earth resistance being taken as fairly low so high fault current? And as to my picture am I right in saying it’s probably been done as if it’s there no body questions it but electrically all they Are doing is making the unistrut live when an earth fault occurs.

There is always a question of if it is better to earth random bits of metal or leave them 'floating' -  the question is what sort of fault you guard against.If the post is well earthed, but you are holding a damaged extension lead with an exposed live core, you may prefer it not to be earthed,  and indoors a similar argument may be made for kitchen sinks and faulty kettles, and advice about earthing sinks or not has varied over the years, with no clear wining strategy, except perhaps recently, to not worry too much but to fit RCDs instead.

If the installation is to be properly protected by equipotential bonding, you have to do everything, a half-and half affair can be worse than none.

You can achieve this indoors, but outdoors the idea of an equipotential zone rapidly becomes impractical, as there is too much stuff in reach that is casually earthed, fence posts, spades, wheel barrows, that cannot sensibly be bonded, and actually cannot sensibly become live either. For this reason kit for outdoor use, mowers, hedge trimmers etc are normally double insulated, and metalwork is self earthed to the ground it stands on by a fairly indeterminate impedance, depending how far it is hammered into the ground or is sitting in a puddle. (though a container sitting on its metal skids in the ground can be a fairly low impedance if the ground is damp and rich in organic matter - so farmland over clay is surprisingly good.)


However, if the metal work in question is part of the support or the enclosure for some class 1 electrics, so perhaps there is a credible fault path that might make it live, then maybe we'd rather it was earthed - again if the box or the light switch was metal, the same connection would be there via the fixing screws and brackets and no-one would question it, would they ?

It is not really clear-cut.

Ideally if we had the choice, the cable, the light switch and an fittings would all be double insulated and no CPC would be  needed - but that is not how most things are made.


Regards Mike


PS also where two SWA cables go into a plastic box, is often a weakpoint in terms of the armour and earthing continuity and current handling capacity, as it all funnels down onto a rather weedy little bolt  stuck into a painted metal or plastic box.

The pirahna nuts, and the related twin hole versions are much better in that way than banjos, which for my money should be omitted from the gland kits and  the nuts recommended instead, as indeed the likes of screwfix are doing.
    

Pirahna nuts in various sizes or Whiska 2 hole earth plate


M.

The so reason for main bonding being such a large conductor is due to the earth resistance being taken as fairly low so high fault current?

Bonding size is only partly down to fault currents from within the installation and its impedance to earth - some extraneous-conductive-parts (think old metallic water and gas pipes or structural steelwork in a multi-premises building) are quite capable of importing fault voltages/currents from other installations or indeed in the case of PME systems quite significant diverted neutral currents without any faults at all.

   - Andy.

So I suppose metalwork bolted to the floor Ie uni strut isn’t going to have a low enough impedance to earth m. But if we was attaching a flylead from a local swa is this supplementary bonding and is it the correct way of doing it?

The OP photo was a standard requirement for the Water Service here. Not sure if it still is. On sites we inspected, controls for pumps were often secured to steel stanchions and whether inside or out, the incoming swa sheath was bonded to the stanchion. I am not sure why, it may have been a leftover from the days when anything metallic within range was bonded or perhaps it was just a carry on from specifications for their areas where explosion risk was present. We often found the SWA glands to be corroded but not sure whether that was due to the harsh environment or the link to the stanchion. 

Setting aside the need to provide main bonding to the stanchion, the tray appears to be neither an exposed or extraneous part. 

As a matter of interest the note to Reg 411.4.2 in IS 10101 (Regs in ROI) recommends taking every opportunity to connect protective conductors to elements that provide effective earthing. I guess the OP photo may get a tick for that!

Is that a galvanized steel coupler between the box and switch?

   - Andy.