RCD omission risk assessment

[pc1966]

From what I've observed doesn't the German schumo socket have dual earth contacts? Although they can be reversed too so it might be for that reason.

Some of that style of plug can be reversed (the ones with the twin side earths) and others cant (the ones with the earth pin) but typically most now accept either.

But I think it is less an issue of the contact and wanting "two pins", more likely the integrity of the CPC connection to the pin and that is where the larger BS4343 (Commando) plugs benefit from having dual screws for retaining the cables.

 

[Simon47]

... more likely the integrity of the CPC connection to the pin and that is where the larger BS4343 (Commando) plugs benefit from having dual screws for retaining the cables.

But it's still one contact, one cable core, one connection inside the equipment - so while it does mitigate one particular failure mode, it doesn't mitigate several others.
Back to one of the earlier situations - telecoms facilities. I would suggest that these are relatively simple - there's options for bonding everything so if (say) a supply cord loses the CPC, there's still bonding so ADS will still work without anything exposing dangerous voltages. Plus it's a relatively controlled environment - the operator has the option of ensuring only properly trained people go in. I'd have no problem doing an RA and specifying additional bonding (if not already present) for every item, plus training of all personnel who enter the facility.

But for some of the others mentioned (factories, bakeries) it's a lot harder. My limited experience in commercial environments is that people are happy to ignore any warning notices or instructions if they get in the way of doing their job. Plus people seem happy to carry on using equipment with obvious faults. For example, one time I was in the local hospital (visiting) I observed someone happily trundling along with a heated food trolly, where the cable gland was something like 6" away from the hole it was supposed to be fitted into, and presumably it was only the electrical connections of the cores keeping the cable in - i.e. it was a really obvious fault, visible from a distance, and not some obscure issue that could have been overlooked (I told the person trundling it that it was dangerous and needed referring to maintenance before it was used again - no idea if it was). I'm sure we've all got tales like that - especially cables pulled out of cord grips/glands. And the least said about some of the stuff we used to do back when I was a youngster working on a local farm - before I knew any better.
So I'd be "rather wary" of omitting RCD protection for some of these situations unless there was some sort of unique plug/socket arrangement that offered a bit more protection than a standard BS4343 or BS1363 type - hence my query about connectors with multiple earth pins.

 

[pc1966]

But it's still one contact, one cable core, one connection inside the equipment - so while it does mitigate one particular failure mode, it doesn't mitigate several others.

True.

Back to one of the earlier situations - telecoms facilities. I would suggest that these are relatively simple - there's options for bonding everything so if (say) a supply cord loses the CPC, there's still bonding so ADS will still work without anything exposing dangerous voltages. Plus it's a relatively controlled environment - the operator has the option of ensuring only properly trained people go in. I'd have no problem doing an RA and specifying additional bonding (if not already present) for every item, plus training of all personnel who enter the facility.

That sort of situation is what I normally deal with and most equipment has multiple earth paths via cables and metal racks, plus I normally bond the rack as well as the supply cable CPC so two paths at least.

But for some of the others mentioned (factories, bakeries) it's a lot harder. My limited experience in commercial environments is that people are happy to ignore any warning notices or instructions if they get in the way of doing their job. Plus people seem happy to carry on using equipment with obvious faults. For example, one time I was in the local hospital (visiting) I observed someone happily trundling along with a heated food trolly, where the cable gland was something like 6" away from the hole it was supposed to be fitted into, and presumably it was only the electrical connections of the cores keeping the cable in - i.e. it was a really obvious fault, visible from a distance, and not some obscure issue that could have been overlooked (I told the person trundling it that it was dangerous and needed referring to maintenance before it was used again - no idea if it was). I'm sure we've all got tales like that - especially cables pulled out of cord grips/glands. And the least said about some of the stuff we used to do back when I was a youngster working on a local farm - before I knew any better.

This might be the reason for the "electrically skilled" aspect not listed, so stuff like that is picked up by anyone with even basic PAT experience.

I have fixed PA amplifies with loose flexes and broken CPC before, many MANY years ago, and not far off the period when you would hear of musicians ding of electric shock on stage. Thankfully RCD use and radio mics/pickups have largely eliminated that, as well as the PAT aspect.

So I'd be "rather wary" of omitting RCD protection for some of these situations unless there was some sort of unique plug/socket arrangement that offered a bit more protection than a standard BS4343 or BS1363 type - hence my query about connectors with multiple earth pins.

The issue is then you need dual CPC to the equipment, etc, and that is very unlikely to be supported. You could tie some 4mm flex in parallel with the power cord and fit at any chassis screw, etc., but I can't see that as being acceptable in general. More generally, if equipment is seriously abused as you see on farms and building sites, both could be visibly damaged an nobody cares!

Now fixed connections that are permitted are not really different, but I guess they are used for items that hardly move so I'm guessing the argument is they are less likely to have damaged connectors, cable glands, etc, than portable-ish stuff.

 

[Simon47]

The issue is then you need dual CPC to the equipment, etc, and that is very unlikely to be supported. You could tie some 4mm flex in parallel with the power cord and fit at any chassis screw, etc., but I can't see that as being acceptable in general.

Replace the 3 (or 5) core cable with 4 (or 6) core, find an alternative internal connection point.
An alternative could be to use a pilot pin (as fitted on larger BS4343 connectors) and use that to monitor (in real time) integrity of the main CPC.

More generally, if equipment is seriously abused as you see on farms and building sites, both could be visibly damaged an nobody cares!

Absolutely. I do wonder how some of us survived - or perhaps it was wearing wellies that helped.

The aforementioned working on a farmer in my youth. We had an old water heater - freestanding "large metal bucket" with a lid to pour water in and a tap on the side. We noticed we got a tingle from it, and when I looked I noticed that the contacts on the earth pin in the 15A socket were splayed out. I squeezed them together so they made contact, plugged in, flicked the switch - and it went BANG. Turned it over, the heating elements were thin things, sandwiched between sheets of mica, and clamped onto the bottom of the "bucket" by flat metal plates. One had come loose, the element had moved and was touching the stud - hence the tingle when the earth was broken, and the bang when the earth was fixed. This was in the dairy - always wet floor, and steel floor tiles, so I guess it was our rubber wellies that saved us !

And then a very long extension lead - needed to get from the few sockets (I think there were only 2 or 3 altogether) to anywhere else. The rubber sheath was cracked, and there was nearly as much tape as showing sheath. But it was 2 core cable, and sticking out from the socket on the end was a coiled cable with a crock clip. None of us knew what it was for, so we tended to clip it onto the frame of the machine we were using - "what else could it be for ?" ! Again, I think luck was our friend here as it mostly got used in dry weather for powering an elevator for hay bales. At some point, it got chopped off as it was a nuisance.
Idea was sound, but since no-one knew why it was done that way, it never got used safely.

 

[Yellowhammer]

A good portion of work I do is on equipment with inherently large leakage from cabs full of VSDs. Recently went to a job up north for a trial of a new industrial manipulator for a very specific task. Arrived and the cabinet for the manipulator is in a shipping container, within an industrial work hall. Container has a CU with overall RCD supplying 2x 16A TPN sockets (of of which is designated for the control cab), lighting, 1363 sockets, 110v sockets and 16A sockets, fed via 32A lead with a plug at each end and a 32A RCCB at the up stream dist. panel.
As soon as cab is powered the RCDs go, indiscriminately because of a lack of selectivity between them.

I overcame the first issue by pulling out the container RCD, putting in a 100A switch, RCBOs for everything, and removing the TPN 16A sockets. Hardwired the cabinet in.

High integrity earth run in 10mm to the supply cab alongside the supply cable to ensure no metalwork rises to the potential in the case of loss of the cable earth, and a procedural process put in place to restrict access to the 32A socket up stream. (Labels stating use for this equipment only, and a requirement to lock the iso off when not in use.) Typed into a standard RA format stating the problem, likely risk, severity, mitigations and then what the risk is deemed to be after and it's done.

Could have hard wired the container straight to the panel, but as usual it requires agreement from a lot of interfaces, vs. just one interface who owns the 32A socket to get this approved and turned around quickly.

We have to do this a lot, due to the equipment. I see no real issue, look at an EIC, right at the start you are given the option to state deviations from 7671, and append an RA to the docs. It's all just about due diligence and proving you took the right steps to maintain the safety of the installation.