RCD protected socket outlets in offices

[Chris Holmes]

Hi All,

I work for an NHS trust.

One of our buildings has just had a major refurbishment including a complete rewire….

We now have floor boxes feeding network snd power to most desks and the power sockets are all twins with a built in RCD.

A contractor who was fitting monitors to the desks was of the opinion that there was no advantage in having these. (And a disadvantage, as people will accidentally hit the test button and cut the power to the desks).

I wondered if maybe they are needed in order to disconnect quickly enough in certain fault conditions?

Or if by having 30ma RCDs at the desk, it it allowable to feed them with something with no / “less” protection (a slower “blowing” RCD or one rated at say 100ma).

Or are they over(non)kill?

Or something else?

[John Rumm]

That may be true - but a bit depends on what else has been done and why.

If they expect lots of electronic and IT kit to be powered, then that is likely to be a high leakage environment (lots of mains filters). That could cause nuisance tripping problems on the circuit. So moving the RCD protection to the individual outlets saves sharing the 30mA RCD trip "budget" between multiple users.

Also possible

Much depends on what is on the upstream supply. If there is no RCD protection elsewhere, then it will need to be at the socket. If there is, then also having it on the socket offers not benefit and some downsides since in the event of a fault condition you now have two devices that might trip - and which trips is not deterministic.

If they can meet the rules for the protection of the cable by other means (depth from surface, earthed screening, mechanical protection etc) to the socket, then they don't need RCD protection upstream.

If you have cascaded RCDs then the upstream one needs to be of a time delay type. Those are only suitable for infrastructure and fire protection - they don't offer individual shock protection. So commonly only used on TT style installs.

Can't really tell without knowing what has been done and why.

[Theo]

I would guess the reason is so that an RCD disconnection only kills that socket, not the whole circuit. So the whole office doesn't go out when the RCD trips. That could also enable the circuit installation to be more flexible (eg no need to maintain a separate lighting circuit that isn't affected by an RCD trip so that tripping doesn't take out the lights).

Or that there is enough equipment connected to the circuit that all the power supply leakage adds up to exceed 30mA, but not for individual sockets.

I could imagine the delta over non-RCD protected sockets isn't massive compared with the installation costs, so in a commercial installation the cost to do it might not be so much.

Another thought is that the Trust has rules over installation works that are intended for patient areas, and the rules also get applied to buildings that don't see patients.

Theo

[Ottavio Caruso]

Am 14/07/2023 um 12:31 schrieb Chris Holmes:

I tend to agree with the contractor, on conditions that all electrical devices plugged in are built on modern specifications.

If one of these devices is cheap Chinese junk with no quality control, there is still a theoretical advantage in sticking an RCD in every socket.

[Ottavio Caruso]

Am 14/07/2023 um 13:36 schrieb Ottavio Caruso:

I was told by one of my previous employers (Not sure if it is actually true) that you can import an electrical device from China, take it out of the package, put it in a different package, mark it as "Made in UK" and "quality controlled" even if it is not QC'd in UK, and that would be perfectly legal.

[Owain Lastname]

AFAIAA RCD protection at the socket on its own only doesn't meet IET Regulations because the British Standard for socket and FCU RCDs wasn't included in the list in the Regulations for compliant RCD protection (for socket circuits under 32A)

So the circuits would need to be RCD protected anyway.

BICBW

Owain

[Fredxx]

Is there a BS standard where the socket includes RCD protection?

Assuming the relevant standards are met, and the installation is equivalent to the IET regs, then there is nothing from the 'designer' of the system to noting a deviation, as allowed by the IET Wiring Regulations. That design would need to be held whoever is responsible for the building works.

BICBW too.

[Alan Lee]

Thats been remedied in Amd. 2. It was an oversight that they have put right.

Only if the cables are buried in walls at less than 50mm. From experience in such places, there will be no buried cable, it'll be ran on trays or trunking, then trunking down the walls.And all socket outlets must have RCD protection now, whether in homes , businesses or public buildings. There was a get out before, where a risk assessment could say RCDs are not required, that has gone, no excuse to not fit them now.

[Owain Lastname]

That will make things a *lot* easier for many small jobs which could otherwise have triggered a CU change.

Owain

[alan_m]

Is it just a standardised implementation for the whole of the trusts properties? Maybe an overkill in a office environment but perhaps a sensible arrangement in wards, examination rooms etc. where you don't want (life saving) equipment being turned off because something else has tripped the whole system.

[SteveW]

Probably good even for an office. Office equipment can have a high leakage due to mains filters, which can make an RCD feeding many prone to tripping.If it trips, many companies will require a check before putting it back on, which may mean calling someone from some distance. Having a load of PCs off and workers unable to work for an hour or two is expensive. A single or a couple of PCs being off likely means that the workers can move to desks unoccupied that day and carry on.

[Chris Holmes]

Thanks for all the suggestions.

Thinking about it, I don’t think any of the circuits had RCD protection prior to the refurb.

And having it just at the “end of the line” seems like a sensible way to go, and probably what has been done.

[Brian Gaff]

Overkill. If regular inspections of cables is done, there would be no issue. Brian

[NY]

I would say that there are advantages and disadvantages of having one RCD per desk.

disadvantage

- extra initial cost (RCD-capable sockets are likely to be more expensive than simple unprotected ones)

advantage

- assuming the desk RCDs are set to trip at a lower residual current and/or trip more quickly than any RCD that may cover a whole floor or building, any fault will probably trip the most local circuit possible, and not take out a whole floor/building

Probably a good idea to have the less-sensitive wide-area RCDs as a backstop, but with higher-sensitive RCDs as locally as you budget allows.

I cannot see any reason why a local RCD will give less protection or be more of a safety risk than a wide-area one.

[Scott]

Is an RCD needed for a ring main at all if all the sockets contain an RCD? I thought all RCDs were 30mA so would it not be a lottery which goes first?

[Scott]

My washing machine used to trip the RCD at the start of the spin cycle, which was by news if I was on my PC at the time. I don't know what inspection I (or anyone else) could have carried out. I got RCBOs for each circuit to limit and mitigate any problems.

[Theo]

The RCD at the consumer unit will protect against somebody drilling into the cable, which socket RCDs would not. This is why, in a domestic setting, you need an RCD at the meter box if the cable run to the consumer unit is long and buried, because this bit of cable is only protected by the main supply fuse.

Theo

[Scott]

Does this mean if there are two RCDs it is a matter of chance which one goes first, or will the one nearest the earth leakage be the first to go?

[John Rumm]

That is a flawed assumption on a number of levels... For shock protection need to have a trip limit of 30mA or less - so it does not matter where it is. If you cover too many outlets with one RCD you increase the risk of nuisance tripping.

Cascaded RCDs don't discriminate unless the upstream one is time delayed (and then it is not suitable for shock protection). In the event of a genuine RCD imbalance, the earth leakage will typically be many times the trip threshold, and hence any or all of a cascaded set of RCDs can trip. More is not better!

The only time the "less-sensitive" RCD makes any sense is in cases where the earth loop impedance is not low enough to clear a fault to earth using a fuse or MCB - so traditionally in properties with TT earthing.

A 100mA trip threshold RCD will allow a fatal shock to be administered without tripping. Its main purpose is for infrastructure protection and fire prevention.

A RCD on the end of a circuit can't protect someone receiving a shock upstream of the RCD. So The circuit would need to be either otherwise protected if it is to forgo the RCD at origin. (so by using earthed metallic protection on the cable, or by the cable being buried 50mm or more from the surface).

[John Rumm]

Yes.

No.

A RCD has no ability to limit earth leakage - all it can do is recognise a situation where there is leakage over a threshold and open the circuit. So if one sees a leakage over the threshold, then they will all see it.

There is a very marginal case where an upstream 30mA threshold RCD might discriminate with a 10mA threshold one, where the leakage is say 15mA - but that is going to be pretty unlikely as well as being an atypical install.

(lower trip threshold RCDs are usually used for protection of livestock, where their significantly wide spread foot spacing and lack of shoes makes them far more susceptible to shock injury caused by voltage gradients in the soil).