32 amp 3 phase socket, 300ma RCD

[timhoward]

I'm interested how you would all approach this one.
A 3 phase distribution board on a caravan site, under lock and key. Only warden and manager have access.
One of the final circuits is protected by a 32A 3ph MCB and it supplies a 32 amp TPN socket adjacent to the board, the socket having it's own integrated RCD.
It took me a minute or two to realise it was failing RCD x1 tests because it was 0.3A not 0.03A.

Talking to the manager, it's used to supply an additional portable shower block using an armoured cable, for a couple of events a year.
The shower block has an all RCBO consumer unit , and the original installer (who I've spoken to) deliberately went with 300ma to provide selectivity.
When I put the front back on, there was a sticker saying "Requires 30ma downstream RCD protection" under the flap that I'd not noticed.

While I understand the rationale, as it stands it doesn't meet 411.3.3 ii) and I'd consider it a C2.
If there was a risk assessment in place stating who has control of the socket, and how it will be used, I might feel differently, but I only have the verbal assurance of the warden of what he will and won't do.
Am I being too harsh?

I'm still considering whether the best remedy is:
-change it for a time delayed 30ma device
-write a risk assessment putting the declared arrangements in writing.
-other?

Comments welcome.

 

[Julie.]

Personally I would write a formal risk assessment , it has to be by or with the involvement of a skilled person (you an al.) And located with the EICR/EIC.

That would be compliant, whether you charge extra (probably) is up to you, but I would point to the exception to 411.3.3.ii as resulting in a C2 if not in place, so either remedial work in writing a risk assessment by/with a skilled person or do it up front by agreement.

 

[UNG]

The potential issue I see is how is the installation in the portable shower block tested and the operation of the RCBO's / RCD verified once it is plugged in

 

[timhoward]

The potential issue I see is how is the installation in the portable shower block tested and the operation of the RCBO's / RCD verified once it is plugged in

It's a good question. The further complexity is that one event uses their own portable unit, another event uses a larger one that is hired in.
I know their own unit is tested annually as I've been asked to test it and have seen the records.
The unit they hire in has it's own periodic testing, and rudimentary testing done by the supply company after delivery ( RCD test buttons and polarity check with plug in tester) and they hand over and say sign here....

I've just skimmed section 6, section 717 and GN8 out of interest and (surprisingly?) there doesn't seem to be a more rigorous test schedule suggested for mobile /transportable units. However annually doesn't seem unreasonable to me, on a par with caravan sites.

The proposed risk assessment could stipulate various control measures, including ensuring test buttons of downstream RCDs actually do something after equipment is connected.
But at the end of the day it would be on the duty holder to ensure any control measures are actually adhered to.

 

[loz2754]

change it for a time delayed 30ma device

Not readily available in this country (last time I looked), as they're not suitable for providing additional protection, and would not comply with the requirement for socket outlets to have 30mA RCD protection.

Regarding a risk assessment, the only valid reason for omitting an RCD where it would normally be required, is if it can be demonstrated that it is not needed. We can't omit the correct type of RCD just because of poor circuit design, or any other reason. Having a label that says this socket must not be used for any other purpose than the intended use may well not be enough to deter someone from plugging any other device into a socket that lacks the proper protection.

 

[timhoward]

Regarding a risk assessment, the only valid reason for omitting an RCD where it would normally be required, is if it can be demonstrated that it is not needed. We can't omit the correct type of RCD just because of poor circuit design, or any other reason

Thanks for your thoughts.
Out of interest, what would be a ‘good’ circuit design for this scenario, as ultimately mobile units arrive with a male socket on them. They tend to have RCD main switch or all RCBO boards these days.
The only alternative I can think of is hard wiring each unit movement?

My thinking RE risk assessment is if socket isolator is locked at all times (which it can be whether in use or not) then it is under control of management. If control measures specify armoured cable should be used to connect loads then impact protection (only real danger?) is met via 300ma RCD and failing that ADS via earthed swa.
If further signage highlight the socket is only safe if used as per RA then I think it’s about as safe as it can be.

Further discussion reveals there is normally a distribution unit the far end accepting the 3 phase female plug and breaking down into 16 amp circuits. This has RCDs too!

The only danger I can think of is someone decides to unplug the female 3 phase plug at the distro and try and use it for something else. The chances of having something unexpectedly needing a 3 phase 32 amp socket seen to be pretty low.

I’m not dismissing that the original design doesn’t meet the regs, after all it’s me that’s raised it in an EICR. But is there a vanilla compliant way to do it that doesn’t involve hard wiring the mobile units each time?

 

[mainline]

-change it for a time delayed 30ma device
Comments welcome.

You can't reliably discriminate between two 30mA devices on time alone.

A time delayed 30ma RCD would not provide a person with the necessary protection against electric shock.

 

[timhoward]

This would have all been fine under 17th edition!
The 32A sockets needing additional protection under 18th should have had an exemption for 3ph distribution circuits IMHO!

 

[UNG]

This is one of those situations where there is no real optimum solution as all the options have a conflict because of equipment, regs or people
Where people are involved distraction is likely to be the biggest problem and unless the designated persons fully understand their obligations at all times, the problems will occur when they are to busy or otherwise engaged and delegate to others who are not fully appraised of the requirements laid down to keep the general public safe
With regard to equipment and regs as the regs appear to become more onerous it is making it more difficult to install a fully compliant installation in some cases and this is highlighted in this thread

 

[timhoward]

With regard to equipment and regs as the regs appear to become more onerous it is making it more difficult to install a fully compliant installation in some cases and this is highlighted in this thread

Agreed!

I've just briefly explored if I can legitimately consider this under section 740, as a temporary supply to a structure.
This would require a 300ma RCD at origin (tick), and at first glance 30ma RCD protection for final circuits, and not specifically distribution circuits that come from a socket (740.415.1)

I guess the socket is still there when the structure is not, which scuppers this idea!

I also dreamed up another totally ridiculous way of complying without a RA.
If the socket were 63amp this doesn't need RCD protection in the first place.
There are then two possibilities:
1 - use a commercial 63amp to 32 amp adapter and everything else as before.
2 - use a new 63amp cable and a 63 amp distribution box
Clearly both options do not in any way improve upon what is already there and are completely daft.

The fact that if installed to 17th edition it would have had no code at all, and uprating the socket's capacity is compliant, taken together is starting to make me think that this should be a C3 not a C2.
A shame the EIC didn't have this as a departure as it isn't a very hard case to make.

 

[westward10]

I wouldn't apply a Code 2 knowing the circumstances.
Sometimes selectivity cannot be achieved with caravans being the most obvious example. Put a 30ma RCD on it.

 

[UNG]

I also dreamed up another totally ridiculous way of complying without a RA.
If the socket were 63amp this doesn't need RCD protection in the first place.
There are then two possibilities:
1 - use a commercial 63amp to 32 amp adapter and everything else as before.
2 - use a new 63amp cable and a 63 amp distribution box
Clearly both options do not in any way improve upon what is already there and are completely daft.

Does a 63A socket need to have a 63A supply
Could not use a 40A or 45A supply to the 63A socket might not need to change the cable then

 

[mainline]

Agreed!

I've just briefly explored if I can legitimately consider this under section 740, as a temporary supply to a structure.
This would require a 300ma RCD at origin (tick), and at first glance 30ma RCD protection for final circuits, and not specifically distribution circuits that come from a socket (740.415.1)

I guess the socket is still there when the structure is not, which scuppers this idea!

I also dreamed up another totally ridiculous way of complying without a RA.
If the socket were 63amp this doesn't need RCD protection in the first place.
There are then two possibilities:
1 - use a commercial 63amp to 32 amp adapter and everything else as before.
2 - use a new 63amp cable and a 63 amp distribution box
Clearly both options do not in any way improve upon what is already there and are completely daft.

The fact that if installed to 17th edition it would have had no code at all, and uprating the socket's capacity is compliant, taken together is starting to make me think that this should be a C3 not a C2.
A shame the EIC didn't have this as a departure as it isn't a very hard case to make.

I think the easiest and most compliant way would be to as @westward10 said to use 30ma rcd.

If the rcd is resettable at the socket outlet, then I can't see it being much of a hassle regarding selectivity.

 

[davesparks]

This would have all been fine under 17th edition!
The 32A sockets needing additional protection under 18th should have had an exemption for 3ph distribution circuits IMHO!

It's not a distribution circuit if it feeds a socket, it is a final circuit.

The fact it is usually used to feed a transportable unit doesn't change the type of circuit, or prevent it from being used for anything else in the future.

 

[davesparks]

I also dreamed up another totally ridiculous way of complying without a RA.
If the socket were 63amp this doesn't need RCD protection in the first place.

Not a ridiculous idea at all, it is how I have dealt with the request to remove an RCD from a 32A socket in the past.
Whoever is using the socket can then connect whatever distribution box they require, which should have the necessary MCBs and RCDs, in to the 63A socket then test and certify the temporary installation. As this is for events it will probably be a BS7909 certification.

1 - use a commercial 63amp to 32 amp adapter and everything else as before.

Any such commercial adaptor will include an MCB and RCD (unless custom ordered without the RCD)

The fact that if installed to 17th edition it would have had no code at all,

The code on an EICR wouldn't change regardless of when it was installed, an EICR is carried out to current standards and reports on everything that is non-compliant and affects safety.

 

[davesparks]

My thinking RE risk assessment is if socket isolator is locked at all times (which it can be whether in use or not) then it is under control of management.

Would it be under the control of management who are fully aware of the dangers and care about electrical safety and diligently ensure it is locked off after every use and the keys properly controlled?

Or would it be under the control of management who would just hang the key on a hook next to the isolator or if a 32 socket was needed for a welder to get a job done or a big water pump if something floods, would just hand over the keys and not ever check that it got locked again?

 

[timhoward]

Would it be under the control of management who are fully aware of the dangers and care about electrical safety and diligently ensure it is locked off after every use and the keys properly controlled?

Point taken and I agree this is the main issue - the level of risk is a function of the control measures which can be completely ignored immediately or in the future if management change.
In this case I believe the present management do care. To obtain keys for the 3 phase distro board that feeds all of this I had to go to the office and sign them out. I've had sensible discussions about other EICR findings and they took swift action with another situation that was dangerous.

So a few ways to skin this cat.
a) change to 30ma RCD. This leads to 3 x 30ma RCDs in a row, far from ideal, but complies. Only comeback on me is either accumulated leakage / nuisance tripping.
b) Write risk assessment leaving it alone. Complies. Safety dependent on people sticking to RA. I'm left exposed if my RA isn't up to scatch.
c) change to 63amp socket. Complies whatever the socket is used for.

I've appreciated all the input. Thanks.

 

[westward10]

You are over thinking this.
Choices, change to a 63A socket or fit a 32A socket which incorporates a 30ma RCD (leave the 300ma RCD in place).

 

[timhoward]

leave the 300ma RCD in place

300ma is integral to existing 32amp socket.

 

[westward10]

300ma is integral to existing 32amp socket.

Well just fit a 30ma RCD or 63A socket. Your job stops at the socket not what is connected from it with regard to the RCD.