Joys of "Park" homes - Can an RCD be described as "Earth Leakage Circuit Breaker" ?

[Dartlec]

Seems to be my month for finding odd bodges by supposed professionals.

Installing an external socket in a shed today at a park home (non-mobile home) and find this label on the consumer unit:



and this behind the door:



It looks like the CU was installed at the factory and then shipped with the house to site, where someone installed it on a TT and left it like it.

A more recent electrician apparently put a bigger earth rod in (8 feet rather than the 2 feet - Ze of 22 Ohm so fairly good as earth rods go) and the gas people discovered a 'reverse polarity' that was fixed.

But (according to the client at least) at no point did anyone tell him that none of the lighting, Central Heating or Oven circuits were protected by anything other than MCBs with a Zs that would certainly not meet any disconnection times.

The label is just plain confusing though. I know in the 16th there was a regulation that socket outlets needed RCD protection, and that's fine. But it ignores the fact that I've never seen a TT system where some sort of upfront RCD wasn't required for all other circuits to meet disconnection requirements. I'd have expected them to have fitted an S type RCD as the main switch at least, but no doubt this saved them some pennies.

I've also never seen a modern RCD described as an Earth Leakage Circuit Breaker - I thought that only applied to the old voltage ones which were considered obselete in the 80s.

Currently putting something in writing to the client so they can brandish it to the site owner, who is apparently not known for his adherence to regulations, and will quote for a new board.

Anyone seen a label like this before?

I'm thinking I'll quote for a Fusebox RCBO board with SPD as they seem good at the price. Though even a Screwfix special dual RCD would be an improvement!

 

[pc1966]

With Zs of 22 ohm it would possibly not disconnect on a 6A MCB, even if it did it could take an hour! While the regulations don't demand a RCD in the TT case, it is practically impossible to meet the disconnection times (if at all) without one, as you say.

It is possible that for a while the RCD approach to measuring leakage was generally referred to as "ELCB" as that was the job it was doing, though it does not operate in the same manner as the old VOELCB devices.

 

[R-fur]

In the 14th edition they had ELCBs which could be voltage operated or current operated.
The current operated ones were the same as RCDs the voltage operated ones tripped when the earth conductors of the installation were more than 50V above earth.

 

[Dartlec]

In the 14th edition they had ELCBs which could be voltage operated or current operated.
The current operated ones were the same as RCDs the voltage operated ones tripped when the earth conductors of the installation were more than 50V above earth.

Ah, that might explain the notice mentioning a 30mA ELCB, which is not relevant to the voltage operated ones. Perhaps they just never updated the label (other then changing to 16th edition)

It's quite possible that the instructions for 'installing' the house originally said that it must be connected via an RCD and the site owner didn't bother. The feed to this house is currently on a 60A Wylex switched Fuse in the meter cupboard between pitches. If that had been installed as a Time Delayed RCD then I guess it would have complied under 16th.

I haven't done any calculations yet, but there appears to be no earth in the meter cupboard for the submain either, so the Wylex fuse is likely only providing L-N fault protection for the submain. The cable coming into the home I think is armoured, though naturally not terminated correctly. The meters are all supplied by the Site Owner himself (He buys it in and resells it apparently), so none of it is the DNO's responsiblity.

It may turn out to be a can of worms, but I'm interested in any one who has worked on well installed Park Homes to confirm what best practise is?

The current situation is:

Site owners cupboard: Wylex 60A Fuse (BS5419 isolator) -->
16mm SWA (I think) - about 15 meters to the home (Not terminated correctly in the home, although that may have been to ensure no earth was exported to it - even though I'm not sure there is one at the source of it anyway..) --->
Customer's board - RCD on sockets and "P.Heating" (unclear what that refers to as there is C heating on a separate feed) but no RCD on 2 lighting circuits, C heating or Oven

I'll be quoting for a new RCBO board for the home itself, probably with SPD. Or a dual RCD board at worst if price is a major factor.

I'd also think that the sub main isolation switch should be changed to a suitable RCD (S type or perhaps 100ma?) to ensure safety of the sub main (with suitable discrimination) so will be pointing out in writing to the customer that the site owner should be doing that or providing certification to show that their installation is acceptable.

 

[pc1966]

Check the Zs at the end of the sub-main, if it is low enough to meet 5s disconnection with that fuse it is fine.

That assumes it is properly glanded at the source end, of course! Otherwise a fault might just start a fire there.

If the supply Ze is too high for that to be practical then having a delay RCD feeding the sub-main is the obvious solution as you suggest. But that is still not an excuse for the sub-main armour not being soundly connected to whatever system earth is present (TN or TT).

 

[Dartlec]

Check the Zs at the end of the sub-main, if it is low enough to meet 5s disconnection with that fuse it is fine.

That assumes it is properly glanded at the source end, of course! Otherwise a fault might just start a fire there.

If the supply Ze is too high for that to be practical then having a delay RCD feeding the sub-main is the obvious solution as you suggest. But that is still not an excuse for the sub-main armour not being soundly connected to whatever system earth is present (TN or TT).

I thought the disconnection time for a TT installation was 1 second? Not sure if it's always been that or if it was 5 seconds in previous editions? Or is that assuming a supplier earth at the switch fuse?

The cable was 2 core 16mm and the armour was just taped up at the customer end - I didn't get into testing because of time, but will definitely investigate if/when I go back.

It's possible there is some sort of supplier earthing in that meter box that wasn't immediately visible - There was no obvious earth to the switched fuse, but the SWA wasn't terminated into it either, so there may have been a good termination with a suitable earth that was hidden. If I go back I'll test to confirm.

Given that apparently the gas man found a supply reverse polarity not that long ago (that was corrected quickly), I'm not sure I'm entirely confident it's as it should be though!

 

[pc1966]

I thought the disconnection time for a TT installation was 1 second? Not sure if it's always been that or if it was 5 seconds in previous editions? Or is that assuming a supplier earth

It is.

But if the sub-main is part of a TN supply then it is 5s.

If the supply is also TT then you wont get anywhere close to those times without an RCD anyway, and if they have an RCD on that you would disconnect everything in the event of the caravan tripping on the non-RCD circuits you mentioned. Quite possibly you would have tripped it testing Ze if using the default MFT approach!
[automerge]1598037095[/automerge]
OK, I think I have misunderstood the setup. So the home is on the TT supply (without upfront RCD, etc) and the sub-main you are talking about is from the home to the shed?

I was talking about the site owner being TN and the home on TT.

 

[Strima]

Can you get 30 mega amp RCDs?

 

[Dartlec]

OK, I think I have misunderstood the setup. So the home is on the TT supply (without upfront RCD, etc) and the sub-main you are talking about is from the home to the shed?

I was talking about the site owner being TN and the home on TT.

Sorry may not have described it clearly. The site owner's cupboard I saw supplies 8 homes on a site of 20, so probably isn't the DNO source. It might be TN at that point but I couldn't verify that visually and only had a brief time to look. The metal Wylex fuse switch isolating the property in question didn't appear to have an earth to it, though I didn't take it apart. The sub main under discussion is from there to the property, which is definitely TT.

I did the standard 2 wire (hi) loop test to get Ze at the home's CU, so if there were an RCD upstream somewhere it maybe would have (should have) tripped.

My guess is that both are TT and that maybe all the homes on the site are similarly unprotected, as they were all built and installed at about the same time...

It might get messy as to who is responsible for each bit too, unless the lease is clear. The physical home is owned by my customer, but they rent the foundations it sits on I gather.

 

[pc1966]

Ah, it gets even messier

All you can really do here is deal with your customer's safety and inform the site owner (in writing, of course) of the potentially serious issues facing the main supply arrangements.

If you fit a replacement CU, for example, that has proper protection for the high Zs TT system then they should be OK. While not strictly necessary if they have all-RCBO or dual RCD board, my own preference is still to have a 100mA or 300mA delay RCD up-front as then there is no single point of failure in the future that would prevent safe ADS on a fault.

That just leaves the sub-main SWA armour as a risk if it is not earthed in a manner that can result in the supply ADS working within the regulation time for shock protection. You or the customer would need to inquire about the responsibility for that.

I presume there is a billing meter? Where is it located?

Usually the demarcation of ownership would be at that point, with the site owner responsible for the supply to the meter and the meter itself, and the customer from the meter's output downstream.

 

[Dartlec]

Ah, it gets even messier

All you can really do here is deal with your customer's safety and inform the site owner (in writing, of course) of the potentially serious issues facing the main supply arrangements.

If you fit a replacement CU, for example, that has proper protection for the high Zs TT system then they should be OK. While not strictly necessary if they have all-RCBO or dual RCD board, my own preference is still to have a 100mA or 300mA delay RCD up-front as then there is no single point of failure in the future that would prevent safe ADS on a fault.

That just leaves the sub-main SWA armour as a risk if it is not earthed in a manner that can result in the supply ADS working within the regulation time for shock protection. You or the customer would need to inquire about the responsibility for that.

I presume there is a billing meter? Where is it located?

Usually the demarcation of ownership would be at that point, with the site owner responsible for the supply to the meter and the meter itself, and the customer from the meter's output downstream.

Yes I can see a can of worms but may leave that to others to worry about once I've made things clear in writing. I didn't do enough of a site survey of the supply to be sure either way whether it's perfectly OK or a complete lash up...

From what I gather the site owner has a single meter for the site from the DNO, but that is not something I saw. I'd assume it's a 3 phase incomer for a site that size - he then supplies his own meters for each home and charges them based on their usage. Presumably he does quite well out of it as he's just moved into an 8m pound house....

It may be that all the meters/isolators in the cupboard I saw were on one phase and that the other properties are split that way, but that would purely be speculation.

I've followed some of the issues that changing CU to metal brought up with TT installations, though I don't see them that often around here now. As I understand it the NICEIC consider a cable gland as acceptable to minimise the risk of incoming tails touching metal enclosures, though several of the manufacturers seem to provide or sell cable clamps internally too.

Do you install the Time delay RCD in a separate insulated enclosure before the board such as the Wylex one often used for isolators, or replace the main switch?

 

[pc1966]

I've followed some of the issues that changing CU to metal brought up with TT installations, though I don't see them that often around here now. As I understand it the NICEIC consider a cable gland as acceptable to minimise the risk of incoming tails touching metal enclosures, though several of the manufacturers seem to provide or sell cable clamps internally too.

Do you install the Time delay RCD in a separate insulated enclosure before the board such as the Wylex one often used for isolators, or replace the main switch?

I think the risk of an input to metal CU short in a properly installed CU is negligible:

• Using a gland avoid both the sharp edge of the case and keeps down any strain on the tails from external movement
• Using the 19-strand Doncaster tails are more flexible and less likely to strain the internal connections (than traditional 7-strand double sheathed cable)
• Checking the incomer terminal screws (and others) with a torque screwdriver as a final once-over after fixing stuff in and pull-testing the connection, just before replacing the cover, should avoid anything coming loose in the first place
• Having an up-front delay RCD means the busbar is already protected (unlikely to short to case anyway)

Others on here might have more inputs on this to take in to account!

 

[Dartlec]

I used the Doncaster tails on my most recent install, and I don't think I'll be using anything else now - they make the bends in the board so much easier!

 

[Robert2254]

Seems to be my month for finding odd bodges by supposed professionals.

Installing an external socket in a shed today at a park home (non-mobile home) and find this label on the consumer unit:

View attachment 60297

and this behind the door:

View attachment 60298

It looks like the CU was installed at the factory and then shipped with the house to site, where someone installed it on a TT and left it like it.

A more recent electrician apparently put a bigger earth rod in (8 feet rather than the 2 feet - Ze of 22 Ohm so fairly good as earth rods go) and the gas people discovered a 'reverse polarity' that was fixed.

But (according to the client at least) at no point did anyone tell him that none of the lighting, Central Heating or Oven circuits were protected by anything other than MCBs with a Zs that would certainly not meet any disconnection times.

The label is just plain confusing though. I know in the 16th there was a regulation that socket outlets needed RCD protection, and that's fine. But it ignores the fact that I've never seen a TT system where some sort of upfront RCD wasn't required for all other circuits to meet disconnection requirements. I'd have expected them to have fitted an S type RCD as the main switch at least, but no doubt this saved them some pennies.

I've also never seen a modern RCD described as an Earth Leakage Circuit Breaker - I thought that only applied to the old voltage ones which were considered obselete in the 80s.

Currently putting something in writing to the client so they can brandish it to the site owner, who is apparently not known for his adherence to regulations, and will quote for a new board.

Anyone seen a label like this before?

I'm thinking I'll quote for a Fusebox RCBO board with SPD as they seem good at the price. Though even a Screwfix special dual RCD would be an improvement!

The TT supply must have an RCD of 100mA Time Delay or a 300mA rcd. The circuits downstream must have an RCD of 30mA for Discrimination ( now called Selectivity )
The RCD is an Earth Leakage Device the older types now banned and obsolete were Voltage operated circuit breakers

 

[westward10]

The TT supply must have an RCD of 100mA Time Delay or a 300mA rcd. The circuits downstream must have an RCD of 30mA for Discrimination ( now called Selectivity )
The RCD is an Earth Leakage Device the older types now banned and obsolete were Voltage operated circuit breakers

Where does it state a TT earthing system must have rcd protection.

 

[Robert2254]

Where does it state a TT earthing system must have rcd protection.

in the Regs 411.5.2. Fault Protection. if your Earth Rod Impedance is low then you can rely on a overcurrent device but the Earth Rod changes all the time depending on the ground conditions, always err on the side of safety. ( my motto ) you do as you like my friend but I am devout RCD man on TT

 

[nicebutdim]

You previously stated

The TT supply must have an RCD of 100mA Time Delay or a 300mA rcd

Given the OP states they had an RA of 22Ohms, I see nothing in 411.5.2 that justifies the above statement.

Don't get me wrong - I'm not suggesting an upfront RCD shouldn't be fitted, but pointing out your suggestion that one 'must' be fitted is erroneous.

 

[Robert2254]

You previously stated



Given the OP states they had an RA of 22Ohms, I see nothing in 411.5.2 that justifies the above statement.

Don't get me wrong - I'm not suggesting an upfront RCD shouldn't be fitted, but pointing out your suggestion that one 'must' be fitted is erroneous.

If you had a high Rod impedance is what I am referring to. If you did have ( for instance 99 ohms )
or less and an overcurrent device would not suffice in protecting your install , how would you have fed a Ring and a Cooker Circuit ?
lets just say it was 20 ohms would you feed the supply through a 10A mcb?, knowing that as soon as you turn on the cooker or the ring had a kettle on or toaster or something similar it would trip. so.......RCD time delay as I already stated

 

[nicebutdim]

If you had a high Rod impedance is what I am referring to. If you did have ( for instance 99 ohms )
or less and an overcurrent device would not suffice in protecting your install , how would you have fed a Ring and a Cooker Circuit ?
lets just say it was 20 ohms would you feed the supply through a 10A mcb?, knowing that as soon as you turn on the cooker or the ring had a kettle on or toaster or something similar it would trip. so.......RCD time delay as I already stated

Perhaps it is a failure to understand on my part, but I'm struggling to see the point you're making here.

 

[Robert2254]

Perhaps it is a failure to understand on my part, but I'm struggling to see the point you're making here.

OK
1.Instead of an RCD you can use a MCB as FAULT PROTECTION "IF" the Rod Resistance is low enough to do so.
BUT supposing your Rod Resistance was 50 ohms? 230/50 =4.6A on Fault. This wouldn't be enough to trip a 5A Breaker or fuse.
Answer? Fit a Time Delay RCD Type S ( time delay 100mA ) as FAULT PROTECTION. Even if you used a 5a Breaker as soon as you turn on a cooker or use some other device above 5A load you will lose your supply.
Short Circuit protection ( L-N ) would be protected by the service fuse ( assume 63A or similar ) Fault Protection via the time delay RCD

2. you answered your own question on your original post.
22 ohms on the rod ( which will change in different seasons according to ground conditions )
230/22 = 10.45A.
Again I ask you, would you put a Low rating MCB on the supply as Fault Protection? NO
I would use a Time Delay or Type S RCD.
I hope this clarifies the matter.
I would fit the RCD to CYA ( cover your arse )
All the circuits in the board would be 30mA rcd protected. ( dual RCD or RCBO's )