Poll: Compliance of Sockets Circuits

[ericmark]

I think the testing is a bit of a red herring, since there really is no way in which an electrician can undertaking any meaningful tests...

When we install a ring final we are told to consider 20 amp load in centre and 12 amp even spread, this is also unlikely to be true most of the time, but it is a valve used to decide ring final length.

So

Equipment having a protective conductor current exceeding 3.5 mA but not exceeding 10 mA, shall be either permanently connected to the fixed wiring of the installation without the use of a plug and socket-outlet or connected by means of a plug and socket-outlet complying with BS EN 60309-2.

this means items plugged in to a ring final have a limit of 3.5 mA. If we like when using 26 amp as the load for working out ring final volt drop have a figure which we consider as reasonable leakage we could come up with a acceptable formula to decide how many RCD's are used, we clearly should allow at least 3.5 mA per socket circuit and we should not exceed 9 mA with a 30 mA RCD so we can say 3 ring finals are the limit.

With items like cookers which are hard wired 10 mA seems to be limit, so cooker on it's own RCD.

So if we have a high integrity consumer unit with three RCBO's two for lights and one for cooker, and supply the rest from the two RCD's it seems unlikely we would have a problem. In fact if the lights are using RCBO's then all the safety arguments go out of the window.

Although the cooker is likely to have some earth leakage, what we have to consider is if to remove from general RCD loads likely to cause a trip or loads where there is a safety or financial implication should it fail, so lights and freezers on RCBO's and rest RCD.

The problem is often size, my old RCD's were 4 module wide, now down to 2 module wide but having a consumer unit with three RCD's a main switch and SPD means 10 ways are taken up before we include any MCB's. But with RCBO only 4 ways taken up. I have 14 RCBO's in an 18 way board, to go to RCD need a 24 way board. I have the room, but when you see some pictures submitted clearly room is a problem. Also although a high integrity board has three neutral bars, never tried fitting three RCD's so not sure you could?

And three RCD's is not two and the question was about using only two. But I look at my aunty's two up, two down house, and with a battery backed light at top of stairs, two RCD's would likely be good enough. She never had a fridge never mind a freezer, and still had gas lights on the wall. One bed room converted to bathroom, originally toilets outside at bottom of yard with door to back ally so muck man could collect the waste.

Just because I live in a 5 bedroom house does not mean everyone lives in a house that big.

 

[ericmark]

An RCBO is an RCD

No a RCBO contains a RCD it is a lot more.

 

[JohnW2]

No a RCBO contains a RCD it is a lot more.

... but that doesn't alter the fact that it IS a 'residual current device', even if it has additional functionality beyond that.

Kind Regards, John

 

[JohnW2]

When we install a ring final we are told to consider 20 amp load in centre and 12 amp even spread ...

Well, you once heard it said at a meeting. Although it sounds like a reasonable enough rule-of-thumb, I'm not sure that I have ever seen it formally stated anywhere. Nor do I see why it is specific to a ring final - did the person who said it also say that, for a 32A radial, one should consider it as being loaded with 20A at the end, with the remaining 12A evenly spread across the length of the cicuit?

In any event, I'm not sure what this has got to do with the matter we are discussing, which is primarily about earth leakage currents (which are not dependent of where loads are connected to a circuit).

So this means items plugged in to a ring final have a limit of 3.5 mA. If we like when using 26 amp as the load for working out ring final volt drop have a figure which we consider as reasonable leakage we could come up with a acceptable formula to decide how many RCD's are used ...

I don't understand that. As above, the leakage currents are nothing to do with the load current - a large load may have a tiny leakage, and a small load may have a very large leakage.

I would also remind you that the question is about 'sockets circuits' in general, not just ring finals.

... And three RCD's is not two and the question was about using only two.

It was, because what stimulated this poll was the suggestion that a dual-RCD CU is 'non-compliant'.

Kind Regards, John

 

[JohnW2]

A few questions for the person (we all know who he is ) who, so far, is the only one other than flameport who has voted "Yes" ...

1... Since you have answered "Yes" (and not "It depends") do I take it that you believe that a dual-RCD CU with three sockets circuits is always 'non-compliant'?

2... If that is the case, is this because you appear to be a believer in 20A radials, in which case three of them would probably not be enough, or would you still regard it as 'non-compliant' even if, say, 6 x 20A radials were arranged across two RCDs?

Kind Regards, John

 

[ericmark]

It seems there is some thing in 18th edition which is not in 17th edition which covers this question. Since I don't have 18th can't say what it is.

This one

says amendment 3 compliant, but this says

with the added inclusion of an SPD to be compliant with the 18th Edition Wiring Regulations BS7671 and comes complete with 10 outgoing MCBs.

• MCBs Supplied
  10 x MCBs (3 x 6A, 2 x 16A, 4 x 32A & 1 x 40A)
• RCDs Supplied
  1 x 63A 30mA
  1 x 80A 30mA RCDs (Type AC)
• SPDs Supplied - 40kA Surge Protection Device

I note no blank in that list, 11 way and 10 MCB's means it needs a blank. Also need to arrange so 63 amp not exceeded and 80 amp not exceeded if powered from 100 amp fuse, if powered from 60 amp fuse it would be OK. I am sure used in a garage this consumer unit would comply with the addition of one blank.

But would it comply if used as the sole distribution unit in a three story house, with a 100 amp DNO fuse? I think not. My last house had a Wylex fuse box and I had fitted a second Wylex fuse box which had been given to me, and when my son was studying for his RAE I decided it needed RCD protection so fitted two RCD's one for each fuse box, also swapped fuses for MCB's. I assume a 60 amp fuse, it's not marked, but the Wylex fuse box main switch was only rated 60 amp, and it was populated with 2 x 30, 1 x 15, and 1 x 5 amp fuses, so could exceed 60 amp, however very unlikely.

But I know every so often the RCD's would trip for no good reason, I assume spikes on the supply? And resetting one RCD would often trip the other one, again assume due to a spike, and I would need to switch off all MCB's then turn on RCD's then turn MCB's back on. Do remember these were old RCD's not electronic late 80's or earth 90's manufacturer.

I had considered either fitting a new consumer unit, or more modern RCD's, the X-Pole claimed to trip 90 - 100% not 50 - 100% and had a built in warning when getting near tripping limit, and there was also auto resetting, now no longer permitted. But I had to move in with mother until her death so all put on hold, and not long after her death moved here, so it never was ungraded, however the point is it would work without tripping for years, then a bout of tripping maybe 6 times in a month, and then years until next trip.

On moving here went all RCBO, but been here less than 3 years, and as said last house could go for years without tripping, so can't really say if better.

But also have a SPD fitted, if with last house it was spikes causing it to trip, then may not get the spikes since SPD fitted? I went to old house yesterday, and noted the array of GU10 lights fitted by my son had loads of failures, I don't get any LED failures, so again is it spikes on supply?

Twin RCD's are only a problem if they trip, but when we fit them how would we know if likely to trip, only time will tell. Both daughters have twin RCD's and neither complain about tripping.

 

[JohnW2]

.... Also need to arrange so 63 amp not exceeded and 80 amp not exceeded if powered from 100 amp fuse, if powered from 60 amp fuse it would be OK. I am sure used in a garage this consumer unit would comply with the addition of one blank. .... But would it comply if used as the sole distribution unit in a three story house, with a 100 amp DNO fuse? I think not.

It does seem odd that they haven't included two 80mA RCDs, given the trivial cost implications.

However, even if there were 2 x 80mA RCDs, I think there would still be a need for common sense (I suppose technically known as 'diversity') - otherwise essentially nothing would be 'compliant'.

It depends on what you mean by "exceeded" in "...so 63 amp not exceeded and 80 amp not exceeded...". Are you talking in relation to the 'ratings' of OPDs (i.e. In), of "likely" peak demand or of average "likely" demand"? - obviously three very different things.

A 'typical' (perhaps very common) arrangement for a domestic installation might have 3 x 32A sockets circuits, a 32A cooker circuit, a 40A or higher) shower circuit, a 16A immersion circuit and 2 x 6A lighting circuits. If you were talking about the first of those three interpretations of "not exceeded" I mention above, then, given that those circuits add up to 196 A, then the 'rated current' of at least one (probably both) of 2 x 80 A RCDs would inevitably be exceeded. However, if one went with the third interpretation, 2 x 80A RCDs (or, indeed, 2 x 63A ones) would be adequate.

Anyway, to trump all that (even if it were all RCBOs), there is the Main Switch. I've never seen a domestic CU with one rated above 100A, yet there must be very few ('main') domestic CUs in which the 'ratings' of the OPDs do not add up to more than that - 3 x 32A sockets, and nothing else, is obviously 96A - so, as I said, on that basis I imagine that virtually no domestic CUs would be 'compoliant'.

But also have a SPD fitted, if with last house it was spikes causing it to trip ...

how can you know, or even suspect, that? In 35+ years of living with residual current devices, I don't think there has been a single occasion on which I have had particularly good reason to suspect that they were caused by 'spikes' - apart from anything else, as I often say, in all those years I have hardlt any RCD/RCBO trips other than those which occurred "for good reason" (i.e. the device doing what it was meant to do).

As I've written a lot recently, at least as far as I am personally concerned (and there's no reason why I should be unique), we seem to be largely discussing "solutions looking for problems".

Kind Regards, John

 

[ericmark]

A consumer unit has a maximum of 125 amp, in real terms 100 amp, over that it is not a consumer unit. The DNO seem to use 60, 80 or 100 amp fuses, my house has 60 amp, likely since when I moved in there was an old Wylex fuse box which had a 60 amp isolator.

It does raise the question if there is a consumer unit with a 63 amp RCD and a 100 amp DNO fuse, who is in error, the fitter of the 63 amp RCD or the fitter of the DNO fuse? And unless the DNO have identified fuse size, then one must assume it is selected to match the consumer unit or fuse box so only if going down in rating is there a problem.

So if a 100 amp isolator with no RCD protection is replaced with two RCD's unless the fuse is marked as to rating they should be either higher than the sum of the MCB's or 100 amp.

As you say at one time we used diversity, but I understand that is no longer permitted for RCD size, not sure it ever was, we can use a 32 amp overload to feed 60 amp worth of appliances, but end of day worse case is the MCB trips or fuse ruptures. Only looking at nuisance tripping not overload.

As to spikes, without an oscilloscope one does not know if there are any, and in the main non electronic equipment is no damaged by spikes, only semiconductors seem to have a problem, which clearly includes LED lamps, my son has it seems had many failures with no SPD I have had none with a SPD so it seems that may be reason, but clearly not proved.

 

[JohnW2]

A consumer unit has a maximum of 125 amp, in real terms 100 amp, over that it is not a consumer unit.

Quite so. As I said, I don't think you will find one with a Main Switch 'rated' above 100A,despite the fact that the Ins of OPDs will very often add up to more than 100A (and often to more than 125A).

It does raise the question if there is a consumer unit with a 63 amp RCD and a 100 amp DNO fuse, who is in error, the fitter of the 63 amp RCD or the fitter of the DNO fuse?

Is anyone 'in error'? Where is the requirement that any sort of device has over-current protection - don't forget (the widely held belief) that OPDs (fuses/MCBs/RCBOs) are there to protect wiring, not to things connected to the wiring.

What you are saying is similar to what you often say about the ';rating' of accessories on lighting circuits - again, where is the requiremen that they should have external over-current protection?

And unless the DNO have identified fuse size, then one must assume it is selected to match the consumer unit or fuse box so only if going down in rating is there a problem.

I don't know what thinking there is behind the DNO fuse although, unusually, one imagines that it is probably there to protect their distribution cables upstream of the fuse. If (as per BS7671 practices) it were their to protect downstream cables, their only interest would presumably be in the 'tails' from cutout to meter - since they have no interest in (or control over) anything downstream of the meter.

So if a 100 amp isolator with no RCD protection is replaced with two RCD's unless the fuse is marked as to rating they should be either higher than the sum of the MCB's or 100 amp.

Is the 'isolator' you're talking about the 'Main Switch'? Whatever, as I've said, with two MCBs it would commonly be impossible to have 'rating' of the RCD greater than "the sum of the (Ins of) the MCBs".

As you say at one time we used diversity ...

The concept of diversity is implicit in all of this. If one assumed that all final circuits might be 'fully loaded' (to the In of their OPD) simultaneously, then it would probably become effectively impossible to design most domestic installations.

... but I understand that is no longer permitted for RCD size, not sure it ever was ...

Where does that come from? As I've said, I am unaware of any such explicit requirement, nor, indeed, any explicit requirement for RCDs to have external over-current protection.

As to spikes, without an oscilloscope one does not know if there are any, and in the main non electronic equipment is no damaged by spikes, only semiconductors seem to have a problem, which clearly includes LED lamps, my son has it seems had many failures with no SPD I have had none with a SPD so it seems that may be reason, but clearly not proved.

That's pure speculation, and one just can't know. I'm sure that things (as you say, primarily semiconductors) will occasionally be damaged by 'spikes', but I don't know whether that happens 'commonly' or 'once in a blue moon' - although I'm inclined to suspect the latter. I've certainly experienced very few unexpected/ unexplained failures of electronic equipment and, although I have countless LED lights, they don't seem to fail any more than one would expect from their claimed life expectancy.

So, again, at least for me, SPDs appear to be a solution looking for a problem.

Kind Regards, John