When was it a requirement of 7671

[studentspark]

To RCD sockets

Am I correct in saying that until the 17th edition the only sockets that had to be RCD protected were ones thought likely to supply outdoor equipment
(pre June 2008)

 

[EFLImpudence]

Yes, you are correct.

 

[securespark]

Yes, that's right.

As far as I remember, RCD's first hit the market around 1974, during the 14th Edition.

Rarely, you'll find an RCD-incomer protected board from that period (usually Wylex metalclad in my experience).

More often you find a stand-alone RCD in the tails.

Around the mid-80's, in the 15th Edition, you find the big builders like Barratt fitted 2 RCD sockets, one in the hall near the front door and one near the back door (often in the kitchen). Sometimes, if it was an extravagent build, there was one in the garage as well.

This was seen at the time to comply.

Later of course, the "16th Edition" board came in, which to most electricians minds, does not strictly comply, but was seen by the manufacturers as a cost-effective way to comply with the regs.

It's a similar story with the "17th Edition" boards now.

To my mind, there's only one route to full compliance regarding RCD protection: RCBO's to all circuits that have buried cables and/or socket outlets <20A.

Unless you have a TT supply, that is....but that's another thread....

 

[JohnW2]

To my mind, there's only one route to full compliance regarding RCD protection: RCBO's to all circuits that have buried cables and/or socket outlets <20A. Unless you have a TT supply, that is....but that's another thread....

Unless I'm missing something, that 'other thread' (for TT supplies) only comes into play if you have some final circuits which are not protected by RCBOs or RCDs in the CU. If all final circuits are so protected, then there's no need for an additional RCD upstream of the CU, so the 'problem' of N-E faults not being cleared by a SP RCBO (which I presume is what you were thinking of) is not actually a problem. Or am I missing something?

Kind Regards, John.

 

[ericmark]

I would to some extent agree it is hard to say it will comply with just two RCD's. What it says in 314 is more a clarification than something new.
Every installation shall be divided into circuits, as necessary, to: (iv) reduce the possibility of unwanted tripping of RCDs due to excessive protective conductor currents produced by equipment in normal operation.

Which means if it does not trip without fault then it's OK, but if it does then there is something wrong with the installation. However until it's all up and running one will not know if it's going to trip. So if the user has loads of items all which leak slightly then it's the installers fault for not creating enough circuits.

There are so many different RCD's it now a problem working out what should be fitted. Some auto reset, some give a warning when near to tripping current, see here some with work with only sine waves others will work with simulated sine wave. Also passive and active.

To my mind after the RCD the neutral needs switching. So all double pole switching. But in most houses all lights are single pole. So get a neutral earth fault on a lamp and it is really hard to trace.

To get a company to fit X-Pole is not a problem. They understand the cost of down time and will pay the extra to keep the plant running. But in domestic it's another story.

First I saw were in mid 1970's in a large repair workshop. By the 1990's they were quite wide spread and there was a big move to get rid of the old voltage type. Although not required by the BS7671 they were required by many health and safety advisers and not fitting them could get one the wrong side of the HSE inspector. However often they were 100ma and 300ma not the 30ma used today.

 

[bernardgreen]

It is a bit ironic that in 1979 or 1980 a meter installer refused to fit the meter on our new PME builder's supply as there were two RCDs in the system. One for the ring final in the caravan we would be living in and one for supply to the site. In a phone call to his manager the explanation of the necessity to still have power and lights in the caravan after an electrical accident on site immediately over-ruled the installer's decision.

 

[ban-all-sheds]

Every installation shall be divided into circuits, as necessary, to: (iv) reduce the possibility of unwanted tripping of RCDs due to excessive protective conductor currents produced by equipment in normal operation.

Note that it says "reduce the possibility of", not "eliminate".

Which means if it does not trip without fault then it's OK, but if it does then there is something wrong with the installation. However until it's all up and running one will not know if it's going to trip. So if the user has loads of items all which leak slightly then it's the installers fault for not creating enough circuits.

No - I'd say it was the installers fault if he was unaware of the issue of leakage currents, unaware of the requirement in 314, didn't ask the client what quantity of leaky appliances he should assume.

 

[securespark]

To my mind, there's only one route to full compliance regarding RCD protection: RCBO's to all circuits that have buried cables and/or socket outlets <20A. Unless you have a TT supply, that is....but that's another thread....

Unless I'm missing something, that 'other thread' (for TT supplies) only comes into play if you have some final circuits which are not protected by RCBOs or RCDs in the CU. If all final circuits are so protected, then there's no need for an additional RCD upstream of the CU, so the 'problem' of N-E faults not being cleared by a SP RCBO (which I presume is what you were thinking of) is not actually a problem. Or am I missing something?

Kind Regards, John.

Hi John, the way I see it is that if all final circuits are protected by SP RCBO, then there is no chance of clearing an NE fault.

Also, I still think that RCD's and RCBO's to 61008 and 61009 do not comply, as the regs requires a 0.2s disconnection time on TT.

BS4293 devices allow compliance, as the maximum operation time for the RCD is 200ms.

 

[EFLImpudence]

the way I see it is that if all final circuits are protected by SP RCBO, then there is no chance of clearing an NE fault.

Also, I still think that RCD's and RCBO's to 61008 and 61009 do not comply, as the regs requires a 0.2s disconnection time on TT.

BS4293 devices allow compliance, as the maximum operation time for the RCD is 200ms.

I think you're getting a bit muddled up, secure.

 

[securespark]

Care to unmuddle me?

 

[JohnW2]

Hi John, the way I see it is that if all final circuits are protected by SP RCBO, then there is no chance of clearing an NE fault.

Obviously true, but I'm tempted to ask 'so what?'. Whilst we can all think of (very rare) hypothetical problems of having an uncleared N-E fault, so long as some circuits unprotectred by RCDs are allowed, that's always going to be a possibility.

[Also, I still think that RCD's and RCBO's to 61008 and 61009 do not comply, as the regs requires a 0.2s disconnection time on TT.

Indeed it does (presumably because 'touch voltages' will usually be higher than with TN), but I think they have that one 'covered' so as to make 61008/61009 OK for this purpose. One of the footnotes to Table 41.1 says that where compliance with 411.3.2.2 (required disconnection times) is provided by an RCD, "the disconnection times in accordance with Table 14.1 relate to prospective residual fault currents significantly higher than the rated residual operating current of the RCD". This presumably means that they accept that with a 61008/61009 device, it will take a residual current approaching 2*I&#916;n to achieve the required disconnection time (61008/61009 devices have a max disconnection time of 0.15s at 2*I&#916;n). Do you agree?

Of course, if one is prepared to install supplementary bonding, 411.3.2.6 appears to allow one to 'ignore' an inability to achieve the required disconnection times!

Kind Regards John