RCD test 6 months

GN3 A2 2022 says that "RCDs are only rated for a finite number of operations, carrying out frequent optional tests can shorten the product's life"

it also says in bold "prior to these RCD tests, it is essential, for safety reasons, that the EFLI is tested to check the requirements are met"

So a combination of less testing, and an aversion to the shock risk to ordinary persons. Note that the metal lid of the CU is often being held up by the other hand whilst the test is performed, which seems bad design to me.

I was once informed by an industrial sparks that he`d been assured by "those who know" that an RCD need only work 6 times before it packed in to be compliant to standards. I reckoned it was a wholesaler myth of the +ve and -ve tests each of x 1/2, x1 and x5 mA testing.

lyledunn said:

I was asked why the change to six months. To be honest, I couldn’t answer with confidence. I did hear that folk might be more inclined to test the RCD at  the twice yearly clock change. Can anyone point me to an authoritative explanation? 

Quite simply, it's a national requirement only (not in HD 60364-5-51 or IEC 60364-5-51), and it's what was decided upon when the original quarterly requirement was looked into in preparing BS 7671:2018.

This was a question on a recent City and Guilds level 4 2396 written paper. The chief examiner’s report indicated that it was poorly answered and encouraged centres to research the reason for the elongated time period. 

So if you are aware of the reason Graham, I would appreciate if you would expand. 

lyledunn said:

This was a question on a recent City and Guilds level 4 2396 written paper. The chief examiner’s report indicated that it was poorly answered and encouraged centres to research the reason for the elongated time period. 

Interesting ... the "when the clocks change" has been cited in other places, but that's now fast becoming a rabbit-hole in itself with the advent of smart-phones and watches, etc. ... I can't remember the last time I changed a clock or watch manually, so I forgot to push my Test Buttons a couple of weeks back 

I think that's the best we can do in terms of "research" ... the reality is, it's never as simple as "just change the words".

I recall that JP asked the question at clock-change time a year (or 18/12) ago. Even if the clocks change themselves (even in the car!) there is at least something to remember.

I still have to change the CH controls, but I think that is all because I haven't used my video recorder for ages.

Naturally, the various wind-up clocks have to be altered as well.

Note that the metal lid of the CU is often being held up by the other hand whilst the test is performed, which seems bad design to me.

the saving grace is that the RCD self-test does not impress any voltage on the CPC or earthed metalwork over and above that already there. (being a Z-link from live out to neutral in or v/v)

Of course a real RCD tester will pull up the CPC towards live, but the time and the current are limited to safe, if painful, values, should a human being be  the main fault path for any reason.

An ELFI (loop) test may involve higher than 'safe' currents for sustained exposure, but I'd hope folk read the handbook and start out small before working up and stop tests if the earth appears disconnected.

The obvious engineering answer would be not to earth the lid and to use insulating hinges, but I suspect on the typical hinge, it sort of makes adventitious contact, but then the fingers probably mostly grip the paint, so the user is not well connected anyway.

I agree it does not sound great, but I suspect that in practice the risk is negligible

Mike.

mapj1 said:

the saving grace is that the RCD self-test does not impress any voltage on the CPC or earthed metalwork over and above that already there. (being a Z-link from live out to neutral in or v/v)

Of course a real RCD tester will pull up the CPC towards live, but the time and the current are limited to safe, if painful, values, should a human being be  the main fault path for any reason.

An ELFI (loop) test may involve higher than 'safe' currents for sustained exposure, but I'd hope folk read the handbook and start out small before working up and stop tests if the earth appears disconnected.

Unless the RCD is tested using the upstream/downstream method, which is the preferred method according to PD IEC/TR 62350, and now recognized in GN3 9th Ed (2022) as RCD test method 2 ... although in this case I can understand why the preference for RCD test method 1 (test to PE) has been recommended for so long in TN and TT systems, as in a given installation Method 2 will require a more diligent approach to risk assessment etc.

Interesting point,  - not one that had come to my attention. Personally I'm of the school of thought that testing the CPC is actually present & continuous , as well as the RCD functions is a good thing, at least with a time/ current limited test (< 30mA for ever, or reduced time with rising currents as per the shock curves.), so "method one" for me. ;-)

However I also appreciate that where an RCD supplies a system with no cpc to provide shock protection - not perhaps in the UK at least officially, but certainly OK in Italy at one point, and very  common in Latin America today,  then another reliable method to assure RCD function without giving someone a shock from things that are doubtfully earthed  is essential. Z method (method 2 ) accomplishes this neatly.

Mike

mapj1 said:

Interesting point,  - not one that had come to my attention. Personally I'm of the school of thought that testing the CPC is actually present & continuous ,

To conduct the test without checking the cpc in the first place can be dangerous ... and GN3 recommends a loop test is done first [as well as the continuity and, if fitted, earth electrode resistance, tests are required by BS 7671 before energization] ... so how many times does a cpc need checking?

It's really a moot point for many 30 mA RCDs, because the fault path for "additional protection" is not necessarily back down the cpc as with fault protection in ADS - the example of the severed lawnmower flex is often used, and in that case there is no cpc.

mapj1 said:

However I also appreciate that where an RCD supplies a system with no cpc to provide shock protection - not perhaps in the UK at least officially,

Don't forget RCDs can be used in IT systems as well.