like for like replacement

[pushrod]

A MWC is used for additions and alterations which do not extend to the provision of a new circuit. A like for like replacement can be recorded just on a receipt or "This certificate may be also be used for replacent such as accessories or luminaires, but not for the replacement of distribution boards or similar items" [BS7671]

So when replacing a faulty RCD on a 16th edition style board what certificate, if any, do you issue and what testing do you do?

I suppose it hinges on is an RCD protecting half a dozen circuits an accessory or a similar item to a distribution board. Even issuing an MWC changes it from a half hour job to a possible major headache if you then take responsibility for insulation resistance testing all the circuits it is protecting. Or would you just do the rcd tests and say the IR tests were not applicable?? Interested to hear any views, cheers

 

[malcolmsanford]

If it was an exact like for like replacement I would do a MEIWC for this as that requires you to do IR tests, EFLI, Polarity, continuity and as it's a RCD replacement then you do the operating times also on the MEIWC.

Issuing a MEIWC I think always looks more professional to the client/customer. It only really takes a few minutes to do. On the EFLI I would open that old nest of vipers and use 1667 so need to check that if the device operates withing the stipulated times.

As it's non notifiable then you don't need to send it off to a scheme if your in one, and to be perectly honest if you never filled a MEIWC out, who really would know, and would it really leave the installation "less safe', of course not. IMO I think it looks more prefessional to issue paperwork.

 

[Guest123]

Straight swap of the above, like for like is an MEIWC for me.

 

[Bigboy247]

I think it is always best to complete some paperwork MWC for any work undertaken. As we are all aware people alter thing after we have done our bit and time breaks electrical items down. So if a problem were to arise in the future you have evidence that the work carried out is correct and tested. As you could end up with your word against theres if it went wrong and someone was injured. In this day of no win no fee always cover you back.

 

[pushrod]

Cheers for the replies, its my brother in law's house and it would be a freebie. Last time i was down there they mentioned they had random tripping so turned off all the protected mcbs except one and tested the rcd thru the sockets and found the rcd was up the spout - ramp testing was the only test that i could get anything sensible out of it and it showed that about 7mA was taking it out so obviously knackered. Want to do it professionally and give a mwc but am convinced that if i start insulation resistance testing all the circuits it is protecting that i will find all sorts of probs (the house is v old and has lots of history) which i don't really want to get into.(only there a couple of days over new year!)

In this circumstance what testing is a MWC actually asking for - IR for all circuits and Zs for all? From what i can remember there are 6 circuits being protected by this rcd. Obviously would do the rcd tests and do Ze, but the others? the IR testing has to miss out the new rcd anyway, and there is only space for one Zs test on the cert - would be using the BS7671 version.

 

[malcolmsanford]

Ahhhhhhhh the good ole frebbie. Well the IR for the circuits must be pretty sound if it's holding in the old RCD, I would personally, ahem. take a IR on the whole insulation 250v L/N to E and see what you get. If you find something really outrageous then you have to check it, after all it is family and you don't want to leave them with a poor installation. You can also note the IR test on the departures part of the form.

Zs will be 1667 end of it.

Now the really naughty buit, sod the cert fit the new RCD and forget it.

 

[Des 56]

I think with certification,we tend to get bogged down with a comparison with whats usual or set in stone and thats not achievable,because there is nothing to compare to, only examples of someone elses take on things

My veiw is that the replacement Rcd is manufactured to a Bs standard and factory set and you need to confirm its operation only noting the times of that operation
To do that you could include the result of the trip times and the Ze of the installation with 1667 ohms as an operating limit

You are not touching the outgoing circuits only dis connecting and reconnecting a bus bar or tail and your test could be continuity of those components only, with a description on the certificate of what it actually refers to



Now the really naughty buit, sod the cert fit the new RCD and forget it.

Now thats probably real life practice, and I believe, naughty but nice sound advise


Getting bogged down with an installation problem that we havn't touched would be avoided by not overstretching the basic requirement to certify what we have done

 

[Sintra]

Hi Phil,

Have to agree with Des on this one.

Test the rcd only at 1x & 5x and note the values and limitations on a MWC.
You haven't touched the outgoing circuits so I personally wouldn't get involved with them.

 

[WDMDL]

Pushrod, with nuisance tripping and finding that the RCD tripped at 7mA on your ramp test, you may find that the RCD isn't actually faulty... it is possible that 15-20mA is leaking away somewhere to start with.
They don't have a Hotpoint CTD series condenser dryer do they? (soggy fluff on heater insulation - common fault)

So yes, you need to IR test anyway, to determine if there is a problem, or it is a faulty RCD.
EFLI, I would record as N/A, stating WHY (if you only replaced the RCD)

Your MEIWC could end up being for something other than replacing an RCD.

Remember, IR has to be REALLY bad to trip an RCD... 0.01Megohms (10K Ohms) would give you 23mA

Simon.

 

[pushrod]

Pushrod, with nuisance tripping and finding that the RCD tripped at 7mA on your ramp test, you may find that the RCD isn't actually faulty... it is possible that 15-20mA is leaking away somewhere to start with.
Simon.

Thanks for the reply but it was just the rcd - when i tested it originally it was from a single socket on its own mcb with everything else turned off so i knew that the rcd was definitely faulty. Whether there is any other contributory faults as well i don't know , but it seems to be ok now (touch wood!)

 

[Sintra]

And that is the correct way to test an RCD ie. at source with loads disconnected.


Sent from my iPhone using Tapatalk