RCBO timing too high

[Richard Burns]

I normally find that RCDs pass both tests, however there are a few that fail (which is worrying since if they can fail from new when might they fail in use?).
I have just had a 30mA RCD test OK on a new install and then when it was actually under use it failed to reset even when the fault was clear.
I find that the trip times are much closer to the regulations limits with the modern RCDs with their electronic control, whereas I find older RCDs trip much faster but at the same time for both 1x and 5x.

If a replacement device fails on testing then I would suspect that there was a problem with the attached circuits and ensure that there was only the incomer attached and test then, if it still fails then it was two faulty units in sequence.

 

[Hawkmoon]

I said it could be considered to be illegal

no okay it isn't, sorry

fair enough

Respect to you - manning up to apologise/admit error is woefully rare on this forum

 

[southsea1201]

@ Richard Burns...Agree with you..I have never had a 61008 fail the test. Only 1 modern one failed for me and that was the RCBO I alluded to in this thread. It is interesting that 4293's always seem to fail for me...has anybody else found problems with them?


It's very interesting to note in the OSG (as much an advisory document as BGB) that when testing 4293's no advice is given to do a 5 times test...it only advises this for for 61008 and 61009. Again I could use this document in my defence.


Also 100% agree about testing at the board with circuits disconnected. On my test sheets I number my circuits as usual (say 1-10) but I have seperate lines headed RCD Module 1 and RCD Module 2 and treat these entirely seperately.

 

[southsea1201]

Sorry if already said and I have missed it.


Isn't the fundamental flaw in southsea's argument that RCDs limit the fault current?

They don't so the time IS important

But Geoff say you were in a "can't let go situation" and your body resistance at that moment in time would allow 22mA through you....time is a factor in electrocution...I'm sure multiple seconds at this very low current would kill you and the RCD wouldn't trip. When I got shocked by an RCD protected circuit I had to actually pull the wire off my finger with my other hand..I reckon I was stuck to it from anything between half and 1 second without the RCD tripping. It could have been longer...it is so difficult to perceive time when you are being shocked.

 

[Geoffsd]

I was referring to your accepting more than 40ms for >150mA and implying that RCDs limited the current (shock).

In your example above re: 22mA, (10 - 11kΩ ?), you are correct in that the RCD would not operate.
RCDs aren't a magic solution to everything.

 

[jamie-spark]

Hi All
just done a c/u change fitted with 8 RCBO's and when testing the trip times 2 of the RCBO times were too high (40a 30ma =40.8 ms . 32a 30ma = >40 ms )on X5 ? changed the 2 x trips and retest exactly the same times ?! disconnect the circuits and retest trip times perfect, ok so retest circuits all ok ??! phone control gear and speak to tech it could be the cooker hood (double insulated ) or the w/machine, they don't like them. in my case it was the (40a) electronic shower (32a) combi boiler .
just letting you know
John

I asked same the IET same question years ago when I was silly and here is the reply I got. Hope it helps

 

[jamie-spark]

Hi Jamie

Please see the reply from Hager technical manager I received earlier. I recall a megger engineer on the national committee mentioned something similar about the meters being affected by this when loads were in circuit

Hi Richard,

Quick reply as I am in and out of company meetings!!

I hope the following answers the questions. Out of interest, please can you tell me the type of connected load.

I am assuming that your comment "it fails to trip at 150ma (x5) within 40 mS" means it trips, but appears to exceed the maximum time?

This is a well known phenomenon attributed to the load (capacitance) influencing the RCD tester.

This is why the correct RCD test procedure prescribed in GNs states:"The load supplied should be disconnected during the test".

The ESC are currently investigating this, please see attached, page 15 "Influence of connected loads".

 

[jamie-spark]

Google
Switched on magazine