RCD Tripping times

[owls]

I've just had a long conversation with a NICEIC inspector regarding which RCD out of a 30mA and a 100mA would trip with a fault current first?

Now maybe i've just gone mental or something but everything i know points me towards the idea that the 30mA would trip first, however the guy from the NIC swore me blind that 100% of the time the 100mA would trip first.

Has anyone else got any experience of this? I'm quite prepared to accept i've gone mental and this guy was correct. It just doesn't make any sense to me.

Thanks

 

[Adam_151]

It could be either if the fault is over 100ma, very much a lottery depending on tolerances, etc.

It would probably be the 30ma; at 150ma the 30 is required to trip within 40ma, the 100 is allowed 200ms, in real life they are quite a bit faster, but there is still a bit of a difference between x1 and x5 times

Could quite easily be both though, I don't think I'd put any money on it though, and if I did it wouldn't be on the 100ma device

 

[JohnD]

As I recall, the reason the higher rated RCDs are, on average, slower is that the current builds up during the 50Hz cycle, for a moment it is nil, then build towards its peak within half a cycle, so, on average, the more sensitive RCDs reach their tripping current a fraction of a cycle quicker.

This is almost invisible inside the tripping period of the device unless you take lots of measurements.

 

[owls]

thank you, you've said exactly what i thought. This guy was so certain though that "100% of the time" the 100mA would go first i thought i was missing something.

i just stood there quoting regs at him about 5 times I delta N disconnection times and all that, he just kept saying yes yes but with experience you realise that the 100mA always goes first.
So with experience the laws of physics change or what?

thanks for the reply

 

[Lectrician]

On a really low impedance L to E fault, either could pop first in my opinion, and TBH, the MCB could chance it and go before the RCD's.

 

[owls]

i'm really pleased with the replies as they make total sense to me, i was worried i'd got a 7671 with missing pages or something!!!!

The guy was so certain (to the point of arguing) that he was right i started to doubt myself

 

[JohnD]

Well, i've got a 100mA whole-house (not time delayed) and some 30mA sockets. If you give me £10 I'll simulate a fault and see what happens. I won't do it for less because i have to go round resetting clocks and timers.

 

[owls]

Well, i've got a 100mA whole-house (not time delayed) and some 30mA sockets. If you give me £10 I'll simulate a fault and see what happens. I won't do it for less because i have to go round resetting clocks and timers.

how about a fiver?

 

[JohnD]

...£10 ... I won't do it for less...

 

[Spark123]

thank you, you've said exactly what i thought. This guy was so certain though that "100% of the time" the 100mA would go first i thought i was missing something.

i just stood there quoting regs at him about 5 times I delta N disconnection times and all that, he just kept saying yes yes but with experience you realise that the 100mA always goes first.
So with experience the laws of physics change or what?

thanks for the reply

He wasn't just saying that doing a 5I test on a 30mA device will almost always also trip an upstream (non time delayed) 100mA device?

 

[Adam_151]

On a really low impedance L to E fault, either could pop first in my opinion, and TBH, the MCB could chance it and go before the RCD's.

With LEC on this, a quick glance over the let through engery specs of modern breakers will tell you that a tripping time of sub 10ms is often acheived (even though the regs assume 100ms to be on the safe side)

 

[dingbat]

Owls, he's talking bolleaux. There is no body of evidence to support what he has said, but there is also no point in pishing him off. Just nod, say "yes, sir" and move on. You know what you know!