RCD tripping times, 0.4 or <200 etc.....

[normanton]

I always been told that an 30m/A RCD should trip within 200m/s (x1) and 40m/s (x5). But something iv never taken much notice of before is on test sheet the maximum disconnection time of a circuit less than 32A is 0.4. So....which is it? within 0.4 or within 200m/s????

 

[spark 68]

200ms for General purpose RCD's to BS4293

and 300ms for General purpose RCD's to BSEN61008, or RCBO's to BSEN61009

@Idn

 

[normanton]

Ok i get that bit. Im just wondering where this 0.4 figure comes into it? :S

 

[drew35]

Anything longer than 0.4 under fault conditions and you soil yourself.

 

[spark 68]

0.4s = 400ms for TN disconnection times<= In 32A, and 5s for >32A In or distribution circuits.

for TT 0.2s and 1s respectively

An RCD/RBO is faster in both cases, 40ms (@5XIdn) = 0.04s

 

[normanton]

So the max disconnection time for a circuit <32A is 400m/s. Yet the RCD has to trip <300m/s? Are you saying an MCB's max d/connection time is <400m/s? If so there is no way to test the tripping time of an MCB as far as i know? :S

 

[spark 68]

Distribution circuits are 5s for TN and 1s for TT, 1s = 1000ms

 

[normanton]

Lets say iv got a circuit. Its TNCS and its on a 10A MCB and has a 63A, 30m/A RCD covering it. If i do an RCD test its should trip out in <300m/s yes?

 

[ackbarthestar]

Circuits supplied by a TN system, as it stands at present, should be designed to disconnect in 0.4sec. It is a comparatively modern idea to use RCDs/RCBOs on all circuits (Its very European you know !)

See 411.3.2

RCDs are used to provide additional protection, but can also be used as fault protection where the normal method is for some reason insufficient to give a disconnection time less than Table 41.1.

 

[normanton]

....and the pennies dropped lol Thanks very much :bucktooth:

 

[normanton]

Actually one more question. How can you test that a circuit disconnects in <400ms (0.4) if it has no RCD on it?

 

[spark 68]

If the RCD is for 30mA or less, and used for additional protection, then this is tested @ 5X Idn and should trip in less than 40ms, (I have missed out the 1/2 Idn,1X Idn and 0 and 180 degrees etc. for clarity), all other RCD's (100mA etc.. used for dist. circuits) are only tested at 1/2 and Idn, does this help ?

30mA or less RCD's are for additional protection of final circuits.

Actually one more question. How can you test that a circuit disconnects in <400ms (0.4) if it has no RCD on it?​

By calculation (max Zs etc.) and the tripping curves in the BGB, you have to just take this at face value, as we lowly sparks do not have the equipment to test this.

 

[ackbarthestar]

Lets say iv got a circuit. Its TNCS and its on a 10A MCB and has a 63A, 30m/A RCD covering it. If i do an RCD test its should trip out in <300m/s yes?

It will depend what the current flow is at the time of disconnection.

To meet the requirements of BS 7671 it must trip out at <=300ms at its IAN, and 40ms at 5 times its IAN

An earth fault current of minimal impedance covered by a 10A MCB will disconnect as fast as an RCD, probably quicker. Some manufacturers have data sheets showing a likely disconnection of 0.01s.

 

[normanton]

Thanks very much for your reply

 

[huntos33]

quick question as well re RCDs, testing a village hall with 30ma up front of everthing rcd trips on 1/2 test With message TRP on the tester fine at x1 28 ms, same as 1/2 on x5 trips with TRP message, checked the booklet with the tester says that message is leekage on the curcuit, but its the same for all 16 curcuits, TT system 194 oms RA sound like a faultly trip ? All insulation test are >500 .

 

[Engineer54]

quick question as well re RCDs, testing a village hall with 30ma up front of everthing rcd trips on 1/2 test With message TRP on the tester fine at x1 28 ms, same as 1/2 on x5 trips with TRP message, checked the booklet with the tester says that message is leekage on the curcuit, but its the same for all 16 curcuits, TT system 194 oms RA sound like a faultly trip ? All insulation test are >500 .

You should be testing at the load terminals of the RCD with the circuit/load cables disconnected. You won't see any natural leakage then... What do you expect with 16 circuits being monitored??

 

[spark 68]

I would do a ramp test and see what value it is actually tripping at with everything disconnected, it could be that on a single up front RCD that you have too much (normal) leakage, which is pushing it over the edge at 1/2Idn, and of course as you know, it should not trip at 1/2Idn.

 

[huntos33]

I would do a ramp test and see what value it is actually tripping at with everything disconnected, it could be that on a single up front RCD that you have too much (normal) leakage, which is pushing it over the edge at 1/2Idn, and of course as you know, it should not trip at 1/2Idn.

The ramp test came in at 18ms with all curcuits/loads which can effect readings off.

 

[spark 68]

The ramp test came in at 18ms with all curcuits/loads which can effect readings off

But what was the mA reading it tripped at ?

 

[huntos33]

But what was the mA reading it tripped at ?

18 ma on ramp sorry not ms, not come across this before hence my ignorance.