does this RCD need replacing

[Andystraughan]

hi everyone

a second opinion would be appreciated

was called to a house in February as RCD was intermittently tripping. some times 3 times a day sometimes none. during my initial look i discovered when the oven was ran for more then 5 minutes the RCD tripped. my inspection then found a faulty cooker element which was replaced and hey presto no problems any more.

until this morning. same problem.

RCD tripped twice on Monday didn't on Tuesday but did 3 times yesterday. hasn't tripped today either.

i went today for 2 hours today as it was all i could spare. my tests were as followed

starting with a circuit at a time

extension ringmain all meggered clear N - E 299 N - L 299 E - L 299
hot tub supply meggered clear N - E 299 N - L 299 E - L 299
shower meggered clear N - E 299 N - L 299 E - L 299
cooker meggered clear N - E 299 N - L 299 E - L 299

i didn't test 3 circuits as its a big --- 6 bedroom house and didn't want to go around unplugging everything - down ring main - up ring main - kitchen ring main.

i then decided to ramp test the RCD. i disconnected every earth and neutral except the extension sockets which i know are clear. also turned all mcbs off excluding extension(where my megger was plugged into) the RCD tripped at 17mA. the mains RCD is a 30 mA 80 amp crabtree starbreaker split load, i then tested the RCD on half times which tripped the RCD immediately. i tried numerous times. is this RCD faulty? i have more follow up tests to do but don't want to purchase a crabtree RCD without factoring in other causes

like i said i was limited to 2 hours so couldn't do thorough test until next week.

any advice would be helpful

thanks

 

[telectrix]

go back and test the RCD with the circuits dissed from it.if it still ramps at 17mA then it's probably faulty. if, on thre other hand, it ramps at say 22mA or over, then you have some leakage on the installation. a mA clamp meter would be useful here. -2

 

[Richard Burns]

I would say that if the RCD has been tripping repeatedly over a long period of time and reset each time then it is probably struggling.
However based on your test results the RCD is faulty, but a test, as Tel says, with no circuits connected would be the clincher.

 

[Andystraughan]

ok thanks i will try with everything disconnected next time. just assumed that as the ring main was clear i could use that to plug my megger into. wasn't 100% sure how to ramp test the RCD as it only has a neutral going through it with it being a starbreaker

thanks for the help

 

[Octopus]

With situations like this its best to do FULL rcd tests when investigating and do them each time you visit.

I had one like this a couple of years back and the inconsistent RCD results confirmed that the RCD needed changing - its not tripped since I replaced it.

 

[tomspark93]

On the flip side there could be a fault on one of the three socket circuits which you didn't get round to testing. I'd do them as well even if the rcd was faulty just for peace of mind if nothing else.

 

[Lucien Nunes]

wasn't 100% sure how to ramp test the RCD as it only has a neutral going through it with it being a starbreaker

All RCDs require both L & N, you just can't get to the L connections on these when fitted, as they plug into the busbar. Take a load circuit out of an MCB, connect the tester to the MCB terminal and switch only that MCB on.

 

[Skelly]

I agree with Telectrix, Dissed circuit and test RCD on it's own. That's the way you should test RCD's anyway and not the circuit.

 

[Risteard]

go back and test the RCD with the circuits dissed from it.if it still ramps at 17mA then it's probably faulty.

A 30mA RCD can legitimately operate from anywhere between >15mA and <=30mA. So 17mA is perfectly acceptable.

But I can't understand how the ramp test has shown 17mA and yet we are told 1/2 I delta n trips every time. Is the dial definitely set to 30mA and not, say, 100mA?

 

[Deleted member 9648]

hi everyone

a second opinion would be appreciated

was called to a house in February as RCD was intermittently tripping. some times 3 times a day sometimes none. during my initial look i discovered when the oven was ran for more then 5 minutes the RCD tripped. my inspection then found a faulty cooker element which was replaced and hey presto no problems any more.

until this morning. same problem.

RCD tripped twice on Monday didn't on Tuesday but did 3 times yesterday. hasn't tripped today either.

i went today for 2 hours today as it was all i could spare. my tests were as followed

starting with a circuit at a time

extension ringmain all meggered clear N - E 299 N - L 299 E - L 299
hot tub supply meggered clear N - E 299 N - L 299 E - L 299
shower meggered clear N - E 299 N - L 299 E - L 299
cooker meggered clear N - E 299 N - L 299 E - L 299

i didn't test 3 circuits as its a big --- 6 bedroom house and didn't want to go around unplugging everything - down ring main - up ring main - kitchen ring main.

i then decided to ramp test the RCD. i disconnected every earth and neutral except the extension sockets which i know are clear. also turned all mcbs off excluding extension(where my megger was plugged into) the RCD tripped at 17mA. the mains RCD is a 30 mA 80 amp crabtree starbreaker split load, i then tested the RCD on half times which tripped the RCD immediately. i tried numerous times. is this RCD faulty? i have more follow up tests to do but don't want to purchase a crabtree RCD without factoring in other causes

like i said i was limited to 2 hours so couldn't do thorough test until next week.

any advice would be helpful

thanks

Why do you want to unplug anything? You are probably looking for a low IR to earth, which is just as likely on a load rather than the circuit. As long as you don't test L-N (which is pointless if the RCD is tripping) and do a soft 250v test initially there is no chance of damaging anything, by disconnecting loads you are removing potential fault sources from your investigations!

 

[tomspark93]

Why do you want to unplug anything? You are probably looking for a low IR to earth, which is just as likely on a load rather than the circuit. As long as you don't test L-N (which is pointless if the RCD is tripping) and do a soft 250v test initially there is no chance of damaging anything, by disconnecting loads you are removing potential fault sources from
your investigations!

This is why I always to an initial IR test of L+N to earth before unplugging items and then do another one once equipment is unplugged. It helps determine if it's the fixed wiring or an appliance.

 

[Andystraughan]

i checked my store and had an older model crabtree star breaker RCCB. i went back to property to try it out. before i did i disconnected every circuit on RCD side and same test results. ramp test trips at 17mA and the RCD trips at half time on an RCD test. when i swapped over the RCD for the older one the results were as follows

RCD 1/2 - 1999
1 x 34.4
5 x 8.8

ramp test it tripped out at 27mA

i have ordered a replacement RCD so it will be sorted by next week

thanks for the help everyone

 

[Risteard]

ramp test trips at 17mA and the RCD trips at half time (sic) on an RCD test.

Once again, how is this possible? Are you sure you have the scale set to 30mA?

A half-current test only utilises 15mA, so if it isn't operating until 17mA then that doesn't make sense.

 

[Andystraughan]

yes tester was on 30 mA setting i made sure of that

 

[happysteve]

Once again, how is this possible? Are you sure you have the scale set to 30mA?

A half-current test only utilises 15mA, so if it isn't operating until 17mA then that doesn't make sense.

Mathematically it might not make sense, but it's probably related to the accuracy of the tester.

Mine quotes test current accuracy of -10% to 0% at 0.5I∆n, and 0% to +10% at I∆n and 5I∆n. That suggests that for the 0.5I∆n test, the test current applied by the tester should not be less than 15mA, and may be upto 16.5mA. However, it says that for the ramp test, it increments in steps of 3mA until the device trips. There could well be some ambiguity at where exactly these 3mA steps are. And no doubt, other testers are slightly different.