Intermitant RCD tripping

[supasparxs]

I had a customer call me out yesterday and here is the script.

Thety move into the house 3 weeks ago,

since they moved in the RCD on the high integrity CU (main switch with 5 MCB's and RCD with 5 MCB's)
has been tripping regularly. (upto 3 times in one day)
The most comon time the RCD has been tripping has been around 3am but it has been tripping during the day and evening.

I look around the premises for an outside light that mayb egetting damp and causing the tripping, however the outside light was on the lighting circuit (on the mainswitch side of the CU)

On the RCD side of the CU I found the following circuits
cooker 6mm cable 32amp MCB
on 500v range on insulation resistance test >500M ohms between Nuetral and earth

power radial
2.5mm cable 16amp
on 500v range on insulation resistance test >110M ohms between Nuetral and earth

power radial (2 radial circuits bunched on one MCB)
2.5mm cable 32amp
on 500v range on insulation resistance test >115M ohms between Nuetral and earth

Trip load on RCD was 27mA @ 7.6ms
this was a 30mA RCD

The customer had a new fridge delivered a bout the time they moved into the house, so I suspect that it maybe the fridges compressor tripping the RCD.
For test purposes I moved the Fridge onto a spare 6amp MCB on the non RCD side of the CU so that we eirther confirm a problem with the fridge or not.


Basically I think the Fridge compressor maybe tripping the RCD

I didn't have my PAT test equipment with me so when I return I will PAT test the fridge

any thorghts anyone?

 

[Guest123]

Hi.

Have you tried an RCD ramp test with all appliances plugged in, then removing 1 appliance at a time to see if there is a major difference when 1 particular item is removed???

Does seem srange for both socket radials showing a small breakdown at the same value??

 

[Deleted member 9648]

This sort of thing can be a nightmare. We got a call to an occasional tripping problem.....went through everything several times over a number of return visits and still the tripping continued. Customer getting more and more irate...(understandably).....to cut a long story short the fault was eventually traced to a capacitor on the printed circuit inside the boiler which had blown (boiler worked fine). The foil wrap inside the capacitor had uncoiled outside it's casing and occasionally touched an earthed part as the temperature inside the boiler casing rose! Nothing showed up on an insulation test and the fault was only traced by leaving circuits disconnected overnight to see if the tripping stopped....ramp tests etc didnt show anything and the rcd was changed twice just in case!
In the end there was an inevitable bill dispute as customer couldnt understand why it had taken so long to find the problem......

 

[supasparxs]

Hi.

Have you tried an RCD ramp test with all appliances plugged in, then removing 1 appliance at a time to see if there is a major difference when 1 particular item is removed???

Does seem srange for both socket radials showing a small breakdown at the same value??

The radings were not exactly the same, they were an improving value.....

as I'm writing this it reminds me of a drying pryo.....

for those who don't know a MICC cable that has a high reading can be dried out by cooking the cable.
if you have a high reading before you seal the pot, you can put 1000 volts down the cable to push the dampness out of the cable.

I think when I return I'll remove all the sockets from a circuit and test each leg. It may also be a biuld up of earth leakage, across all the circuits so putting one of the radials onto a RCBO so letting it have a separate 30mA to work at.

This sort of thing can be a nightmare. We got a call to an occasional tripping problem.....went through everything several times over a number of return visits and still the tripping continued. Customer getting more and more irate...(understandably).....to cut a long story short the fault was eventually traced to a capacitor on the printed circuit inside the boiler which had blown (boiler worked fine). The foil wrap inside the capacitor had uncoiled outside it's casing and occasionally touched an earthed part as the temperature inside the boiler casing rose! Nothing showed up on an insulation test and the fault was only traced by leaving circuits disconnected overnight to see if the tripping stopped....ramp tests etc didnt show anything and the rcd was changed twice just in case!
In the end there was an inevitable bill dispute as customer couldnt understand why it had taken so long to find the problem......

That one sounded like a nightmare.

I know RCD's either work or don't and they on the whole never go wrong (well at lease I've never heard of one going wrong)

I did explain to the customer that intermittent faults are extremely hard to find because they are not always there.

So what I was planing to do is move each radial onto the Main switch side of the CU, so that I can track down the rouge circuit. Once I have the faulty circuit take it from there so either find the actual fault or fit an RCBO.

I have explain to the customer that this is against the regs, however I feel for test purposes it would be a prudent cause of action. I did explain as they were "under supervision" it would be permitted. However the circuit would require returned to a RCD controlled state in the end.

 

[wildgoose]

Just a thought here about a technique that may help you - I've tried it in the past and it worked for me!!

Please beware that you need to be around when you do this and don't forget to connect things up again after - preferably carry out the testing when the house is empty!!

Disconnect the main earth wire, also eliminate any likely parallel paths from bonding earths.

Connect a milliammeter in series with the main earth connection- there may be some odd readings occuring even when the CCU is totally isolated from the supply, so suggest you null-out these spurious reading from the meter.
s
Now switch- in each fuse and measure individual earth leakage contributions. If you know which is the offending circuit, then activate the loads on that circiut one at a time.

I've had one installation where there was several contributors in this case you had a difference load causing the RCD to cutout when you switched them on in a different order - just concentrate on the major contributor otherwise it will send you loopy!!

Hope this helps

 

[wildgoose]

An electricains Nightmare!!
Advice: Open the CCU and find out if there are any circiuts showing low readings on the insulation test. Alway connect the L&N together, then test them with respect to earth - it's easier and you don't risk screwing up the electronics!!

Had a similar problem last week at a large social club which had a unexplained tripout during a big wedding reception. It caused chaos because the CCU had an RCD incomer, so it tripped both the kitchen and the bar (6 circuits in all). Tried really hard to find a problem but couldn't see anything, so converted to a standard incomer with separate RCBO's for every circuit (thankfully no crossed neutrals).

If it happens again, then at least only one breaker will trip and it will also be easier to track down because there are so many items plugged in (about 20 items in the bar area alone).

Hope this gives you some more ideas.

 

[daiplayer]

Hi
havent read all the above posts but i had the same problem recently.

Rcd was tripping regularly, buy showing no faults on insulation, rcd or any other tests.

I decided to perform a voltage log and discovered that the voltage was ranging from 195 -240 v,

I then phones the dno to which they attended the site within 2 hours and discovered a damaged overhead line. i later found out these fluxuating voltages can send the rcd haywire, and they havent had any trouble since.

 

[supasparxs]

I must be going do-lally, I can't remember writing this thread.

I traced the fault to a radial off the ring main which was beside the CU. (3 cables in the MCB)
so I rewired the leg bringing it into the ring and that sorted it.

a real pain, but it took 3 visits

it was a strange one

 

[ian.settle1]

Remember seeing something like this before it turned out to be a dog cocking its leg
that caused it to trip.

 

[La Poste]

I decided to perform a voltage log and discovered that the voltage was ranging from 195 -240 v.

Excuse my ignorance but how do you perform a voltage log?

Thanks.

 

[trev]

It's a function on some testers, I have it on my Metrel. In a nutshell you connect the tester to the circuit/s in question and it records the voltage over a period.You can usually set it to record at given intervals

 

[daiplayer]

Excuse my ignorance but how do you perform a voltage log?

Thanks.

Well on my metrel 61557 instaltest, it has ( as stated) as voltage logger.

I can set the voltage meter onto place i like , and set the meter to record the voltage @ any given time interval, eg. every 1 second, 5 seconds , 2 minutes, etc.

When stopped it will then display the max found, minimum or average.

Never used it before tbh but came in handy at this time, saved sitting and watching a volt meter :-D

 

[thomasn]

Hi nothing to do with your query just the comments on pyro. Pyro is self sealing and will only allow moisture for approx 100 mill, cut this off and you are in good cable. Also there is no point in testing pyro until both ends are terminated, after terminating, do ir test,if slightly low wait 24 hrs and retest you will find cable is up to scratch with no need to apply 1000 volts which doesnt produce any heat anyway. I've seen loads of guys over the years trying to condemn pyro or micc when ther is nothing wrong with it.

 

[Engineer54]

Hi nothing to do with your query just the comments on pyro. Pyro is self sealing and will only allow moisture for approx 100 mill, cut this off and you are in good cable. Also there is no point in testing pyro until both ends are terminated, after terminating, do ir test,if slightly low wait 24 hrs and retest you will find cable is up to scratch with no need to apply 1000 volts which doesnt produce any heat anyway. I've seen loads of guys over the years trying to condemn pyro or micc when ther is nothing wrong with it.

It'll go a bit further than 100mm mate, more like a good couple of feet. All depends how long the end(s) have been open without the compound seal on during storage. It takes a time mind, and the worst case is, when stored in high humidity or damp conditions. It was always prudent to use a bit of potting compound on the ends of stored coils of MICC.
As you say, very easy to chase the moisture out or even normal stripping back for terminations, especially at CU's and FA panel ends...

 

[thomasn]

I agree a coil that has been lying for a long time may have problems but a cable run that has been laid in for a cou;ple of weeks prior to termination will be ok as you will probably take more than 100 mil anyway when terminating.Don't take my word for it but if you ever have the oppotunity (and the time ) test a supposedly bad length of cable, terminate and retest, wait 24 hours and retest, I promise you, you will be suprised