RCD keeps tripping

[rossdargan]

I have the following electric box:

Photo 13-02-2012 08 38 37 by Ross Dargan, on Flickr

and the trip on the right hand side (the main switch) keeps tripping - sometimes once a month, sometimes once a day - it's very random, here are some of the most recent times (I have a UPS which is why the times are so accurate).

08/01/2012 07:15:51
19/01/2012 05:40:12
24/01/2012 01:22:52
31/01/2012 19:11:05
01/02/2012 04:06:39
05/02/2012 11:25:41
07/02/2012 00:23:14

As you can see there isn't really a pattern - and since then it's only tripped once more.

I have had an electrician in who has checked all the wiring and confirmed that it's fine, so he things it's probably an appliance now.

His only remaining advice is to unplug everything and try and figure out whats causing it - but given how it can go months without tripping this approach could take years.

I was wondering if I could instead use several off the shelf RCD socket (the type you would use to protect a lawn mower) to try and figure out the appliance causing the issue - I would attach every appliance that stays on all the time to an RCD and when one trips hopefully I will have identified the issue.

My question is will the above work, or will the RCD on the box be more sensitive? Is it more likely that the RCD on the box is broken - should I get that replaced first?

Any other suggestions would be really appreciated as this is starting to get expensive!

Ta

Ross

 

[holmslaw]

..

 

[rossdargan]

I'm a developer so spend a lot of time fixing (and I suppose writing) faults just like this A complete PITA!

If they will both fault thats fine as it will identify the faulting application, it's only going to be an issue if only the consumer box trips.

The boiler could be an issue to disconnect - I can turn it off at the wall but I know thats not what you mean.

The dishwasher I can, but the Washing machine is new (we had the fault before we had it) so I'm currently ruling that out.

We have some external lights that I have no idea what they are connected to. The sparky disconnected one of them (it appears that was adding a bit of "noise" to the line), but he didn't touch the other ones (annoyingly as I asked him to - but because he couldn't see a switch for them he assumed they are disconnected - Think I'll get him back for them).

I'll ask him about the insulation resistance test, I don't know what that means so I can't say. I could ask him to PAC test some things to, but he said it would be a waste of money.

 

[holmslaw]

..

 

[rossdargan]

Ah, I think he did do that - I will ask him to confirm however. I'm starting to suspect it could be the boiler, I might see if I can get that tested.

 

[riveralt]

If you are a developer and have a ups system in place it might be worthwhile considering having a separate RCBO driven or even non RCD board dedicated to you most important systems through designated sockets.

The normal test for a suspcious RCD would be a ramp test - this will give an indication at what mA range the device will trip. Although they are marked at 30mA - ramp tests will often show they trip at around 26mA.

Identification of the problem is a pain primarily because any RCD that has ramp test below 20mA will invariably cause you problems especially in your case because a single RCD will effectively sum up all the earth leakage on each of your circuits and trip when the ramp test figure is past.

Consequently if say your fridge leaks 5mA, your TV 5mA, your oven 10mA and your hot water system 15mA - you could have three appliances operational and not get a trip or have just two operational and the RCD trips.

As Holmslaw indicated water based appliances kettles, irons etc tend to cause most trips followed a close second by the element in electric ovens.

PAC test -Do you mean Portable appliance test.

 

[rossdargan]

lol, yes PAT not PAC. oops.

The server the UPS is protecting is not super critical, but it's running a lot of virtual machines and I need it to shut down cleanly. Designated sockets would be useful (the fridge is the biggest issue I have when I lose power), but that sound pretty expensive to get that installed!

So if I put an RCD on the kettle say, and the fault was with the TV is it possible that the RCD on the kettle would also trip? I never thought of the kettle, but I will keep that turned off. The oven normally is (but its wired into the wall - I guess this could still cause me issues?)

Another possibility is to get my consumer box changed - is it possible to get one so every circuit has it's own RCD - at least that way I could narrow down what's causing the issue - and the rest of the house wouldn't be effected.

 

[ban-all-sheds]

Yup - you use RCBOs instead of an RCD behind a bunch of MCBs.

They aren't cheap (but cheaper than they used to be)