Over the last four days I have been experiencing random RCD trips on the mains supply.
We are fed by overhead line with an earth spike in the front garden.
The RCD is on the incoming line after the 100A fuse, meter and radio time-switch, but before an "house" wiring. It has no link to earth (so it's looking for a neutral/live imbalance). 30mA if I remember correctly.
As I cannot find any common cause in the house, so is it possible for neighbours (sharing the same pole) or even a cable fault on the "board" side to affect my RCD?
How regularly does it trip? Does it trip with all the circuits switched off? If so, you need to get a sparkie in for a periodic inspection. They will have the equipment to diagnose insulation resistance faults.
BTW, the whole house RCD should be a 100mA time delay type, with an additional 30mA RCD for socket circuits only (or RCBOs). 30mA whole house types are not a good idea for the reasons you've found (excessive frequency and consequences of spurious trips).
My error, I was doing it from memory. The whole house trip is 100mA, there are no other RCD trips on the house breakers. All other circuits are protected by Wylex mcbs.
There is no rhyme, nor reason to the trip. It was fine from 18:00 last night until 13:00 today (when it went three times in 15 minutes). It then went again about 15:00, then 17:40, then 18:15.
What was the local weather doing? At our church we have just such a system (TT earthing with 100mA incoming RCD) and it is guaranteed to trip if there is lightning in the vicinity. I'm not talking strikes a few yards away, I'm talking about last week (last Tuesday?) when we had thunder storms locally with about a 4 or 5 second gap between lightning and thunder clap.
In fact I'm working up a proposal for the church building committee that we look to upgrade to TN-C-S earthing which might alleviate these problems. Tripping didn't used to be a big issue, but now that we have a fire alarm, and the heating is on a timer...
If this is not a likely cause then you may well have an insulation resistance problem somewhere, and the best way (only sensible way) to find that is with the right kit.
that all appliances when unpluged, but turnned on messure at least 1M ohms beween the Live and Neutral to Earth. Beweare don't touch the test leads when testing, it will hurt!!
It is thunderstorm season and an RCD can't tell the difference between incoming and outgoing faults. If the leg current differs it simply switches things off. If the condition continues when the weather is fine then you have a problem somewhere.
You don't do an insulation resistance test on IT equipment during PAT testing unless it claims to conform to BS EN 60950, in which case it has been designed to withstand the 500V test.
Also, the 'pass' resistance for Class I IT equipment, if you do measure it, is 0.5Mohms (Class II IT equipment is 1Mohm).
Youve got no discrimination, so one fault anywhere, on any wiring or any appliance, shuts the whole lot off. This is why single whole house RCDs are generally a bad idea, although they are needed for TT installs.
The best option is to replace the socket circuit mcbs with RCBOs - you may need to get single width ones to fit in. Then any one fault will only kill that one circuit. And also then you have a means to work out which item is at fault, by unplugging and moving half the appliances on that ring to another ring, and narrowing it down like that.
IMHO it would be a better way to spend your money than a tester, because as well as finding the fualt it will also prevent such trouble in future.
This is a TT installation, so *every* circuit needs an RCD of some description. Yes, a 30mA fast trip is no longer recommended for overall protection, but an overall 100mA delayed followed by 30mA fast for sockets certainly is.
Your RCD on every circuit is also allowed, but at £30 a pop for an RCBO (assuming the OP has a CU with MCBs and an RCBO is a straight swap) you'll be looking at a *lot* of money to swap every MCB for an RCBO, and then there's the tediousness of trying to trace the fault which, as it is an intermittent one, could be a *very* lengthy procedure.
For something of the order of £60 to £100 (two or three RCBOs) a local electrician could bring his test kit down and will probably trace any fault reasonably easily, *if* there is a fault. As I and Peter Parry have already pointed out though, this kind of symptom may be related to the unusually active thunderstorms we've had lately.
Come to that, insulation testers aren't hugely expensive either these days and if that's all the OP wants he could pick one of those up for the price of a three or four RCBOs, but unless he knows how to use it, and has a use for it after this problem is solved (or can easily get rid of it), it doesn't seem to me like the best way to spend money.
That's not strictly what is meant by discrimination when talking about circuit breakers.
You can have discrimination with a whole house RCD by making the whole house RCD a 100mA time delayed one and having one or more 30mA RCDs 'downstream' from it protecting individual circuits. A fault on one of the circuits protected by one of the 30mA RCDs will then trip that RCD and not the 100mA time delayed one.
Absolutely, but IIRC the OP only has the one. I wonder if I misread... could well be.
Yes, I wound not recommend that at all. What I suggested is replacing only the socket ring mcbs. 3 of those would be around =A3100.
The simplest solution in 99% of cases is to use a multimeter. Once breakdown occurs the resulting path drops in R, so it can almost always be found with low v testing. =A35-10.
If I were the OP thats where I'd start, test all appliances with a =A310 meter. Its the quickest cheapest way to find the immediate problem - but the OP will still be left with a vulnerable installation, which is best sorted out.
Yes... but if the OP has only one RCD, its a fault thats bound to recur. Appliance leakage faults do happen, and it isnt a good idea to let every little minor fault take your whole house power out. In my mind that is as much a fault as the leaky appliance, and can be solved for around =A3100.
No, you read correctly. I was agreeing with you! What the OP described was a single 30mA RCD covering the whole installation. As I said, that configuration is no longer recommended (though it is allowed I believe) and the series configuration of 100mA TD - 30mA is the current favourite. Actually my personal favourite around here is an upgrade to TN-C-S which removes the need for an overall RCD at all.
And just how do you propose to square that with discrimination issues you raise elsewhere? If the OP keeps his current 30mA overall RCD then additional (i.e. series) ones on the sockets circuits aren't going to discriminate at all, and a fault on a sockets circuit is very likely to trip both RCDs.
In order to avoid this situation, the OP needs to do one of three things:
1: Swap the overall 30mA for a 100mA time delayed variety. Not vastly expensive (>£30) but as there probably isn't an isolating switch before this device he's either going to be in fuse-pulling territory (didn't strike me as the sort to appreciate that) or is going to have to call in an electrician.
2: Split the installation into two separately-fed groups. Group 1 fed via the existing 30mA RCD, group 2 bypassing this RCD and using individual RCDs/RCBOs on each circuit. In the OPs case this is almost certainly going to mean he has to buy a new CU. £60 upwards. Again, fuse-pulling territory.
3: Discard the existing RCD in favour of a 100A switch and then individually RCD/RCBO protect every single circuit. £30 per circuit + a few quid for the switch.
Granted the best place to start would be by unplugging appliances, but chances are that it isn't an appliance fault, especially one which is detectable by your cheap, extreme-low-voltage-low-current tester. If it were, the symptoms would be much more repeatable. I'm with the others here, it's either a problem which has now gone away and was due to the thunderstorms we've been having this last week, or it's a fixed wiring fault.
If the fault is fixed correctly it isn't going to recur. The problem of the RCD tripping may well recur, for other faults :-)
I reckon it's likely to be rather more than £100, but that's by the bye.
Some time ago I was called to a house where they had been having seemingly random RCD (overall 30mA) trips for some time (years IIRC) but which had recently excalated to "every time we turn the washing machine on". Very obviously the fault was with the washing machine, or with the circuit which fed it (it was a separate radial).
Not so; the fault was actually some old cable in the sockets ring which had been badly jointed when the fireplace was re-done. The insulation breakdown in this section of cable wasn't enough to cause a permanent fault, but enough to sensitise the RCD such that switching on the washing machine (or occasionally the vaccuum cleaner) would trip the RCD. I would never have found it without a device capable of pushing enough voltage through the cable. There had been several days of wet weather immediately prior to the fault becoming more-or-less permanent and I suspect that aforementioned joint, being near (or possibly inside) an unused chimney, was rather damp.
I couldn't retrieve the joint without dismantling the new fireplace so instead I isolated that section of the ring and ran a new section to the socket affected, simultaneously making-good a spur-from-spur problem that the B&Q fireplace installers had created.
Years of constant tripping had also damaged the RCD mechanics and it failed the RCD test quite spectacularly, so that had to be replaced too.
I found no problem with the radial feeding the washing machine, and as far as I'm aware there have been no RCD trips since.
Thank you all for you most knowlegable and useful contributions.
I do know what a Megger is, I can use it and it was looking as if I was going to have to buy one. It would have been cheaper to do that (even if it sat on a shelf for the next five years, or sell it on Ebay!) than call out an electrician (I don't know any tame electricans locally!)
However, through a number of repeated trips and a very patient wife, we narrowed it down to either the upstairs socket ring or the downstairs socket ring. Having tripped out the upstairs ring and the then RCD subsequently tripping about 10 minutes later meant it must be the downstairs ring. We're not positive, but since unplugging the kettle at 11:30 this morning, there have been no more trips (and no more tea!)
The wiring in the house, whilst not old, is not that modern. It has Wylex plug in mcbs, the sort that directly replaced the wired fuses. Looking at various catalogues, RCBO's only seem to be on a DIN rail fitting.
So, if I want to make a change how about a split load consumer unit.
I currently have:
40A Shower
32A Cooker
32A Socket ring - upstairs
32A Socket ring - downstairs
6A Lighting main - upstairs
6A Lighting main downstairs
If I 30mA RCD protect the first 4 and leave the lights unprotected (by the 30mA RCD) then any faults will at least leave the lights on. The other alternative is to put in a new consumer unit and use 32A RCBOs for the two socket rings, with no extra (other than the 100mA main RCD) protection.
Mysterious intermittent trips are also often caused by neutral to earth shorts, sometimes on circuits unrelated to the one apparently causing the fault.
Did you try a visual inspection on the sockets of the downstairs ring? A bit of looseness - maybe in the N-to-E department, so no fuse blows - can cause your symptoms (bit of resistance -> bit of heating under load
-> bit of touching -> pop).
Or, as we had in an intermittent-use community-owned property - meeces/ratsen chewen throughen cable... simplest was to rip out that circuit and relay with fresh cable!
HomeOwnersHub website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.