Nuisance RCD tripping - identifying the cause?

[gjscott75]

A question I'm sure the experienced electricians will have heard a zillion times... how do you (inexpensively) identify the cause of nuisance RCD tripping? I've spent the last few hours reading all over the 'net for answers and my head is now buzzing.

As briefly as possible, I know pretty much nil about electrics and I've had problems ever since moving into a new flat about 4 years ago with intermittent (maybe about 4 or 5 times a year) nuisance RCD tripping. Well, I say RCD, sometimes it's a specific MCB which trips instead (the one labelled 'sockets'). Very rarely, resetting the RCD immediately causes the 'sockets' MCB to trip - doesn't happen very often though - mostly just resetting the RCD gets things going again first time. At only 4 or 5 times a year I could just about live with it. Nothing in particular seemed to cause it to happen - it's not like it happens when a particular item is turned on or off. Sometimes it happens during the day, sometimes it happens during the middle of the night when nobody's awake, sometimes it happens when nobody's at home (even so, obviously, there are things which are running 24/7 whether there's anyone around or not - fridge freezer, UPS, server, mobile chargers, set-top boxes etc).

Today, however, things took a definite turn for the worse... the RCD tripped about a dozen times throughout the day. Resetting got things going again, but sometimes for only a couple of minutes, sometimes for a couple of hours. It did occur to me it might be related to the weather (100mph+ winds locally!), or that may just be coincidence, but it inspired me to try to get to the bottom of why this happens at all. Reading about it online it seems like it can become quite expensive to get an electrician in if it only happens as infrequently as it does (well, infrequent apart from today)... and there's no guarantee to ever get to the bottom of it. I'm making the assumption that the cause is something plugged into a socket (as it's only ever the 'sockets' MCB or RCD that trips). If I were to buy a plugin RCD and plug one item into it at a time, either waiting for the next trip or for the problem to disappear, would this be a valid test? Apart from the fact this could take years to narrow down to an exact item which causes it given that it (normally) happens so infrequently, I'm guessing the plugin RCD would either need to be rated less than 30mA or react faster than the one installed in the consumer unit (otherwise the one in the consumer unit might react first meaning that I'd never get the trip happening at the plugin version). So, I guess the question is, is this a valid test and does even such a plugin RCD exist? Or is there in fact a better way to do get to the bottom of this that isn't going to mean I'm keeping an electrician on the payroll for months as the problem will almost certainly never happen when s/he's here (it's not that I don't want to pay to get a pro in - I'm happy to do so, but I don't like the idea of the possibility of paying a pro and still never getting to the bottom of it before I run out of money).

Many thanks,
Graeme


Snapshot of my consumer unit

 

[PrenticeBoyofDerry]

In your position it's a process of elimination.
If as today it tripped 12 times, you should be removing all loads from the circuit and plugging back in one by one to see what activates the RCD.

But 5 times over a year it would be difficult to eliminate by that process.
Do you have any outside equipment, anything that is vulnerable to moisture?
Using a plug in RCD would only be useful on sockets that were not already RCD protected.
If the RCD is tripping quite quickly after reset, it could be something is getting warm, an insulation resistance fault!
It could be an advantage if you had some testing done on the RCD, to check that it was tripping in time and at the right value. It may also be worth going around all the sockets on the circuit that trips and doing a little investigation, for loose or damaged conductors within the accessories. Isolate circuit first though!

 

[TheGayOwl]

Graeme, you seem reasonably intelligent!

Soooo, a RCD monitors current that flows through the live-thru the appliance-and back thru the neutral. If a certain amount of this current goes AWOL then the RCD will trip. Ideally a RCD will 'think' there is 3 amps going out via live, so I want pretty much 3 amps coming back to me via the neutral. If there is an imbalance (of usually 30mA (millli amps) or more) then the RCD should usually trip. This 30mA of current are most likely to be going to earth.

Several appliances could be leaking a small amount of current each, that occasionally adds up to over 30mA thus causing the trip. Prime candidates for this are white goods; kettles, fridges, washers etc.

Has the server got an intentional functional earth? I do not know much in this area but some PC related hardware have to 'dump' some current to earth to prevent static build up due to hard drives etc.....or something like that!!

As the problem is intermittent then it may be a hassle to locate the definate source of your troubles. Try googling the servers tech spec to see if it does have a functional earth. If it does it may require a dedicated circuit, buts thats another story.

Good luck.

And follow PrenticeBoys advice, if you are getting regular tripping then you may find your problem pronto....ish !!

Others on here will know better.

 

[gjscott75]

Sadly, this was all going on whilst I was at work so it was my GF who was at home and running back and forth to reset the RCD otherwise I may have gone round unplugging everything on the circuit and then plugging everything back in one at a time - if it gets as bad as this again, I'll certainly try it (although like I say it's never, ever been this bad - 4 to 5 times a year is 'normal' for it, but never a dozen or more times in one day). Of course, it's all calmed down now - it's not tripped for about 5 hours now.

Don't have any outside equipment - the closest I've got is a server, UPS and some network kit in the loft (which is dry, I checked when I got home from work in case the high winds today loosened a few tiles and caused rain to get into the loft).

Thanks for the tip about the plugin RCD - I won't bother with that theory then

I did read somewhere earlier that UPS's, perversely, could be the cause of RCD tripping, so I've started my testing by unplugging the UPS and running the kit in the loft straight from the electrical socket rather than via the UPS. I'll see how long it lasts before tripping again (hopefully forever, and then I can get onto APC and give 'em an earful). If not then I guess I'll have to get a pro in.

Thanks for the advice.
Cheers,
Graeme

 

[Alarm]

I`ve seen a UPS do this but when it switched over, not in standby.

 

[ericmark]

Spikes on the line will cause cheap RCD's to trip. The winds are likely to cause some spikes so yes likely weather related.

As we start to use more expensive RCD's their ability to hold in with the odd spike increases. Also the less on each RCD the less residual leakage so the bigger the spike required to trip it.

So there are a number of ways to reduce tripping. RCBO's because there is less on each RCD tend to trip less than a RCD on a group of circuits. However it can also mean one is unaware it has tripped.

The Moeller X-Pole RCD is specially designed to both warn of impending tripping and reduce the tripping due to spikes.

This

Untitled

• ericmark
• 6 Dec 2008
• 34

RCD is designed to auto re-set. However at £400 each I think I can manage to manually re-set.

The electrician with his meters can test the RCD and measure the leakage but this is only really any good when there is a fault. There has been a move to install filters on the incoming lines to stop spikes but as yet still rather rare.

 

[gjscott75]

Graeme, you seem reasonably intelligent!

Don't believe a word of it - when it comes to electrics I don't have the first clue.

Soooo, a RCD monitors current that flows through the live-thru the appliance-and back thru the neutral. If a certain amount of this current goes AWOL then the RCD will trip. Ideally a RCD will 'think' there is 3 amps going out via live, so I want pretty much 3 amps coming back to me via the neutral. If there is an imbalance (of usually 30mA (millli amps) or more) then the RCD should usually trip. This 30mA of current are most likely to be going to earth.

Several appliances could be leaking a small amount of current each, that occasionally adds up to over 30mA thus causing the trip. Prime candidates for this are white goods; kettles, fridges, washers etc.

Has the server got an intentional functional earth? I do not know much in this area but some PC related hardware have to 'dump' some current to earth to prevent static build up due to hard drives etc.....or something like that!!

As the problem is intermittent then it may be a hassle to locate the definate source of your troubles. Try googling the servers tech spec to see if it does have a functional earth. If it does it may require a dedicated circuit, buts thats another story.

Good luck.

And follow PrenticeBoys advice, if you are getting regular tripping then you may find your problem pronto....ish !!

Others on here will know better.

Hmmm - I'll check out the server spec to see if it mentions anything about a functional earth (it's a Dell PowerEdge T610 - plugged into an APC SmartUPS 1500). I work with servers all day long at work, and indeed with my own at home, but I've never really considered that they have any special requirements for power... at least at work I've got proper electricians providing the power connections - which is probably just a well.

I did read that kitchen appliances are normal candidates as the cause of the behaviour I've seen - but then all this stuff is on a different circuits to the 'sockets' MCB (dunno that it's particularly legible in the screengrab). I was assuming that it was unlikely to be related to anything in the kitchen or the shower... as I said, if it's the RCD that trips, sometimes (and only sometimes) resetting it immediately causes the 'sockets' MCB to trip (but never the other MCB's protected by the RCD) - my logic is that this means it's more likely to be something plugged into a socket somewhere causing the problem, either that or I've got two problems - one causing the 'sockets' MCB to occasionally trip, and another causing the RCD to trip (this is, of course, based on my non-existent experience so could be entirely flawed logic).

It did actually cross my mind that dedicated circuit on the non-RCD side of the CU just for the server/network kit in the loft might be an idea (perhaps not a quick or particularly cheap idea, but a potential solution all the same). But then I thought that it might not be a good idea after all as there's possiby a higher than normal chance of the kit getting wet (if the roof were to leak, for example). I'll keep it in mind, though...

Thanks for the advice.
Cheers,
Graeme

 

[gjscott75]

Spikes on the line will cause cheap RCD's to trip. The winds are likely to cause some spikes so yes likely weather related.

As we start to use more expensive RCD's their ability to hold in with the odd spike increases. Also the less on each RCD the less residual leakage so the bigger the spike required to trip it.

So there are a number of ways to reduce tripping. RCBO's because there is less on each RCD tend to trip less than a RCD on a group of circuits. However it can also mean one is unaware it has tripped.

The Moeller X-Pole RCD is specially designed to both warn of impending tripping and reduce the tripping due to spikes.

This

Untitled

• ericmark
• 6 Dec 2008
• 34

RCD is designed to auto re-set. However at £400 each I think I can manage to manually re-set.

The electrician with his meters can test the RCD and measure the leakage but this is only really any good when there is a fault. There has been a move to install filters on the incoming lines to stop spikes but as yet still rather rare.

Ooof - 400 quid! Like you, I think I could live with manually resetting (well, as long as it never gets as bad again as it did today). Interestingly, many moons ago, I posted in these very forums about my UPS complaining about high line voltage. Scottish Power stuck their monitor on my feed and conceded that it was indeed a little high and tweaked as much as they could tweak (which brought it down a few V).

Looking at my UPS stats over the last 24 hours, it didn't record any particular spikes and ranged from around 245v - 254v, but it does seem like a bit more than a coincidence that I got around what would 'normally' be 3 years worth or RCD trips on the same day that the wind was tearing roofs off buildings around the city.

It was a while ago now, but the engineer that I spoke to way back when SP were investigating the high line voltage basically suggested that the local sub-station was pretty old and over-capacity and not to hold my breath to ever get a super-clean 230v feed any time soon - which is part of the reason for the UPS (to clean up a potentially dirty input and also avoid the server losing power without shutting down gracefully) but, as I mentioned further up, I did read a something on my trawling through the 'net earlier that UPSs can actually exacerbate RCD issues - so I'm running without the UPS now to see if it makes a difference.

Anyway, thanks for the advice.

Graeme

 

[TheGayOwl]

I did read somewhere earlier that UPS's, perversely, could be the cause of RCD tripping

It is perverse, but keeps them in business, so to speak!

Alarm also points to this as a possible cause....tho he mentions when swirtching over.

Your theory about when resetting RCD occasionally causing MCB to trip is probably linked I would say, others with more time and a more logical mindset (!) may offer their wisdom on this.

When it is having a moment where it is tripping fairly regularly try turning off the kitchen sockets MCB and see if it continues.

Is the loft area susceptible to moisture in any way. Is there ducting from bathroom extractors that may be leaking moist air.

I like Erics idea re spikes as you say it has been much more prevalent in the high winds of late.

It could also be a dodgy RCD and nothing appliance or server etc related. May be sage to employ a spark for an hour or two to test various appliances for excessive earth leakage (that could cause the RCD to trip) and to ascertain the sensitivity of the RCD.

Regards.

 

[EFLImpudence]

if it's the RCD that trips, sometimes (and only sometimes) resetting it immediately causes the 'sockets' MCB to trip (but never the other MCB's protected by the RCD) - my logic is that this means it's more likely to be something plugged into a socket somewhere causing the problem, either that or I've got two problems

If you do have two faults, it would be a huge coincidence that they would happen at the same time.

The RCD detects earth-leakage - from either Live or Neutral.
The MCB detects overload and short circuit.

So - leakage from Live to Earth would seem to be the cause, (Neutral to Earth cannot trip the MCB and from Live to Neutral won't trip the RCD).
Depending on the amount of leakage, this would sometimes trip the RCD first and sometimes trip the MCB.

That in itself may not help you find the fault but it must be an occasional short-circuit between Live and Earth.

I would suspect an appliance which moves - WM, pump etc. or a worn lead.

 

[andyt22]

Just come across this post and thought I'd mention that UPS units REALLY DO cause a lot of nuisance tripping. I have been using UPS units in my house for the past 15 years with no problems and currently have 12 of them. In July, as part of a kitchen refurbishment, I replaced the main consumer unit with a split-load unit with half the load on a RCD.

Since then we have had no end of nuisance tripping and one UPS in particular always tripped the RCD as soon as it was plugged in. It wasn't until I went round with a PAT tester that I discovered all the UPS units had neutral-earth leakage currents that varied between 1mA and 15 mA. It was the sum total of the leakages from all the UPS units that was the culprit.

This was fixed by installing dedicated circuits for the UPS fed from the non-RCD side of the consumer unit. Thought I'd mention it to save someone else tearing their hair out over RCD tripping...

Andy