Why is the RCD tripping?

[HappyHippyDad]

I have just popped out to a frustrated customer who has had intermittent tripping of a 30mA RCD for years, on average 5 - 10 x p/month. It used to be 10 x p/month a year ago, now it is around 5 times p/month. This is a dual RCD board (Harok??) with 2 socket circuits and 2 lighting circuits on the affected side. There is alot of electronic equipment plugged in.

I did a x5 and x1 test (at a socket). The RCD tripped but the tester just came up with 'trp' meaning it tripped too quickly for a reading. The ramp test did the same. I then tested at the RCD and it tested as 24mA and 19ms. The customer was adamant the RCD was 'too sensitive', but the above tests seem to show it is not (although 19ms is a little quick).

Global IR results showed 1.7Mohm N-E.

I clamped the Main earth (TNCS). It jumped up to 26mA for <1s and then stayed at around 3mA. It did this every time I tested. It is a poor quality earth clamp meter. He had switched off most of his electronic equipment when I tested, but not unplugged, therefore N still attached.

He just wants me to change the board, and throw money at me as he is so fed up. However, that's likely not going to solve the problem, hence having a chat with you guys.

Would an RCBO board help? I.e if it is a build up of earth leakage this option would help. It wouldn't help if all the earth leakage was just on one circuit though.

Do you think it is natural earth leakage from electronic equipment? If so, what is the solution? I'm not too sure how to test each item of equipment for earth leakage.

Perhaps it is something else. Is 19mS too quick to trip (I think it is on the low side, but not abnormally low)?

Perhaps an intermittent fault which didn't show up when I tested. However, he said it had tripped just before I came and tested.

My instinct is on accumulated earth leakage, but again I'm unsure of the solution or testing for this.

Cheers all ?

 

[telectrix]

RCBO board. i suspect it's a combination of accumulated leakage. the 19mA trip will probably come better with no circuits connected.

 

[pc1966]

As @telectrix said, go for an RCBO board. It will probably fix it for good if it is related to accumulated leakage. Even if it is not fixing an underlying problem, at least you will know which of the 4 circuits is responsible which would speed fault investigation, and the annoyance value will be significantly less.

Also gives them an SPD, and while many are sceptical of their benefits I certainly think we will see more issues of cheap & nasty consumer electronics failing on smaller surges that simple or well-designed electronics would laugh off.

Something like the Fusebox RCBO board with SPD is not that expensive, total cost would be dominated by your work.

 

[buzzlightyear]

go for an RCBO board. It will probably fix it for good if it is related to accumulated leakage.

Is all well swapping a board but will not solve the problem with out being fixed.
Me find the fault before swapping.

 

[Mikegh]

Is all well swapping a board but will not solve the problem with out being fixed.
Me find the fault before swapping.

In fairness swoping out for rcbos is a method of solving nuisance trips as you're likely reducing the leakage on each device

 

[Mikegh]

Or at least it can be part of the solution

 

[7029 dave]

I would IR test each circuit individually to see each reading per circuit.

 

[timhoward]

Also maybe ramp test at the RCD with one circuit off at a time? See if the "trp" is only with one circuit OFF.
As above, if the customer wants a new board then an RCBO board is a step in the right direction anyway. A Fusebox board would be ideal.

 

[loz2754]

Testing for cumulative leakage is best done by clamping the line and neutral tails together, not the main earthing conductor, as there can be all sorts of stray currents flowing in the main earth.

 

[Mikegh]

Testing for cumulative leakage is best done by clamping the line and neutral tails together, not the main earthing conductor, as there can be all sorts of stray currents flowing in the main earth.

Is that accurate when you clamp the 2 together, I presume it is

Obviously main earth's may have leakage from non RCD protected circuits so isn't likely accurate

 

[loz2754]

Yes it's basically simulating what the RCD senses - an imbalance between line and neutral. Probably even better to clamp the cables of the RCD in question if it's a split load board.

 

[UNG]

I clamped the Main earth (TNCS). It jumped up to 26mA for <1s and then stayed at around 3mA. It did this every time I tested. It is a poor quality earth clamp meter. He had switched off most of his electronic equipment when I tested, but not unplugged, therefore N still attached.

Clamping the main earth is not an accurate method of assessing the earth leakage you need to clamp the live and neutral feeding the RCD together and any inbalance will be read by the clampmeter as the true earth leakage as this will include any parallel earth paths

 

[HappyHippyDad]

Many thanks for all the replies.
I've gone with the RCBO board and will report back when it's fitted ?

 

[Mikegh]

Hopefully , you haven't a massive number of circuits on the affected side

The lighting may not be involved and it could depend on the balance of leakage on the 2 socket circuits if this is an issue

Some testing may be prudent as posters advised

 

[Markyd]

do a full test before fitting new CU...could save a LOT of headache afterwards

 

[LastManOnline]

Many thanks for all the replies.
I've gone with the RCBO board and will report back when it's fitted ?

I think you, ve made the right choice with the rcbo CU. I would also add that the 19ms trip time is very normal for the rcd in my experience

 

[pc1966]

Is all well swapping a board but will not solve the problem with out being fixed.
Me find the fault before swapping.

You are quite right that any faults should be investigated before swapping stuff. But in this case the IR results are OK-ish (low, but not out of a EICR pass) and it seems other checks like clamp for leakage, etc, have already been done but nothing pointing to a definite cause.

Having two socket circuits and two lights off one RCD is not a very good idea in these days of many electronic gadgets, both from an accumulated leakage point of view and from a 'damn its dark suddenly' point of view. So the suggestion of an RCBO board would be what I would think of as dealing with accumulated leakage as well as minimising the consequences of an RCD trip.

Yes, it might not actually solve the problem if there is some wiring or accessory that has a deeper intermittent fault, but it would allow the source to be narrowed down considerably to the specific circuit that trips, and from that point it might be feasible to loot at other diagnostic routes like PAT testing of appliances on one socket group, or pulling all lamps and IR testing at 1kV on lights, etc, to provoke an actual measurable fault.

Calling out a sparky another half dozen times to try and pin things down it probably costing more than an RCBO board and a single follow-up call!

(My reply is for general interest, not any criticism of yourself)

 

[LastManOnline]

do a full test before fitting new CU...could save a LOT of headache afterwards

This is one of those unusual cases.
Firstly to have an rcd tripping regularly for several years without been sorted by one of the sparks who have investigated is very rare. My gut instinct in this situation would also be accumulated earth leakage. What's throws me a little is the IR test results. Just under 2 Mohms is extremely low and usually indicative of electrical issues. However it is often the case that during the test there may still be some equipment or appliance still connected giving an artificially low reading. Whatever the case its clear that in this situation some concrete action has to be taken and the change to an rcbo CU will certainly provide more answers than are presently available. And let's face it, its a good investment regardless

 

[7029 dave]

Lets put ourselves in the customers shoes, internment RCD tripping over a very long period of time. We dont know the full story of any intervention over this time.
Changing to RCBO's the expense of this for arguments sake £600 and the fault although narrowed down to hopefully one circuit, BUT it still tripping you are not going to be happy.

IMO drilling down on each circuit with IR testing is the way forward, and start investigating the the circuits with the lowest IR.

 

[Mikegh]

Lets put ourselves in the customers shoes, internment RCD tripping over a very long period of time. We dont know the full story of any intervention over this time.
Changing to RCBO's the expense of this for arguments sake £600 and the fault although narrowed down to hopefully one circuit, BUT it still tripping you are not going to be happy.

IMO drilling down on each circuit with IR testing is the way forward, and start investigating the the circuits with the lowest IR.

Wait til u change from an RCD to rcbos and it starts tripping 2 after lol

 

[LastManOnline]

IMO drilling down on each circuit with IR testing is the way forward, and start investigating the the circuits with the lowest IR.

You are assuming that there is "a fault". An accumulation of "intentional earth leakage" is not a fault in the conventional sense. Its what happens when the design of the installation is no longer adequate for the practical needs of the installation. It's a situation that we will encounter progressively more as time goes on due to the inevitable increase of electronics in the home.
Taking a conventional approach to this particular problem is in my view not what's required.

 

[pc1966]

Lets put ourselves in the customers shoes, internment RCD tripping over a very long period of time. We dont know the full story of any intervention over this time.
Changing to RCBO's the expense of this for arguments sake £600 and the fault although narrowed down to hopefully one circuit, BUT it still tripping you are not going to be happy.

If this was a simple case of a reported tripping fault and the first action is to suggest a board change then yes, that is going to be a big disappointment for the customer as it probably won't fix any real fault. They would have every reason not to be happy.

But here they suffer from a total outage on both socket circuits and light circuits 5-10 times a month and this has been happening for over a year with (presumably) several spark's visits not identifying anything as a definite culprit.

In this case suggesting a board change and explaining carefully to the customer that it might be a total cure if the actual problem is accumulated leakage from lots of electronics, but at the very least it will reduce the scope of any outage to just the one circuit, and then help diagnose the source of a genuine intermittent fault by limiting the search to just the wiring and accessories there.

Of course they will hope/expect a total cure, but if not at least they will have far less annoyance than the current situation and can look forward to it being solved completely in the following weeks, not years as gone by so far.

 

[Mikegh]

If you called out to a random installation you'd be taking a few global readings at least before moving onto rcbo board

Larger installations probably need it anyway

Will distribute leakage if its accumulated from appliances and narrow down faults if they occur

Customer is also protected against widespread disruption to supply

 

[7029 dave]

You are assuming that there is "a fault". An accumulation of "intentional earth leakage" is not a fault in the conventional sense. Its what happens when the design of the installation is no longer adequate for the practical needs of the installation. It's a situation that we will encounter progressively more as time goes on due to the inevitable increase of electronics in the home.
Taking a conventional approach to this particular problem is in my view not what's required.

They have low insulation 1.7 megohm on a global reading associated with problems, in my view this needs investigating.

 

[Mikegh]

They have low insulation 1.7 megohm on a global reading associated with problems, in my view this needs investigating.

It's not that low ,It's well above the range of tripping
an rcd

I'm assuming global n-e is a reading taken at the rcd terminals

I'm not clear why there's only n-e or is equipment left connected for the test, first time I heard the term used

 

[pc1966]

They have low insulation 1.7 megohm on a global reading associated with problems, in my view this needs investigating.

It is suspiciously low, but I presume that @HappyHippyDad will be doing per-circuit testing as part of the board change and at that point we will find out if it is 8 circuits in the 13M region, or 7 are > 200M and one is the 1.7M lurking like a leaky toad for investigation.

 

[timhoward]

I do see the logic in the board change with due communication to customer.

It's not that low ,It's well above the range of tripping an rcd

Tangent, but I'm glad you've pointed this out as I've heard more than one person say "I won't bother with IR tests as the RCD would trip if it's a problem!". I think a 0.02Mohm IR results in a current of about 11ma, so even lower IR needed to reach trip current of a 30ma RCD.

 

[pc1966]

Tangent, but I'm glad you've pointed this out as I've heard more than one person say "I won't bother with IR tests as the RCD would trip if it's a problem!".

There are so many reasons that is a stupid attitude!

• L-N insulation faults can start fires, not detected by RCD
• Low IR can point to future problems due to water ingress or degraded cables
• Low-ish IR can reveal 'live' metalwork that is not able to trip an RCD until someone touches it, at which point if the RCD is faulty then someone probably dies.

While the regs have values in the 0.5M to 1M region as a 'pass' that is too low for comfort unless there is a good reason to explain it, for example an old warehouse with dozens of dusty light fittings all adding up to the ~1M result.

Hopefully we will get an update at some point with the results of testing the new board and if that really solved the problem, and if not how it was finally pinned down.

 

[Mikegh]

There are so many reasons that is a stupid attitude!

• L-N insulation faults can start fires, not detected by RCD
• Low IR can point to future problems due to water ingress or degraded cables
• Low-ish IR can reveal 'live' metalwork that is not able to trip an RCD until someone touches it, at which point if the RCD is faulty then someone probably dies.

While the regs have values in the 0.5M to 1M region as a 'pass' that is too low for comfort unless there is a good reason to explain it, for example an old warehouse with dozens of dusty light fittings all adding up to the ~1M result.

Hopefully we will get an update at some point with the results of testing the new board and if that really solved the problem, and if not how it was finally pinned down.

I do believe you have misinterpreted Tim's post

The 1Mohm figure is being talked about in the context of an rcd trip not as a good IR value

 

[timhoward]

I do believe you have misinterpreted Tim's post

The 1Mohm figure is being talked about in the context of an rcd trip not as a good IR value

I think it's ok pal, I know the comments weren't aimed at me!

 

[buzzlightyear]

I think it's ok pal, I know the comments weren't aimed at me!

but the gun is .lol.

 

[Mikegh]

I think it's ok pal, I know the comments weren't aimed at me!

Yes its actually me doing the misinterpreting