Electrician detecting why RCD trips

[tony sayer]

Dunno meow.. never seen one!.

Why aren't they so widespread in their use?...

[Andy Hall]

Doesn't matter. They can soon be made up by the interested parties.

[The Natural Philosopher]

If you have live-to-earth capacitance, a large voltage spike (or dip) can cause large transient earth currents. A large live to earth capacitance is exactly what you get with tons of RFI filters scattered around the wiring, and indeed, even the several hundred yards of mains cable in the average house has a substantial amount in its own right.

[Pete C]

There's a couple of good papers on the subject here:

In the second paper, it states that 66% of fatal domestic falls on stairs or steps are the over 65 age group, and gives a breakdown of the causes of falls in this age group:

"Many people fall at the last step thinking they are already at the bottom of the stairs. This is sometimes due to poor lighting, impaired eyesight or because the victim is carrying something. (21%)

Accidents happen in other people?s houses (for example when the victims are visiting their children) where surroundings are unfamiliar. (12%)

Carrying something is frequently a contributory factor, particularly on stairs. (11%)

Slippers are often loose and lead to a trip or stumble and then a fall. (9%)

Alcohol is recorded as contributing to 6 per cent of the falls on stairs and steps. It is likely that it contributes to a substantially higher proportion where the fact is not mentioned.

Falls happen when victims are visiting the toilet. This may be because they are in a hurry and are suddenly on the move after a period of inactivity or are drowsy during the night. (6%)

People trip over all sorts of things including children and pets. (5%)

Poor lighting or failing to turn a light on is sometimes a factor and may be coupled with visits to the toilet during the night. (4%)

Poor eyesight makes a contribution, often coupled with one of the other factors. (3%)"

The only category that mentions absence of light gives a figure of 4%, so IMO it's reasonable to infer that a relatively small number, maybe

40 deaths per year involve absent lighting on steps or stairs.

I'd expect the vast majority of these are people leaving the light off to avoid blinding themselves when visiting the loo at night!

cheers, Pete.

[Andy Wade]

ELCBs installed in houses were never of the fault voltage monitoring type with high-impedance coils. They were somewhat misleadingly described as "fault-voltage operated" devices but had low-impedance coils and operated when the current through the coil was 15 - 30 mA, just like a 30 mA RCD (except that the leakage current had to flow in the earthing conductor). They were connected in series with the main earthing conductor which went to a single earth electrode [*]. Separate earth monitoring electrodes were not employed - that would make for a much too "technical" arrangement for a house where it would be expected to work for decades without maintenance or supervision.

The main problems with ELCBs were (a) that the operating coil could easily be short-circuited (or partially so) by extraneous earth paths [**] and (b) that overlapping earth resistance areas from adjacent installations would lead to nuisance tripping.

Anyway, your misguided idea to resurrect such devices is (thankfully) a complete non-starter since BS 7671 and the IEC standards on which it's based do not recognise their use for indirect contact protection.

[*] Reg. 413-11 from the 15th edition says: "where protection is afforded by fault-voltage operated protective devices, all exposed conductive parts and associated extraneous conductive parts protected by any one such protective device shall be connected by protective conductors to an earth electrode via the voltage-sensitive element of the device. [**] e.g. suppose the overhead supply has been upgraded for PME by the DNO and your neighbour, having recently re-wired, has a TN-C-S installation. Your old TT installation still has a ELCB coil between your main earth terminal and the earth rod, but has had main bonding added "to improve safety." Now, the gas and water mains in the area are still metal, so your main bonding now connects your main earth terminal to your neighbours PME terminal via their main bonding and earthing conductor. If effect your installation has unwittingly become TN-S as there is now a metallic earth path back to the neutral star-point of the distribution transformer. Your ELCB will now do absolutely nothing to protect your installation, but its coil is sitting there monitoring the current flowing to earth via your electrode. If you're on the end of a long line there's a fair risk of nuisance tripping as a result of voltage drop in the neutral of the main cable driving current through your earth rod.

ELCBs are history, RCDs were a big step forward.

[John McLean]

If you have a better design system for earth fault protection, then please advise me, and the rest of the world. Suffice to say that I totally disagree with your statements.

The voltage operated, earth leakage system was largely developed in South Africa, to reduce the number of accidents which were occurring in the gold and coal mines, they were pioneers in this field, the fact that they continue to do so, is proof that they are effective, In addition, their houses have had earth leakage protection for a far longer period than this country and their tripping current is 25mA as I recall, I have been grateful for this E/L protection a couple of times when fault finding inside our own household appliances, (although we shouldn't do it: - there are times when it is more convenient although risky). There was a problem however, with frequent tripping of these ELCB's during thunderstorms and lightning ground disturbances. South Africa has one of the highest isoceraunic levels, in the world. (Voltage operated, earth leakage circuit breakers are not now, the flavour of the month in the UK, due I understand, to the fact that they require a functional earth connection to operate).

One serious incident that I can recall, was when I was the factory engineer in a steel rolling mill in South Africa, we had earth leakage installed on

525Volt, three phase welding outlets scattered through out the plant. I had a routine maintenance system, where the E/L protection on these outlets was routinely tested and the results recorded, using a simple potentiometer and meter to earth, and carried out by the shift electrician. One day, a welder suffered a serious electric shock, when using a large bullet type, welding machine. On investigation, I found that the earth leakage detection system had been disconnected for the socket outlet, on inspection of the last and recent recorded test, the test was signed off as being satisfactory. After grilling of the relevant electrician, he admitted that the E/L had been disconnected some time previously since it had been repeatedly tripping! Had this action resulted in a fatality, he would have been incarcerated; or at the least de-bollocked. I would also have been penalised, being the responsible person. Ring any bells?

How would you test; fault find or monitor the effectiveness of your; or any system?

I would suggest that the main reason why RCD's trip and correctly so, is not due to "flawed systems" or fixed wiring, but because of leakage on appliances and probably moisture to boot. I have used my own system of drying out appliances, many times in the past, usually where heating elements are involved, the initiated can probably guess the corrective action employed.

Do you have records here; or is this a personal experience? More evidence would be required to justify your above statements.

There is nothing novel about this arrangement, apart from the ELCB, see the IEE OSG pages 21 and 22. What nuisance trips would be experienced with ELCB' s, arising from ground fault currents?

This could also indicate that the use of RCD's is effective?

I am a strong supporter of the principle of Part P, for the simple reason, that it must reduce the many unsafe installations that occur. (I disapprove of the method however and would have preferred licensed electricians).

An indication of an effective government is the courage to take unpopular decisions. As someone once said: Democracy isn't the best form of government etc., The wrong person to ask is the man in the street, if it will save him or her money, then they will support it. The hoi polloi are not usually concerned with the technical aspects of an installation; they look at the bottom line.

Jaymack

[Owain]

Being DP, they remove the common at-installation fault of misconnecting the circuit neutral to the inappropriate busbar.

SP MCBs will also hold off an upstream RCD in the event of an N-E fault, meaning that even if the MCB of the affected circuit is switched off by the user, the remainder of the RCD protected installation cannot easily be brought back into service.

Owain

[John McLean]

tips here; or just wishing to advertise your site?

[meow2222]

It sounds like you may have not read the whole thread.

NT

[Peter Parry]

It could were it not for the fact that with the exception of deaths in gardens (Cutting mains leads with lawnmowers and hedge trimmers) the number of people killed by electricity has remained pretty much the same since before the introduction of RCD's.

That lets this lot out then - they are simply interested in who bungs them the most money.

[meow2222]

I already have. You even quote it in this post of yours.

interesting, my other source of info on this reports 2 core wiring with no E conductor, no rcd or elcb, one 15A plug and one fuse per house is the norm. I guess there is no one standard across the board.

I think the main perceived advantage with RCDs is they protect against shocks more than ELCBs. Its a shame their downside seems to outweight the upside.

Methods for testing ELCBs and RCDs are nothing new. With a modern design of ELCB we could expect to see an inbuilt tester, like RCDs have.

FWIW a test routine would apply a votlage to the sense coil, and check impedance between the 2 earth terminals.

As to the other part of your question, its kinda long, and has been addressed elsewhere.

Which if you follow the newsgroup regularly you'll know are mostly not risks. Some cooker element leakage is both normal and safe. Ditto immersion heaters. We've covered all this stuff before.

As I mentioned, following the newsgroup is a record all of us here can readily see.

hence we know how it would behave.

has already been answered fully in this thread

Their introduction has not reduced electrocution to any great extent.

Ah.

In England people are very keen indeed on the notion of complete safety, and will happily vote to spend large sums of money on things with only the remotest risk. When it comes to paying people to replace their rcd no-one is interested, for the simple reason there are more important things to spend the money on.

NT

[meow2222]

In fact there were 2 types of ELCB, voltage operated and current operated. The voltage operated ones were preferred because they didnt insert impedance into the cpc, which is a risk in itself.

that only applies to the current operated ones you describe. It is not applicable to the v-ELCBs I'm tralking about.

This has been addressed in this thread already, and should not be a problem to any significant extrent today..

You're missing one of the points of this discussion. If the case for ELCBs is established, information spreads, and regs follow knowledge, albeit slowly.

This tells us what? That a) someone incompetent can make a balls up of an old installation b) that the curernt operated elcbs you mention have their issues, which is of no significance to the subject here.

If you can make a solid argument I'm all ears.

NT

[Andrew Gabriel]

The inbuilt button doesn't count as a test (for either device).

The testing procedure for an ELCB is to use a 1kW 50V isolated transformer supply to try and force 50V differential between ground and CPC. The test passes if either the ELCB trips, or you cannot get 50V on the CPC due to sufficient low impedance parallel earth paths that the ELCB is not required. (This is from memory; the procedure is in the 14th edition wiring regs, which I don't happen to have to hand.)

Of course, you won't be able to buy an ELCB test transformer anywhere nowadays.

[Alistair Riddell]

No, you clip it to anywhere that dangerous voltage might appear in the unlikely event of a fault occurring, e.g. exposed metalwork (yes allowed on Class II applicances if properly insulated from mains voltage) or low voltage / signal inputs / outputs.

[Peter Parry]

They were? Apart from not actually answering the question you assume a level of competence utterly alien to most Part Peeers. I watched an "electrician" (NICEIC Company) this Friday trying to find out why the domestic air conditioner he had attempted to wire up wasn't working. He decided to "test the mains" which was done with a normal Fluke multimeter. He put one lead of the meter in his pocket "for safety" and prodded everything in sight with the other before announcing the problem was "low volts" and the house needed rewiring. His boss (who had been "in a meeting" for several days) was around within 10 minutes of being told this with a quote for an "upgrade" of only GBP1,500.

When it was pointed out that if you have three wires numbered 1, 2 and 3 and three terminals marked 1, 2 and 3 then connecting 1 to 1, 2 to 2 and 3 to 3 was likely to be more successful than connecting 1 to

2, 2 to 3 and 3 to 1 the answer was that the "skilled electrician" had "Meggered" it and found there were low volts (48VAC apparently) so it was illegal to do anything other than rewire everything "because it was a Part P issoooe".

Managing Director of NICEIC approved company and his oik were told where to reposition their bodies and no, they wouldn't be paid. Put wires in right holes and everything worked. Amazing.

Quite frequently; the more learned contributors are a useful source of information.

[Andy Hall]

Pity the meter wasn't on one of the current ranges and then a sudden need to blow his nose.

These people need a short sharp shock.....

[meow2222]

I dont see any of this as an issue though. If ELCBs are reintroduced the test gear will turn up on the open market as usual.

If someone wanted to diy an elcb outside the regs, which is not what I'm suggesting, it wouldnt be hard to figure out how to test it.

NT

[Peter Parry]

This one is in for a long hard shock - the unit was being fitted in the house of an elderly lady who at the moment is unable to walk and I was along simply as mobile eyes for her. The oik concerned should have glanced at her mail before trying to baffle her with his electrickery colouring book. Noticing that after her name on the envelopes came BSc, C Eng, FIEE might have given him a clue.

Mind you I notice that the NIEIC is now vying with Corgi for who can be the most dishonest :-

"Tough new laws now effectively ban garden enthusiasts from installing their own electrics in ponds, sheds, garages and greenhouses, warns consumer electrical safety organisation the NICEIC.

From 1st January 2005 all electrical work in the garden must be carried out by a government-approved electrician, such as one registered with the NICEIC. The new law aims to reduce the rising number of deaths caused by faulty electrics...According to the new law, gardeners are allowed to maintain and replace equipment. However, because the risk of electric shock is far greater outside, a government-approved electrician who can provide a certificate...must undertake all other work."

"The new building regulation Part P, effective since 1st January

2005, requires most electrical work in the home to be carried out by a government-approved electrician...The NICEIC welcomes the government's decision to finally clamp down on the cowboys who cause these deaths and is advising homeowners to make sure they only employ government-approved electricians."

[tony sayer]

Yep, but that still doesn't explain why a transformer with poor regulation would cause more tripping than a higher rated one.

You must have a lot of capacitance around Philosophers Manor

[meow2222]

It does. Normal capacitive leakage at 50Hz is low. Now add a 5kHz 30v transient and leakge goes up to 5000/50 x 30/240 x as much briefly, ie

12x. RCDs have filters in to reduce tripping on such leakage transients, but they can only do so much.

NT