More 'RCD Type' uncertainty/confusion

[JohnW2]

I am becoming increasingly confused and frustrated by this ‘RCD Types’ business.

Most of what I can find to read about the different types says much the same as 531.3.3 of BS7671, majoring on what sort of residual currents will trip them but saying nothing directly about the issue that so often gets discussed here, namely the potential of DC components of current to impair the functioning (tripping) of the device if a AC (or pulsating DC) residual current arises.

However, in relation to Types A, F and B, 531.3.3 includes:

NOTE 1: For RCD Type A, tripping is achieved for residual pulsating direct currents superimposed on a smooth direct current up to 6 mA.

NOTE 2: For RCD Type F, tripping is achieved for residual pulsating direct currents superimposed on a smooth direct current up to 10 mA.

NOTE 3: For RCD Type B, tripping is achieved for residual pulsating direct currents superimposed on a smooth direct current up to 0.4 times the rated residual current (IΔn) or 10 mA, whichever is the highest value.

Firstly, it is not totally clear (at least, not to me) whether the “superimposed smooth direct current” (in all of those cases) is a “residual” current or a balanced (L& N) one. Since they have not included the word “residual” in relation to it, I’m inclined to think that they probably do not mean ‘residual current’, but I can’t be certain.

Next, those three notes only refer to tripping being ‘achieved’ with “residual pulsating direct currents”, and hence say nothing at all about tripping in relation to any other type of residual current.

In particular we are told nothing about how tripping in response to AC residual currents (which has been the basis of nearly all of our discussions, and also flameport’s experiment/ demonstration) for any of the types of RCD (including Type AC) is affected by superimposed DC currents (be they ‘residual’ or not).

Even in relation to tripping as a result of “residual pulsating direct currents”, we are only told that this is ‘achieved’ if that residual current is superimposed on a smooth DC current “up to” 6mA for Type A, 10 mA for Type F and 12 mA for (30mA) Type B. Does that mean that these devices will not ‘achieve’ a trip due to “residual pulsating direct currents” if the “superimposed smooth DC current” (be it ‘residual’ or not!!) is more than those (very small) figures? If so, are they actually fit for any purpose? ... or am I misunderstanding what they are saying?

So, as above, I find this all about as clear as mud, and therefore continue to be frustrated and confused. Can anyone help me understand - or, as I have asked many times, point me towards some genuinely useful reading material?

Kind Regards, John

 

[ericmark]

I read with interest "a smooth DC current" and I consider how unless as with John Wards test rig where a three phase RCD has smooth DC feed through an unused pole smooth DC can arrive at the RCD? The only way I can see of getting DC is solar panels or a battery. The latter likely electric car charging, and in both cases, if it goes faulty. To my mind the test rig was cheating as it was doing some thing which would not naturally happen.

It did show even with 50 mA which is a lot of smooth DC it still tripped at 70 mA, even at 250 mA smooth DC it still tripped at 210 mA with a 30 mA type AC RCD, John Ward tries to say 50 mA is rather small, however as you have pointed out even the type B is only guaranteed to take 12 mA so he was testing the type AC with 20 more smooth DC than a type B can handle, so really 250 mA is rather high.

The RCD which he shows does work actually works the wrong way around, it sends a current to the release on the MCB to trip it, I used devices similar to shown back in 1992 in the building of Sizewell 'B', the version we had we could set the tripping current and delay, and it was used with a moulded breaker so we would have main supply set 5 amp at 1 minute, next set 1 amp and 30 seconds, and next was standard 100 mA type S and finally in the cabin a 30 mA at 40 mS, however we found a fault would trip all 4 devices, so whole idea of discrimination simply did not work. OK the fault was a nail knocked into portacabin wall through cable to hang up one ones coat, but whole idea of time delay was the 30 mA should have tripped without tripping the other devices.

The other problem was testing, with a 500 amp supply it covers a lot of small devices, possibly 50 or more portacabins, at average 30 amp each, and to arrange a test is not easy, telling all to log off their computers first, so it would be tested when fitted, and likely not tested again as by time the next test was due, it had been decommissioned. And it was not unknown to find the small cable between the sensor and the moulded breaker had been damaged. Or the solenoid inside the moulded breaker had rusted, so would not work. The two integral units 4 module wide each fitted to my old house in around 1992 are still working, nearly 30 years latter, there is no icon or any other marking to say type AC, A, F or B, all it says is 30 mA. I know they still work, as just before moving one tripped costing a freezer full of food.

So in real terms we use type AC unless some item has a label saying should use type A, it was said LED bulbs are a problem, however if you look at most LED bulbs they use a capacitive drop circuit which means can't cause a DC load. The problem is when doing an EICR you would need the instructions for every bit of fixed equipment, this link to podpoint charger is typical, where it says

As of the 1st January 2019 either a Type B RCD must be used or a Type A with 6mA DC protection included in the Pod-Point (see detail on packaging to determine what protection is required),

so by time an EICR is due what are the chances that the packaging still exists? Other than some thing like this mid position valve where 13 kΩ plus resistance of motor means less than 17 mA DC is pulsed, most items will only cause DC if there is a fault, I did consider if the three port valve was reason for the warning use type A with Bosch boiler, however it seems likely over 6 mA pulsed DC so would need type B to be sure which you can't get as a single modular width RCBO so it would mean you could not use the valve in any home using RCBO's, and the three port motorised valve has been used well before we started fitting 30 mA RCD's which seems to point to there not being a real problem.

 

[JohnW2]

I read with interest "a smooth DC current" and I consider how unless as with John Wards test rig where a three phase RCD has smooth DC feed through an unused pole smooth DC can arrive at the RCD? ....

I think they are simply using "smooth DC" to refer to one of the components of the composite waveform. In other words, when they talk of "residual pulsating direct currents superimposed on a smooth direct current", they re probably simply trying to describe a waveform in which there are unidirectional pulses which do not go down to zero.

It did show even with 50 mA which is a lot of smooth DC it still tripped at 70 mA, even at 250 mA smooth DC it still tripped at 210 mA with a 30 mA type AC RCD, John Ward tries to say 50 mA is rather small, however as you have pointed out even the type B is only guaranteed to take 12 mA so he was testing the type AC with 20 more smooth DC than a type B can handle, so really 250 mA is rather high.

As I wrote recently, I am far from convinced that what he did was relevant to the issues we discuss.

For a start, as I said, by just having it flowing through one pole of the RCD, what he was applying was a residual DC current, such as only would occur in practice if DC were flowing through one earth fault. He then introduced an AC residual current (which, in practice, would quite possibly require a second fault to earth) to see if that tripped the RCD.

As I said, if he wanted to find whether DC (not residual DC) flowing through the RCD would prevent the RCD tripping normally in response to an AC residual current (which is the issue we always discuss), I think he should have put both conductors of the DC circuit ('flow' and 'return') therough the two spare poles of the RCD.

As I explained, my main problem is that the descriptions in BS7671 (and everywhere else I've seen) do NOT seem to address that issue we discuss - i.e. the extent to which the response of the various types of RCD (including AC) are impaired by the presence of a DC component of current flowing through the RCD (usually, in discussion, due to some equipment other than that in which the earth fault {which should trip the RCD} arises).

As I said, although not stated explicitly, the three 'Notes' I posted would seem to suggest that Types A, F and B RCDs might not trip (at least, are not guaranteed/required to trip) in response to "residual pulsating DC" (presumably due to a fault to earth, maybe together with other faults) in the presence of superimposed ('smooth' = 'non-pulsatile') DC currents greater than 6, 10 or 12 mA respectively - which is what lead me to ask whether they were 'fit for any purpose'. As you say, John's DC currents were not only unbalanced (not 'residual') but were considerably higher than those figures.

It gets even more complicated in the case of a Type B RCD, since ...

(iv) RCD Type B: RCD for which tripping is achieved as for Type F and in addition:
... (e) for residual smooth direct currents, whether suddenly applied or slowly increased, independent of polarity

In other words, it appears that a 30mA Type B RCD should trip in response to a DC residual current of 30mA, even if no AC current is flowing at all - so the test DC residual currents used by John would presumably trip a 30 mA Type B RCD even if no AC current was flowing through it at all.

As I said, I still don't understand all this. In particular, I cannot find any information of 'typical performance', let alone any 'requirements', as regards the extent to which (for any type of RCD, including AC) tripping in response to an AC residual current (e.g. due to an L-E fault) is impaired by there being constant DC current (of varying magnitudes) flowing through the RCD. That is primarily the issue/'risk' we keep discussing, yet I can find no information about it at all!

Kind Regards, John

 

[ericmark]

It gets even more complicated in the case of a Type B RCD, since ...
In other words, it appears that a 30mA Type B RCD should trip in response to a DC residual current of 30mA, even if no AC current is flowing at all - so the test DC residual currents used by John would presumably trip a 30 mA Type B RCD even if no AC current was flowing through it at all.

Kind Regards, John

Very good point, the type B should have tripped before he even connected up the RCD tester as both 50 mA and 250 mA exceed the 30 mA rating, so we may as well ignore that test.

I also agree it needs two independent faults to cause the RCD not to trip when required. There was another guy who talked about RCD's in railway premises and how at night they tripped but during the day when trains were running they would not trip, but we are not that hitech where I live, most the trains still use steam.

 

[JohnW2]

Very good point, the type B should have tripped before he even connected up the RCD tester as both 50 mA and 250 mA exceed the 30 mA rating, so we may as well ignore that test.

Quite so.

I find it pretty extraordinary that in relation to an issue which could be pretty important, is so difficult to find any useful factual information!

Kind Regards, John

 

[Swwils]

Several EU countries phased out type AC RCD because they see other loads (air con, solar) far more commonly, so have a impetus to detect the respective DC fault currents.

I would guess that the UK just adopted EN 60364-5-53 2015 and so AC stuck around since they are in there.

Shove a pulsed waveform into a AC RCD and it won't trip within safe limits. Not to say that AC RCD are unsuitable for some scenarios.

An example of a scenario for a type A RCD being appropriate: fault on the output side of the bridge rectifier in a microwave will produce a pulsed fault current - generated at the supply frequency.

Type F is more tricky (not designed to detect smooth dc residual currents) so there is reliance on the manufacturer to describe what is required - something like a single phase inverter used in motor speed control will vary with the characteristics of the equipment and they should determine if Type A, F or B is required really.

 

[aptsys]

Under what scenario would you see a DC bias on the AC supply without the item in question going bang? The low source impedance of most domestic AC supplies suggests to me that this would not be possible.

 

[Swwils]

I'm not an expert but don't some scenarios result in protective conductor current?

This is where it falls fowl of 531.3.2 - Operational leakage current (protective conductor current) for the circuit should not exceed 30% of the RCD sensitivity.

 

[JohnW2]

Several EU countries phased out type AC RCD because they see other loads (air con, solar) far more commonly, so have a impetus to detect the respective DC fault currents. .... Shove a pulsed waveform into a AC RCD and it won't trip within safe limits.

That may well be true, but nearly all of the discussions, here and elsewhere, seem to relate to the ability of DC components of current though an RCD to have the ability to impair the RCD's ability to trip in response to an (AC) residual current due to something else. Whilst I don't doubt that such is, at least qualitatively true, I have been struggling for a long time to find any information about this - particularly the amount of DC component of current required to significantly impair the RCD's ability to trip in response to an AC residual current.

In any event, as I've said, the material I posted, relating to Types A, B & F, seemed to imply that such devices may only (be required only?) to trip in response to a residual pulsating DC current if it is superimposed on a 'smooth' (I take to mean 'constant') DC current of no more than 6-12 mA - which seems pretty 'restrictive'.

Not to say that AC RCD are unsuitable for some scenarios.

Indeed. As I've commented before, BS7671 says:

For general purposes, Type AC RCDs may be used.

... albeit without any indication of what they mean by "general purposes".

An example of a scenario for a type A RCD being appropriate: fault on the output side of the bridge rectifier in a microwave will produce a pulsed fault current - generated at the supply frequency.

I'm not sure I understand that. Provided that the rectifier is intact, the "fault current" (in the sense of the current flowing through the fault) will be 'pulsed DC' (since alternate half-cycles of the input will be inverted (at a 'pulse frequency' double that of the supply frequency), but the current in the L supply conductor (hence passing through the RCD), and hence the (L-N) residual current seen by the RCD, will remain as sinusoidal AC, won't it?

In any event, as above, nearly all the discussions here are not about the ability of RCDs to trip in response to residual pulsating DC currents but, rather, the ability of DC components of current to impair the device's ability to trrip in response to residual AC currents - and it's that which I struggle to find any information about.

Kind Regards, John

 

[JohnW2]

I'm not an expert but don't some scenarios result in protective conductor current?

Of course, but I'm far from sure about DC protective conductor currents (when the supply is AC).

Kind Regards, John

 

[Swwils]

Oh I see what you mean.

Have we all made the good faith mistake of thinking that appliances will be designed properly!

I did find this image:

From: http//www.kynixsemiconductor.com/News/57.html

Where type B use is discussed.

 

[JohnW2]

Oh I see what you mean. Have we all made the good faith mistake of thinking that appliances will be designed properly!

Maybe. However, I think I might struggle to even deliberately design something powered by AC which resulted in a DC current in a protective conductor, wouldn't you?

Apart from anything else, a current has to have a 'complete circuit/path'. If there were a DC current 'going into' the protective conductor, where would be the 'rest of the circuit'? - i.e. where would it be 'coming from' - the L of the (AC) supply?

Kind Regards, John

 

[Swwils]

I can think of stuff that does have residual which contains a smooth DC component though!

Maybe something with rectified AC could fault in a way discussed; microwave was my first thought, but yes would expect it to go bang. Maybe it won't go bang before being able to do some damage to unlucky someone.

Schneider made the following doc with nice tables of the international standards and how fluxgate works.

No idea how an electrician will be able to navigate this without hundreds of standards docs for each appliance.

 

[JohnW2]

I can think of stuff that does have residual which contains a smooth DC component though!

A 'smooth DC component" in the (primarily AC) supply (L & N), yes - that's very easy. In fact, with things around in my youth which had half-wave rectification straight off the AC supply, there would be a very large 'DC component in the 'supply' current. However, as I said, I'm not at all sure how one would get a DC 'fault current' leading to a DC L-N difference in the supply conductors (i.e. a DC residual current).

Maybe something with rectified AC could fault in a way discussed; microwave was my first thought, but yes would expect it to go bang. Maybe it won't go bang before being able to do some damage to unlucky someone.

Whether it goes bang surely depends upon the magnitude of the fault current? Don't forget we're talking of a situation in which we would like an RCD to trip with only 30 mA of 'fault current' - and, if it were remotely that low, any bang would probably be extremely quiet!

However, as I recently wrote, although I may be missing something I don't see how a fault ('to earth') downstream of a ('intact') bridge rectifier could result in anything other than a sinusoidal AC current in the supply L, albeit less than the sinusoidal current in the N, hence an L-N imbalance to be seen by the RCD. Am I missing something?

Kind Regards, John

 

[Swwils]

Isn't that the rub though?

I found a video that describes it perfectly!