ECN Forum
Posted By: renosteinke GFCI vs. RCD - 11/23/06 06:44 PM
Another thread, about portable generators, has me a bit confused.

I was under the impression that an "RCD" was the same as our "GFCI" (here in the USA).

A GFCI, or "ground fault circuit interrupter," is a device that -in simple terms- measures the current going out through the 'hot' wire and also measures the current combing back in the neutral. If the two figures are not very nearly the same (within 5ma), the device shuts off the power.
So, in the generator question, A shock would result in some of the current not returning to the generator, which would trip the GFCI.

I thought this was what the RCD did, but some have said that the RCD would not operate in this situation.

If the two are the same ... ho can it fail to operate?
Posted By: gideonr Re: GFCI vs. RCD - 11/23/06 09:20 PM
A GFCI is nearly equivalent to an RCD, but a generator is a special case, it might not have been grounded. Current must have a return path, which is usually ground for mains systems which would trip the GFCI/RCD, but with generators that aren't grounded the only return path is back to the generator, which if it goes around the GFCI/RCD and not by the neutral will trip the GFCI/RCD. If the shock current returns by the neutral then neither will trip.

Gideon.
Posted By: renosteinke Re: GFCI vs. RCD - 11/23/06 09:57 PM
I agree that a fault that returns via the neutral will be seen as 'just another load.' A GFCI will not se that as a fault.

Yes, I think that current can flow into other places, away from the generator. After all, is not the wire "hot" to many things, besides the neutral that goes back to the generator?

If so, would not such current leakage be detected by the RCD?
Posted By: gideonr Re: GFCI vs. RCD - 11/23/06 10:10 PM
RCDs are usually 30mA trip, though occasionally 100mA trip are used.
Posted By: Alan Belson Re: GFCI vs. RCD - 11/23/06 10:26 PM
Reno I think you are right. Setting myself up to be shot down in flames, but, if the genset replaces the mains supply and is wired up correctly, RCDs or GFCIs should react the same way, surely? The device[s] should have no way of sensing any difference.

Alan
Posted By: Wolfgang Re: GFCI vs. RCD - 11/23/06 10:31 PM
Two things mixed up here:

The main difference between the typical RCD and a GFCI is that the first is typically passive. It doesn't contain much electronic stuff and would work at 20 VAC as well as at 230V as long as more current is passing than allowed and written on it. On the other hand this principle limits trip minimum to 30mA for 3phase and 10 for 1phase devices.

The generator issue is something else. The very small gennys often have isolated conductors. There is no "hot wire" measured against the ground. It can be compared with "separating transformers" which we use here in Germany f.i. in tv repair shops. This issue is limited by a maximum power of some kW and a maximum length of the isolated mini-"grid". Otherwise a capacitive coupling (correct word?) could produce dangerous voltages against ground also with an isolated system. These limits are well defined in the IEC regs as far as I remember.
Posted By: pauluk Re: GFCI vs. RCD - 11/24/06 11:25 AM
Quote
I was under the impression that an "RCD" was the same as our "GFCI"
Effectively, yes, although as already stated our general-purpose RCDs are usually 30mA or 100mA trip rather than the 4 to 6mA trip of the American GFCI.

Current practice is to use a 30mA RCD where necessary (e.g. for receptacles used to power equipment outdoors). 100mA RCDs are still available and commonly used where RCDs have to be cascaded. A typical example would be a system where the main grounding of the installation is just to an earth rod rather than to the supply. The 100mA RCD would then be a delayed-trip type to provide protection to the entire system, with a 30mA RCD then used to provide the higher level of protection as needed.

10mA RCDs are also available, but rare.

Quote
if the genset replaces the mains supply and is wired up correctly, RCDs or GFCIs should react the same way, surely?
The key point is "wired up correctly." It's all going to depend on if and how the system is grounded.

Scenario #1. Genset frame is grounded by way of a rod, and one side of the 240V output is also bonded directly to the frame/ground. In this case the source will behave exactly like a normal (TN) supply, and the RCD will "see" any fault which exceeds its trip level.

Scenario #2. The 240V winding is not bonded at the genset (as is common on many small generators sold here), then somebody installs a bond to ground on the load side of the RCD to reference the system to earth. The RCD is now useless, since it cannot detect an imbalance in the installation's line/neutral current.

Scenario #3. The 240V output is completely floating, with no bonds anywhere. Again, the RCD would be ineffective. A bolted ground fault on either side of the installation would -- effectively -- just turn the system into scenario #2 above.

Quote
The main difference between the typical RCD and a GFCI is that the first is typically passive. It doesn't contain much electronic stuff and would work at 20 VAC as well as at 230V as long as more current is passing than allowed and written on it.
I'm not sure about the RCDs usually available in Germany, but in the U.K. there are plenty of active RCDs which incorporate an amplifier to increase sensitivity to the required level.

Quote
It can be compared with "separating transformers" which we use here in Germany f.i. in tv repair shops. This issue is limited by a maximum power of some kW and a maximum length of the isolated mini-"grid". Otherwise a capacitive coupling (correct word?) could produce dangerous voltages against ground also with an isolated system.

Capacitive is the right word. [Linked Image]

Your "separating transformer" (direct translation from German?) is what we would refer to as an isolating transformer. Of course, this is exactly the principle used in our bathroom shaver outlets in Britain: The transformer is located right at the outlet so that the shaver supply is completely floating, and there is not enough wiring between the transformer secondary and the shaver for capacitive effects to become significant.



[This message has been edited by pauluk (edited 11-24-2006).]
Posted By: aussie240 Re: GFCI vs. RCD - 11/28/06 12:40 AM
Quote
this is exactly the principle used in our bathroom shaver outlets in Britain
Paul,
Is it compulsory to have isolated sockets in UK bathrooms? So there's no 'proper' socket to run a portable heater then?
Here, we just have the normal non isolated 240V 10A outlet, although I did replace mine with the RCD type of GPO.
Posted By: Trumpy Re: GFCI vs. RCD - 11/28/06 09:56 AM
In accordance with AS/NZS 3000:2000
All habitable areas require RCD protection, with a 30mA RCD, within a domestic installation.
This includes lighting circuits and socket-outlet circuits under 32A.
It does not include Air Conditioning circuits or circuits dedicated to alarms or TV amplifiers.
Refrigerator circuits are also exempted, where these are dedicated circuits.
Heaters in Bathrooms that afford Double Insulation, are also exempt, where they are supplied from thier own circuit.

[This message has been edited by Trumpy (edited 11-28-2006).]
Posted By: pauluk Re: GFCI vs. RCD - 11/28/06 11:41 AM
Quote
Is it compulsory to have isolated sockets in UK bathrooms?
Yes, the only socket permitted in a bathroom under the IEE Regs. is a xfmr-isolated shaver type. It's been that way since at least 1955, if not longer. Of course, as I've pointed out before there has never been any compulsion to follow the Regs. in residential wiring, so it's not compulsory in the legal sense, and although regular sockets in bathrooms are rare, I've seen a few.

However......

This seems all set to change with the 17th edition of the Regs. due out in 2008. If all the proposed changes go through, then it will be permitted to install a regular outlet, subject to distance requirements and it being RCD-protected.

See IEE/BS7671, 17th edition



[This message has been edited by pauluk (edited 11-28-2006).]
Posted By: djk Re: GFCI vs. RCD - 11/28/06 11:46 AM
In Ireland we require as a minimum 30mA RCD protection on all socket outlets <32A too. As well as on waterheater, electric shower, central heating etc.

Only exemptions are stoves, cookers etc. However, in most instances modern hobs and ovens will not trip a 30mA RCD unless they are actually faulty. Older ovens and solid plate hobs sometimes do.

It would be very unusual to find a fridge/freezer here that isn't connected to a circuit protected by an RCD and I've never heard of domestic fridges, even the largest types, causing RCD trips.

There are some changes to the bathroom regulations here too, similar to those in the UK in accordance with CENELEC and EU regulations.

It adds RCD protection at 30mA (minimum) to all circuits within the bathroom so lighting circuits, fans, shaver sockets (the special type via the isolating transformer) must be connected to an RCD.

In practice, it doesn't mean a lot as Ireland has obtained exemptions so that the ETCI can continue to ban the use of sockets/switches in most cases anyway.

All it does is add extra layers of RCD protection to the existing requirements so that Irish regs comply with harmonised EU regs.
Posted By: Trumpy Re: GFCI vs. RCD - 11/28/06 03:43 PM
Dave,
Quote
It adds RCD protection at 30mA (minimum) to all circuits within the bathroom so lighting circuits, fans, shaver sockets (the special type via the isolating transformer) must be connected to an RCD.
Explain to me how an RCD protects a person when there is an isolating transformer in the circuit?.
Is the RCD a Core-Balance type?.
Posted By: djk Re: GFCI vs. RCD - 11/28/06 04:46 PM
They just whack the whole bathroom on an RCD.
Those shaver sockets are on the same circuit as the lights normally.

I suppose if surface mounted there's an incrediably obscure chance that someone could get water down the back of the socket / combined light/socket fitting and contact the actual branch circuit rather than the output of the transformer.
Posted By: mikesh Re: GFCI vs. RCD - 11/28/06 05:54 PM
Could someone translate RCD? Acronyms are fun.
I have seen the term with respect to European Electrical systems but what does it stand for?
Posted By: djk Re: GFCI vs. RCD - 11/28/06 06:06 PM
RCD = Residual Current Device
Posted By: pauluk Re: GFCI vs. RCD - 11/28/06 10:37 PM
Quote
Explain to me how an RCD protects a person when there is an isolating transformer in the circuit?

Well, it's in case somebody gets bored while sitting in the bath (trying to think of another song to sing!) and decides to unscrew the shaver outlet and start tinkering around with the primary connections....... [Linked Image]

Quote
Could someone translate RCD? Acronyms are fun.

{.....}

RCD = Residual Current Device

Also formerly known as:

RCCB = Residual Current Circuit Breaker.

And before that, it was known as a current-operated or current-balance ELCB.

ELCB = Earth Leakage Circuit Breaker.

Personally, I think the old ELCB designation is much more descriptive.

There was another type known as a fault-voltage or simply voltage-operated ELCB. In times past, this was the type assumed when somebody talked about an ELCB in a domestic setting. The voltage ELCB worked on a rather different principle though, and has been considered obsolete for years. The IEE Wiring Regs. here ceased to recognize it in 1981.
Posted By: djk Re: GFCI vs. RCD - 11/29/06 12:24 PM
There are some Siemens branded RCDs here from the 1970s that are labled ELCB but are most definitely not voltage operated. They're standard 30mA RCDs.
Seems the term ELCB was still in use when they were installed though.
Posted By: Texas_Ranger Re: GFCI vs. RCD - 11/29/06 08:12 PM
I'm surprised they even bothered to relabel them - In Greece I saw some of similar vintage that were labeled: "Siemens Fehlerstrom-Schutzschalter, Made in Germany"...
Posted By: kiwi Re: GFCI vs. RCD - 11/30/06 08:39 AM
G'day Pauluk, I also would like to return to calling them "ELCBs" or "RCDs". "RCCB" takes much longer to say and is just annoying.

Does anyone know how this change of terminology happened ? I don't know because I left the country for a few years, came back and BAM! We're calling them RCCBs.

"Earth Leakage" is a much better way of describing the device. And ELCB is easier to say than RCCB !

ElCBO is also so much easier to say than RCCBO.
Posted By: Texas_Ranger Re: GFCI vs. RCD - 11/30/06 11:33 AM
German name back in the days prior to harmonization used to be "FI"
F stands for fault and I for current. Full name "FI-Schutzschalter" (fault current protection switch).
The voltage-operated version used to be called FU.
Posted By: pauluk Re: GFCI vs. RCD - 11/30/06 12:40 PM
Quote
Seems the term ELCB was still in use when they were installed though.
I'm trying to remember when RCCB came into use as the replacement term for a current-operated ELCB. I don't think it was until the early 1980s.

Quote
Does anyone know how this change of terminology happened ?
I'm not sure why the change was ever proposed in the first place. And although RCD might be easier to say than RCCB, that was an even more ridiculous change. Why replace the more descriptive term circuit breaker with the vague device anyway? [Linked Image]

Although RCD is now the "official" name, I know that at least a couple of years ago Crabtree was still labeling some units as RCCB.

Quote
ElCBO is also so much easier to say than RCCBO.
Just to confuse things even furthe, over here the combination overcurrent/earth-leakage breaker is now generally referred to as an RCBO (only one C).
Posted By: Wolfgang Re: GFCI vs. RCD - 11/30/06 12:59 PM
The only advantage of the renaming is that the English abbreviations are taught in other countries also. (Theoretically) every German electrician should know the terms RCD (common), MCB (rare) and RCBO (also rare).
Posted By: djk Re: GFCI vs. RCD - 11/30/06 05:51 PM
I've seen Hager units here multi labled in French, German and English

RCB, FI, Differential

Siemens Ireland seems to have always had its own labeling though. The details on the old RCD are

Siemens - Made in Germany
Current Operated Earth Leakage Circuit Breaker
30mA
60A
Diagram of the circuit and test instructions explaining how to find an earth leakage fault by unplugging all appliances connected to the sockets, resetting the breaker and plugging them back in one by one.
Calling an electrician / having the faulty appliance professionally repaired or disposed of.

Other than that there's a 2 pole RCBO for the shower and 10A RCBO covering lighting in the bathroom and outdoors. (Modern ABB units, much smaller than the old siemens one)

[This message has been edited by djk (edited 11-30-2006).]
Posted By: Trumpy Re: GFCI vs. RCD - 12/01/06 06:52 AM
One thing that has to be remembered about a plain vanilla Residual Current Device.
That's all it does, is sense an imbalance between Phase and Neutral in a circuit.
A RCBO, has two functions, it senses like the RCD does, but it has the features of a Circuit Breaker built in to it.
The term RCCB and ELCB mean the same thing.
However, Voltage Operated ELCB's were a different kettle of fish, they were only used where the basics of earthing could not be provided and used to trip if the voltage got above 22VAC, with reference to earth (cold water pipe).
They were used here in places at (usually) the ends of long spur lines, where the ground was so rocky that if you went to sink a rod, the thing would either bend and come up a foot or so away from where you were driving it or on the first hit, would disappear into oblivion below.
They used to be really bad things, trip at night when the soil temperature dropped below a certain value.
Considering that they were sealed by the local Power Company here, the Home-owner was not allowed to reset them.

{Message edited to remove a few typo's} [Linked Image]


[This message has been edited by Trumpy (edited 12-01-2006).]
Posted By: Trumpy Re: GFCI vs. RCD - 12/01/06 02:09 PM
BTW Dave,
I think I must have mis-understood you, in your post about the Isolating Transformer and the RCD.
We have plates here that have both for bathrooms in hotels.
As in a Shaver socket, that would be the Isolating Tranny and anything else would be fed via the RCD.
However, I'm interested to know if a Selective Type RCD would be used up-stream of the 30mA unit in the bathroom?.
Type S being 100mA?.
I know that Aussie uses them, and we should do, but what is common in Ireland and the UK?.
Posted By: pauluk Re: GFCI vs. RCD - 12/01/06 05:32 PM
Quote
One thing that has to be remembered about a plain vanilla Residual Current Device. That's all it does, is sense an imbalance between Phase and Neutral in a circuit. A RCBO, has two functions, it senses like the RCD does, but it has the features of a Circuit Breaker built in to it.
There's certainly some confusion in the less enlightened DIY circles here as to the difference.

In this regard, the American GFCI fitted to a panel to protect one branch circuit is thus equivalent to the RCBO here (except for the lower trip current).

Quote
However, Voltage Operated ELCB's were a different kettle of fish, they were only used where the basics of earthing could not be provided and used to trip if the voltage got above 22VAC, with reference to earth (cold water pipe).
For those across the pond who are having trouble visualizing how these devices worked, have a look at this diagram , which is in the U.K. power systems thread in the reference area.

This shows how the voltage ELCB was wired to provide whole-house protection. Note that the protective grounds from throughout the house are not grounded directly but are instead connected via the operating coil of the ELCB. The rod connected to the other side of the coil to provide an earth reference had to be located outside of any overlapping resistance gradient of other rods/pipes.

Quote
Considering that they were sealed by the local Power Company here, the Home-owner was not allowed to reset them.
What was the idea behind that? [Linked Image]

The British ELCB was/is part of the homeowner's apparatus, and thus fully accessible for reset (or for testing with the button which deliberately connects the line to the coil through a suitable resistance).

Although obsolete for a good many years, there are still plenty of voltage ELCBs still in service in rural areas.

See this thread for some pictures of a typical old voltage-operated ELCB as used in domestic wiring:

Voltage-Operated Earth-Leakage Circuit-Breaker


[This message has been edited by pauluk (edited 12-01-2006).]
Posted By: Kenbo Re: GFCI vs. RCD - 12/05/06 11:27 AM
Reno

Here is a diagram of the RCDs used here in the UK

Sory it is on this web site (Paul) [Linked Image] but it appears to be the best diagram I can see to explain how it actualy works. (I use the same one to explain it to my students)

RCD diagram
Posted By: renosteinke Re: GFCI vs. RCD - 12/05/06 03:24 PM
Thanks, Kenbo!
Posted By: pauluk Re: GFCI vs. RCD - 12/05/06 05:05 PM
Quote
Sory it is on this web site (Paul)

Well, at least that website is good for something! [Linked Image]

Let's go into a little more history, starting with the typical domestic installation of the 1960s. The situation is complicated in the U.K. by the different types of earthing arrangements employed (see link to U.K. power diagrams above).

In areas served by TN-S (installation grounded by a completely separate conductor right back to the transformer) or TN-C-S (grounded to the neutral, American-style), ground-fault protection normally relied upon the regular fuses or MCBs (overcurrent circuit breakers). In rural TT areas (installation grounded only to a local rod) though, the loop impedance could be far too high to blow even a 5A fuse on a lighting circuit, nevermind a 30A fuse on a ring or range circuit. So in houses fed by TT, it was necessary to fit a main ELCB to provide ground-fault protection to the whole installation. This was normally the voltage-operated ELCB described above at that time, but current-operated ELCBs could also be used, often with 300 or even 500mA trip in the earlier days.

Since that time, the voltage ELCB has become obsolete, the current-operated ELCB became the RCCB/RCD and the trip current reduced, first to 100mA and then to 30mA which is the norm today.

Under the current Regs., RCD protection at 30mA is required for any outlet which is likely to be used to feed equipment outdoors (it is also needed on any other circuit in which the loop impedance and operating time of the overcurrent device cannot provide disconnection within the specified time, but that would be rare in a domestic setting).

So, we have some circuits which need RCD protection and some which do not. In TN-S/TN-C-S houses, that gives several options. We could fit an individual RCBO for each branch circuit which requires RCD protection. Unfortunately, even though costs are dropping the RCBO is still relatively pricey here, so for anything more than a couple of circuits it can soon become very expensive.

The alternative, and more common solution to date, has been to use a split-load board. There is a simple main isolator switch (not circuit breaker) for the whole board, then circuits which do not require RCD protection can be put on MCBs straight onto that main bus. The bus also feeds to a sub-main 30mA RCD, then circuits requiring RCD protection are placed on MCBs on that RCD bus. Thus a ground-fault on a circuit still knocks out more than just that one circuit, but not the whole house.

In rural TT areas -- where all circuits need RCD protection simply due to the high loop impedance -- it's still common to find just a single main 30mA RCD which protects the entire system (or a 100mA RCD installations dating back to the 1980s). That's far from ideal though, and it has led to wider use of the split-load board with two RCDs.

This works in a similar way to the split-load board on a TN-S/TN-C-S system, but with a 100mA delayed-trip RCD in place of the main isolator switch to guarantee that a fault on the 30mA bus will trip that sub-RCD and not the main one.

Below is a typical split-load board, main switch on the right, 30mA RCD protecting the left-hand part of the board. For TT use, we'd swap the main on the right for the 100mA delayed RCD.

[Linked Image]
Posted By: Trumpy Re: GFCI vs. RCD - 12/07/06 12:56 AM
Paul,
Quote
What was the idea behind that?
This happened (so I'm told by an older Faultsman) because a guy up on a farm tried to bypass the unit one day back in the 60's and nearly burned himself to death after he tried to tie the wrong wires together to re-instate the supply to his house, the pole fuse down the road never tripped because of corrosion.
After that, all of the units sported covers over the entire unit, except for a red neon that showed if the unit had tripped.
Our Deputy Chief Fire Officer also confirmed this story, apparently half the house was lost as well.
Posted By: pauluk Re: GFCI vs. RCD - 12/07/06 02:18 PM
Sounds like another case of one person doing something stupid resulting it a rule being put into place which makes life difficult for everybody else. [Linked Image]

When you say the pole fuse didn't blow, do you mean a fuse feeding just the one house or one on a whole section of distribution?

Over here, the voltage-ELCB was downstream of the individual house main fuse, i.e. Supply -> Main fuse -> Meter -> ELCB -> Distribution panel.

The modern main RCD (where used) is in exactly the same place in the system, except that it may be a separate unit or it may be incorporated as the main switch in the panel itself. The old voltage ELCBs were always a completely separate unit.
Posted By: Trumpy Re: GFCI vs. RCD - 12/07/06 04:21 PM
Paul,
Quote
When you say the pole fuse didn't blow, do you mean a fuse feeding just the one house or one on a whole section of distribution?
I'm stunned that you haven't been listening all the time that I've known you. [Linked Image]
The standard NZS 3431 pole fuse pertains to a 230V wire link rated at 60A.
That actual standard these days could be referenced to wooden buildings.
Once they started fitting pole fuses with HRC fuse links to BS 88 things changed here.
A pole fuse in my language is the fuse that if removed, will disconnect that installation.
Obviously a 3 phase installation will have 3 fuses.
There is a certain error here, where these are the LV fuses in our system and yes they are on poles, but the HV fuses are known as DDO's or Dominion Drop-Outs,these are also often on the same pole but higher up feeding a step-down transformer, usually 11kV/400V.
These fuses are known in the trade as Tops.
If you go to a callout here, you will be told by the control room that either 1, 2 or 3 tops have blown.
3 Tops is a real worry, it normally means a Temporary sub-station to bypass the faulty transformer.
Never again will I try and close DDO's onto a bad transformer like that, it rained fire that night.
33,000V of angry electricity.
Posted By: pauluk Re: GFCI vs. RCD - 12/07/06 04:28 PM
I thought you did mean the individual installation fuse, but the "down the road" part made me stop and think.

I guess with it being a farm though, "down the road" could mean a 400 yards down the farm's private entrance road. [Linked Image]
Posted By: RODALCO Re: GFCI vs. RCD - 12/30/06 03:23 AM
In The Netherlands they call them "Aardlekschakelaar"
which is translated into English
"Earth leakage switch"
They were 30 mA on non earthed circuits and 500 mA on earthed circuits bearing in mind that The Netherlands does not use the MEN system and has the earth wire separated from the neutral and connected to it's own earth mat.
These Aardlekschakelaars were usually fitted to protect a batch of 2 to 5 circuits at the time.

edited for typo's

[This message has been edited by RODALCO (edited 12-29-2006).]
© ECN Electrical Forums