Will an RCD offer any protection in a building with NO earth circuit?

[GeeGee50]

Hi Westward,

I needed to come back to this post because things are slowly starting to click in my head (thank you to all you have been considerate to educate and reply so far, I really appreciate it).

There is a few its and buts here. If the frame is not earthed, say it is floating in the air and completely isolated then there is nowhere for any current to go and as such the RCD shouldn't sense an imbalance.

Another scenario, again the frame is not earthed by way of the electrical system but has a tenuous contact to earth. For instance a washing machine sitting on a damp floor which is likely to allow leakage to flow hence you get an imbalance.

I now see why you said there's a few IFs and BUTs in that video.

Current Leakage

For example, in the video example already provided (i.e. live wire touching metal frame, no earthing), it states the following (around the 3:00 mark):

'When the current will flow through the RCCB and then through the load, there will be leakage current in the body, but as there is a difference between the phase current and neutral current, the RCCB will trip"

I incorrectly interpreted this to mean there is current being lost by simple virtue of it flowing through a metal frame (yes dumb I know). I now understand that to mean that current is literally 'leaking' to the earth (thanks to your 2nd scenario above).


No Imbalance

I also think I now understand why you've been saying why there's no imbalance, esp. given the reply from @Simon47 below.

If the live supply wire comes free from it's terminal and touches the frame then two things happen :
1) The supply current will stop. There's no longer a path from live to neutral via the device load, and no path to earth from the frame, therefore when you apply Kirchoff's law, there can be no current in the live. No current in the neutral, no current in the live, therefore no current imbalance and the RCD will not trip.

However, this contradicts the video which states that when a Live wire touches a metal frame, the current can still flow back to the RCD, causing it to trip. I get the video is making an example, but is this a plausible real-world scenario?

The scenario would be:

• Frame is not earthed
• Live touches metal frame
• Current flows through metal frame
• Current flows from metal frame to neutral
• RCD doesn't trip as there is a complete circuit and no loss of current

Can this situation exist in the real world (@westward10 / @Simon47)?

Cheers

 

[Simon47]

If the load is still intact, and there is a L-E (frame) fault, then it helps to separate the current flows which we'll assume for now are independent.
1) The normal load current. For this, the currents in L&N will balance and not trip the RCD.
2) The current due to the fault. This flows from whatever has contacted the frame (we'll assume the L supply) and down to the protective earth connection. The size of this depends on the circuit impedance. If the frame is perfectly isolated from ground, nothing will flow. If it's well connected then a large current will flow and the supply over-current protection (fuse or circuit breaker) will blow/trip. In between these two extremes, some current will flow and if it's large enough then either the RCD or overcurrent protection (or both) will trip.
In a properly wired UK system, any fault current will NOT flow back through the neutral connection - it's expressly forbidden to have the circuit neutral share a connection with the earth connection (at least, once past the supplier's service head and you'll find some threads here and over in the DIY-Not forums about "lost PEN" faults (loss of the protective earth and neutral, PEN, conductor) on the supplier side. Since the earth connection does not go through the imbalance sensing coil of the RCD, any current flowing in the circuit's earth conductor will not balance that in the live conductor and (if large enough) will trip the RCD.

Now, you are not in the UK, and i gather some countries have what we might describe as "interesting" ideas regarding electrical safety. As you have no earth wiring, it sounds like Cambodia is one of them. If exposed conductive parts are "earthed" by connecting them to the supply neutral (which we'll assume is earthed, making it a TN-C system) then things do get more interesting.

If it's as simple as having a 2-pin plug & socket, and the appliance case is connected to neutral, then a L-case (frame) fault will not trip an RCD - there will be a current from L, through the case, and back out of the neutral wire. Since (again, apply Kirchoff's current law) "what goes in must come out", L & N currents will still balance and the RCD will not trip - depending on the nature of the fault, the overcurrent protection may trip, or the case may become live to some extent. Only if someone touches the case while also touching something "earthy" will a current to earth flow - and that should trip the RCD. On that basis, having an RCD will be safer than not having one as long as you don't then lower your guard because you consciously or subconsciously think "it's safe now I've got an RCD".
Possibly the worst case fault in this situation would be a broken neutral somewhere together with another fault. That could make the case of every appliance go live even when they are switched off ! That's the situation I mentioned above about loss of the PEN on the supplier's side - there's scope for the "earth" connection in a property to become live, not normally a problem inside as we have equipotential bonding, but a problem once we export that equipotential zone outside (such as to a metal tap or an EV on charge).


As to whether you can usefully do much about it, that depends on local custom, local regulations, and your personal situation. For example, if you are in a position to create your own earth electrode (a.k.a. earth rod), then you could potentially convert to a TT system in which case you would need RCDs to provide safety, but they would probably protect better than you have now.
Try this page which will probably help, or this page.

 

[RetiredNY]

However, this contradicts the video which states that when a Live wire touches a metal frame, the current can still flow back to the RCD, causing it to trip. I get the video is making an example, but is this a plausible real-world scenario?

The scenario would be:

• Frame is not earthed
• Live touches metal frame
• Current flows through metal frame
• Current flows from metal frame to neutral
• RCD doesn't trip as there is a complete circuit and no loss of current

Can this situation exist in the real world (@westward10 / @Simon47)?

Cheers

If frame is not earthed, and live touches metal frame, current will be present on metal frame and RCD will not trip. If Neutral is connected to metal frame, RCD will not trip but this is a short circuit condition and then the breaker will trip.

If neutral has been connected to frame as the device has been in use, and the electrical service to the premises is neutral grounded to earth, and the utility, then the RCD would trip whenever a person came in contact with the frame and earth, creating an alternate return path to neutral at the premises utility and/or system grounds as long as the milli-amp leakage is enough.

 

[Simon47]

If frame is not earthed, and live touches metal frame, current will be present on metal frame and ...

VOLTAGE will be present, current may or may not flow depending on whether a circuit is formed. Let's not keep confusing matters by using wrong terminology. This is wrong (voltage and current are different things), not just a matter of a difference in opinion on terminology (e.g. "plug top" vs. "plug", or "transformer" for what is generally these days a switch mode supply).

 

[RetiredNY]

VOLTAGE will be present, current may or may not flow depending on whether a circuit is formed. Let's not keep confusing matters by using wrong terminology. This is wrong (voltage and current are different things), not just a matter of a difference in opinion on terminology (e.g. "plug top" vs. "plug", or "transformer" for what is generally these days a switch mode supply).

In the scenario I was responding to (see post above mine) poster says live touches frame, frame is not earthed, current flows through neutral. Poster was describing a dead short.

 

[mainline]

In the scenario I was responding to (see post above mine) poster says live touches frame, frame is not earthed, current flows through neutral. Poster was describing a dead short.

A live touching a metal frame that is connected to neutral or earth doesn’t mean it’s going to be a short circuit and cause the ocpd to trip, it’s going to depend on the impedance of the circuit.

It could be for instance that the disconnected neutral side of a lighting circuit contacts the metal frame.

 

[RetiredNY]

A live touching a metal frame that is connected to neutral or earth doesn’t mean it’s going to be a short circuit and cause the ocpd to trip, it’s going to depend on the impedance of the circuit.

It could be for instance that the disconnected neutral side of a lighting circuit contacts the metal frame.

I assumed neutral was solidly connected to metal frame, in this hypothetical example.

 

[mainline]

I assumed neutral was solidly connected to metal frame, in this hypothetical example.

And as already said that still doesn’t mean it’s going to be a short circuit and trip the breaker.

 

[Simon47]

However, this contradicts the video which states that when a Live wire touches a metal frame, the current can still flow back to the RCD, causing it to trip. I get the video is making an example, but is this a plausible real-world scenario?

The scenario would be:

• Frame is not earthed
• Live touches metal frame
• Current flows through metal frame
• Current flows from metal frame to neutral
• RCD doesn't trip as there is a complete circuit and no loss of current

Can this situation exist in the real world (@westward10 / @Simon47)?

OK, "live touches frame" is a fault - it can happen and that's why we design in features to deal with it.

"Current flows from metal frame to neutral" is a different issue.

Here in the UK, that would need a separate fault as the frame(case) would be earthed. So we've be looking at two faults for that to occur - and the device would have stopped working with either one of them. Otherwise it would need a fairly improbably fault - e.g. power cord pulled such that all 3 cores get disconnected, L&N both tough the frame together, but E does not touch the frame.

However, I believe in some countries that's not the case - I suspect your's may be one of them. If you only have 2 wires, then the only way to "earth" the frame will be to use the neutral and rely on neutral being tied to earth "somewhere". So in that situation, yes a current could flow from L to N via the frame - and because that would be balanced it would not trip an RCD.

Such systems are inherently dangerous though. If the neutral connection fails outside of the equipment (and it could be outside of your house), then the frame will be connected to live via the internal load. That means a single (and not uncommon) fault can make the case live - there will be no current as the neutral is broken, so the RCD will not trip. If you touch the case and provide a path to earth then an RCD should trip and protect you - so adding it will provide protection from some faults, but not others.

 

[GeeGee50]

If Neutral is connected to metal frame, RCD will not trip but this is a short circuit condition and then the breaker will trip.

Cool, tx.

A live touching a metal frame that is connected to neutral or earth doesn’t mean it’s going to be a short circuit and cause the ocpd to trip, it’s going to depend on the impedance of the circuit.

It could be for instance that the disconnected neutral side of a lighting circuit contacts the metal frame.

Ok, noted.

I assumed neutral was solidly connected to metal frame, in this hypothetical example.

You assumed right.

If frame is not earthed, and live touches metal frame, current will be present on metal frame and RCD will not trip.

I'm assuming you mean where the metal frame is not connected to Neutral?

If so, I still don't understand why the RCD would not trip.

Here is my logic:

• Metal frame is not earthed AND frame is not connected to Neutral
• Appliance is on
• Live touches metal frame
• Live-neutral circuit is broken and therefore a complete circuit does not exist
• Current is unable to make its way back to RCD
• RCD notices imbalance and trips <------------ is this correct?

GeeGee

 

[GeeGee50]

OK, "live touches frame" is a fault - it can happen and that's why we design in features to deal with it.

"Current flows from metal frame to neutral" is a different issue.

Here in the UK, that would need a separate fault as the frame(case) would be earthed. So we've be looking at two faults for that to occur - and the device would have stopped working with either one of them. Otherwise it would need a fairly improbably fault - e.g. power cord pulled such that all 3 cores get disconnected, L&N both tough the frame together, but E does not touch the frame.

Yes, agreed in terms of earthing. I do understand and appreciate why earthing is critical for safety.

My scenario is purposefully leaving out any earthing of the frame as I want to understand how an RCD will behave (even if such a setup is highly unlikely and not advised).

 

[brianmoooore]

If a complete circuit does not exist then the RCD will not trip.
If you use your body to create a complete circuit, by simultaneously being in contact with the conductive ground and the frame of the faulty appliance, then the RCD will trip if the resulting current is high enough. This will be extremely painful, but in most cases won't be fatal.
If the current flowing is insufficient to trip the RCD, the effects will range from a barely perceptible tingle to a continuous pain that will encourage you to let go as soon as possible.

 

[RetiredNY]

If so, I still don't understand why the RCD would not trip.

Here is my logic:

• Metal frame is not earthed AND frame is not connected to Neutral
• Appliance is on
• Live touches metal frame
• Live-neutral circuit is broken and therefore a complete circuit does not exist
• Current is unable to make its way back to RCD
• RCD notices imbalance and trips <------------ is this correct?

GeeGee

OK, there is no current in your example. In order to have current flow, a circuit must exist. Voltage would be present on the metal, un-earhted, un-neutral referenced frame, but there is no current flow. RCD sees no imbalance because current needs to flow through both the neutral and the live conductors it is monitoring. Should, however a person who is earthed reach out and touch the live metal frame, then current will flow through the person to earth, and not return through the neutral, so the RCD sees current flow, but no equal current return, that creates an imbalance or a magnetic field in the RCD coil, which generates a micro-voltage to trip the RCD off.

 

[GeeGee50]

If a complete circuit does not exist then the RCD will not trip.
If you use your body to create a complete circuit, by simultaneously being in contact with the conductive ground and the frame of the faulty appliance, then the RCD will trip if the resulting current is high enough. This will be extremely painful, but in most cases won't be fatal.
If the current flowing is insufficient to trip the RCD, the effects will range from a barely perceptible tingle to a continuous pain that will encourage you to let go as soon as possible.

tx mate...

I think I'm missing the finer technicalities of how an RCD operates.

In my mind (and in my scenario), there IS a complete circuit initially BEFORE the live wire touches the frame.....so at the instance the live touches the frame, then the current that went out via the RCD will fail to come back into the RCD....doesn't that meet the definition of an imbalance?

 

[RetiredNY]

tx mate...

I think I'm missing the finer technicalities of how an RCD operates.

In my mind (and in my scenario), there IS a complete circuit initially BEFORE the live wire touches the frame.....so at the instance the live touches the frame, then the current that went out via the RCD will fail to come back into the RCD....doesn't that meet the definition of an imbalance?

Your scenario leaves out any earthing of the frame. So it's assumed the frame has no connection to earth or neutral. The live touching the frame doesn't complete any circuit, therefore there is no current flow for the RCD to measure. Therefore, all's right with the world from the perspective of the RCD.

Now along comes poor Joe, who is barefoot on a moist kitchen floor, and touches the live frame. Now some current flows through the RCD's coil, but it doesn't return through the circuit's intended path, the neutral, to cancel the current the live is experiencing. This current will be detected by the RCD if it exceeds it's trip point.

 

[GeeGee50]

Your scenario leaves out any earthing of the frame. So it's assumed the frame has no connection to earth or neutral. The live touching the frame doesn't complete any circuit, therefore there is no current flow for the RCD to measure. Therefore, all's right with the world from the perspective of the RCD.

Now along comes poor Joe, who is barefoot on a moist kitchen floor, and touches the live frame. Now some current flows through the RCD's coil, but it doesn't return through the circuit's intended path, the neutral, to cancel the current the live is experiencing. This current will be detected by the RCD if it exceeds it's trip point.

Yes, correct assumption. That is, my scenario has always been:

• appliance frame not connected to earth
• frame not connected to negative

OK, so everyone keeps saying an RCD needs a complete circuit to work (and I'm not disputing that).

However, It still makes no sense to me why an RCD can't detect when a perfectly working circuit is suddenly lost (with regards to my scenario), especially if it's lost whilst current is running through the circuit.

I've drawn a diagram to better illustrate my point:

 

[mainline]

Yes, correct assumption. That is, my scenario has always been:

• appliance frame not connected to earth
• frame not connected to negative

OK, so everyone keeps saying an RCD needs a complete circuit to work (and I'm not disputing that).

However, It still makes no sense to me why an RCD can't detect when a perfectly working circuit is suddenly lost (with regards to my scenario), especially if it's lost whilst current is running through the circuit.

I've drawn a diagram to better illustrate my point:



View attachment 112956

Yes, correct assumption. That is, my scenario has always been:

• appliance frame not connected to earth
• frame not connected to negative

OK, so everyone keeps saying an RCD needs a complete circuit to work (and I'm not disputing that).

However, It still makes no sense to me why an RCD can't detect when a perfectly working circuit is suddenly lost (with regards to my scenario), especially if it's lost whilst current is running through the circuit.

I've drawn a diagram to better illustrate my point:



View attachment 112956

In the 1 st scenario the return current in both the L and N are the same (balanced) so the rcd won’t trip.

In the 2nd scenario there is no current flow at all because the frame doesn’t connect to E or N so the rcd won’t trip.

 

[GeeGee50]

In the 1 st scenario the return current in both the L and N are the same (balanced) so the rcd won’t trip.

In the 2nd scenario there is no current flow at all because the frame doesn’t connect to E or N so the rcd won’t trip.

So what happened to the 10A active in the circuit before the circuit was broken?

 

[James]

It stopped flowing the moment you broke the circuit.
in the same way that the current (Amps) flowing through your lounge light stops flowing the moment you turn off the light switch.

 

[GeeGee50]

It stopped flowing the moment you broke the circuit.
in the same way that the current (Amps) flowing through your lounge light stops flowing the moment you turn off the light switch.

hmmm

so it stopped flowing, ok.....but what about the current that was already present in the circuit after the circuit was broken??......it just disappears into thin air?