Neutral to Earth

[ForkEnigma]

how many ohms can you have neutral to earth without opening a 30mA RCBO.

My thinking is If=V/R so, If=230/0.03 = 7.6Kohms

So 7.7kohms will keep it closed

 

[James]

the above formula would be correct for a live to earth fault.

N to E fault depends on the loads that are being drawn at the time (creating an imbalance in the voltage between N and E and therefore a current flow between them)

 

[ForkEnigma]

the above formula would be correct for a live to earth fault.

N to E fault depends on the loads that are being drawn at the time (creating an imbalance in the voltage between N and E and therefore a current flow between them)

the above formula would be correct for a live to earth fault.

N to E fault depends on the loads that are being drawn at the time (creating an imbalance in the voltage between N and E and therefore a current flow between them)

A dead short of lets say 0.2 ohms N-E without current will not open RCBO, with a load it will open.

Now if the fault was >7666 ohms, with current will the RCBO open.

 

[James]

Think about how an rcd operates.

It is a simple comparison between live current and neutral current.

An imbalance between those measurements of >30mA will cause it to trip.

It is not possible to give a resistance between N and E that will cause it to trip or not trip.

A fairly low resistance fault could be tolerated if there is zero load on the entire installation
As loads increase, not only on the circuit protected by rcd but also others, a combination of parallel paths and differing volt drops in parts of the installation can cause it to trip.

 

[ForkEnigma]

Think about how an rcd operates.

It is a simple comparison between live current and neutral current.

An imbalance between those measurements of >30mA will cause it to trip.

It is not possible to give a resistance between N and E that will cause it to trip or not trip.

A fairly low resistance fault could be tolerated if there is zero load on the entire installation
As loads increase, not only on the circuit protected by rcd but also others, a combination of parallel paths and differing volt drops in parts of the installation can cause it to trip.

I'm not talking about imbalance, which we know is how an RCD operates. So forget that for now.

I'm talking about a Neutral to Earth resistance that is greater than 7666 ohms. On load will it open the RCD ?
[automerge]1594834846[/automerge]
Anything less than 7666 ohms on load will open the RCD.

 

[James]

I'm not talking about imbalance, which we know is how an RCD operates. So forget that for now.

I'm talking about a Neutral to Earth resistance that is greater than 7666 ohms. On load will it open the RCD ?

depends on the load, that is what creates the imbalance

 

[ForkEnigma]

depends on the load, that is what creates the imbalance

A load of anything above 30mA will open a RCD when the fault is below 7666 ohms, but regardless of load above 30mA if the resistance is greater than 7666 ohms then RCD stays closed

 

[Mark Wright]

an impedance of 7666 Ohms would be the absolute maximum impedance like equivalent max Zs for a RCD if you ever had an impedance of 7.6 kiloOhms there is a serious problem.

 

[James]

Like I said in my first post, 7.6k is the maximum live to earth fault value that can be tolerated

 

[ForkEnigma]

an impedance of 7666 Ohms would be the absolute maximum impedance like equivalent max Zs for a RCD if you ever had an impedance of 7.6 kiloOhms there is a serious problem.

Indeed sir, it shows degradation of the insulation.
Max Zs is 1667 ohms on a 30mA which is touch voltage of 50V.

In doing EICRs its impossible to disconnect everything, especially with lighting loads.
[automerge]1594835934[/automerge]

Like I said in my first post, 7.6k is the maximum live to earth fault value that can be tolerated

James the OP states N-E

 

[James]

so to trip the rcd Il needs to be different from Ina by greater than 30ma
Ina= In-If
If is dependant on both the load Rn, and the parralel resistors Rfault and Rn (nutrual conductor resistance.)
the 7.6k value you are mentioning would cope with a load of zero ohms or a dead short.

 

[Mark Wright]

I think if you get values above 1kohms you need to check your earthing conductor, supply or earth rod etc.

 

[ForkEnigma]

Like I said in my first post, 7.6k is the maximum live to earth fault value that can be tolerated

Surely you mean minimum

 

[Mark Wright]

any impedance under 7.6k will allow 0.03Amps you would need a maximum insanely high impedance of above 7.6k for there to be a problem so 7.6k is the maximum impedance that can be allowed etc.

 

[ForkEnigma]

I think if you get values above 1kohms you need to check your earthing conductor, supply or earth rod etc.

......WHAT, we are talking resistance between N & E! Should be >.5 meg ohm right? To comply with BS7671, Which is much higher than what we are talking about.

 

[Mark Wright]

I think you mean insulation resistance not resistance there different.

 

[ForkEnigma]

I think you mean insulation resistance not resistance there different.

Ok, i feel really stupid now.

 

[Mark Wright]

Why what did you do take care

 

[James]

now here is where it gets a little complicated.

depending on the other loads Rext and the resistance of the bond between E and N, the voltage of the N line can vary.
if it goes up, some of the current from Rext will pass back through the rcd and down the parallel path through Rn and Rfault SUBTRACTING from the current passing through the neutral line on the rcd.
this will INCREASE difference between Il and Ina and help the rcd to trip.

 

[ForkEnigma]

View attachment 59486
now here is where it gets a little complicated. give me 5 to edit

Oh i like it!
[automerge]1594837207[/automerge]
Not sure of the R bond though
[automerge]1594837229[/automerge]
loving the diagram, much respect

 

[James]

Oh i like it!
[automerge]1594837207[/automerge]
Not sure of the R bond though
[automerge]1594837229[/automerge]
loving the diagram, much respect

As we know, N is at 0v and bonded to earth at some point in the system, either at the point of supply or further up the supply chain. this resistance can vary immensely depending on what type of supply you have.
[automerge]1594837978[/automerge]
p.s. I am enjoying the technical chat, whilst having a beer.
certainly beats posts about peoples lights not working after they have (male chicken'd) about with them.
I am not saying I am right, just trying to put some science into the conversation so we all know what we are looking at. polite criticism is welcome, i have not done all the parallel path calculations but from what i can see, it is not as simple as Rfault must be below "x" for the rcd to stay latched.
it truly does depend on many other contributing factors.

 

[Lucien Nunes]

I'll take a step back and work up to James's (correct) observation.

At the tripping threshold, Rins = V (from faulted conductor to earth) / IΔN.
I.e. the resistance of faulty insulation that will just allow the tripping current to flow, equals the voltage across the faulty insulation divided by the tripping current. It is this voltage that we need to find, since we know the tripping current.

Considering first the case of an L-E fault, the voltage across the faulty insulation is the nominal supply voltage Uo (typically 230V) plus or minus its usual tolerance, minus any voltage drop in the conductors. These variable factors are typically a small fraction of Uo, therefore we can reasonably use Uo/IΔn to discover the minimum insulation resistance; it's the familiar 230/0.03=7.7kΩ and it is not going to vary much.

Considering an N-E fault instead, the voltage across the faulty insulation is now much lower than Uo because neutral is nominally at earth potential. The small voltage that does exist to drive current through the fault is the sum of various voltage drops along neutral and earthing conductors that form a loop between the fault and the point at which N and E connect. This is the origin if it is a TN-C-S supply, the substation if it is TN-S. Since the resistances of these cables and the magnitudes and directions of the currents flowing through them are unknown, possibly unknowable (noting that this might include current from other installations flowing through the DNO's cables and/or local earthing and bonding conductors) there is no predictable voltage from which to calculate the resistance.

We can merely estimate the order of magnitude, based on typical N-E voltages of a few volts when 1-2% of Uo is dropped along neutral conductors. At 2% of 230V the fault resistance would need to be lower than 153Ω to cause a trip, but with minimal load on it might be an order of magnitude or two lower still. There will be situations, as mentioned above, where even a hard short does not cause a trip because the fault loop voltage is so near zero that the resistance of the circuit conductor is sufficient to prevent 30mA flowing.

 

[ForkEnigma]

Let's not go back to the origin of a supply. As we know with any TN system, neutral is connected to earth, this is not what an RCD would monitor. Also Neutral is a live conductor, and will have 230V passing through it.
From the load side of the RCD is a value of R (N-E) is what I'm looking for.
153 ohms is quite low, so are you suggesting that an 200 ohm would NOT open the RCD with 30mA flowing through circuit?

I'm going to do a practical experiment on this, with some resistors that I have in my kit.

Thanks for a rather in depth reply, although I've got to keep reading it over, its rather hard to digest at 5 30 am in the morning.

Peace brothers

 

[James]

Wear some safety glasses, a small mistake could lead to exploding resistors.

 

[ForkEnigma]

Wear some safety glasses, a small mistake could lead to exploding resistors.

I feel a you tube video coming. Lol

 

[davesparks]

Max Zs is 1667 ohms on a 30mA which is touch voltage of 50V.

For a TN system it is 7667 ohms not 1667, that is for TT systems.

 

[ForkEnigma]

For a TN system it is 7667 ohms not 1667, that is for TT systems.

Where the maximum Zs value for a protective device cannot be satisfied in a TN system, Regulation 531.3. 1 permits an RCD (including an RCBO) to be used to provide automatic disconnection of the supply. Table 41.5 would therefore apply and the maximum tablulated Zs value for a 30mA RCD is 1667 ohms

 

[Lucien Nunes]

neutral is connected to earth

Also Neutral is a live conductor, and will have 230V passing through it.

230V with respect to what? A voltage cannot 'pass through' a conductor, it is the potential between two points. Revisit PC1966's post and mine to get a clear picture of the source of the voltage that appears across the N-E fault. That will demonstrate why we need to consider the cables back to the origin and beyond, and why the minimum resistance to cause a trip cannot be predicted and could be as low as zero ohms.

Zs>7.7kΩ: RCD won't trip (at least, cannot be relied on to trip, it's allowed to trip down to 15mA)
7.7kΩ>Zs>1.7kΩ: RCD trips, touch voltage >50V so cannot be used as fault protection
Zs<1.7kΩ: RCD trips, touch voltage <50V, everyone's happy.

 

[ForkEnigma]

Revisit PC1966's post

Excuse me what post is that? Certainly not in this thread, and not one I have visited before.

A disconnected neutral has the potential to be live with 230V to earth.

As we all know a neutral is a live conductor.
[automerge]1594893423[/automerge]
Zs>7.7kΩ: RCD won't trip (at least, cannot be relied on to trip, it's allowed to trip down to 15mA) 7.7kΩ>Zs>1.7kΩ: RCD trips, touch voltage >50V so cannot be used as fault protection Zs<1.7kΩ: RCD trips, touch voltage <50V, everyone's happy.

Indeed sir, but dosnt quite answer the OP.

 

[Lucien Nunes]

Sorry, I meant James's post #19.

The neutral is a live conductor but under normal operating conditions it is near earth potential. On a single-phase system, that is the only factor that makes it a neutral, so it's only meaningful to consider Rins of a neutral conductor, when it is intact and near earth potential as intended.
[automerge]1594903135[/automerge]
I've drawn out the voltage divider explicitly here, for the very simple case that there is one circuit on a TN-S supply with an insulation fault at the end, in either the line or the neutral. In this case we can calculate the minimum value of Rins based on the voltage at each of the two possible fault positions.

 

[Mark Wright]

What I don't understand is if you have a circuit in series the voltage will be different and in parallel the the current will be different at the end also obviously the load will use some so how can what's going out on the live be monitored coming back on the Neutral it won't be the same?
[automerge]1594906338[/automerge]
@op JW has a video call RCD neutral earth fault not watched it yet.
but..... there is another video by bigclivedotcom iirc he takes an RCD apart and inside is a Resistor used for the test button when you press the test button its not just mechanical the Resistor actually lets a current flow to trip the RCD - he can actually see the value of the resistor -check that video out.

 

[DPG]

What I don't understand is if you have a circuit in series the voltage will be different and in parallel the the current will be different at the end also obviously the load will use some so how can what's going out on the live be monitored coming back on the Neutral it won't be the same?
[automerge]1594906338[/automerge]
@op JW has a video call RCD neutral earth fault not watched it yet.
but..... there is another video by bigclivedotcom iirc he takes an RCD apart and inside is a Resistor used for the test button when you press the test button its not just mechanical the Resistor actually lets a current flow to trip the RCD - he can actually see the value of the resistor -check that video out.

I think the RCD test functionality is fairly well known. But always worth a reminder.

 

[Lucien Nunes]

how can what's going out on the live be monitored coming back on the Neutral it won't be the same

Current can't just disappear. If it goes down the line conductor every bit of it ought to go through the load(s) and come back up the neutral. Any that doesn't has escaped the circuit, probably to earth, which is exactly the leakage that the RCD is looking for.

 

[ForkEnigma]

p.s. I am enjoying the technical chat, whilst having a beer. certainly beats posts about peoples lights not working after they have (male chicken'd) about with them.

Yeah good fun isn't it. Trust me there will be a lot more! I too was having a beer when I started this thread.

 

[davesparks]

Where the maximum Zs value for a protective device cannot be satisfied in a TN system, Regulation 531.3. 1 permits an RCD (including an RCBO) to be used to provide automatic disconnection of the supply. Table 41.5 would therefore apply and the maximum tablulated Zs value for a 30mA RCD is 1667 ohms

I don't have a regs book to hand but that sounds like the pre-18th requirements

 

[ForkEnigma]

I don't have a regs book to hand but that sounds like the pre-18th requirements

The physics have not changed, and the values are the same in the 18th