UK TT undetectable PEN fault

[Hakuhou]

Reviewing earthing methods and come across this.
Suppose a TT system is installed where the installation earth penetrates the PEN, not completely servering it but enough that it effectively creates a second circuit (see image).
Am I right in thinking that this is effectively undetectable?
The RCD doesn’t trigger, and the overcurrent wouldn’t trigger.
So the entire installation becomes energized, waiting for a path of lower resistance?
Also would this keep re-energizing the installation even if a discharge event occurs (ie person touches installation)?
Lastly would the DNO have something in place to detect something like this?
From what I can tell you couldn’t even tell this were occurring when looking at the electricity meter of the installation.

 

[westward10]

When you say penetrates but not completely severing do you mean a physical contact or that the separation of the electrode is not enough to isolate it from anything connected to the PEN?

 

[Hakuhou]

When you say penetrates but not completely severing do you mean a physical contact or that the separation of the electrode is not enough to isolate it from anything connected to the PEN?

Thinking about this, let me rephrase.

The installation ground is installed in a manner that it becomes a secondary ground for the PEN, but without breaking the connection the PEN has with the source earth.

In my head the only way this can happen is if the secondary ground somehow splices the PEN.

But from the way you’ve phrased the second option it sounds like you can have the situation I’ve illustrated in the above with cutting into the PEN.

 

[DefyG]

Looks like its converted to TN-C-S

 

[Tom256]

Your installation Earth electrode is showing on the supply side, upstream of the RCD. I assume this is meant to be a TT installation with separated earthing from the TN-C-S. Have you shown it this way for a reason?
You have added to the MULTIPLE in PME.
Earthing systems need to be robust and separated from different systems of earthing. TT earthing in the vicinity of services that may be bonded to TN-C-S systems can indeed be an issue which is what Westward may be pointing you towards (?)

As a side: A PME system can have additional earth electrodes, indeed such is now an advisory point in BS 7671. This does not make the PME a TT system.
Have a read of 411.4.2

 

[Tom256]

Are you asking into PEN faults and how they can go undetected?

 

[mainline]

I suppose you would know if something was wrong with the rod as the Ra would be very low.

 

[Hakuhou]

Your installation Earth electrode is showing on the supply side, upstream of the RCD. I assume this is meant to be a TT installation with separated earthing from the TN-C-S. Have you shown it this way for a reason?
You have added to the MULTIPLE in PME.
Earthing systems need to be robust and separated from different systems of earthing. TT earthing in the vicinity of services that may be bonded to TN-C-S systems can indeed be an issue which is what Westward may be pointing you towards (?)

As a side: A PME system can have additional earth electrodes, indeed such is now an advisory point in BS 7671. This does not make the PME a TT system.
Have a read of 411.4.2

The original picture illustrates a TT installation

I’ve edited it:

Such that installation earth electrode has been installed badly in a manner in which it becomes an additional ground for the PEN, and links back to the installation, while keeping the PEN intact enough that current still managed to return on it back to the source.

This causes the same conditions for a PEN fault with the difference that the PEN is partially connected back to the source.

@mainline notes the resistance would be much lower than expected.

@Tom256 i understand that PEN faults go undetected as the RCB won’t detect a difference in the phase and return current (since there is none).

A PEN fault would also cause the installation to immediately become energized.

Is it correct that in this case the installation would only become intermittently energized depending on the resistance of the installation earth vs the source earth?

 

[Tom256]

Just pause a second, the PEN conductor is not partially connected back to the source in the arrangment as you have shown it in your diagram.

 

[Hakuhou]

Just pause a second, the PEN conductor is not partially connected back to the source in the arrangment as you have shown it in your diagram.

Making it clearer

 

[Tom256]

Well not sure that made it clearer (possibly just a me thing, I would accept that as a very real possibility).

A breakdown of the CNE conductor can of course occur yet with parallel paths formed from multiple sources such may not be detected. In your case the breakdown is because it has formed an additional connection to earth.
The supply sees no difference between neutral and earth currents flowing in the CNE conductor. The sum of which of course can change under loads and resistive paths.

One purpose for the standard approach to what we do with Earthing electrical systems is to have control over how a system will react under fault.
Once that is moved away from untold possibilities unfold as to what will happen to current and voltage under fault. A fault in one premises can cause havoc in another. Voltage gradients over ground, resistance values and touch voltages rising, with currents finding every which way back home.

 

[Tom256]

A PEN fault would also cause the installation to immediately become energized.

Just seen your edit after I posted.

A PEN fault is a rather subjective term and you need to consider this may well now affect multiple installations.
If you are talking about a CNE conductor break in the cable supplying the installation you may not detect a voltage on extraneous metalwork in the property.
Consider what might happen under fault. The Neutral currents divert (I know you are using earth as your reference but remember where the issue is on your supply) and only a disconnect in the earth at [a] premises may show a dangerous voltage in another.
Indeed sometimes a person will report a break in supply only to have it return for no apparent reason. The reason could be an earth being used was temporary disconnected in a property elsewhere.
Detecting such needs a clamp meter. Voltage indicators may not pick this up.

It is understanable to forget yet important to keep in mind with these examples that current will take every path open to it not just that with the least resistance

I refer you back to my comment on having control of what occurs under fault in my post number #11

Loads may still be working as they always do so, ask yourself how is it possible that everything is working?

 

[Hakuhou]

In your case the breakdown is because it has formed an additional connection to earth.

This is the expression I was after.

Indeed sometimes a person will report a break in supply only to have it return for no apparent reason. The reason could be an earth being used was temporary disconnected in a property elsewhere.

i need to think about how that would happen

Loads may still be working as they always do so, ask yourself how is it possible that everything is working?

As in when this occurs with PEN fault?

 

[Tom256]

In your example, due to poor installation practice, your intended TT installation now has a connection with the CNE (Combined Neutral Earth) conductor.
Perhaps they drove in the rod and clipped the supply cable. The supply cable underground is using a CNE conductor which is now in contact with the rod. The rod is not touching the line conductor. The CNE conductor is not severed but is damaged and with time may well fail and go open circuit.

The requirement was for a TT system but is that now what you have? What earthing system do you now have? Is the intended installation earthing now compliant ?
Should the CNE go open circuit many things could happen at this point, some all-together not expected. What may not happen is an obvious loss of supply to the installation.

 

[Tom256]

Just searched youtube and found a video for you that may help with your understanding. I have not watched it through but from my quick scan through it covers one of the points you have raised:

View: https://www.youtube.com/watch?v=z7han_twcnw