240V GFCI, No EGC

[Lioneye]

Thanks all.

 

[LarryFine]

Yes. Standard 120/240 200A resi service at main structure. Neutral, GEC, bonding to metal piping, IBT, EGC to branches in house all verified. They could install a 2-pole GFCI and terminate the device neutral to the grounded terminal. The 240V branch circuit to the well is definitely 2-wire only.

Then you're golden. Put the GFCI breaker in the panel with the neutral pigtail on the neutral bus, and the entire circuit and well are protected. More correctly, people are protected from the circuit and well casing.

If a 4ma to 6ma current flows from the circuit to earth, the GFCI will trip regardless of where the nearest electrically-grounded surface is.

 

[winnie]

Just a small addendum: GFCI devices are commonly described as 'comparing' the current on the circuit wires and tripping if there is a difference. This is a fair description of what is going on, however it isn't how the GFCI is implemented internally, and thus can sometimes cause misunderstanding. If you have some sort of balanced fault then the GFCI might see exactly the same RMS current flowing on both hot and neutral (or hot and hot for 240V) and yet still trip.

The way a GFCI functions is that all of the circuit conductors pass through a single current transformer. Because of Kirchhoff's Current Law, any current flowing 'out' on one circuit conductor must be exactly balanced by current flowing 'in' on the other circuit conductor(s), unless some current is flowing someplace else (eg. a ground fault, or a fault to another circuit). The _net_ current seen by the GFCI is thus zero when the protected circuit is functioning properly. The 'sum of all currents' measurement is what the GFCI is looking at.

-Jon

 

[anthonysolino]

Just a small addendum: GFCI devices are commonly described as 'comparing' the current on the circuit wires and tripping if there is a difference. This is a fair description of what is going on, however it isn't how the GFCI is implemented internally, and thus can sometimes cause misunderstanding. If you have some sort of balanced fault then the GFCI might see exactly the same RMS current flowing on both hot and neutral (or hot and hot for 240V) and yet still trip.

The way a GFCI functions is that all of the circuit conductors pass through a single current transformer. Because of Kirchhoff's Current Law, any current flowing 'out' on one circuit conductor must be exactly balanced by current flowing 'in' on the other circuit conductor(s), unless some current is flowing someplace else (eg. a ground fault, or a fault to another circuit). The _net_ current seen by the GFCI is thus zero when the protected circuit is functioning properly. The 'sum of all currents' measurement is what the GFCI is looking at.

-Jon

the CT, being both wires pass through it, the current negates its self I believe? and thats what the device monitors ?

 

[synchro]

Just a small addendum: GFCI devices are commonly described as 'comparing' the current on the circuit wires and tripping if there is a difference. This is a fair description of what is going on, however it isn't how the GFCI is implemented internally, and thus can sometimes cause misunderstanding. If you have some sort of balanced fault then the GFCI might see exactly the same RMS current flowing on both hot and neutral (or hot and hot for 240V) and yet still trip.

The way a GFCI functions is that all of the circuit conductors pass through a single current transformer. Because of Kirchhoff's Current Law, any current flowing 'out' on one circuit conductor must be exactly balanced by current flowing 'in' on the other circuit conductor(s), unless some current is flowing someplace else (eg. a ground fault, or a fault to another circuit). The _net_ current seen by the GFCI is thus zero when the protected circuit is functioning properly. The 'sum of all currents' measurement is what the GFCI is looking at.
-Jon

An example of what Jon is saying how two wires with the exactly the same RMS current IRMS could trip a GFCI is the following: equal L-N loads on a 2-pole breaker without a load neutral terminal, and with it fed by two phases of a 208V wye. In this case the sum of these two currents is not zero even if their individual RMS values are identical. Instead, their sum is actually the same amount of current IRMS but at +-120 degrees respectively from the two currents, and this summed current returns through the neutral connection to the wye. Obviously the breaker would immediately trip in this situtation and there would need to be a GFCI breaker that included the neutral for it to work properly.
I'm just mentioning this extreme example, which won't happen with the proper equipment and connections, just to drive home the point that Jon mentioned. In the example mentioned the two currents going through the breaker have the same RMS value but they will trip it because the two currents do not sum to zero (that would require them to be at 180 being degrees from each other).

 

[LarryFine]

An example of what Jon is saying how two wires with the exactly the same RMS current IRMS could trip a GFCI is the following: equal L-N loads on a 2-pole breaker without a load neutral terminal, and with it fed by two phases of a 208V wye. In this case the sum of these two currents is not zero even if their individual RMS values are identical. Instead, their sum is actually the same amount of current IRMS but at +-120 degrees respectively from the two currents, and this summed current returns through the neutral connection to the wye.

In the example mentioned the two currents going through the breaker have the same RMS value but they will trip it because the two currents do not sum to zero (that would require them to be at 180 being degrees from each other).

I disagree. Without a neutral reference, the 2-wire circuit will be seen as a single, symmetrical sine wave. The voltage being 208v instead of 240v is the result of the 120-degree difference, but there is still only a single, symmetrical current.

 

[synchro]

I disagree. Without a neutral reference, the 2-wire circuit will be seen as a single, symmetrical sine wave. The voltage being 208v instead of 240v is the result of the 120-degree difference, but there is still only a single, symmetrical current.

The example I mentioned it is actually a 3-wire circuit because of the two L-N loads, with a load on each of the two phases (e.g., L1-N, L2-N). If it was only a line-to-line load then I agree with your comment.

 

[LarryFine]

I gotcha. I thought you meant a GFCI breaker with no load neutral terminal, like a QO260GFI.

Still, L1- N loads will have a symmetrical current, L2-N loads will, and L1-L2 loads will. They should all still total to zero.

Added: Wait! You did say a breaker without a load neutral terminal. Of course it will trip with any line-to-neutral load!

 

[synchro]

I gotcha. I thought you meant a GFCI breaker with no load neutral terminal, like a QO260GFI.

Still, L1- N loads will have a symmetrical current, L2-N loads will, and L1-L2 loads will. They should all still total to zero.

Added: Wait! You did say a breaker without a load neutral terminal. Of course it will trip with any line-to-neutral load!

And even balanced line-to-neutral loads on the two phases.

 

[LarryFine]

And even balanced line-to-neutral loads on the two phases.

Yep, but then all of the above applies to a GFCI with no neutral terminal on 120/240v 1ph, too.

Regardless of the source characteristics, all CCCs must pass through the CT in the GFCI.

 

[synchro]

Yep, but then all of the above applies to a GFCI with no neutral terminal on 120/240v 1ph, too.

Regardless of the source characteristics, all CCCs must pass through the CT in the GFCI.

Which is why I said:

I'm just mentioning this extreme example, which won't happen with the proper equipment and connections, just to drive home the point that Jon mentioned.

 

[wwhitney]

Yep, but then all of the above applies to a GFCI with no neutral terminal on 120/240v 1ph, too.

Regardless of the source characteristics, all CCCs must pass through the CT in the GFCI.

On 120/240V 1ph, if the two L-N loads are perfectly balanced, there would be no neutral current. So a GFCI that is only connected to the two ungrounded legs but not the neutral would not trip.

Whereas on a 120/208 3-wire circuit, even if the loads were perfectly balanced, there would be a neutral. So a GFCI that is only connected to the two ungrounded legs will always trip when the load current exceed the 4-6 ma threshold.

Cheers, Wayne