Livestock Water Tank

[kwired]

I agree with what you are saying. W/out a grounded source, it takes 2 faults for a shock hazard. Chances are that 1st fault is after the GFP and if it is, the GFCI will not see the 2nd fault as a problem. You would need a path for current that does not go back to the GFCI in order to trip it. So, if you had 2 faults after the GFCI, the GFCI will not work.

This is all theory, because almost ALL of what we work on is a grounded source. Grounded multiple times (which is part of some of the problems we have with NEV).

At least this is how I understand it. Maybe Larry will stop by and offer support.

Where you may run into this the most is with a portable generator that has no neutral to frame bonding, or other non utility source that isn't bonded/grounded.

That generator may even have factory mounted GFCI receptacles (I happen to have one like this). There is for the most part little shock hazard when used as "stand alone" system. You will need a second ground fault for there to be a problem. Not impossible but probably pretty rare.

Now take same generator and properly connect it to a premises wiring system and you should be running an EGC from the premises wiring system to the generator frame, and the premises wiring system should be grounded. In that situation the GFCI receptacle on the genset does have a EGC and will then trip on the first ground fault that occurs on load side of the GFCI.

So it isn't that GFCI's don't work on an ungrounded system, it is that it takes different circumstances to create a fault that is going to create potential shock hazard.

 

[oldsparky52]

You will need a second ground fault for there to be a problem. Not impossible but probably pretty rare.

It takes more than a 2nd ground fault, it takes a parallel path for current to the source before the GF, and you just don't have that in an ungrounded circuit.

 

[kwired]

It takes more than a 2nd ground fault, it takes a parallel path for current to the source before the GF, and you just don't have that in an ungrounded circuit.

second ground fault allows a parallel path, but to trip the GFCI it would have to flow around the GFCI and not through it. So yes there is potential for stray current on items not intended to carry current and if you get in that path or across potential created by that path then there is shock hazards.
You can also get shocked by putting yourself between the two protected conductors and not trip it whether a grounded system or not.