GFCI breaker in panel vs spa panel

[ajdelange]

Unfortunately it is NOT cheap insurance, it's a false sense of security.

It's not intended to be. It (GFCI) is intended to be suspenders in addition to your belt. If the floor and your feet are dry and you are wearing your Crocs (which I do as a fashion statement, not for protection from spurious currents) you don't need GFCI. It is there for when things go wrong that shouldn't go wrong.

You should check your GFCI EVERY TIME.........

You should do it periodically. Doing it every time you use it cannot, of course hurt and were I brewing once a month checking it each time would indeed be good practice.

Checking grounding is as simple as having a push button that routes a significant load through the ground from one leg. If the ground carries the load (120 volt load), it works, and will protect you. There are a number of clever ways to do this.

It's much simpler than what you propose and does not lead to objectionable current in the system. Simply stick one probe of your ohm meter in the grounding conductor hole (round) of an outlet near (can be the same) the one into which the appliance or rig is plugged and touch the other to the brew kettle, stand, sheet metal of the appliance or whatever you are testing. If the impedance is less than 1 Ω clearly your appliance is properly grounded. If you want to get fancy stick one probe of the meter into the earth (round) and the other into the wide slot (after having first verified with the voltmeter or bug that the outlet is wired correctly) of the outlet. This measures the impedance of a round trip from the receptacle to the panel half of which is the impedance of the grounding wire. You can, then, deduct half of what you measure here from what you measure from your appliance.

If you measure (appliance to system ground) 1 Ω (don't correct as the fault current has to flow to the fault point through the phase as well as back to the panel through the grounding connector) then it is clear that a phase to grounding conductor (not phase to neutral) fault will draw 120/1 = 120 amperes which is plenty to trip a breaker up to 120 amps. If it is, conversely 4 Ω then the fault current will be 120/4 = 30 amps which will not trip a 50 amp breaker. The grounding circuit in such a case needs attention (remove corrosion, tighten terminal screws etc.). Now note that you are still safe from the POV of the wires in the walls as the phase and grounding conductor are both carrying currents below the level of the protecting breaker. But lots of power is being dissipated in that extra 3 or so ohms of fault. That point will get hot and as it is probably a screw terminal possibly hot enough to start a fire but in a box, not the walls.

As the discussion here is mainly on GFCI's it is pertinent to assume that the poverty in the grounding system might occur by a corroded or loosened grounding ring at the kettle electrode. In this case we might assume that the 3 Ω is across that ring and then suppose a phase to the kettle wall fault. As objectionable current of about 30 amps is flowing across that ring the voltage on the kettle wall will be 90 volts WRT house reference. If you touch the kettle in this case and you are an IEC standard bloke (without your Crocs and wet feet) on a wet slab floor then you might have 90/3000 = 30 mA or more, as in any of the cases which can reduce human impedance below the IEC measurements, leakage through you. This, in a nutshell, is what GFCI is for.

We note again the belt and suspenders aspect of it. We had to have
1)A problem with the grounding system and...
2)A phase to grounding system fault
in order for the GFCI to step in and save the day.

 

[Owly055]

Your VOM will tell you if you have a ground, but not if it's sufficient to carry a large load. A fuse for example will show virtually zero resistance, though it may be a 1/2 amp fuse. Hence the suggestion of having a significant load.

H.W.

It's not intended to be. It (GFCI) is intended to be suspenders in addition to your belt. If the floor and your feet are dry and you are wearing your Crocs (which I do as a fashion statement, not for protection from spurious currents) you don't need GFCI. It is there for when things go wrong that shouldn't go wrong.

You should do it periodically. Doing it every time you use it cannot, of course hurt and were I brewing once a month checking it each time would indeed be good practice.

It's much simpler than what you propose and does not lead to objectionable current in the system. Simply stick one probe of your ohm meter in the grounding conductor hole (round) of an outlet near (can be the same) the one into which the appliance or rig is plugged and touch the other to the brew kettle, stand, sheet metal of the appliance or whatever you are testing. If the impedance is less than 1 Ω clearly your appliance is properly grounded. If you want to get fancy stick one probe of the meter into the earth (round) and the other into the wide slot (after having first verified with the voltmeter or bug that the outlet is wired correctly) of the outlet. This measures the impedance of a round trip from the receptacle to the panel half of which is the impedance of the grounding wire. You can, then, deduct half of what you measure here from what you measure from your appliance.

If you measure (appliance to system ground) 1 Ω (don't correct as the fault current has to flow to the fault point through the phase as well as back to the panel through the grounding connector) then it is clear that a phase to grounding conductor (not phase to neutral) fault will draw 120/1 = 120 amperes which is plenty to trip a breaker up to 120 amps. If it is, conversely 4 Ω then the fault current will be 120/4 = 30 amps which will not trip a 50 amp breaker. The grounding circuit in such a case needs attention (remove corrosion, tighten terminal screws etc.). Now note that you are still safe from the POV of the wires in the walls as the phase and grounding conductor are both carrying currents below the level of the protecting breaker. But lots of power is being dissipated in that extra 3 or so ohms of fault. That point will get hot and as it is probably a screw terminal possibly hot enough to start a fire but in a box, not the walls.

As the discussion here is mainly on GFCI's it is pertinent to assume that the poverty in the grounding system might occur by a corroded or loosened grounding ring at the kettle electrode. In this case we might assume that the 3 Ω is across that ring and then suppose a phase to the kettle wall fault. As objectionable current of about 30 amps is flowing across that ring the voltage on the kettle wall will be 90 volts WRT house reference. If you touch the kettle in this case and you are an IEC standard bloke (without your Crocs and wet feet) on a wet slab floor then you might have 90/3000 = 30 mA or more, as in any of the cases which can reduce human impedance below the IEC measurements, leakage through you. This, in a nutshell, is what GFCI is for.

We note again the belt and suspenders aspect of it. We had to have
1)A problem with the grounding system and...
2)A phase to grounding system fault
in order for the GFCI to step in and save the day.

 

[ajdelange]

Just too interesting not to post a little more data. My fridge has stainless steel drawers and doors. The impedances from the doors to system ground are both about 4Ω each which is low enough to render a 15 amp breaker sufficient to protect against a phase to door fault. The drawers run on nylon wheels and have rubber gaskets to seal them so their impedance to ground is ∞. A phase to drawer fault would not be protected against. But is there likely to be a phase to drawer fault or phase to door fault. Not likely. The place where the fault would be likely to occur (though I hope it won't) would be in the compartment where the compressors and condensers are located and the impedance to ground for the metal parts of that compartment is 0.4 Ω.

Another interesting aspect of this is that if I stand on the kitchen floor in bare wet feet and touch the phase wire in an outlet close to the fridge I would not get a shock. If I did the same thing while reaching into one of the fridge drawers I would not get a shock either but if I did it while grabbing the handle of one of the doors I would. Were the adjacent receptacle GFCI (it isn't nor does it have to be as it is just, by inches, over 6 feet from the sink) it would trip in the cases where I would get a shock. That is what a GFCI is for.

 

[passedpawn]

Your VOM will tell you if you have a ground, but not if it's sufficient to carry a large load. A fuse for example will show virtually zero resistance, though it may be a 1/2 amp fuse. Hence the suggestion of having a significant load.

H.W.

That's right. The ability of the grounding system to carry current is referred to as bonding in regulatory parlance. In fact, machines I've designed were required to have ground bonding test performed on every one before they went out the door. We alligator-clipped the leads from he tester and ran 10A (something like that) through the grounding for some time to be sure all the green/yellow wires inside had properly grounded the metal panels, from one end to the other.

 

[ajdelange]

Your VOM will tell you if you have a ground, but not if it's sufficient to carry a large load.

Clearly you haven't much experience in these matters but when an electrical installation is done one of the checks an inspector wants to see is the ground fault impedance. It uses exactly the principles I laid out above but it is done for the service, not for each circuit. An ohm meter (but one designed especially for the application) is used to measure this impedance.

What does carry a large load mean? Whatever it means it is immaterial. What is required is that the grounding circuit can pass enough current to trip the breaker in the event of a phase to grounding system fault. The loop impedance is sufficient to tell you that and the Fluke in your toolbox will measure that impedance (or the resistive part of it anyway).

 

[ChocolateMaltyBalls]

It's not intended to be. It (GFCI) is intended to be suspenders in addition to your belt. If the floor and your feet are dry and you are wearing your Crocs (which I do as a fashion statement, not for protection from spurious currents) you don't need GFCI. It is there for when things go wrong that shouldn't go wrong.

This is the salient point for any discussion about the merits of GFCI in the home brewing context. It doesn't matter how many years you've been a professional electrician or if you never changed a light bulb. What matters is you build your system with proper gauge wire, good connections, good grounds, a good diagram, etc. You can do everything by the book or even better and still get fried because it's what's outside of your control that will kill you. GFCI is the last line of defense.

The point is to have a GFCI and never need it, but if you ever need it.... then you'll REALLY need it.

 

[ajdelange]

That's right. The ability of the grounding system to carry current is referred to as bonding in regulatory parlance. In fact, machines I've designed were required to have ground bonding test performed on every one before they went out the door. We alligator-clipped the leads from he tester and ran 10A (something like that) through the grounding for some time to be sure all the green/yellow wires inside had properly grounded the metal panels, from one end to the other.

No, I don't think that is right. What we need to be assured of is that the fault loop impedance is low enough to allow a fault to trip the breaker. If you use a generator to run 10 A of current from the end of the grounding system (the series parallel connection of the green wires) most distal to the point where this network is bonded to the system reference ground and the system reference ground then you want to know the voltage it took to push that 10 amps as the ratio of the two is the impedance which is what you are actually looking for. The supply system voltage divided by that impedance is the current that would flow in the grounding system if there were a fault. If that current is greater than the breaker's rating you are OK.

I may be misunderstanding what you said as you made no reference to measuring the voltage required to push the 10 amps but it sounds to me as if you implemented an ohm meter with a power supply and a voltmeter and if that's the case what you did makes sense.

Also I think we need to emphasize that we are talking about the adequacy of the internal wiring of the building and thus discussing fault loop impedance in the case of a fault between phase and the grounding system conductor. There is more to it than that (I said it's complicated) as there is also the possibility of a ground fault (where a feeder wire falls off the pole and lies in the muck of a farm yard. Here the fault path is though the earth back to the grounding rod and into the panel. If the feeder is on a 100 amp breaker and the impedance of the grounding rod is 2Ω then 120 V can only drive 60 amps through this loop and the 100 amp breaker for this service would not open. You would be pushing 60 amps into mother earth to warm her up.

Actually, I just remembered something a colleague told me years ago. He kept his boat at a marina year round and had an electric hookup to keep a small heater in the cabin to keep things from freezing. Ice moved in and displaced pilings supporting the docking to the point that the wiring beneath the docking went into the water. The grounding system was not adequate to carry enough fault current to trip the feeder breakers and they thus warmed the bay for the month or so before the operator figured he ought to do something about this. The point of my friends story was that the marina wanted the boat owners to foot the huge electrical bills caused by the marina's incompetence.

 

[passedpawn]

No, I don't think that is right. What we need to be assured of is that the fault loop impedance is low enough to allow a fault to trip the breaker. If you use a generator to run 10 A of current from the end of the grounding system (the series parallel connection of the green wires) most distal to the point where this network is bonded to the system reference ground and the system reference ground then you want to know the voltage it took to push that 10 amps as the ratio of the two is the impedance which is what you are actually looking for. The supply system voltage divided by that impedance is the current that would flow in the grounding system if there were a fault. If that current is greater than the breaker's rating you are OK.

I may be misunderstanding what you said as you made no reference to measuring the voltage required to push the 10 amps but it sounds to me as if you implemented an ohm meter with a power supply and a voltmeter and if that's the case what you did makes sense.

Well OK, I left out exactly what the meter was doing. It applied 264VAC (I think), limited to 10A, and ensured the ground system was (from memory) under 100mohms. Something like that. Whatever IEC601 required, but I think that's it.

 

[ajdelange]

There you go. You were looking for that 0.1 Ω ground loop impedance. Thus the loop fault current would be 10 amps/volt. For example, in a 120V system that would be 1200A and any system with breakers less than that would be protected.