On Tue, 17 May 2005 00:30:12 GMT ehsjr wrote: | snipped-for-privacy@ipal.net wrote: |> On 13 May 2005 09:32:29 -0700 Ed wrote: |> |> | Seems not many understand what role the nuetral plays. First in single phase it |> | should correctly be called the "grounded conductor", then it is a current |> | carrying conductor (but at low after load voltage) and misunderstanding of how |> | it functions is one of the largest causes of electrical problems. The neutral is |> | a transformer balancing conductor and its function has little to do with it's |> | single connection to earth ground. Poor "grounded conductor" connections or |> | inadvertent connections to an earth ground (other than the main bonding |> | jumper)cause voltage level problems that shorten the life of all components and |> | extent the trip time of circuit breakers. If you think that in an electrical |> | panel that the white wires can correctly share the same bus connections as the |> | bare wires then you do not understand what is really happening to the |> | electricity and the function of the grounded conductor. |> |> You are correct that the concern really involves whatever conductor is |> the grounded conductor. In a 2-wire system, one conductor is grounded, |> but it's technically not the neutral because there isn't a neutral. |> |> However, the original concern is the voltage available on this |> conductor due to its use as a current carrying conductor and it's |> distance from the one point where it is bonded to ground. The |> conductor does not have zero impedance, so it will have some voltage |> when load is applied. Part of figuring out that voltage involves the |> impedance of that conductor and its effective voltage drop under the |> effects of the load and other impedances in the system. This voltage |> would not be that high, maybe 1 or 2 volts in typical cases. My |> concern is it some cases it could get high. That and the fact that it |> is a grounded system (low voltage lighting with exposed conductors is, |> by comparison, an ungrounded system), could create some shock hazard at |> some points where ground contact is also available. |> |> If it is perfectly safe, then we really don't need a separate grounding |> conductor. But we would rather have the frames of appliances connected |> to the grounding conductor rather than the grounded conductor. So |> which of those 2 is the mroe dangerous, and to what degree? |> | | Under normal circumstances, neither one is dangerous. | Introduce a defect or defects, and all bets are off. | | The egc requires 2 defects (one to energize it and one to | disconnect it from ground) to become dangerous, while the | neutral can become dangerous with one defect.
So why not connect the EGC to ground in multiple places to greatly reduce the chance of the 2nd defect? Doing so near the load (e.g. connect the EGC to earth or building frame) would greatly reduce the voltage difference at that location between equipment frame and other things people may be in contact with. Of course one problem with this is that the return path for faults won't be in parallel with the supply line.
My original issue is with the groundED conductor (which happens to be the neutral conductor in Edison style 3-wire circuits). I assert that it is also dangerous, though because it is grounded at a couple places, not to the degree the line conductors are dangerous. Still, I believe the level of danger justifies interruption of the groundED conductor in certain locations where people may readily form a lower impedance fault path to real ground (the places where GFCI protection of receptacles is required).
Most GFCI receptacles do interrupt both conductors, so they do meet the need as I see it. GFCI breakers, however, do not interrupt the grounded conductor; it merely passes straight through the current sensor. What I believe is needed is a GFCI breaker that does interrupt all conductors (except EGC). One workaround I have found is to place a contactor after the breaker. The normally-open contactor would be energized from the line coming from the breaker, closing 2 to 4 poles. When a ground fault happens the breaker would sense it and disconnect the line conductor(s). Then the contactor, no longer being energized, would open all the conductors, thus interrupting the grounded neutral at some point after the breaker opens the circuit. The contactor would still have to be applied to all conductors, not just the neutral, to be sure that when the line is energized, the line wires are not connected by themselves first.
Personally, I still dislike the idea of specifically depending on high fault currents to interrupt circuits. Much of the wiring requirements are as they are to ensure low impedance fault return paths. I believe that this approach is a legacy of the era of fuses. The first circuit breakers were designed principly to be replacement of fuses. But now we have newer technology in the form of ground fault and arc fault breakers. But we are only using them in limited circumstances. I believe if we use them in all circumstances, we will no longer need low impedance fault return paths in order to properly interrupt faulty circuits. These high current faults can and do result in high energy releasing arcs, often resulting in fires and serious burns. It is (still) my opinion that with most faults being line to ground, a high impedance grounded system can be made safer. Note that I am absolutely not referring to the typical industrial use of high impedance to avoid interruption of the circuit due to a single fault. Instead, what I am referring to is the use of a high impedance grounded system to reduce the fault energy, while at the same time detecting AND interrupting such faults immediately. There would still be overcurrent protection so that any line to line faults, as well as circuit overloads, would still also be immediately (or in due time as apprpriate for the overload) interrupted.
The problem with the above is that too people people have a fixed box of thinking about a "high impedance ground system" in terms of the way it is used in an industrial setting. There are certain dangers and risks with such a design in that setting, which have to be balanced with the dangers and risks of a non-orderly shutdown. As such, a single fault does not trigger an interruption. But the suggestion I am making is one that does not apply to the industrial setting, and would trigger an interruption as soon as any fault is detected.