In sci.engr.electrical.compliance HorneTD wrote:
| An isolating switch on the other hand would not need to be rated for the | calculated load because it is never opened with current flowing.
But there can be current flowing between the service neutral and ground. It might not be substantial, but the isolation switch would need to be able to interrupt it safely.
| Isolating switches are always labeled to indicate their purpose and that | they must not be opened prior to opening the Service Disconnecting Means | (SDM). One good example of a switch that might be suitable for use as | an isolating switch but not as the SDM is a switch that is rated for the | voltage and current but is not motor rated or is not suitable for use as | service equipment. A common use for isolating switches is to | de-energize a service rated transfer switch for servicing or to renew | internal or external fuses. Such transfer switches are available with | and without fuse holders or other OCPDs. The service OCPDs may be | located on the line or load side of the switch. When they are located | on the line side of the switch there is a need for an isolating switch | on the line side of any fuses to permit safely renewing the fuses. The | plant wireman or an electrical contractor opens the SDM, clears the | fault, closes the SDM to the alternate source of power, opens the | isolation switch, renews the fuses, re closes the isolation switch, and | operates the transfer switch to the original position to reconnect the | primary power source to the load. When you want to work on your RF | filters, you open the SDM and then open the isolating switch.
I understand what you mean by an isolating switch. So why would such a switch be used instead of something else? Lower cost? Would this be the "unfused safety switch" like I see in a Square-D catalog?
| If your ground fault detector is to protect you from an external power | cross then I only have one more suggestion to make. You can install the | service away from the building in question. If the service is overhead | than you would have to install a yard pole. If you have another | substantial structure on the property, you can install the Service | Equipment in that building. You then have your SDM located away from | the building that will be protected by RF filtering. The main power | switch at the subject building then becomes the Building Disconnecting | Means (BDM) that is required by NEC 225.31. The rule we have been | struggling with does not apply to building disconnecting means but only | to service equipment. The SDM would consist of a four-pole breaker | equipped with Ground Fault Protection. The breaker would open all four | wires of the feeder including the Equipment Grounding Conductor. The | utility neutral at the service equipment is never disconnected from the | service Grounding Electrode Conductor (GEC). At the building the BDM | would be the main breaker of the panel or a separate breaker enclosure | or switch. That building would also have a grounding electrode system | which would be connected to the BDM enclosure and to its EGC buss bar | but not to its neutral buss bar. Since this BDM is not service | equipment, it can open all conductors of the feeder upstream from the | point of termination of the building?s GEC if that is desirable.
So by having the SDM at some distance from the BDM, the ground is not a viable conductor to worry about. The building would then be a subfeed of the remote SDM for purposes of separate neutral and protection ground (EGC), unless I decide to also do a separately derived system.
| The obvious draw back to this approach is the cost of; two sets of | service equipment, the ground fault detection equipment, and the shunt | trip on the service breaker. That combination of equipment is going to | cost you a pretty penny. The biggest of those costs will be the four | pole circuit breaker and it's enclosure that is functioning as the | Service Equipment. I can see no way to avoid needing the four pole | breaker if you want to provide automatic disconnection of the buildings | GEC and its EGCs from the utility neutral. Since the building will need | to have other utilities that are bonded to ground at neighboring | premises you will need to locate the Network Interface Device (NID) for | those utilities at the remote service equipment. If you use coax or | shielded telephone cable between the NID and the building, you will have | established a sneak current path for the stray high voltage you appear | to be concerned about. I'm not even sure that the protectors on the | telephone lines at the NID and at the building will not provide a | current path about which you may be concerned. That would only happen | in the event of a high voltage cross with the secondary of the utility | transformer, the service entry conductors, or the telephone drop or | lateral.
Suppose I feed all power through a transformer at the building, where the BDM is just before the transformer primary. Would I even need to feed the neutral conductor from the remote SDM to the building, given that the transformer (240 to 240/120) won't connect to it at all? If I can leave the neutral out of the feeder, then can I also leave it out of the breaker? If so, then that would get me to a 3-pole breaker, which is more readily available.
| This possibility of a high voltage cross with the utilities? wires does | not seem likely enough to take such expensive precautions.
I've seen a case where it happened. And I fully understand why the fire department just stood there watching the house burn down.
Still, I'm more concerned with the voltage across the terminal block in the RF filter enclosure. One thing I thought of doing there is to use not one, but two ground conductors in the enclosure. Extra filters would need to be used, but I could work on them one at a time, and never have both ground conductors disconnected at the same time. Or I could do two whole RF filter systems in parallel with 4 isolating switches to take one or the other out of service for maintenance.