NEC 230.90(B) says:

Not in Grounded Conductor. No overcurrent device shall be inserted in a grounded service conductor except a circuit breaker that simultaneously opens all conductors of the circuit.

NEC 404.2(B) says:

Grounded Conductors. Switches or circuit breakers shall not disconnect the grounded conductor of a circuit.

Exception: A switch or circuit breaker shall be permitted to disconnect a grounded circuit conductor where all circuit conductors are disconnected simultaneously, or where the device is arranged so that the grounded conductor cannot be disconnected until all the ungrounded conductors of the circuit have been disconnected.

It would seem to me that disconnecting the grounded neutral wire is permitted by the code where there simultaneously all ungrounded conductors are disconnected. I cannot find any provisions where this is otherwise prohibited. And in some cases it is required (areas of explosion risk). Are there any known hazards to doing this?

FYI, quotes are from the 2005 draft version.

 

It all sounds good to me. Do you see a specific problem?

 

I'm looking at several possible problems in a special house design. The biggest one is dealing with some radio frequency grounding issues simultaneously with separate power feeds from the utility and from a generator. Being able to switch the neutral would be a help. I just wanted to confirm that I didn't miss something in the code, and that this is allowed by NEC (convincing the local AHJ once I'm building the house might be another matter).

In order to be sure I don't have RF issues propgating back out the service entrance, I may want to put an RF filter on the power line. That's not a certainty, but I want to make sure I design things leaving that option in. One possible design is:

utility -> meter -> main disconnect -> RF filter -> main panel

where the neutral will be bonded to ground in the main panel. I need to have a disconnect ahead of the RF filter (so I can cut power and do maintenance on that filter), and ground behind it (possibly bonding neutral to ground in the filter enclosure itself rather than the main panel, if that's needed). Since such maintenance work could be disconnecting the ground from the neutral, I'd feel better about being able to disconnect the neutral with the other conductors in the main disconnect.

 

Skapare;

In the arrangment you propose, the Service (Main) Dis-connect is in an enclosure seperate from the "Main Panel". You state; " Neutral will be Bonded to Ground in the 'Main Panel' "

Art 250.28 reads----" For a Grounded System, a Main Bonding Jumper shall connect the Equiptment Grounding Conductor(s) and the Service Dis-connect enclosure to the Grounded Conductor (Neutral) within the Service Dis-Connect enclosure"

Art. 250.142 (B), Load-side Equiptment reads----"A Grounded circuit conductor shall not be used for Grounding (purposes) on the Load-side of the Service Dis-Connecting Means". The Neutral must be "Ground-isolated" at any and all points beyond the Service Dis-Connect"

Art 250.24 (5) reads---"A Grounding connection shall not be made to any Grounded Circuit conductor on the Load side of the Service-Disconnect"

Art. 230.75 reads----"Where the Service Dis-connect does not dis-connect the Grounded Conductor, other means shall be provided for this purpose in the Service equiptment."

 

PATTBAA:

Maybe you can help me figure this out, then. Here is what I have as requirements:

There will be separate entrances for two power sources, with separate disconnect switches for each. One will be utility power. The other will be a generator fed from a separate building.

Radio frequency filtering needs to 100% isolate all RF energy to prevent it from entering the utility feed (this is not required of the generator feed). This means the filtering attachments must be made to every conductor on the customer side (after the meter) of the service drop. Further, for reasons of safety, it should also be made after the (first) disconnect so that maintenance of the RF filtering can be done with all power off.

A solid grounding system must remain regardless of whether power is supplied from utility, or generator, or both (there may be some circuits that are generator-only ahead of a transfer switch), or neither. The grounding conductor must always have a direct path to a grounding electrode designed to RF quality (which is essentially the same as for power purposes but with lower impedance, better anti-corrosion, and radials fanning out).

The grounding conductor must not be a means for RF to bypass the RF filtering. If the neutral of the supply is bonded to the ground on the line side of the filter (whether on the load side or line side of the disconnect), then RF on the grounding conductor can enter the line side neutral by that means (even though neutral would not be bonded to the grounding conductor on the load side).

Note that the concerns about RF going back to the supply do not apply to the generator, since I would have maintenance control over that wiring and can modify it to de-resonate instead of filtering. I won't have that control over the utility service drop, which is why the RF filter would be there.

The utility drop does not have a separate grounding conductor. It's just a grounded neutral and two hots (three hots if it were three phase). If I could put the RF filtering on all of those wires, that would address the RF aspect. I can't put it on the utility wires before the meter because that is under utility control. And I can't put it before the disconnect because I want the safety of not working with live wires when maintaining that filter (and I don't want to have to call the power company out to pull the meter to disconnect).

Disconnecting the neutral is for a different reason than RF, but it needs to play with the RF issue. Because wiring problems on other service drops could result in voltage relative to ground on the service drop neutral, it would be dangerous to work on the grounding system (since in needs to be maintained to RF specs) unless the neutral can be disconnected (in addition to all other conductors).

My original idea was to bond the neutral to the ground on the load side of the RF filter, which would be on the load side of the first entrance disconnect (which would also disconnect the neutral for safe maintenance of the grounding system).

 

My 2005 draft copy reads:

Load-Side Equipment. Except as permitted in 250.30(A)(1) and 250.32(B), a grounded circuit conductor shall not be used for grounding non current-carrying metal parts of equipment on the load side of the service disconnecting means or on the load side of a separately derived system disconnecting means or the overcurrent devices for a separately derived system not having a main disconnecting means.

Exception No. 1: The frames of ranges, wall-mounted ovens, counter-mounted cooking units, and clothes dryers under the conditions permitted for existing installations by 250.140 shall be permitted to be grounded by a grounded circuit conductor.

That seems to be saying to me that you must not use the neutral as a ground wire, except in certain legacy situations. I don't see how it prohibits bonding the neutral to the grounding wire somewhere between a service disconnect and the point where circuits are first branching off.
My 2005 draft copy reads:

Load-Side Grounding Connections. A grounding connection shall not be made to any grounded conductor on the load side of the service disconnecting means except as otherwise permitted in this article.

FPN: See 250.30(A) for separately derived systems, 250.32 for connections at separate buildings or structures, and 250.142 for use of the grounded circuit conductor for grounding equipment.

It's not clear what "a grounding connection" means here (for example connecting equipment to the grounding wire, vs. bonding the neutral to the grounding wire). But if it does refer to bonding the neutral to the grounding wire, there are still some exceptions. So maybe I need to make sure one or more exceptions applies to me. Maybe if I run the service entrance to a detached garage first (which also happens to be where the generator will be), I can put the RF filter on the feeder (3-wire, no ground so the neutral can be bonded to the ground at the house) from the garage to the house. Or maybe I need to use a transformer to create a separately derived system.
This seems to say if the service disconnect does not disconnect the neutral (the grounded conductor) that some other means to do this shall be provided. But what other means? Another disconnect switch? Typical home installations have no means at all to disconnect the neutral. Does that mean they are all in violation of this section? What the 2005 draft says is:

Disconnection of Grounded Conductor. Where the service disconnecting means does not disconnect the grounded conductor from the premises wiring, other means shall be provided for this purpose in the service equipment. A terminal or bus to which all grounded conductors can be attached by means of pressure connectors shall be permitted for this purpose. In a multisection switchboard, disconnects for the grounded conductor shall be permitted to be in any section of the switchboard, provided any such switchboard section is marked.

But that doesn't seem to be much different in intent.

 

It is to be noted that the "execption" for Grounding appliance frames to the Neutral is for existing, not "new", installations.

Art 250.32, (B), (1) includes this explicit requirement which would seem to apply to the design and installation of a new Feeder between two seperate structures----" An Equiptment Grounding Conductor SHALL BE RUN with the (Feeder) Conductors-----"

250.32 (B), (2) permits a Ground-connection to the Neutral conductor of a Feeder between two seperate structures IF there are no "continuous metallic paths" (i.e. metallic pipes) between the structures. This would permit a Neutral-to-Ground connection for an existing Feeder between structures, circumstances permitting.

To design a "new" Feeder on this basis for the purpose of eliminating the EGC in the Feeder circuit would not be justified because of the possible Code-compliant complications that might occur- an inspector might "insist" on the presence of the EGC for a "new" installation.

I believe the intent of Art.230.75 is to prohibit routing the Neutral Conductor directly thru the Service Dis-connect enclosure, avoiding any Neutral terminations in the enclosure, such termination beings a "means" for dis-connecting the Neutral.

Skapare; can you provide us with a WSA that has "info" on the R-F filter equiptment?

Good Luck & Enjoy the Experience!!!!!!!!!!

 

Not to worry; I have no intentions to leave any appliance without a proper separate equipment ground. The RF situation also requires it.
But the grounding wire would still be attached to the grounding electrode in the 2nd building. What I would then have to do is put RF filtering on the ground wire of the feeder (as well as all the other wires). But I don't know what issues might exist with that. But it is essential to block the RF somewhere before it gets to utility wires.
Well, you certainly don't want an increased exposure to the neutral being disconnected while the other wires are energized. There's also a lesser hazard, but one that exists, of having one of the hots disconnected while the other hot wire, and the neutral, are energized.
Sorry, but I don't know what you mean by that. But the RF filter isn't yet a piece of equipment. It is something I will be designing, building, modifying, and/or replacing, as needed. In addition to that, damping filters could be attached to any wiring anywhere if it is found to be resonant and causing problems (I can't do that to utility wires, so I have to just block all RF along the conductor and hope induced RF won't be strong enough to cause problems there).

I may still resurrect a previous plan to put all the RF control entirely in a separate (detached and some distance away) ham shack. Still, powering it can be complicated due to grounding issues. Running a ground that is grounded at the ham shack and result in ground electrode corrosion issues that affect RF (while generally not being an issue for power). The possibility of running the ham shack entirely from natural gas (piped in non-metal) generated power does exist.

But still, how can I completely and totally disconnect from utility power, while still having a grounding system that has no path to any utility wire?

 

Skapare;

I refer you to Art.110.2, Approval, which reads in part----"The equiptment required or permitted---shall be aceptable only if approved"

Please give serious consideration to the connection and installation of any equiptment that is not Code-approved.

 

I'm not sure how to explain this. But RF filtering does not actually interrupt the current carrying conductors, per se. Further, it is something that does have to be periodically modified to make it deal with the variations that exist in the system. RF filtering isn't a product or piece of equipment, per se; it is a process that gets applied. Yes, there will be some piece of equipment. That would be a large stock metal enclosure with a terminal block. If those pieces need to be UL listed or whatever, then surely they are available with such listing. Components of RF filtering can be had with UL listing, too. But it is still necessary to have the RF filtering in its own enclosure put together in a way where those components can be organized and constructed in the way that properly blocks the RF problems as they actually exist. But in some cases, specific RF filtering components do have to be custom built; there are too many possibilities to make commercial products for it for smaller markets.

I'm looking for constructive suggestions here.

At this point, I am designing the house, and need to make decisions like whether the service drop will go to the garage first, or directly to the house, and how much room to put in a utility room where the service comes into the house. But to make the decisions so they can be final, I really need to work out many details of the electrical system, including the RF filtering. But I don't really want to play a game of trying to find alternative paths and being continually blocked every which way I go. So constructive suggestions will be far more helpful.

 

Skapare,

I suspect that you are chasing a phantom here. It would be _impossible_ to have a grounding system that has no path to any utility wire, even if you had no electrical connection to the utility supply at all. Your radio gear is connected to your earth electrode system, and the utility transformer is also grounded to its earth electrode. This means that there is always going to be _some_ connection between the two systems. Even if it is not a metallic connection, you will have a connection of on the order of 25-100 ohms DC resistance...and I have not clue how to characterize the AC resistance or RF impedance of this connection.

I do not believe that you can do any better than to make absolutely certain that no RF shows up as potential difference between any of your utility feed wires, and that no RF shows up as common mode on the entire set of utility feed wires. This requires proper filtering between all the feed wires and neutral (and ground), and a neutral that is well bonded to earth at your main disconnect, and probably also requires that you carefully design the service drop so that the loop area between the service triplex and earth doesn't act as some sort of antenna to pick up your radiated RF.

If you place your filtering equipment in the service triplex prior to the ground point, the connection will still be _after_ the transformer ground point; there is still the connection that you say you want to avoid. If there are other houses around you that share the transformer, then each of them will have a neutral to ground bond.

The only way that you could entirely avoid this sort of connection is if you have some sort of separately derived system, eg a transformer, motor-generator set, generator, etc. With an SDS, you are required to reground the secondary, otherwise you would have a floating, ungrounded electrical system. But when you control your own SDS, you also control all of the secondary grounding. The requirements for transformer grounding still mandate the bond between neutral and ground either in the transformer itself or in the disconnect after the transformer, but now you get to consider (and select) the transformer itself as part of your filtering equipment. This is a rather expensive route to isolating the ground circuit, but in theory it should be possible.

Just my opinion. The best I can say about RF grounding is that it is a black art Have you researched how commercial broadcast stations handle their service entrances?

-Jon

 

If the only "connection" is through the earth, and the utility is well outside the range of the radials, then the ground resistance will be sufficient to prevent problems. Re-radiating from the power lines themselves can be minimized by the distance (and as long as they are, only affects the lower frequencies, anyway).

There is more than one issue here. A good RF setup requires some things. Safety requires some things. Usually they don't conflict. It's when they do that things have to be worked out very carefully.

How do you suggest putting the RF filtering on the service triplex? If I understand you correctly, that would mean the RF filter enclosure would be connected ahead of the meter. Aside from the safety issue, I think the power company would have a fit (if I ran the power company and someone did that, I would have a fit).

The SDS route has been considered, and is still possible. But I also want to make sure I fully explore every possibility that does not involve an SDS. I've already decided that if I do a separate detached ham shack, it will be SDS. But a separate shack would mean a cheaper transformer since I'd only need to do 5 to 10 kVA. Putting the whole house on a transformer would be more expensive (but the added safety of having it could sway that decision, too).

Commercial broadcast stations don't generally have these issues for two reasons. One, they have more clout with the power company, and if the power lines are resonant, they can simply get a service drop length change or them add some filtering to change the effective RF length. The other is that they generally are running one frequency or just a few frequencies. Once done, it would not need to be re-done. In my case I could be running frequencies from 3 kHz to 24 GHz. That makes the RF more complex to deal with and can readily require changing any filters anywhere in the wiring (for example if a circuit is found to be resonant and I add a filter to deresonate it, I might inadvertently make it resonant on another frequency I use at a later date, and have to go back and change it).

I'm still trying to figure out if I need to attach a ground on the primary side of the transformer if I go with a separately derived system. Ideally, I'd take 2 hots, but not the neutral at all, from the meter, through a 2-pole main disconnect, through the RF filter box, and then to the transformer primary (which must be a 2-wire connection). If a ground is required there (I've heard some utilities require the meter be grounded), then I may have to find a way to back that off some distance (like having the meter at the pole and run underground from there).

 

I'm sorry, I did not mean to imply trying to install the filtering prior to the meter or main disconnect by my use of 'triplex', but you are correct in that triplex generally points to the service wires.

What I was trying to say was that I believe that if you properly filter the circuit conductors relative to each other and to ground, then the fact that you have ground bonded to neutral _after_ the filter should not result in significant signal coupling to the utility _in excess of the signal coupling via other ground paths_.

However I suppose that I should stress that I am not experienced with the black art of rf grounding

Good luck!

For transformers you do not need to bring a neutral to the transformer. However you do need to bring the equipment ground to the transformer, and you won't be able to locate the transformer prior to the main disconnect.

-Jon

 

The RF filtering on the power conductors will make those conductors appear, to RF, to be a high impedance path. The RF will then reflect back to the source, or to other points of higher impedance. The ground provides an alternative path. So the ground should be on the load side of the RF filter.

To work with the filter, it will be necessary to the conductors at a terminal block. That's something that should only be done with power off. Having the power company shut off the service just isn't practical for this. So the RF filter needs to be on the load side of a disconnect.

Disconnecting the neutral isn't an RF issue. But it is a safety issue. With the ground on the load side of the RF filter, and the neutral bypassing the disconnect, there can still be current flowing between the neutral and the ground. Breaking that conductor by hand is then a hazard because once broken, the voltage can rise very high, sufficient to put 10 milliamps or more through a person.

I know of a case where it was found that 5 amps was flowing in from the service drop to ground. I suspected the problem was a broken neutral on the pole secondary wires somewhere between where the utility transformer grounded the neutral, and where the service drop neutral connected to other service drop neutrals. Current imbalance in other customer service drops was seeking a way back to the transformer and ground was it. And there was surely a path in that neighbor's ground connection. But since earth ground has some resistance, a parallel path in other service drops through their earth grounds helped provide a lower resistance back to the transformer. That means the current in this situation can use my ground as a path back to the transformer. And when that path is broken, the voltage drop across the various return ground paths is also the voltage drop I get across that conductor break.

I got an interesting suggestion from someone today. Use a 4 pole main disconnect breaker. Let the 4th pole feed ground from the load side to the line side, and attach ground to neutral on the line side per code. It's a partial solution, but I still need ground on the load side (yes, even if power is disconnected). Now to find a 225 amp 4-pole circuit breaker.

Another suggestion was to use a transformer and create a separately derived system, which would be grounded on the secondary side, while the primary side would still have overcurrent protection (and hence a disconnect), but would not need a neutral at all.

 

I am confused here. The ground is well bonded to ground potential; that is why you have all of the buried radials and the like. The filter components in-line with the current carrying conductors act as high impedance to RF...but I presume that the filter also includes shunt components (eg. capacitors) to ground, acting as an extremely low impedance. The net result is that any RF that makes it past the filter will only show up as common mode signal. And since the entire filter circuit is bonded to ground, this common mode signal would be the local ground potential, eg changes in the local ground potential, with no potential difference seen past the filter.

Seems reasonable

I believe that this risk depends upon several bad problems happening at once. It requires multiple ground to neutral bonds (as would happen with separate customers), and it requires good conductivity between the various grounds. (as would happen if the various customers shared underground metallic piping). But given that there is current flowing in the ground from various sources, and that the situation that you describe would make the neutral conductor the path between two grounding electrodes, I agree that there is a safety hazard. Another approach to solving this safety issue would be to provide a shunt on the neutral around the filter components, rather than disconnecting the neutral.

I don't believe that you can place a breaker in your EGC, although this is in the rather interesting situation that the only situation in which the EGC would be disconnected is when the supply is disconnected...given the grounding system that you describe for the ham shack, I suspect that the breaker would not survive the first nearby lightning strike.

Here is yet another suggestion. Please remember that I do not know much about the RF filtering issues. Place your main disconnect on a pole at the edge of your property. Run a feeder from this pole to your home, underground. The pole is a structure, making your home a _detached structure_, so that the feeder to your home would not need to have a separate EGC. The pole would have to have a local grounding system (minimum two ground rods, etc.) and ground would have to be bonded to neutral. The chassis for the filter would be grounded. But the three circuit conductors would be isolated from ground, and would go into your underground conduit without an EGC, and without much exposure to the line side of the filter. At your home you would rebond ground and neutral, as is appropriate for detached structures with no EGC and no metallic paths between the two structures.

-Jon

 

If this apparatus must be arranged so it can be isolated from ALL circuit conductors, and from ALL Ground-connections, and must be Ground-isolated when dis-connected, I presume you will have to use a non-metallic Wiring Method such as PVC conduit for enclosing the circuit conductors that supply the power.

A 4-pole dis-connecting means implies that there will be an Equiptment Grounding conductor in addition to the 3 circuit conductors, and that the Neutral is Ground-isolated and not bonded to any surface required to be Grounded.

Should you Bond the Neutral to the EGC, you will established 2 parallel paths for Neutral current. If you Bond the Neutral to all surfaces required to be Grounded,as is permitted for certain specific appliance connections, you wouldn't need a seperate EGC.