US Electrician qualifications for a Brit

On Mon, 16 Oct 2006 03:27:03 GMT Tom Horne, Electrician wrote:

| You can wire up an isolation transformer but you will almost inevitably | have to violate one or more code rules to achieve full isolation. I | wired up a dry transformer on a dairy farm to separate the farms wiring | from the Multi Grounded Neutral (MGN) of the medium voltage distribution | system. The power company threatened to cut off service until the | public service commission of the state government got involved and | pulled out an old tariff for ungrounded delta service. The rule is that | if the service is tariffed then the Utility must provide it. When they | tried to have this "outdated" tariff rescinded the entire dairy coop | system apposed it. So far that is the only NEC compliant way I have | found to actually break the earthing pathway between the MGN and the | customers premise. The US NEC requires that the grounded conductor of | any wye connected transformer be brought to the service equipment | enclosure (Customer Service Unit) and bonded to it. That conductor must | also be grounded / earthed at the customers end. This means that it is | inevitable that stray MGN currents will be flowing over the customer | premise grounding electrode system. The reason that that practice | continues is that the cost of enlarging the utility MGN or installing a | separate insulated neutral in the medium voltage distribution system | gives the utilities' management nightmares. With the increasing | population density the inadequacy of the present neutral system will | become more and more apparent as the MGN becomes more heavily loaded and | stray currents increase. | | There is a special type of utility transformer that is specifically | designed to supply dairy farms that accomplishes the prevention of | utility neutral current flow on the secondary grounded conductor without | violating the National Electrical Safety Code that governs there work. | I have no idea how it works.

It breaks the connection between the primary and secondary, but has a "spark gap" between them to allow a lightning strike discharge to go to ground. That supposedly prevents internal damage to the transformer when there is a surge from a lightning strike.

I believe it would be adequate to have the primary and secondary of the transformer separately grounded at some distance between the electrodes. The distance needs to be sufficient to effectively block MV circuit voltages from entering the LV drop. On either side of a single pole is probably not sufficient.

The correct title for the code that governs their work should be: National Electrical Economic Code. It tends to focus more on the "safety" of equipment than of people and animals. Someone obviously put "Safety" in the title as a selling point.

| Edison may turn out to have been right about the dangers of AC current | after all. Edison abandoned ground return for electric current fairly | early in the development of the Edison electric system.

Ground return is certainly bad. Except in a few isolated cases, utilities don't use that. But the problem is, they effectively have ground return to some extent due to the way they try to cut their costs.

I do have one "design" for a safer system. In a separate building well distant from where the ground currents can be an issue, put in a large motor driven pump system. This pump will then drive a circulating loop of non-conductive fluid through a non-conductive piping system located underground running a significant distance to a generator.

Otherwise the only safe electrical system is independent of the grid, neither drawing from it, nor supplying to it.

wrote in message news: snipped-for-privacy@news1.newsguy.com...

Yep. It's enough to blow fuses on the primary when one phase is grounded in some cases, guaranteed to blow fuses in a phase crossover (phase lines looped over each other, quite a common occurrence in overhead).

As you note in other posts, the utilities don't run separate grounds. When they have L-L or L-L-L connections, no neutral / ground is run. If they ran the ground, voltage rise on the secondary would be less of a problem. But it's the absence of a connection to the much lower impedance ground present in a Y connected system where the utility neutral is run along pole to pole.

I have an anecdotal story about just this kind of failure. One of my in laws farm irrigation systems suffered a primary to case or secondary failure in one of the three transformers feeding the service. The primary voltage is 12.5/7.2kV, with most services connected Y. The irrigation service is connected Y to center tapped delta. The Y center point is not connected to anything else. The branch line to the irrigation service does not have a neutral run along with it - just the three line conductors. Two irrigation services are powered from the tank. The transformer pole has one ground rod, one of the irrigation services has one ground rod, and my in laws service has two ground rods. The second ground rod was added when we replaced the panel to upgrade to a larger pump. A owl tangled up in the phase conductors several poles away. We know it was a owl because his smoking carcass was on the ground right aftarwards. He tangled two of the phase conductors around each other. One fuse at the lateral dropped right away. After a couple of minutes, one of the transformer cans on the pole popped and one more fuse dropped. When the can blew, the whole irrigation system went hot. There was arcing, hissing and sizzling along the irrigation pipe throughout the farm. I'm sure anyone in the river near the suction line was would have been fried. This continued for about 15 minutes until the third fuse finally dropped out. The only damage to my inlaws equipment was to a float control that ran a small pump to keep a cattle trough full. There pump motor was definetly saved by proper ground.

On Mon, 16 Oct 2006 16:35:02 GMT Matthew Beasley wrote:

| As you note in other posts, the utilities don't run separate grounds. When | they have L-L or L-L-L connections, no neutral / ground is run. If they ran | the ground, voltage rise on the secondary would be less of a problem. But | it's the absence of a connection to the much lower impedance ground present | in a Y connected system where the utility neutral is run along pole to pole.

When no neutral / ground is run, then what all does the LV secondary get connected to? Just the winding center and the electrode at the pole?

| I have an anecdotal story about just this kind of failure. One of my in | laws farm irrigation systems suffered a primary to case or secondary failure | in one of the three transformers feeding the service. The primary voltage | is 12.5/7.2kV, with most services connected Y. The irrigation service is | connected Y to center tapped delta. The Y center point is not connected to | anything else. The branch line to the irrigation service does not have a | neutral run along with it - just the three line conductors. Two irrigation | services are powered from the tank. The transformer pole has one ground | rod, one of the irrigation services has one ground rod, and my in laws | service has two ground rods. The second ground rod was added when we | replaced the panel to upgrade to a larger pump. A owl tangled up in the | phase conductors several poles away. We know it was a owl because his | smoking carcass was on the ground right aftarwards. He tangled two of the | phase conductors around each other. One fuse at the lateral dropped right | away. After a couple of minutes, one of the transformer cans on the pole | popped and one more fuse dropped. When the can blew, the whole irrigation | system went hot. There was arcing, hissing and sizzling along the | irrigation pipe throughout the farm. I'm sure anyone in the river near the | suction line was would have been fried. This continued for about 15 minutes | until the third fuse finally dropped out. The only damage to my inlaws | equipment was to a float control that ran a small pump to keep a cattle | trough full. There pump motor was definetly saved by proper ground.

Would this have happened if the system was designed as I suggested, where:

1. The distribution supply secondary is WYE, with the center point solidly earthed.

1. A grounding-only wire, NOT used as a neutral runs along the poles of the distribution, originating at the WYE center point, and is also earthed at periodic intervals.

2. An optional neutral may be run, which may also be earthed only at poles where the grounding-only wire is NOT earthed.

1. Each LV customer tap primary is connected L-L for single phase or delta for three phase, if there is no neutral.

2. If there is a neutral, LV customer tap primary may be connected L-N for single phase and WYE for three phase, but may also be connected L-L or delta. The neutral is NOT earthed at this pole or within 10 meters of any ground wire earthing.

1. All three phase LV customer tap secondaries are always WYE with the center point connected to the grounding-only wire described in #2 and is also solidly earthed with one or more electrodes at that pole.

2. All single phase LV customer tap secondaries are always center tapped with that tap connected to the grounding-only wire described in #2 and is also solidly earthed with one or more electrodes at that pole.

1. No center tapped delta.

Yes.

It would need to have at least arresters so that it couldn't develop high voltage with respect to local ground.

For both 4&5, there still is the problem of ferroresonance & backfeed.

#6 & #7 You still have secondary neutral currents going through the ground. This will eliminate some of the problems, but not all.

What's the problem with delta secondaries if grounded?

|> When no neutral / ground is run, then what all does the LV secondary |> get connected to? Just the winding center and the electrode at the |> pole? | | Yes.

Sounds like what I want ... in single phase.

|> Would this have happened if the system was designed as I suggested, where: |>

|> 1. The distribution supply secondary is WYE, with the center point |> solidly earthed. |>

|> 2. A grounding-only wire, NOT used as a neutral runs along the poles |> of the distribution, originating at the WYE center point, and is |> also earthed at periodic intervals. |>

|> 3. An optional neutral may be run, which may also be earthed only at |> poles where the grounding-only wire is NOT earthed. | | It would need to have at least arresters so that it couldn't develop high | voltage with respect to local ground.

That's why it is earthed. Would it be better if it were not earthed?

|> 4. Each LV customer tap primary is connected L-L for single phase or |> delta for three phase, if there is no neutral. |>

|> 5. If there is a neutral, LV customer tap primary may be connected L-N |> for single phase and WYE for three phase, but may also be connected |> L-L or delta. The neutral is NOT earthed at this pole or within |> 10 meters of any ground wire earthing. | | For both 4&5, there still is the problem of ferroresonance & backfeed.

These are common for dry transformers. Why wouldn't it be a problem with them, too?

Sounds like it will be necessary to always have a 5 wire distribution to be safe.

And how would single phase backfeed?

And what if, instead of one of the phases going out, the neutral does?

|> 6. All three phase LV customer tap secondaries are always WYE with the |> center point connected to the grounding-only wire described in #2 |> and is also solidly earthed with one or more electrodes at that pole. |>

|> 7. All single phase LV customer tap secondaries are always center tapped |> with that tap connected to the grounding-only wire described in #2 |> and is also solidly earthed with one or more electrodes at that pole. | | #6 & #7 You still have secondary neutral currents going through the ground. | This will eliminate some of the problems, but not all.

But not very much at all feeding into the customer ground.

|> 8. No center tapped delta. |>

| | What's the problem with delta secondaries if grounded?

Perhaps none if D-D. Y-D would have backfeed. In any case, it's not part of the design I gave because the intent of it is to have a smaller L-G voltage.

"Tom Horne, Electrician" wrote in message news:b_CYg.10014$ snipped-for-privacy@newsread4.news.pas.earthlink.net...

Some confusion here- you speak of a separate ground wire -not carrying current -on the MV system and which can be uses for grounding the secondary side of the MV-LV transformers. Sorry, the picture is not clear. Lets consider the system that is used as above -where one leg and neutral of the MV (taken as the primary-say 7200 L-N ) is taken down an alley (or goes underground) to feed one or more single phase transformers. The neutral is tied to ground at the transformer. Service tap neutrals are also tied to ground at that point. They will also be tied to ground at the premises. However the service and the MV neutrals are tied together at one point only- at the transformer. If the LV secondary (240/120V) is run along parallel to the MV for some distance (to feed more than one home as is sometimes the case, there is the choice between using a common wire or using a separate LV neutral. I have no problem with the latter if that is what you propose. A separate LV neutral grounded only at the transformer carrying only secondary neutral current. and not common to several transformers "may" be safer but other factors enter the picture -hence the maybe. However, in that case the LV is grounded at the premises as well so there is a loop involving a parallel ground path which is common to both the MV and LV sides and there is no guarantee that some of the MV current won't prefer the LV neutral to the earth path.

----------------------------- >>

----- Ground return on DC is more problematic than on AC and the current distribution in the ground is quite different. However, Steinmetz was against the use of grounded systems. Practical problems that came about as systems grew, dictated that the benefits of a grounded system outweighed the costs.

------------ The present practice is to use (2) with taps being line to neutral. This means that 3 phase services don't have to be run everywhere and there is a cost savings. Transformers are also cheaper. As for safety- it is questionable whether there is any real safety penalty compared to (1). Now using (1), in effect, you are running a ground wire back to the substation and grounding it there. This wire will be coupled to the primary and since loads can be quite unbalanced, there can be appreciable voltages induced in the wire. These can be much higher than those due to neutral current. This can be avoided by use of multiple ground points but in the case of a poor ground at a customer entrance there is the possibility (however rare) that induced voltages from hundreds to thousands can occur- particularly in the case of a fault on the primary. Such voltages could exceed the voltages produced by a high common neutral current (which is actually much less of a problem) . There are safety concerns with both systems but how many safety problems have been caused by this considering the extensive use of the common HV-LV neutral point/ conductor throughout North America for the last

|> The substation transformer secondary should be WYE. The center point is |> earthed at the substation. Now there are two different ways to run that |> circuit: |>

|> 1. Run 5 wires, identified as A,B,C for the phases, N for the neutral, |> and G for the grounding wire. Loads can then be connected to any |> combination of A,B,C,N as needed. |>

|> 2. Run 4 wires, identified as A,B,C for the phases and G for the |> grounding |> wire. Loads can then be connected to any combination of A,B,C as |> needed. |>

|> With design #1 you can have L-N taps for customer service transformers. |> With design #2 you are limited to L-L taps. In all cases the secondary |> would be WYE. | ------------ | The present practice is to use (2) with taps being line to neutral. This | means that 3 phase services don't have to be run everywhere and there is a | cost savings. Transformers are also cheaper. As for safety- it is | questionable whether there is any real safety penalty compared to (1). Now

The difference between #1 and #2 is not as you describe. #2 has no neutral. For what you describe we need a new entry, #3, with A,B,C,N. Still 4 wires, but also electrically different.

| using (1), in effect, you are running a ground wire back to the substation | and grounding it there. This wire will be coupled to the primary and since | loads can be quite unbalanced, there can be appreciable voltages induced in | the wire. These can be much higher than those due to neutral current. This | can be avoided by use of multiple ground points but in the case of a poor | ground at a customer entrance there is the possibility (however rare) that | induced voltages from hundreds to thousands can occur- particularly in the | case of a fault on the primary. Such voltages could exceed the voltages | produced by a high common neutral current (which is actually much less of a | problem) . There are safety concerns with both systems but how many safety | problems have been caused by this considering the extensive use of the | common HV-LV neutral point/ conductor throughout North America for the last | 80 years? There are the "stray-voltage" situations met in some dairy barns | but, the cause claimed, has, as far as I know not been proven technically | (although proof in court is easier).

Based on what I have seen, the "stray-voltage" is a proven and real problem. Break the connection between primary and secondary neutral at the transformer and the problem goes away.

Electricity takes every return path it can find, in inverse proportion to resistance. The neutral wire running down the pole is best, and most of the current will go that way (on single phase branches) or via the other phases (as soon as balancing loads on other phases are reached). But some will take the earth path back. It's not much, but the higher the voltage, the more significant it is.

Just how much this will affect things will also depend on the quality of the distribution circuit design and condition of maintenance. Some power companies do better than others. Some do quite horrible.

Ham radio operators searching out RFI problem have found a great many of these problems are on the power lines, so they are usually checking them in the course of tracking them. Even when the actual RFI cause is elsewhere, they do frequently find multiple unrelated problems on power lines. One very common one is a broken ground connection. I found two of these in my neighborhood when I was a kid. At least I knew not to touch them. My dad called the power company to report them, and they never, ever got fixed. Numerous ham radio RFI reports I've read include the hams coming across the very same thing.

There are many unexplained technical problems with equipment like computers that I'm wondering if this could be related to. Transients one the MV lines could translate to transients on the ground wires at levels too short for people to see or realize what they are feeling, but could be disasterous to a tiny solid state component. I don't have the resources to acquire and operate the test equipment to verify this. But from a technical understanding, it is very plausible.

The key thing is, these problems don't result in power outages, at least not right away, and as such, there is very little incentive for the power company to address these issues. They give lip service to safety, and you see that in actual spent dollars on things like TV ads telling kids not to touch downed power lines. But the shabby maintenance of some power lines tells a lot more to me about where their real incentives and motivations are.

| Some confusion here- you speak of a separate ground wire -not carrying | current -on the MV system and which can be uses for grounding the secondary | side of the MV-LV transformers. Sorry, the picture is not clear.

Which part is confusing?

| Lets consider the system that is used as above -where one leg and neutral | of the MV (taken as the primary-say 7200 L-N ) is taken down an alley (or | goes underground) to feed one or more single phase transformers. The neutral | is tied to ground at the transformer. Service tap neutrals are also tied to | ground at that point. They will also be tied to ground at the premises. | However the service and the MV neutrals are tied together at one point only- | at the transformer. If the LV secondary (240/120V) is run along parallel to | the MV for some distance (to feed more than one home as is sometimes the | case, there is the choice between using a common wire or using a separate LV | neutral. I have no problem with the latter if that is what you propose.

It is not, but only because I didn't address that specifically.

| A separate LV neutral grounded only at the transformer carrying only | secondary neutral current. and not common to several transformers "may" be | safer but other factors enter the picture -hence the maybe. However, in | that case the LV is grounded at the premises as well so there is a loop | involving a parallel ground path which is common to both the MV and LV sides | and there is no guarantee that some of the MV current won't prefer the LV | neutral to the earth path.

The system I described for my design would have 3 wires going do that alley. Those would be the line wire Lp, the neutral wire Np, and the ground wire Gp. If they were insulated according to color standards used in buildings for LV circuits, they would be black, white, and green.

The transformer primary would be connected to Lp and Np. Thus Lp and Np would be carrying current. Gp would not be carrying this current, though it can acquire a proximity charge. But that charge can bleed off to ground with very small levels of current.

On the secondary side of the transformer, we have a center tapped winding, and two line wires. I'll call these As, Bs, Ns. Ns is the center tap. There is also a wire running between the transformer and an earth electrode I'll call E. Ns, Gp, E, and the transformer case, are all connect together.

The path between Ns and Np is solidly metallic only all the way back to the substation where the transmission to distribution transformer connects to both of them at it's wye secondary center point. The other path between Ns and Np is through earth between separate electrodes space distances apart (for example alternating between whether Np or Gp is earthed on even and odd poles).

Suppose you have a building fed with 480Y/277. On each floor of the building is a dry transformer stepping that down to either 208Y/120 three phase or

|> Edison may turn out to have been right about the dangers of AC current |> after all. Edison abandoned ground return for electric current fairly |> early in the development of the Edison electric system. | ----- | Ground return on DC is more problematic than on AC and the current | distribution in the ground is quite different. However, Steinmetz was | against the use of grounded systems. Practical problems that came about as | systems grew, dictated that the benefits of a grounded system outweighed the | costs.

There is a difference between a groundED system, and a ground return system. That difference is a bit fuzzy in practice considering that the neutral wire is not zero resistance, and maintenance is imperfect.

At the voltage Edison was working at (110/220), using earth to return the neutral would have not performed very well, even over the short distances in the part of New York he was serving. Ground return on a 2 wire system, AC or DC, can be done at higher voltage and managed despite the ground resistance by adjusting the voltage to match the load. But having taps on that ground return line _and_ having them be not quite in balance over two opposite polarities or phases, would be more challenging even today. Edison had no chance of making a ground return successful with existing technology of his day.

If the neutral is grounded only back at the substation and lightning hits the line, it would develop high voltage with respect to local ground. Even induced voltage from a nearby strike could result in equipment failure. By placing an arrester neutral to ground, it would limit the neutral to ground voltage rise.

Usually they are fed with three phase protection. Phase loss between the source and transformer is rare. Single phase protection is prefered by many utilities for distribution, and loss of a single phase - either open or grounded is more common.

The impedance of backfeed on single phase would be small, just the load current. In single phase the major danger would be ferroresonance.

With the current multipoint grounded wye systems, it becomes SWER (Single Wire Earth Return). With no grounding except back at the substation, the neutral could be subject to ferroresonance.

In most cases, but not always. The voltages are lower but currents higher in secondary systems. Neutral drop can be significant if there is significant neutral current. In a properly designed customer's system there should be, but that's not always the case.

D-D and Y-D with no connection to center point are about the same. I absolutely agree on the standardization point - the delta services just add to the number of "standard" voltages.

|> That's why it is earthed. Would it be better if it were not earthed? | | If the neutral is grounded only back at the substation and lightning hits | the line, it would develop high voltage with respect to local ground. Even | induced voltage from a nearby strike could result in equipment failure. By | placing an arrester neutral to ground, it would limit the neutral to ground | voltage rise.

If the neutral is not grounded, I can believe that. But consider two points:

1. In my design I do suggest grounding the neutral to earth at as many as every other pole. The other poles would be where the non-conducting groundING wire is earthed.

1. What happens to the line wires that are not earthed? They are not grounded anywhere. Would they not develop that high voltage?

|> |> 4. Each LV customer tap primary is connected L-L for single phase or |> |> delta for three phase, if there is no neutral. |> |>

|> |> 5. If there is a neutral, LV customer tap primary may be connected L-N |> |> for single phase and WYE for three phase, but may also be connected |> |> L-L or delta. The neutral is NOT earthed at this pole or within |> |> 10 meters of any ground wire earthing. |> | |> | For both 4&5, there still is the problem of ferroresonance & backfeed. |>

|> These are common for dry transformers. Why wouldn't it be a problem |> with them, too? | | Usually they are fed with three phase protection. Phase loss between the | source and transformer is rare. Single phase protection is prefered by many | utilities for distribution, and loss of a single phase - either open or | grounded is more common.

But that loss of a single phase does not propogate into the premise dry transformer?

|> |> 6. All three phase LV customer tap secondaries are always WYE with the |> |> center point connected to the grounding-only wire described in #2 |> |> and is also solidly earthed with one or more electrodes at that |> pole. |> |>

|> |> 7. All single phase LV customer tap secondaries are always center |> tapped |> |> with that tap connected to the grounding-only wire described in #2 |> |> and is also solidly earthed with one or more electrodes at that |> pole. |> | |> | #6 & #7 You still have secondary neutral currents going through the |> ground. |> | This will eliminate some of the problems, but not all. |>

|> But not very much at all feeding into the customer ground. | | In most cases, but not always. The voltages are lower but currents higher | in secondary systems. Neutral drop can be significant if there is | significant neutral current. In a properly designed customer's system there | should be, but that's not always the case.

What about this setup:

The single phase customer drop has 2 phase lines at 240 or 480 volts. It reaches the service meter placed on a pole just a short distance away from the customer building, then runs over to the building where it enters through a disconnect and immediately to a dry transformer inside. The secondary of the dry transformer derives 120/240 and its center tap is earthed by 2 paths going out either side of the point where the power comes in.

The service drop neutral comes to that pole with the meter, and is earthed there. But it is not extended across to the building.

What I've been told before is that the transformer in the building would be subject to lightning induced ground gradient voltages by a nearby strike. But I would think that to be small if the distance between the pole and building is not too great (for example 3 meters).

|> |> 8. No center tapped delta. |> |>

|> | |> | What's the problem with delta secondaries if grounded? |>

|> Perhaps none if D-D. Y-D would have backfeed. In any case, it's not |> part of the design I gave because the intent of it is to have a smaller |> L-G voltage. |>

| | D-D and Y-D with no connection to center point are about the same. I | absolutely agree on the standardization point - the delta services just add | to the number of "standard" voltages.

It seems all the dispute with my design was how I described it could be connected at the MV distribution side. But I think my design still has merit (other than for the fact that there is no chance in hell the world would make any changes today). Light (as in incandescent illumination _and_ small appliances) loads would be connected to 24 volts wired L-N (maybe 24-0-24) and everything else would be connected to 288 volts wired L-L coming from single phase 144-0-144 or three phase 288Y/166. The 24 volt (24-0 or 24-0-24) system would be separately derived from a 288 volt branch circuit. There might be more than one such system if there is a lot more incadescent lighting. I think I would go with 72 Hz.

I do have a few different designs for the ultimate receptacle.

Now I just need to get my time machine working :-)

--------- At one time single phase rural lines did not carry a neutral- and used the earth as the return- grounds were at every transformer along the line and typical individual loads at 7200V were in the 5KVA range. While this worked quite well, it has some disadvantages and now a neutral is used. The problems that I was thinking of were not related to ground impedance or low voltage but to unidirectional current and corrosion as well as the different distribution of current in the ground. DC is a "shortest path" return while AC tends to follow the line. >

I see the system that you propose. In normal circumstances there is essentially a primary loop tied to ground at the substation (if not then it is floating- not a good idea) but not elsewhere (? this assumption may be wrong in terms of your statement about grounding of Np and Gp on alternate poles -If so -then I'll deal with it later-suffice it to say that, if both the NP and GP are multigrounded any advantage that you may have is gone). You are extending the idea of a separate ground wire as used in residential systems to include transformer tanks etc. The use of a separate ground came about initially because of problems with power tools used outside where a line to case fault could leave the case "live" Polarised plugs didn't deal with this because there is always the chance of an incorrect connection so the grounding receptacle came into use. Fair enough. In the case of MV equipment, transformer tanks and groundable equipment is not exposed to the general public or to "home electricians" so the need for a separate ground is not apparent. As to currents- in normal use there will be no current in the MV neutral and current in the LV neutral and the ground wire Gp would be secondary unbalance current- fair enough. There will be a voltage induced in the open ended MV neutral but this will generally be quite small.

For the present situation with a grounded neutral, there will be a current in the neutral, the ground and the service drop neutral- how much and what is the relative phase with respect to the normal LV service drop current? A rough estimate is 3% if all the ground current was through the customer's ground rod, less if otherwise. Note that this would be subtractive. This implies a complete failure of the ground at the transformer- that is the MV neutral open and the ground rod at the transformer out also open. Is this a probable hazard? Note also that single phase transformers intended for L-N operation have the Neutral tied internally to the tank already and have a single MV bushing. The secondary center tap is also tied internally to the tank.

Now consider a line to ground fault on the MV side (recalling that the Y point at the substation is grounded) - In your proposed scheme there will be a ground return current which will involve the Gp line and any service drop- just as there would be if one tied a connector from the hot to the green in one's home. If the conventional scheme with a grounded MV neutral (at all transformers and ideally but not necessarily in between) and no Gp the same situation would occur. No gain from the separate ground. The separate ground would however require something more than a fuse at the transformer in that a fault between MV and the tank (fat bird or squirrel) would produce a ground fault which would not involve the MV neutral if it was not grounded to the tank. If you want to replace a simple protective device with a more complex device you lose the KISS principle. Are there safety gains to be had from the proposed scheme- possibly, under certain conditions but these imply both failure of the most robust part of the system connections as well as exposure of consequence to people or other equipment. Can the changes be justified-questionable. The use of a separate grounding wire in domestic conditions came about because it was justified.

You mention alternate pole grounding of Np and Gp- which effectively puts them in parallel so the result is simply that you have two grounded neutrals leading back to the source. Then there is no advantage or disadvantage except redundant conductors and a slightly lower ground path impedance. The alternate pole idea really doesn't change this much from a single wire with a ground point at every pole. For AC ground "resistance/meter" is frequency dependent while inductance/m is resistivity dependent. Ground rod resistance is generally much greater than the resistance between grounding points so distance between alternate ground points doesn't mean much.

in your building example. I would follow the code whether I like it or not and if I didn't like it for perceived safety reasons I would raise questions. I have not looked at a code book for over 10 years and it has been much longer than that when I did any building service design.

I note that the actual ground as a return is not present. Is it better to use a green ground wire tying all LV neutrals together and to ground? It has an advantage where people can come into contact with equipment and the building ground as with the outside lights. I should think that tying the LV neutrals to the green wire would be the best. The building situation is different than the distribution systems that we have been discussing where earth ground is omnipresent and distances are larger. There I would not recommend a separate ground back to the supply as the advantages are not present.

Don Kelly snipped-for-privacy@shawcross.ca remove the X to answer

----------------------------

On Wed, 18 Oct 2006 05:23:00 GMT Don Kelly wrote:

| I see the system that you propose. In normal circumstances there is | essentially a primary loop tied to ground at the substation (if not then it | is floating- not a good idea) but not elsewhere (? this assumption may be | wrong in terms of your statement about grounding of Np and Gp on alternate | poles -If so -then I'll deal with it later-suffice it to say that, if both | the NP and GP are multigrounded any advantage that you may have is gone). | You are extending the idea of a separate ground wire as used in residential | systems to include transformer tanks etc. The use of a separate ground came | about initially because of problems with power tools used outside where a | line to case fault could leave the case "live" Polarised plugs didn't deal | with this because there is always the chance of an incorrect connection so | the grounding receptacle came into use. Fair enough. In the case of MV

I believe the first requirement of a grounding receptacle was in the laundry room. But it would certainly make sense to have them for outside outlets as soon as possible. I know the first house my parent built around 1969 had non-grounding outlets outside. Two extra front ones were put up where we planned to have Christmas lights and those were controlled by an inside switch. The only grounding outlet originally planned for the whole house was one in the laundry room dedicated to the washing machine. I did get my bedroom wired on a dedicated circuit with grounding outlets in anticipation of ham radio needs.

| equipment, transformer tanks and groundable equipment is not exposed to the | general public or to "home electricians" so the need for a separate ground | is not apparent. | As to currents- in normal use there will be no current in the MV neutral and | current in the LV neutral and the ground wire Gp would be secondary | unbalance current- fair enough. There will be a voltage induced in the open | ended MV neutral but this will generally be quite small.

Why would secondary unbalance current go back to the primary ground?

However, in the abnormal case, e.g. various faults, you can have many possible dangerous scenarios.

| You mention alternate pole grounding of Np and Gp- which effectively puts | them in parallel so the result is simply that you have two grounded neutrals | leading back to the source. Then there is no advantage or disadvantage | except redundant conductors and a slightly lower ground path impedance. The | alternate pole idea really doesn't change this much from a single wire with | a ground point at every pole. For AC ground "resistance/meter" is | frequency dependent while inductance/m is resistivity dependent. Ground rod | resistance is generally much greater than the resistance between grounding | points so distance between alternate ground points doesn't mean much. | | in your building example. I would follow the code whether I like it or not | and if I didn't like it for perceived safety reasons I would raise | questions. I have not looked at a code book for over 10 years and it has | been much longer than that when I did any building service design. | | I note that the actual ground as a return is not present. Is it better to | use a green ground wire tying all LV neutrals together and to ground? It has | an advantage where people can come into contact with equipment and the | building ground as with the outside lights. I should think that tying the LV | neutrals to the green wire would be the best. The building situation is | different than the distribution systems that we have been discussing where | earth ground is omnipresent and distances are larger. There I would not | recommend a separate ground back to the supply as the advantages are not | present.

What about a service drop that has only L-L, originating from a transformer with the center tap ground there, as usual for 120/240, going to a second transformer, where the primary is not earthed or if earthed, is not so close to the building. There are number of ways to do this, but the idea is to metallically isolate the building from the MV distribution 100%. I realize the transformer could be more exposed to lightning induced damage this way.

Many people have success in powering their homes exclusively from natural sources, usually solar. Perhaps that's what the dairy farmers need to be doing, at least for the buildings housing the milking operation.

I might even consider that for the house I'm designing now.

| Don Kelly snipped-for-privacy@shawcross.ca | remove the X to answer

I can't find the "X".

Just curious Phil:

Have you followed the work of Donald Zipse? He is on a crusade to eliminate the multi-grounded neutral Y distribution in the US. You might want to get some of his papers and read them. He is proposing some of what you are.

| Just curious Phil: | | Have you followed the work of Donald Zipse? He is on a crusade to eliminate | the multi-grounded neutral Y distribution in the US. You might want to get | some of his papers and read them. He is proposing some of what you are.

I have already read some of his papers.

Snip

If the utility transformer has a grounded conductor then the neutral must, by code in the USA, be run to and bonded to the service disconnecting means. The circuit from the derived system source; the dry transformer in this case; must have an equipment grounding conductor that is run from the bonding point of the derived system to the load to be served.

| If the utility transformer has a grounded conductor then the neutral | must, by code in the USA, be run to and bonded to the service | disconnecting means. The circuit from the derived system source; the | dry transformer in this case; must have an equipment grounding conductor | that is run from the bonding point of the derived system to the load to | be served.

I know this.

But what if it does not have one? What is the actual hazard of not doing the connection?

FYI, it is one of my quests to find a lower cost way to effectively evade this rule. A high cost design I have involves driving a generator using a fluid turbine. The fluid loop is run over some distance and is powered by a pump driving by the utility power source. Another high cost design is using natural gas instead of electricity as the utility source, and running the generator from. I'm just looking for something better, like an isolation transformer that can handle the ground voltage difference.

Why? you haven't yet actually made a case for your desired non-metallic bonding between neutrals of the MV and LV system. Yes, in a building distribution you could run a separate ground for the LV- tied back to the building ground and it would work, but, beyond code violations, is there any point? If there is a MV fault to ground, it will then likely involve the LV ground system so there is no advantage there. If it doesn't, then there may be a problem with differential "ground" voltages .

As for the removal of the X, - cross my heart, my address is not a penitentiary.

| Why? you haven't yet actually made a case for your desired non-metallic | bonding between neutrals of the MV and LV system. Yes, in a building | distribution you could run a separate ground for the LV- tied back to the | building ground and it would work, but, beyond code violations, is there any | point? If there is a MV fault to ground, it will then likely involve the LV | ground system so there is no advantage there. If it doesn't, then there may | be a problem with differential "ground" voltages .

With a connected neutral back to the source on the MV side, most of the MV return current will go that way. A small amount will go by way of ground. Some of that will do so over the LV ground. That by itself is not an issue. But it illustrates the bigger issue. What if the MV neutral is the one that breaks and falls to ground somewhere? Tell me what mechanism will cut out the whole MV circuit immediately.

With service drop transformers connected L-L, that neutral won't matter to them. But as you and others point out, that's not the common practice. So the neutral upstream of the break will have to find another path to return its current.

Sure, I'm paranoid. But having seen deplorable conditions of maintenance of utility wiring, and the fact that I cannot go fix these things myself beyond the meter, I just don't want any part of it, and that includes the grounding of the MV circuits.

So I'm looking for an isolation transformer that has high voltage isolation capability.

Maybe what I might do is have the utility land the service drop to the first of three wooden poles. After the meter and disconnect, I'll take the 240 volts over to the next pole and energize an ungrounded polepig in reverse. Then take the MV from that over to the third pole with a second polepig taking it back down to 120/240. Crazy idea? Sure. You have a better one?

You might not share my motivation to be doing this. But I think I have made it clear what I want to do. I'm still open to other technical ideas on how to do this. But I'm not really open to debating why. So this is the last of trying to explain that.