What common arrangements are made inthe US re the provision of supplies to domestic and commercial premises? I am aware that there are a number of "funny wonders" but lets say if it were a new supply. Paul, might you have diagrams?

I'd say the most common arrangement is TN-C-S or PME, i.e. the neutral is bonded to earth at several points, the final bond being at the main distribution panel. Each of these systems has to have a local ground rod as well (unlike some older Austrian installations that were only "earthed" to the neutral) TT systems like used in many rural areas of Europe are illegal.
In simple words, at the subscriber's customer unit the incoming neutral, the earthing (or grounding) elctrode conductor and all branch circuit neutrals as well as abonding jumper to the earth bus bar go to the nutral bus bar.

Yes, the TN-C-S system is standard for 120/240V 3-wire residential systems, and for 120/208 and 277/480V 3-phase commercial.

The neutral busbar in an American panel has provision for a bonding screw to connect it to the metal casing. At the main panel this bond is put in place, and both the incoming supply neutral and the GEC (Grounding Electrode Conductor) are connected to the neutral bar. As Ragner has said, both neutral and ground wires from circuit cables can be landed on this busbar.

Where a unit is employed as a sub-panel, the grounding bond has to be removed and separate earth and neutral busbars employed. The feeder from the main panel also needs to be 4 (or 5) wire to keep the neutral and ground separate.

As you can see from this, the only major differences between this and British PME are that in the U.K. we have the bond placed before the meter and keep neutral and earth separate in the main panel, and that the American system has a local ground rod as well. The latter is quite reasonable when you consider that in a typical American neighborhood there will be a larger number of smaller transformers feeding the houses, instead of the one big xfmr with a 3-ph/4-w network that's typical here.

The other main point to note is that range and dryer outlets fed from the main panel were not requirted to have a separate EGC, but the frames could be grounded to the neutral (i.e. a 3-wire connection from panel to receptacle, with the white serving as both neutral for the 120V loads and as a protective earth). Ranges and dryers fed from a sub-panel still required a 4-wire system.

This system has now been outlaws under the 2002 NEC, and all new range and dryer outlets must be 4-wire.

I don't have any diagrams I can post easily, but maybe one of our other members can point you toward something.

Pauluk, that was a good description. The UK system avoids some net current problems we have here, and to me it makes more sense to keep the neutral and ground (earth) separate until the Tformer.

Instead, with neutral bonded to "ground" at the main panel, if there is a metal water supply system, some of the neutral may return to the Tformer by way of the water pipe to a neighbor's house and back their neutral. Once again, if impedances differ, neutral can be shunted all over the neighborhood, creating the magnetic fields I am called on to deal with.

Of course the rod to earth does not enter into this circulation since the impedance to earth is relatively so high.

Another problem is that the bonding screw you spoke of is sometimes turned in in a subpanel out of ignorance, thus putting neutral on many earthing paths throughout the house, and in commercial buildings all over the place. Best to keep the two separate until the Tformer, but the US is stuck with its system.

Karl

I forgot to say that the local ground rod for the building is there only for lightning protection (if such exists) or for accidental high voltage surges from storms.

However, many US electricians carry in their heads the belief that it plays the main or sole role in clearing a fault (opening a breaker or fuse) in the case of a wiring fault. Hence there are many serious discussions about how many rods to use, and how to measure their impedance, etc. It has a lot to do with the multiple meanings of the word "ground", which some are trying to correct in the Code (NEC).

Karl again

Karl,

We have some drawings of U.K. systems in the technical reference area which ECN member Nick kindly put together in AutoCAD for me some time ago:
https://www.electrical-contractor.net/ubb/Forum15/HTML/000074.html

The underground systems (now designated TN-S) as originally installed in most urban areas did indeed have a ground on the neutral only at the star point of the distribution transformer. The ground-fault path was by way of the outer armor on the distribution cables.

Our PME system (otherwise known as TN-C-S) can give rise to similar problems of parallel neutral paths as in the States due to bonding to all water services, gas pipes, etc.

The TT system, very common in rural areas, does actually rely on the earth as the ground-fault path (and would thus not be allowed under the NEC). The relatively high impedance limits fault currents, but necessitates the use of an RCD (GFI) for all circuits in the house.

Karl,
Confused about function of earth rod at building. Would have thought that primary purpose was to keep neutral at or about earth potential. Function of star point connection to earth at xtrfmr may well serve to limit voltages impressed from primary through capacitive coupling.

Pauluk, yes the GFCI provides the needed safety. We can't use that on the service cable here due to the water-pipe alternative for neutral. The GFCI would trip constantly.

Lyledunn, yes, the ground rods of the system as a whole keep the voltage at a more or less constant relation to earth, but this does not rely on any particular rod; the rod at the building would not be needed as the others are doing their job.
Of course, neutral is ony at ground potential when no current is flowing, but since 30% or more of primary neutral flows back to the substation by way of the earth, there is always a small neutral/earth potential, sometimes negative.

Did I muddy those waters? If so, set table on fire.

Karl

Once again, Pauluk, yes, I had seen those excellent schematics and was basing my comments on them.

I did not see how the grounding to water pipes would carry neutral when it was isolated from ground in the building. ??

Karl

Current on the pipes isn't a problem with TN-S and TT systems, but it's a definite possibility with PME/TN-C-S where the pipes are bonded to the neutral.

Newer water/gas installations use PVC service lines these days, which obviously reduces the current flow, but there are still many old systems with metallic pipes criss-crossing a neighborhood.

The possibility of parallel neutral paths via pipework is why the PME system is not permitted by our code in certain situations, e.g. at gas ("petrol") stations.

Originally, the majority of distribution in Britain was TN-S in urban areas and TT in rural areas. The neutral was grounded only at the transformer.

The TN-C-S/PME system has been in use since the 1930s, but was only employed in areas where soil conditions made it difficult to get a sufficiently low resistance ground connection. The PoCo had to install multiple ground rods along the route and actually get the permission of the Secretary of State for Energy before they would allow customers to ground their systems to the neutral.

In recent years, however, there has been a move toward converting all the main LV distribution lines to PME by installing the extra ground rods. In the area served by my local PoCo, the conversion is complete, so all customers can switch to TN-C-S if they wish, although I'd say that 95% of the rural homes in this area are still TT.

Hi Paul

By my name you can gather i live in Cornwall,,

if you ask for any alteration on a service down here in the sticks,you now seem to automatically receive a PME supply,,

not sure why the electric companies are changing over to PME from TT..saying that though every pole round here now seems to be earth staked and bonded to neutral..

just a foot note though,, went to a pair of cottages to carry out a test,, TT supply.. found no earth reading on the earth rods,, these had been rewired not more than 5 years previous.. the RCDs just didnt operate,, had to put three rods per house to achive a suitable earth reading,, i dont belive the earth reading would have been reduced to such a level in that time.. i belive the last spark just knocked them in and hoped for the best..

Are you sure about the TT supply for petrol stations?? all the ones i have come across are PME

I am slowly getting used to UK systems. I have this question: I assume TT was used before GFCIs were invented. How was a fault cleared back then, or was everything on a 3 amp fuse?

A study done in 5 midwestern states here showed a ground rod impedance average of about 113 ohms. What do you typically measure on yours?

Karl

Yes, TT has been employed for a considerable time, using protection in the form of an ELCB (Earth Leakage Circuit Breaker).

There were early types of current-balance ELCBs, which would trip on an imbalance of about 500mA or so. In more recent times these were made more sensitive, and until a few years ago 100mA was pretty much the norm. These days, 30mA is considered standard, so they still don't offer quite the same level of protection as the typical American GFCI. As the typical device now feeds an entire house, any attempt to increase sensitivity further would lead to problems of nuisance tripping.

The committees here have an unfortunate habit of messing around with terminology for no particular reason, and the current-operated ELCB has been renamed twice in comparatively recent times. First it became an RCCB (Residual Current Circuit Breaker), then more recently still the name was changed to RCD (Residual Current Device). At least one manucturer appears to have resisted this latest change and still uses RCCB on some products.

We also had a device known as a voltage-operated ELCB, which inserted a solenoid trip coil between the house grounding and the local rod.

These devices are obsolete now, but were very common in rural residential properties up until the 1970s. There are still quite a lot in service in my area.

Click here to see how they were wired.


[This message has been edited by pauluk (edited 08-29-2003).]

Old TT systems solely relied upon the 6A or 10A fuse in the distribution panel.
In the 1950ies old installations were usually upgraded by stringing a red wire from the water pipe to 1 or 2 kitchen receptacles, just across the wall, using the metal water pipe (back then usually lead or in newer work copper) as a ground rod. New wiring back then used a real ground rod and grounded receptacles everywhere. Residential RCDs came up along with MCBs in the mid 1970ies, usually 500 or 100mA.

Thanks for the clear descriptions. In USA I often see many amps flowing from the neutral/ground bond and on to the water pipes, to complete the circuit by way of the neighbor's pipe and neutral. So you save yourself a lot of problems by bonding only at the Tformer.

Of course we are allowed to insert an insulating spacer in the water pipe 10' outside the building. This will force the neutral current to travel in the neutral conductor. This is not a common practice, however, and sometimes meets some resistance (mental impedance).

Karl

lyle,
What a question you ask!.
This very topic has got me into trouble more than once.
Just ask Bill.

Paulcornwall,
PME not permitted on petrol filling stations-Institute of Petroleum. Reason, diverted neutral current through metal work associated with underground tanks could be explosion risk.Where you come across such a situation you should advise client to convert to local TNS or TT by employing locally grounded Isolating transformer.
Karl,
Typical ground resistance where I live (Foothills of the Mourne Mountains in Ireland)is in the order of many hundreds of ohms! Sometimes more money is spent on the earthing system than on all the other parts of an installation.

Sorry, I forgot about the impedance question.

Impedance values aren't too bad around here, I can often register 50 or 60 ohms. But I know that in some parts of the country it can be difficult to get values anything like that low. It's these difficult areas which were the places in which PME was originally employed.

Karl,
One point you may not be aware of is that the typical residential ground rod in Britain is generally only about 4 ft. these days, as opposed to the 8 ft. rods used in America.

Tex,
Did Austria really rely on just a 10A OCPD and a local rod to clear an earth fault? The loop would need to be 22 ohms or less even to get 10A flowing, nevermind enough to open the fuse in a reasonable period of time.

What sort of grounding systems were used way back?

Yep, really, at least in residential wiring. I read that out of a 1958 DIY book.
Grounding methods largely depended on the PoCo, some of them have been using TN-C-S for ages (maybe even TN-C, I'm still waiting for an answer on that question), others solely used TT until 1998, like Wienstrom, even in urban Vienna. I think the ground rods were either driven in the basement or put into the concrete foundation. The oldest RCD I've ever seen was at a school that focused on electrical engineering. It was a _huge_ standalone box on the wall, feeding a single receptacle. Maybe the size of a shoe box.

I can't imagine trying to use TT with no sort of ground-fault protection beyond a 10A fuse! (Regulars will know that I don't care for the TT system at the best of times.)

If the loop impedance isn't very low, then current will flow back via the ground but the whole grounding system of the building will float up to near 220/240V. The ground rod might just as well not be there at all.

In fact you'd be better off with no grounding -- At least then it would be only the frame of the faulty appliance/device which gets energized, not everything that's bonded to it throughout the house.



[This message has been edited by pauluk (edited 09-02-2003).]

Over here in Ireland bonding to a water pipe to ground an installation is absolutely illegal as there is a strong possibility of the watermains being made of plastic resulting in a pretty poor grounding arrangement.This has been the situation since the 1960s. Even where a water service enters the house as a copper pipe the main feeding it might be plastic.

semiflexible plastic pipes are increasingly being used in internal plumbing here too further complicating bonding requirments as it is still necessary to bond the metal parts of the plumbing to ground and obviously to ensure that the plumbing doesn't carry a current in the event of something energising the pipework (e.g. a faulty immersion heater) With plastic pipes there is a possibility that the path of least resistance may be via the tap/shower and the person touching it if bonding is not carried out properly.
Particuarly considering that Irish houses use a plastic tank in the attic as a head tank for the hot and cold water systems other than the kitchen. So you could have a completely electrically isolated body of water in the attic feeding the bathrooms!!! If the water heater had a ground leak to the water system the only path to ground might be via the shower if grounding / bonding is not correctly carried out.

Local grounding's always done through an earthrod, these come as a large concrete block the top of which is suitable for blending into paving slabs/bricks.

If anyone's interested in seeing all of the permissable connections in domestic LV situations click on this PDF http://www.esb.ie/main/downloads/energy_home/National_code_of_practice_3rd_Edition.pdf

ESB network's code of practice.

also Pictures of irish service entrances, main fusing arrangements, metering arrangements, etc etc..

It even has pics of how to wire up a non-metered service for a payphone & associated lamp & on street CCTV cameras


[This message has been edited by djk (edited 09-02-2003).]

Let me add a comment about North American practices {and typically worldwide} that may be apparent. [I hope I haven’t overlooked someone already detailing this here.] “Bonding/grounding” of interior metal water and gas piping serves two ends. To an increasing degree, it’s use as a grounding electrode where an insufficient length of it has direct soil contact, is becoming less common with the increase in non-metallic piping, both underground and above ground.

Nowadays, the primary reason for “bonding/grounding” is to limit potential difference between metal piping and other conductive objects in and around buildings. Continuity is also conducive to rapid clearing of overcurrent-protective devices (breakers/fuses.)




[This message has been edited by Bjarney (edited 09-02-2003).]

Outlawed 01/01/01, but I installed such a system myself (or at least extended it) some time ago.
Original wiring was like this: 10A Edison base MCB (awfully old no-blow type, was stuck in the "on" position). Feeder to kitchen 2 black cloth covered wires in plaster. At wall sconce NYIF cable tapped off to receptacle, 1.5mm2 red wire in conduit from water line somewhere near the sink or maybe the toilet tank on the other side of the wall to the receptacle. From this point cables with ground fed the 1500W 5l HW boiler and another receptacle. I rerouted the wiring, but the grounding remained the same until we replaced the feeder. And we installed a 30mA RCD and new breakers right at the beginning. Old apartments were usually like that. All outlets apart from 1 or 2 in the kitchen ungrounded, these few bonded to the water pipe. If there was an infrared heater in the bathroom it was probably grounded as well. And even though it's illegal I think hardly anyone bothered changing old wiring. Even where water risers in apartment buildings are replaced with plastic the grounding usually remains that way, i.e. the ground isn't a ground any more but just a useless wire.

Since it seems that in some countries at one time the earth rod was the only means for hopefully opening a fuse in case of a fault, I am wondering how any electrician could be so optimistic.

This is a question I have for US electricians also, since so many still assume that the connection to earth provides this safety. One would think that everyone learned V=IR and could understand that a rod/earth impedance of, say 50 ohms would allow only 2+ amps at 120 V or 4.8 amps at 240 V. How can that blow a fuse or trip a circuit breaker?

Does anyone on this forum understand the thinking involved?

Karl

That's an interesting document from the ESB in Ireland. It suggests that all new services are PME (TN-C-S). Is that right Dave?

The grounding in Austria sounds as though it was very haphazard in the past. (Maybe that's why they chose red for ground: It might actually be energized to 220V by a fault! )

Karl,
I think the issue of grounding/earthing in general is an area which is grossly misunderstood, both by the layman and by some electricians.

It goes hand-in-hand with another problem that we've discussed at ECN in the past: The belief that electricity always tries to return to ground. It doesn't.

If people would abandon that misguided concept and accept that current will simply try to flow in a complete circuit back to its source, then they would see that the earth can just be a big conductor in a part of that circuit.

Paul,
Yes, that is what I try to get across: the fault current is seeking to complete the circuit to the transformer (following voltage gradients). I have referred to the belief that the earth is an electrical sponge as the picture some people have, perhaps because they see lighning strikes disappearing into the earth, but even here there is a circuit being completed between earth and atmosphere.

But what I am still mystified about is how V=IR has escaped the attention of so many electricians and also engineers.

Karl

Paul,

Yup is the standard even on old installations, except in absolutely exceptional circumstances.

The ETCI (ElectroTechnical Council of Ireland) and ESB are pretty strict on neutralising requirements contractors have no flexibility or options the ESB specify after a pre-connection site survey how it will be done and the bond is sealed.

[This message has been edited by djk (edited 09-03-2003).]

Karl,
The use of Ohms law is made with the UK requirement that the earth electrode resistance multiplied by the nominal operating current of the RCD should not exceed 50V for normal installations and 25V for special locations like farms and construction sites.

I find the ESB's use of that term quite amusing. Maybe it's because we apply the term neutralizing in electronics to a totally different concept.

For whatever reason, some people seem to believe that ground connections work differently and don't obey the same laws of physics as the rest of the circuit. It's as if they think that driving a rod into the ground at any point makes an automatic zero-ohm link to any other ground rod.

Lyle mentioned the 50V requirement for RCD-trip x ground impedance here. When the voltage-operated ELCB was installed here, it was also a design requirement that the breaker must trip with 50V across its operating coil. In many cases, the ELCB tripped at considerably less than that.


[This message has been edited by pauluk (edited 09-04-2003).]

The ESB just like words ending in "ise"

Neutralise
Energise
De-Energise

so then you can sound really complicated and talk about the neutralisation and energisation of the 0.23 kV 80Ampere supply, which was previously de-energised, that occured at 12:02:39.

that way you charge four times the price


[This message has been edited by djk (edited 09-04-2003).]

Lyle, I was talking about the former use of the rod only, but thanks for the formula. At 30mA for the RCD you could have a rod with 1,666 ohms impedance! Pretty safe system.

Paul, perhaps the belief that a rod to earth brings everything to the same potential is based on the fact that this is true as long as no current is flowing. Nasty old current!

Karl

Yes, current is really nasty: It's current that heats up the wires and cause fires. It's current that kills when it flows through your body. It's current that causes electromagnetic fields. Electricity would have been so much simpler without it!

[This message has been edited by C-H (edited 09-05-2003).]

LOL!

So it works just fine until it's switched on!

Re the 50V limit, this is another case where changes to the rules over the years have not necessarily tightened the requirements as one might expect.

Allow me to quote from the 1966 IEE Regulations (14th Edition):

The trip currents have become much lower, however:

And since those long superceded regs were published there has been a lot of international research on the effects of electric current on the human body, not to mention the technical advances in RCD design.