What is "stray voltage"?

Yes. Not only that, each separate grounding wire connected to the building’s steel work and plumbing should terminate at a single point connected to the dirt.

I am tempted to jump in here but there has been a lot discussed already and much of it is slightly off target. To begin with, no one is using the correct NEC terminology. There is a grounded conductor and a grounding conductor. The grounded conductor is commonly (and incorrectly) referred to as the neutral. It and the grounding (safety ground) conductor are connected together at ONE POINT only. That point is in the main service panel where the electrical service enters the building. All subsequent sub-panels (and there can be many of them) do not have the grounded and grounding conductors bonded together. This is very important in an outbuilding where there is bound to be a sub- panel.

The driven ground rods are 10’ by code and often a group of three several feet apart. The resistance from the ground rods and the earth should be measured and additional ground rods driven until a certain value is met. Very few single family homes have done this.

A large building will have a ring of heavy wire buried all around the building to serve as the ground. Even the re-bar in the concrete is attached to it.

It’s hard to describe without a diagram (and I’m too lazy to look for one).

This is true for many parts of the U.S.:

There’s a pair of wires running around your neighborhood that has a high voltage between them. I think it’s around 7500 VAC for most places. One wire is connected to ground (the dirt) in various places. The other is not.

When the two wires go by your house, a transformer is installed to reduce the voltage from 7500 VAC to 240 VAC, and the 240 VAC is run to your house. (The transformer might feed just your house. Or it might also feed one or two other homes.) The two wires that have 7500 VAC between them connect to the primary winding of the transformer, and this results in 240 VAC showing up on the secondary winding of the transformer. (There’s also a center tap on the secondary winding. Which is nice, because it means 120 VAC and 240 VAC are delivered to your house.)

There is electrical isolation between the primary and secondary windings of the transformer. If the isolation were perfect, then there would probably not be a need to connect the center tap of the secondary to the earth. But the isolation for a real transformer is not perfect; the electrical resistance between the primary and secondary is not infinity, and there is always some capacitive coupling. In addition, if there’s a fault it could result in a low-impedance path between the primary and secondary windings, which would really be bad.

So for a real transformer, if the secondary were not referenced to earth ground, its common-mode voltage could “float up” to an appreciable fraction of the primary voltage. That’s dangerous. So to keep this from happening we connect the center tap of the secondary winding to the earth. We do this at the transformer and at your house for good measure. In fact, the more locations it’s connected to the earth, the better.

Oh, and one more thing: the secondary of the transformer has three terminals: L1, L2, and the center tap (N or neutral). To keep the secondary from “floating up” to a high voltage, it really doesn’t matter which terminal (L1, L2, or N) is connected to earth ground. The best choice is N, though. That way, every 120 VAC circuit in the house will have one conductor connected to earth ground, which makes things safer.

Large buildings often have multiple (‘static’) ground points, the multiple ground points connected to the frame, and may have each ground point bonded to a grid electrical ground/neutral.

The grid electrical / static ground should then be bonded to the building electrical ground at one point.

Multiple earths, multiple connections between the supply, the frame, and the earths, only one connection on or around the main board bonding the building power neutral to electrical ground and to earth.

This is the footprint of a large building. I think that for a very tall (>10) building, the frame is isolated from the electrical ground at all points other than at the foundation / substation.

The frame should be at approximately neutral potential everywhere (no stray voltage), because (1) there is no current in the frame: it’s the same voltage as the bonding point, and (2) the electrical neutral is large and low resistance, so that even when current flows through it, the upstream end is still at approximately the voltage of the bonding point.

When the frame is hit by lightning, the common electrical ground helps equalize the frame voltages, the multiple earths help dissipate the lightning strike.

Leslie Harvey was the singer for Stone the Crows. In 1972 he grabbed a microphone that was not properly grounded, and he was electrocuted and died.

This has happened several other times. Most recently a young singer named Agustin Briolini died while performing in a club and touched a hot mic.

Lots of famius musicians have gotten serious shocks on stage due to improperly grounded equipment. George Harrison, Keith Richards (probably where he got his superpowers from), Moby, Kesha, Nic Lowe… The last one is interesting as apparently the shock actually stopped Lowe’s heart, but a roadie who tried to kick the gear away from him accidentally kicked him instead, and that supposedly revived him. I have no idea if the story is true, but I’ve seen it in several places.

Roadies: Ground those instruments!

Where do big buildings move the current to?

I live in a high-rise in a big city with loads of skyscrapers. They get hit by lightning all the time but no one is zapped because they are leaning against the building when it gets hit.


(real pic of three buildings in Chicago being hit by lightning at the same time…left-to-right: John Hancock Tower, Trump Tower, Willis Tower (formerly the Sears Tower)…they are closer in size than is apparent in the pic due to the angle of the photo and distance to each building)

I don’t know what the norm is for most large buildings, but in the lab where I work, a reliable, low-impedance ground is of major importance for sensitive instrumentation. A few years ago I learned about a ground mesh (see slide #14 here) that was installed for this purpose. A mesh provides more contact surface area between your ground element and the earth, and distributes that contact over a much larger patch of earth. In our case, the mesh is maybe half the size of the parking lot.

I also learned about ground enhancement material (GEM). If the soil has decent conductivity, and you’re just dealing with a safety ground, then pounding a ground rod directly in the soil is adequate. For soils with crummy conductivity, or a need for a better quality ground (see previous paragraph), you can use GEM, which is basically conductive concrete. A 5/8" diameter rod a few feet long provides very limited contact area if you sink it directly into soil, which can result in unacceptably high electrical resistance. So instead you bore a post-hole, fill it with this stuff, and set your ground rod in it. Now instead of your 5/8" rod, you’ve got a 6" diameter cylinder making contact with the surrounding soil - ten times more contact area, reducing contact resistance by a factor of ten.

The simple explanation for grounding is that you want everything that humans will be coming into contact with to be at the same potential, because if a person is touching two things at different potential, then a current will flow through that person. And the best and easiest way to ensure that things are at the same potential is to connect them all with good conductors. And one of the things that people will be coming in contact with is the planet, and there’s not much you can do to change the potential of the planet, so you want to make sure everything else is at the same potential as the planet, which you do by connecting everything else to the planet with good conductors.

Um, no, maybe not well explained - the neutral conductor is connected as you would expect but what you have is a neutral return conductor that goes back to the earth electrode in a combined system.

Farms typically have a 2 wire overhead supply (IT supply system) - with no earth supplied with the incomers, to get the earth reference back to the star point of the supply transformer whose star point becomes the neutral return and is also referenced at the transformer end to earth via its own earth electrode.

The consumer supplies their own earth electrode - which will be a grounding rod at their end - but if this does not produce a good enough earth reference then the rod needs to be much longer, or be a large area grid plate sunk into the ground. Farms are well advised to barrier the area in the close vicinity of the earth electrode to protect livestock from stray voltages that arise - especially during fault conditions.

What can happen is that the area in which the local earth reference is located dries out significantly it can raise the potential of the local earth reference.

This in turn has an effect on the neutral return since it now floats in voltage - what can happen here is that the neutral was earth referenced to a different point (back at the supply transformer star earth point), it can have a lower potential than the local earth electrode.

Is there any issue with having current running between multiple grounds?

When I was a cable tech, they had somewhat recently* changed policy from using our own grounding rod to tying into the existing ground rod for the electrical system. The reason given was if they were at different voltages, which could happen because of different soil chemistry, or just different amounts of moisture content, then you would actually have a current running between the grounds.

Not enough to shock anyone, and really not enough to damage equipment, but enough to interfere with the signal and potentially cause static.

*I think that this change actually happened many years prior, but we were told to check on this and update it to code whenever we worked on a house.

Yes, this causes “group loops.”
Instead of all the grounds being at the same potential (zero), some are higher than others, which means that equipment that is connected to both grounds (like cable TV, or telephone or data lines) can get current in the grounded shield of the cable. This causes “hum bars” on old-fashioned TVs, and can cause data corruption on RS-485 lines. Most data systems can handle several volts of ground difference, but it’s always nice to eliminate it.

BTW - here’s someone who was killed by stray voltage.

Yep, my old apartment building had this problem. The (crappy) electrical system and the cable TV drop had slightly different ground potentials. I had the cable box connected to my home theater system, causing a tiny amount of current to flow between the electrical and CATV grounds.

The difference was tiny, maybe tens of milivolts, and didn’t affect most things, but when listening to (vinyl) records it caused a very audible hum owing to the phono pre-amp stage which works in the mV range.

I found a ground-isolating surge protector that fixed the problem. It has co-ax pass-throughs which are isolated from the CATV ground to ensure that everything stays at the same potential. No more hum, but a real pain to track down.

Of course buildings tend to be wired up to code and any subsequent changes made by properly qualified people.

Farmers, in my experience, are reluctant to pay someone to do anything they believe they can do themselves. Given the state of a typical small farm cowshed - wet concrete, metal all over the place, electrical equipment, often with trailing leads… it’s hardly surprising that ‘stray voltage’ is a problem.

I live in Chicago and unionized workers (plumbers, electricians and so on) are SUPER expensive. It is painful to pay for their services (recently had a toilet installed…$400…took less than an hour).

But, I will say, they do great work (in my experience). I get wanting to save money but sometimes it is best to just pay to have it done right the first time. It’ll save money and hassle in the long run that can be well worth that initial cost.

As this thread shows, experience can count for a lot. You may think you can do it because all you need to do is screw this thing into that. Later you learn the hard way there was a lot more to be considered.

Nitpick, but RS-422 and RS-485 use differential transceivers and are thus fairly immune to ground loops and common-mode noise. RS-232, on the other hand, uses single-ended transceivers and must have a common, low-impedance ground.

The common-mode range of RS-485 input receivers is finite. Most will support up to 12v, but any more than that will cause problems.

One should also be careful to distinguish between “stray voltage”, “feral voltage”, and “wild voltage”. Stray voltage is voltage that was once domesticated, before escaping. Feral voltage was never domesticated, but was induced from domesticated or stray voltage. Wild voltage, meanwhile, has no domestication at all in its history.