Trying to better understand the physics of electrical earth ground

[FrankJ777]

I'm trying to better understand the physics of how Earth ground works.

In circuit analysis and other electronic courses they usually present a conceptual picture like below where the Earth is viewed as a path that completes a circuit? In this conceptual view, the current travels on the transmission line and the return path for the current is the Earth ground. This fits very well into our idea of current always traveling in a circuit.

However, in my thinking the flow of charge just goes from high potential to low potential (or vice versa), and the return path is just conceptual in this case.

Watching this video, WCLN - Physics - Electrical Ground , they explained for Earth ground the Earth acts as a sink with a relatively neutral potential. Because of it's large mass it can except excess negative charge, in the case of a negative ground, and in the case of using a positive ground it can give up negative charge. So this bypasses the conceptual Earth return path.
Just to make sure I'm understanding what really going on, I put together this silly example in the picture below. Ignoring the obvious physical limitation, would this concept work? If I'm understanding how Earth ground works, it seems that you could have a single conductor transmission line with a power source grounded on a large mass like Earth and a load grounded on physically separate large mass like Mars, and because both massive enough to provide a relatively neutral ground, the "circuit" would still flow?
So in other words, I guess I'm asking if the two "earth grounds" really require a path between them?

 

[Dale]

Whenever you have a “floating” conductor like this the thing to do is to connect each “floating” conductor to ground with itself capacitance. Self capacitance is the capacitance of an object with respect to a sphere of infinite radius surrounding the object. Both Earth and Mars can be considered to be capacitavely connected to this same infinite sphere.

Earth’s self capacitance is about 700 uF and Mars’ would be similar. Two 700 uF capacitors in series would give an equivalent capacitance around 350 uF.

With the resistance that would give you a RC circuit whose time constant you could calculate. But probably you would also need to include the resistance and inductance of the wire.

Of course, none of it is realistic, but you could do something similar in a vacuum chamber with spheres of rock.

 

[etotheipi]

In circuit analysis and other electronic courses they usually present a conceptual picture like below where the Earth is viewed as a path that completes a circuit? In this conceptual view, the current travels on the transmission line and the return path for the current is the Earth ground. This fits very well into our idea of current always traveling in a circuit.

Note that your first diagram is an example of a "single-wire Earth return" system. As far as I am aware for steady power transmission (i.e. not a one-off discharge) you always need some means of completing the circuit, be it a neutral wire or, in this case, the Earth. Maybe @anorlunda can advise...

 

[Paul Colby]

And of course someone should note the obvious that actual power distribution systems don't rely on a ground for the current return. Your electric bills would be enormous.

 

[kuruman]

I think of the Earth ground as an infinite reservoir of charge that can provide or accept any number of charge carriers without changing its potential. As an analogue think of the Earth's oceans as an infinite reservoir of water. You can pump water into a pipeline from New York harbor, send it across the USA and dump it in San Francisco Bay. While this happens, the drop of the ocean level is negligible and the water that is dumped in San Francisco does not have to return to New York to maintain the water flow in the pipeline. In the case of Earth ground, electric charge carriers replace water mass carriers but the idea is the same.

 

[anorlunda]

Arrgh. Grouding is one of my least favorite subjects.

First, regarding your first sketch. There really is something called single wire Earth return power distribution. It is rare, but it does exist. See below.
https://en.wikipedia.org/wiki/Single-wire_earth_return

On the other hand do ground the wires at the customer site for safety purposes in case of accidental open or short circuited paths. But in normal circumstances and normal voltages, the Earth return path is much higher impedance than the wire return in a closed circuit. Therefore, the current going into that ground connection is nearly zero. Assume it is zero.

So yes there is such a thing as single wire Earth return power distribution, but when the return path is provided by wires (the usual case), and there are no wiring faults or errors, then the ground connections can be ignored.

But then there is the case of long power transmission lines. Those lines do have a capacitance between the wires and the ground. It really does shuffle electrons in the ground back and forth, but the effect is distributed over hundreds or thousands of miles and perhaps 1 million square meters of ground before it becomes significant.

What you imply in your second sketch with Mars, is that a single wire grounded at a single point, will have a huge capacitance because the planet is large. As @Dale said, there is a self capacitance of the ground and we have typical numbers for it. However, the significance of that is frequency sensitive. For DC and at power frequencies, that is negligible and we wave our hands and call it zero. In fact the rules of circuit analysis explicitly forbid consideration of charged "piling up" at any point in a circuit.

At higher frequencies, kilohertz into megahertz range, the self capacitance of a grounded point may become significant. You could slosh electrons in the ground like water in a bathtub. But at the same time, those long wires begin to act like antennas and radiate power away in all directions, and other strange effects begin to show up.

My usual advice is that when the assumptions of circuit analysis are no longer valid, then forget circuits altogether. Do not draw circuits like in your post #1. Instead, you must use Maxwell's equations and the actual 3D size and orientation of everything to describe what happens. Trying to understand exotic cases while thinking of normal circuits is evil.

Lightning hitting the ground, voltage gradients in the ground during a solar storm impacting Earth, and people electrocuted while swimming near a marina, piezoelectric effects near earthquake faults, Saint Elmo's fire, ground loops in the power wiring of computer farms, and particle accelerators, are other examples of real physical things that can't be adequately explained using circuits.

 

[FrankJ777]

Power is just one case where this interests me. I've read that the telegraph systems used a single conductor with an Earth return. Not sure about the old wired telephone system, but I think i was simular. Anways grounding is an important issue, so I'm trying to have a better understanding of what I'm trying to achieve.

 

[Paul Colby]

So yes there is such a thing as single wire Earth return power distribution,

What of the ohmic loss of dirt?

 

[anorlunda]

Back when I worked for G.E., grounding was one of those things called "The Art and Science of .." That's why I hate it. I like Art and I like Science, but Art-and-Science makes me dizzy.

What of the ohmic loss of dirt?

There may be a wide range of values, and especially whether the ground is dry or wet. The Wikipedia article linked says that in Alaska, they need to put the grounding rod below the permafrost because permafrost has high resistance.

Single wire power distribution sacrifices power efficiency for the benefit of lower capital costs. Copper wire is very expensive. Places like Alaska, Brazil, Australia and New Zealand, may have customers in very remote areas who use very few kWh per month. For them, initial capital cost is more important than operating cost. The Wiki article says.

SWER also reduces the largest cost of a distribution network: the number of poles. Conventional 2-wire or 3-wire distribution lines have a higher power transfer capacity, but can require 7 poles per kilometre, with spans of 100 to 150 metres. SWER's high line voltage and low current also permits the use of low-cost galvanized steel wire (historically, No. 8 fence wire).[9] Steel's greater strength permits spans of 400 metres or more, reducing the number of poles to 2.5 per kilometre.

so I'm trying to have a better understanding of what I'm trying to achieve.

What are you trying to achieve?

Here's some more links for a range of applications.
https://www.electronics-notes.com/articles/antennas-propagation/ground-wave/basics-tutorial.php

https://www.sciencedirect.com/topics/engineering/return-path

https://borderlandsciences.org/journal/vol/53/n01/Vassilatos_on_Ground_Radio.html

https://www.allaboutcircuits.com/technical-articles/an-introduction-to-ground/

https://en.wikipedia.org/wiki/1-Wire

http://denethor.wlu.ca/pc320/lectures/singlewirebeam.pdf

 

[FrankJ777]

OK. So I was wondering if in general i should think of a grounding system like below, with a capacitance to ground?
Or also I was wondering given the explanation by @Dale if I should just consider both ground point as capacitivly coupled to the same imaginary infinate sphere. Even if both points are on Earth that seems like it would be a better disciption of what's happening.
Also, to be sure, when we're talking about ohmic losses, I think we don't meen the total resistance from the ground and point A throuhg the Earth to the ground at point B do we? If so I think we would be treating the Earth return path as a big resistive path. If that were the case then the longer the Earth return path, the more resistance would be added to the circuit. But I don't think that that happens, or does it?

 

[tech99]

It is also interesting that if you place a copper plate on the Atlantic on the US side and another on the Europe side and join them with a wire, then you have a circuit. This was done with telegraph cables and is also how power is fed to modern telecom cables for powering the submerged repeaters. So the current enters the Atlantic in the USA and finds the copper plate in Europe. How does this happen? Because the current in the water is attracted to the cable and follows it.

 

[anorlunda]

If that were the case then the longer the Earth return path, the more resistance would be added to the circuit. But I don't think that that happens, or does it?

As I said before, it depends on the frequency. Is your application communications? Is it power transmission at 50 or 60 Hz? Is it DC power transmission?

 

[DaveE]

I like Art and I like Science, but Art-and-Science makes me dizzy.

Yes, LOL. Me too!
I always preferred designing a good ground system to avoid dealing with the question "How bad is a bad thing and in which ways is it bad?" But even good system designs can have tricky signal grounding issues. Sometimes the "Art" part is picking the lesser of evils without spending your whole career analyzing it.

 

[DaveE]

Also, to be sure, when we're talking about ohmic losses, I think we don't meen the total resistance from the ground and point A throuhg the Earth to the ground at point B do we? If so I think we would be treating the Earth return path as a big resistive path. If that were the case then the longer the Earth return path, the more resistance would be added to the circuit. But I don't think that that happens, or does it?

For ohmic measurements, the longer path will also allow the current to spread more widely in the perpendicular directions. Longer paths will also become wider paths, so the resistance doesn't keep growing with distance as much as you would expect. Of course this depends on lots of non-uniform details in the ground (soil type, water, etc.)

 

[sophiecentaur]

I think of the Earth ground as an infinite reservoir of charge that can provide or accept any number of charge carriers without changing its potential.

The Earth is big, agreed, but, as Capacitors go, it's really not that big. It's a mere 0.7milliFarads (treating it as a conducting sphere), which is a value of Capacitor that you can go to a shop and buy. 0.7mF means you can charge it up with 1Coulomb (just 1A flowing for 1s) and the Potential will increase by 1/0.7 = about 1.4V.
The Net Potential would also depend on the total population using their 'Earth' connections which could well average out and produce the effect of a much larger sphere.

 

[cmb]

Yes, there is a physical electrical DC connection, if that is your question.

Sink two grounding rods and try putting a current between them, it's quite a low resistance (well, it should be, if they are good grounding rods).

It's a way I test how good my grounding rods are, I will sink two, measure the resistance between them to make sure they are 'good', and then use them in parallel as 'the Earth rod'.

(PS .. good way to sink copper rods is to put a hose on one end and run water through it, as the rod sinks down the water softens and washes soil and stuff out of the way as you hammer it in.)

But consider that the return conductor can be very long indeed, and if so then it'll start behaving differently if you are ramping up and down a current. Connect via Mars, and even if there were a return path for charges, it would not behave like a scale model circuit!

So, when you are thinking about RF 'grounding' over a long distance, think of it more like a sea of electrons washing back and forth, trying to find its own level on two beaches many miles apart, and it is not simply like a DC current when a RF wave passes in the air over the Earth's surface, and it causes ground waves at HF frequencies.

...
Can I throw in a slightly different, related idea, though, that I've never really fathomed, and is this actually 'why' there is a conductive path?

Is the Earth electrically neutral, or slightly positive or slightly negative? I will give a definition, important here; by that I mean, are there as many electrons as there are protons in the Earth's molecules?

We can only test 'relative' potential so if we all lived on the top dome of a massive Van de Graaf generator at megavolts to somewhere else (for the sake of an analogue) how could we tell?

How can we measure how many 'excess electrons' there are on/in Earth? Is it intrinsically neutral, if so why is there a very significant electric field vertically in the atmosphere?

If we visited a planet that had only half the number of electrons to all of its molecular protons, would the chemicals of our bodies begin to fall apart as the valency electrons got sucked out of them?

 

[anorlunda]

Is the Earth electrically neutral, or slightly positive or slightly negative?

The Earth is approximately neutral. However it is also bathed in the solar wind, which brings both positive and negative particles. The visible effects of that are the aurora. So I guess that the real answer is that the neutrality of Earth is time variable.

If we visited a planet that had only half the number of electrons to all of its molecular protons, would the chemicals of our bodies begin to fall apart as the valency electrons got sucked out of them?

The Coulomb Force is much stronger than gravity. However when matter is exactly neutral, the attractive forces balance the repulsive forces and the net force at a distance is zero. If you had 50% unbalanced charges, I think the answer is that the planet could have never formed in the first place. Seeing so many objects in the universe that are not flying apart gives evidence that they are all approximately neutral.

 

[DaveE]

How can we measure how many 'excess electrons' there are on/in Earth?

'excess electrons' isn't the point. A glass of saltwater is very conductive and has no 'excess electrons', same with a Cu wire.

'excess electrons' isn't common because of the strength of the coulomb force. However, that is what happens with the static electricity when you shuffle your feet on the carpet or rub a rubber balloon. It's also what happens with lightning, which is a pretty dramatic rebalancing of electric charge.

'excess electrons' makes me think of batteries and such, much more than conductors.

 

[John Kovach]

What is missing is from the discussion is that Earth ground was never intended to have global significance. It is a local common reference for electronic equipment vs. the chassis ground for a specific piece of electronics. It is for safety purposes. Electronic instruments in a lab can have different potentials and represent a shock hazard if they do not share common ground. All your ground fault circuits are connected to a common Earth ground. Your neighbors house might be at a different electrical potential than your own. You have environmental factors that make a global Earth ground impossible: soil moisture, thunder storms, radiation from antennas, cosmic rays...it's a regular smogasborg of things that can influence the potential difference between two Earth sites. Yes the ground plain is important in radio transmissions, but again, it is a local circuit and does not include consideration of the neighbors yard down the street. We have local grounding hooks on the flight line to provide an electrical ground for aircraft (that sit on rubber tires) before they are serviced. Earth ground is a local reference.