I never considered earth to be particularly conductive. It's just dirt, after all.

However, I've seen "earth ground" conductive stakes driven into the ground in order for electricity to be grounded, because it will find its way down there.

However, it never made sense to me why earth would even provide such an effect: why electricity would bother to flow to dirt out of all the conductive goodness inside the circuit?

What characteristics of earth/electricity makes the current flow right into the ground?

  • 1
    \$\begingroup\$ In modern installations, all current flows through wires. The earth ground provides a voltage reference only. Only in fault conditions will current flow through earth to that stake. GFI breakers may trip when that happens. I think it makes sense to have a 0V reference, and Earth is probably a good choice because a lot of stuff in your house is connected to it anyway. But the conductivity of earth is too variable to use as an intentional part of your circuit. \$\endgroup\$
    – mkeith
    Dec 24, 2014 at 5:55
  • 1
    \$\begingroup\$ See also Single Wire Earth Return for a way of using the Real Earth. Although the regular grid might work with no Real Earth connection at all, it makes sense to reference the power system to the real earth. At some point you will be outside cutting the lawn with an electrical machine, or the power grid will be struck by lightning. \$\endgroup\$
    – tomnexus
    Dec 24, 2014 at 5:59
  • 9
    \$\begingroup\$ Electric current doesn't flow in dirt, but it flows reasonably well in the salty water that makes the dirt moist. \$\endgroup\$
    – markt
    Dec 24, 2014 at 6:49
  • \$\begingroup\$ "why electricity would bother" // Charges are not alive. Stop thinking in terms of "electricity chooses the path of least resistance". \$\endgroup\$
    – alejnavab
    Feb 7 at 22:31
  • \$\begingroup\$ @alejnavab, seems a pointless comment to add, 7 years after three good and detailed answers were posted...? \$\endgroup\$
    – TonyM
    Feb 8 at 8:27

3 Answers 3


In a nutshell

Electricity is not supposed to flow through ground stakes in normal conditions. It doesn't mean its resistance is high, it's actually surprisingly small. That branch of the circuit is simply not closed normally.

In details

A ground is a reference point. You could litterally take any net in your circuit which is supposed to stay at a steady voltage and call it ground. After all voltage sources create a difference of potential (called a voltage) between two nets, regardless of what their potentials are - if they're both fixed externally, there will be a conflict and bad things, but if one of them is fixed the other potential will change accordingly. Generally the ground is taken such that we work with positive supplies predominantly, e.g. ground on the - terminal of a rectifier bridge. It doesn't mean all the current flow through that, it's only a reference.

The Earth has mainly a person protection role. No current is supposed to flow in the Earth because the actual supply circuit is isolated from the Earth, however what if this isolation is compromised (wires eaten by rabbits, children shoving their fingers in sockets...)? Everyone is indirectly connected to Earth (no isolation is perfect), which means that that circuit will now be closed and the only thing that will limit the current going through whatever is closing the circuit (e.g. people) is its internal resistance. Depending on the environment, that resistance can be sufficiently low to kill someone; refer to this thread about what voltages are considered safe. To prevent that, every enclosure is connected to Earth (a Earth-R-Earth circuit has a near-0A current), and the electric supply has a residual current device that compares the current going in and out, and cuts off the supply if there is a leak (through Earth).

The Earth is used for an equi[reference]potential supply The electricity provider needs to protect its people too, so the upstream supply is also referenced to Earth. Just like everywhere else. So what happens if the Earth is not a good conductor and its potential is not homogeneous? Users could be in contact with 2 different Earths, which can be a high difference of potential (=voltage). Thankfully, moist in dirt and water patches are good conductors, but above all the equivalent cross section of this fictive conductor is massive. Except during short upsets such as lightning, it has an excellent homogeneity in potential. Why use another conductor for ground which will use more copper and actually be less effective if we can use what's under our feet?

The Earth is also useful as a protection against lightning: lightning is just like any dielectric/isolator breakdown, it occurs where the resistance between the charged cloud and the Earth is minimal (see this amazing GIF). High trees, towers etc., and we can't risk relying on luck alone so highly conductive spikes are used to attract lightning, and the Earth is used to dissipate that energy. Loosely said. Normally lightning has enough current flowing to create through Earth and across human legs a voltage high enough to kill them, so it is spread out more evenly.

As usual, I'll warmly welcome anyone correcting me if not accurate.

  • \$\begingroup\$ sweeet! Thanks so much for the detailed description! \$\endgroup\$ Dec 25, 2014 at 4:29
  • \$\begingroup\$ What does "Earth-R-Earth" mean? \$\endgroup\$
    – jonathanjo
    Jan 14, 2020 at 14:37
  • \$\begingroup\$ Earth on both ends of a resistor \$\endgroup\$ Jan 27, 2020 at 22:59
  • \$\begingroup\$ The answer doesn't make sense to me. If ground is high resistance, then anything attached to ground will have a changing voltage level (because it's open circuited). Thus ground is not a reference. \$\endgroup\$ Dec 2, 2020 at 21:12
  • \$\begingroup\$ Can you expand on what does not make sense to you exactly - because as a simplistic example, the other end of a pull-up tied to +5V is referenced to 5V, its voltage does not change much... \$\endgroup\$ Dec 8, 2020 at 18:48

Dirt without moisture (water content) is NOT a very good conductor and NOT a good ground. So you are right, dirt is not a very good conductor, by itself, and therefor by itself not a good ground. Dry sand & rock are poor grounds too.

Frozen earth isn't a good ground either which is why it is necessary to use grounding stakes long enough to reach below the "frost line". Ice is not a good conductor. In fact ice is approximately an insulator

In fact PURE liquid water is not a conductor at all. It's the mineral contaminates in water that conduct electricity, or rather the ions in those minerals. Salt water has lots of free moving ions, so salt water is very conductive.

Weirder still, pure dry salt doesn't conduct electricity either.

It's all about FREE MOVING ELECTRONS in a conductor and in a solution FREE MOVING IONS.

In pure water there are no free electrons and no ions.

In pure salt there are ions but Sodium Na+ and Cloride CL- ions can't move, ionically bonded to each other.

The polarity of water molecules enables water to dissolve many ionically bonded substances. So liquid salt water contains free moving Na+ and CL- ions. Any other ionic mineral dissolved in water would probably make a good conductor too, in general.

If you freeze that salty water, two things happen. First ice (solid water molecules) tend to exclude the ions Na+ and CL- from the ice crystals. Second, even if the ions are not excluded they can not move enough anymore to transfer charge. Therefore no current.

So grounding into the earth only works well when the soil is right. Often that is moist soil creating a water solution containing free moving ions.

Read this article: "Achieving an Acceptable Ground in Poor Soil" It talks about how in Hawaii the volcanic soil is a problem for creating a grounding system.


  • \$\begingroup\$ Isn't the frost line issue more about freeze unfreeze cycle moving the spike out? \$\endgroup\$
    – Passerby
    Oct 16, 2021 at 0:33

OK so this is how I view it. There is no such thing as absolute voltage but only a relative voltage. Which means if you have +5 volts somewhere in the system it means it's 5 volts more than the voltage at some other point or point of reference.

But instead of making a random point as reference and measuring voltages with respect to this point we bring in the concept of Earth ground.

The earth ground may be just earth or dirt physically or mostly a point in the circuit system where we consider the point potential to be zero volts and every other potential in the system is measured with respect to this point.

  • 1
    \$\begingroup\$ You may also think of the zero potential in terms of charges. The potential at infinity due to a Columbic charge system is considered to be zero so the potential at any point is the difference in potential between the position concerned and 0 potential at infinity. \$\endgroup\$
    – Timmy
    Dec 24, 2014 at 12:12

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.