Why do I measure 120V between phase/neutral and ground?

Question

In a house initially without earthing/grounding, I've make a new one using 3 copper rods 1.5m long, spaced 6 meters apart and connected by 35mm² bare copper cable. But something is wrong.

A socket tester says "live/grd reverse, missing grd". It is a 3-light test device, and the 3 lights shine.

A multi-meter measures:

• Around 220V between phase (aka live, aka hot) and neutral (we are talking about a mono-phase 230V house in Spain).
• More or less 125V between phase and my ground cable, same between neutral and ground.

If I simply place one of the multi-meter probes in the phase or neutral and the other one touching the floor or metal in contact to soil, the result is the same, around 120V.

In addition, I've made the same test in several houses around my own one. In none of them does the earthing seems ok, same measures than in my own one and socket tester reporting error.

Any suggestion about what could be the problem?

Background Information: In Spain, neutral must be connected to earth at power plant / medium to low voltage conversion plant. Earthing/grounding is done in each house. It is forbidden to connect neutral to ground in the installation of particular houses.

Low voltage distribution is usually done in 3-phase, 400 V / 50 Hz, a single house connecting to one phase and neutral, resulting in 230V. That is (if I remember correctly), each phase in the 3-phase distribution is 230V from neutral, with angular displacements of 0, 120 and 240 degrees. That implies a voltage between any pair of live phases of 230V*sqrt(3)=400V

Note: This house had, a lot of years ago, power supply of 125V (in fact 127V). As two of the answers says, this could be related with the issue. To be noted that 127V*sqrt(3)=220V, the value multi-meter is measuring nowadays between hot and neutral.

Answer 1

I see a few possibilites here.

1. You don't have the supply type you think you have. Just because 230V/400V is the most common system that doesn't mean it's the only system in use. I know there are parts of Europe that use 220/127 three phase systems with loads connected between two phases rather than connected phase to Neutral but I don't know about spain specifically. If the voltage was at the lower-end of the tolerance range than such a system could easilly be consistent with your measurements.
2. There is a fault in the supply system, and the connection from neutral to earth at the transformer is broken.
3. There are voltage gradiants across the ground such that your local ground is at a different potential from the ground at the transformer.

Personally I think 3 is unlikely because of the symmetry of the situation, so that leaves possibilities 1 and 2.

I would say talk to your electricity supplier, but my experience with talking to large organizations as a regular customer has been that it's virtually impossible to get to talk to someone who is technically competent rather than a low level call center worker. It's probablly not a bad idea to talk to them anyway but don't be surprised if you don't get anything useful out of them.

Failing that I would suggest looking for a competent local electrican, ideally someone who has worked in your local area for many years and knows of any weird local practices.

Answer 2

You have a TT system there

In that system, you don't do any bonding between neutral and earth. You bond your house's earthing to earth locally, which you are doing, and that's it.

This is not what is done in North America. There, neutral is forced to earth at every utility connection aka billing electric meter. (well, it's at the first disconnect switch past the meter, but those are supposed to be near each other). Don't confuse the two systems. They have different design goals and rationale.

The most important thing - and this is where the British TN-S and TN-C-S systems fail dismally - is that the dirt near your house needs to be quite close to electrical "earth" potential, for safety. You don't want a situation where you plug in a table saw and the table saw is 120 volts from the actual earth. So if neutral and earth aren't the same, it's more important that earth match the voltage of the actual ground.

It's not a complete failure for neutral to not be near earth. Consider the Philippine system - they switched from North American split-phase to European style 3-phase after they became independent. On the old North American territories, they outlawed use of neutral to get 110V, and deliver 220V center-ground to houses. All circuit breakers are 2-pole. Then you have the British 110V construction site power, which is center-grounded, each pole 55 volts to ground. North American NEC has an entire article (647) on wiring that exact system.

Add to it that I bet your sockets are symmetrical. That is, you can pull out a plug, rotate it 180 degrees and it'll still plug in just fine. North America and Britain can't do that (note even on 2-prong plugs, USA has a slightly taller neutral slot). As such, neutral polarization is not so important, since any plug-in appliance has a 50% chance of one poliarization or the other.

What is concerning is that on a 3-phase supply, if earth is offset 120V from neutral, that means your phase is 120V to earth, but other phases are 280-320V from (your) earth depending on the phase imbalance! However, depending on the voltage gradients across the soil, that difference may not exist.

Regardless, it seems like some sort of electrical fault is occurring from the supplier end, and I would report it to them.

Although - I don't know your situation but the symmetry of it (earth dead nuts halfway between live and neutral) almost sounds a deliberate attempt to center-earth in Philippine style. However if so they really need to tell you that, since you need 2-pole circuit breakers on all circuits.

Answer 3

The symptoms are the same as you would see with a broken earth wire in your house.

Just about all electronic equipment has interference filter capacitors connected from phase to ground and neutral to ground. These will pass enough current for a digital multimeter to show about 120 V on a broken earth wire, but much less than the trip current of an RCD and probably indetectable to the average person wearing socks.

As you have observed the same in neighboring houses, I agree with Peter Green's no. 2 hypothesis.