140522-1026 EDT
I did not say too much in my original post because I wanted to see the responses.
Yesterday I decided to make some random magnetic field measurements around the house and yard. This was with a magnetic sensor I bought about 50 years ago and it is simply a many turn coil inside of a non-magnetic, but metalic enclosure, possibly brass. The physical size of the electrostatic shielding tube is 1.125" x 2". The center of the coil is about 3/4" from the non-cord end. DC resistance is 570 ohms. I don't know the wire size and number of turns in this sensor but there will be more than 1500 turns. This is based on a coil of somewhat similar size.
I don't have a current transformer large enough to fit over the conduit, but readings from such a probe should be close to 0 current. At the moment I won't describe why I can determine that there is current flow in the conduit, but a clamp-on current transformer would read near 0.
Answers to some of the questions.
In the conduit there are three 0000 copper wires, two hots and one neutral. The conduit is 2". In the meter base the neutral is bonded to the enclosure. In the main panel the neutral is bonded to the enclosure. Thus, the conduit and the neutral wire between the meter and the main panel are parallel current paths.
Grounding of the system occurrs only at few points. My water supply line (better than any ground rod), no ground rod (was not required), a power company ground rod at the pole where my service goes into the ground (this pole originally supported the transformer), a power company ground rod at the pole for the new location of the transformer, and any grounds in my neighbor's home. Our two homes and two street lights are the only loads on the transformer. It is a 50 kVA unit now. There is no power company primary side neutral or ground other than what is at the substation.
With a coil type or other magnetic field sensor one can only get relative estimates of current in an adjacent conductor. In specific cases this can be calibrated in amperes, but just placing the magnetic sensor near a current conductor does not provide a measurement in amperes. In a coil type sensor the coil output voltage for a given magnetic field intensity is proportional to frequency. In a Hall type sensor the output voltage is frequency independent until higher frequencies.
Comparing my old commerical magnetic sensor, and said 1500 turn coil, there was about 10 times the voltage output. This might imply 15000 turns in the commericial unit. Probably not that many. Geometry can be a factor.
My main panel is the service entrance. A few individual circuits are fed from it. But primarily it feds 5 subpanels in different areas. These are basement, pantry, laundry-sewing room, seond floor, and garage.
Some numerical measurements.
0.25 mV --- water line at floor entry point ( 1.25" copper tubing about 150 ft to street).
0.00 mv --- gas line (it is copper to the street),
2.00 mV --- neutral in main panel.
2.5 to 0.7 mV --- next to conduit, meter to main panel.
0.4 to 0.2 mV --- next to conduit, meter to soil.
0.15 mV --- next to wood cover over GEC at pole where transformer was originally located.
0.19 mv --- next to wood cover over GEC at present pole transformer location.
0.3 to 0.1 mV --- general background level in yard and under delta primary lines.
These measurements are meaningful in a relative sense, but not as absolute current measurements.
This thread is partly related to a thread where the originator placed a clamp-on current probe around a conduit and was getting a substantial current reading, and was initially assuming this was an indication of current flow in the conduit. It could possibly be, but most likely was a result of parallel neutral paths somewhere that diverted current from the neutral in said conduit. The magnetic field type of measurement I made here would not have clearly indicated whether or not there was current flow on the conduit. A voltage drop measurement along the conduit might be a better test.
.
