A while back I came home and some lighting circuits were dim. I checked the voltage at the main box before the main breaker and one power leg to neutral was about 150 volts but the other power leg to neutral was about 90 volts. The local utility sent a lineman out and after finding no bad connections, he concluded (rightly) that the neutral wire from the pole was cut. My question is: why didn't both read the same since the neutral was connected to the house ground? Thanks.

The resistance of the soil around the ground rod is very high in comparison to the other condutors. This means that you can pretty much ignore the ground in the calculation.

The voltage in either leg is simply a result of the load on it. If you have no load on one leg, you'll have 240V to neutral on the other. If the load on both legs is equal the voltage vill be 120V on both legs.

The question is: Does the ground rod keep the grounded conductor at ground potential, or does the ground rod keep the ground at the grounded conductor's potential? Contrary to conventional wisdom, I hold the latter to be true.

This is the rationale behind NEC250-2(d): The earth shall not be used as the sole equipment grounding conductor or fault current path. For low-voltage systems, earth return is a very poor way to limit potential difference and facilitate rapid operation of overcurrent devices

Does this mean that when we ground devices, tools, machines, etc. that maybe things are not nearly as safe as we assume because of the relatively high resistance of the soil around the grounding rod?

ds — As I read the Code, it means just the opposite—never expect an earth return to substitute for metallic bonding.

DS,
We have a discussion going in the non-U.S. area at the moment which includes TT vs. TN-C-S grounding systems.

Many homes in Britain and Europe have exactly the situation you describe: The buildings sole ground connection is a rod, with no bonding to the neutral. The overall impedance of the circuit in such an installation means that even a direct phase-to-ground short doesn't result in enough current to open a normal OCPD, so everything has to be GFI-protected.

Grounding just to a rod without such protection is practically ineffective, as a ground fault will just result in a couple of amps (or less) flowing back to the supply through the earth and the frame of the faulty appliance rising to supply potential.

Such installations are banned in America by the code section BJ quoted above.

Someone cut the noodle at the Transformer???
Funny how far some jokers will go to be A@SES!!!

Nevertheless, everyone has given excellent descriptions of the odd voltage readings.
Whatever the loads (the connected equipment) are, resulted in an array of series connected stuff (unless loads are meant to be L-L / 240 VAC connected). The loads with higher Impedance resulted in a higher voltage across them.
Eventually everything worked out to be 150 VAC and 90 VAC to the center point.

If the PoCo's mid point ground bond was still connected at the Transformer, that would result in a slight trickle of current finding its way back to the Transformer via the Earth ground. This would probably raise the L-G voltage even higher than 150 VAC if there were less or lighter loads connected - or if voltage was checked with a Hi input Z meter, but who knows?!?!?!

Open Noodle situations really suck!

Scott35