The Kaufmann Experiment
This experiment, which was conducted in the early 1950s (Kaufmann, 1954), illustrates the importance of routing the ground conductor with the power conductors. The experiment setup (Figure 17) illustrates the typical arrangement that can be found on any job site. A current source was connected to the phase conductor and to pairs of possible ground return paths, which included a #4/0 conductor run with the phase conductor, and a #4/0 wire run 1 foot (0.3m) away from the rigid conduit and structural building steel. In a comparison of the relative impedances of 100 feet (30 m) of the rigid steel conduit vs. the insulated #4/0 ground conductors routed external to the conduit, 95% of the fault current flowed on the conduit, and only 10% flowed on the equipment grounding conductor routed outside the conduit. The impedance of the conduit was nine times less than the impedance of the grounding conductor routed external to it.
However, when the #4/0 equipment grounding conductor was routed with the phase conductor inside of the conduit, 80% of the fault current flowed in the equipment grounding wire, and only 20% flowed in the conduit. This experiment proves conclusively that the fault current that will flow through a ground conductor will be much higher when it is routed with the phase conductor. In the event of a fault, the circuit-protection device will be tripped much more quickly because of the high fault current, which minimizes the duration of the hazard. The results of the Kaufmann experiment are a primary reason why electrical codes require grounding conductors to be run with phase conductors.
When the building steel was compared to the rigid conduit, 95% of the fault current flowed on the conduit; only 5% flowed on the building steel.