Abridgment of arcing grounds and effect of neutral grounding impedance

This paper was written to review and extend the theory of overvoltages due to the arcing grounds because of the increasing tendency to use impedances between the neutral point and the ground, thereby losing the advantage of the solidly grounded neutral. The “third-class conductor” theory of Steinmetz is touched upon very briefly and is considered as not applying to transmission line conditions. The theory when the phenomenon is controlled by normal frequency arc extinction, as presented by Peters and Slepian, is reviewed, and the maximum voltage for this analysis is found to be
$3{1\over 2}$
E, where E is the normal line to neutral voltage. The theory when the phenomenon is controlled by oscillatory frequency arc extinction as originated by Doctor Petersen is given in detail but in a modified and extended form. The maximum voltage for a single-phase circuit when no damping is considered is found to be 6 E. The analysis for the three-phase circuit is newly developed for the case in which there is an impedance between the neutral and ground and the maximum voltage is found to be 7.5 E when the effect of the damping factors and capacitance between lines is neglected. The method of determining the various reductions or damping factors is outlined. The effect of a neutral grounding resistor is discussed and it is pointed out that a surprisingly high value of resistance can be used without incurring the possibility of dangerous overvoltages. It is shown that the use of reactance is more liable to result in overvoltages than resistance but that relatively large values of reactance can be used in conjunction with resistance. The Petersen Coil is usually considered as causing the arc to go out by giving a balance of lagging and leading currents in the arc. It is brought out in this paper that there will be no voltages built up when the Petersen Coil is used, whether or not the arc goes out. The relation of the overvoltages on a non-grounded and an effectively grounded system is outlined, and a criterion for determining whether or not a system is effectively grounded is proposed.