Fault location using impedance-based algorithms on non-homogeneous feeders

Impedance-based algorithms like the positive-sequence reactance and Takagi methods are derived assuming a homogenous line conductor. This assumption is violated in practical distribution feeder circuits. The objective of this paper is to demonstrate that these methods can still be effectively applied to non-homogenous feeders by using the line parameters of the most commonly occurring conductor type in the circuit. This approach is first tested on a modified IEEE Test Feeder by simulating two cases, namely, the reference case and the case with a homogenous feeder. For the positive-sequence reactance method the maximum absolute error in both cases is about 8%, while for the Takagi method it is 6.92% and 9.53% in the homogenous case and reference case, respectively. The proposed approach is then demonstrated using ten fault cases on utility distribution feeders. The average absolute error obtained for the positive-sequence reactance and Takagi estimates is 10.23% and 9.34%, respectively, which corresponds to a median error value of 0.16 miles and 0.15 miles in the location estimates.