This work deals with the known problem of forecasting overall capacity and viability of a complete distribution feeder as the medium for Power Line Communication (PLC). The estimation involves previewing the input impedances and power loss for each coupling arrangement on every position pair at the feeder, calculated for frequencies between 9 kHz to 10 MHz. When we deal with PLC in primary distribution feeders, the time changes on the stochastic parameters of the channel model are less important than in secondary distribution, making structural modeling a good choice. The presented algorithm is based on the existing theory of multiconductor transmission lines (MTL) and nodal analysis. Models from literature for distribution transformers, high frequency ground losses, noise on overhead feeders and for coupling units are considered. All the cited models and theories are integrated in an algorithm that is capable of predicting the behavior of the PLC signals in a complete feeder. The parameters of the lines, connecting topologies, wires and loads are read from tables generated by the actual data base systems currently adopted by energy suppliers. The results can be applied to predict the optimized allocation of the power spectrum with the minimum necessary power use, the possible achieved capacity and consequent technical viability. Also, to reduce the cost of installed network, the number and the cost of transceivers and repeaters and no less important, preventing possible interferences with other systems. Finally, we present the results for a real overhead feeder.