In this paper we are dealing with the electrical distribution network planning problem where a network configuration has to be specified in order to meet demand, to satisfy the operating constraints, and to minimize investment, operating, and power-loss costs. The iterative procedure includes several heuristic algorithms to generate a radial network, to choose a set of open feeders to meet the reliability constraint, and to solve the reconfiguration problem in order to reduce the power-loss costs. Furthermore, in order to compare the reliability of potential solutions, a predictive reliability assessment measure is established.