This work describes the self-healing systems and their benefits in the power distribution networks, with the objective to evaluate the necessary level of automation. The methodology is based on graph theory and genetic algorithms. There are benefits for consumers, that will profit from a more reliable and stable system, and for the utility, that can reduce costs with team field and financial compensations paid to consumers in case of violation of continuity indexes. A real distribution network from the state of Sao Paulo will be used as a case study for the application of the methodology.