Optimizing traditional urban network architectures to increase distributed generation connection

Distribution networks have always been planned for end users and without many concerns about distributed generations. Their function is to supply all the consumers ensuring good continuity of supply and service quality. Following this objective, distribution network operators converge to a small amount of architectures and operation rules depending on the degree of service quality and the initial investment desired. An increasing connection of distributed generations to the distribution network raises technical problems and brings planning rules to be no more optimal. Business as usual solutions currently used are not economic and may finally become insufficient. A deep review of existing planning tools should be required to find efficient solutions to increase the distributed generation interconnection. This paper proposes a heuristic method able to optimize current French architecture features (amount of feeders, choice of consumers connected to them) in order to increase distributed generation penetration. A real urban network has been used to validate this methodology.

► Heuristic algorithm to automatically build the French urban distribution network.
► Nonlinear and contradicting criterion (PL balancing and L minimization).
► Generalization to several HV/MV substations (linear programming).
► Optimal topology features to increase distributed generation penetration.
► Architectures evaluation (costs, reliability, distributed generation penetration).