Some contributions to the analysis and automation of practical distribution networks

As the distribution network operating environment evolves and the prospects of customer choice approaches, the importance of efficient operations and reliable service becomes paramount. Distribution automation is a system which remotely monitors, coordinates and controls the distribution network by utilizing control, computer and communication technology; this system can be used by energy suppliers to improve and maintain efficiency and reliability in distribution networks. This thesis strives to provide some analytical foundations to and develop some application functions for distribution automation in practical distribution networks.; A basic function of distribution automation is power flow analysis whose solution yields the distribution system steady-states. In this thesis, analytical results on the radial three-phase power flow solution with detailed system modeling were obtained. Specifically, the existence and uniqueness of the radial three-phase power flow solution were shown. Also, the monotonic properties of voltage magnitudes with respect to changes in load were established.; The analytical results obtained in this thesis were applied to justify physical implementations of distribution automation systems. They also justified previous distribution application problem formulations and led to new simplified problem formulations. Thus, the implications of the analytical results provided a foundation for distribution automation.; With these foundations in place, a new application function, called load capability, was designed to determine how much more load a network can sustain given an arbitrary load variation pattern before incurring an operating constraint violation. A new problem formulation was presented, and a fast, effective solution algorithm based on three-phase power flow and two estimators to the allowable load margin was developed. Simulation results were presented on a real 394 bus network.; While load capability addressed how much load any given network can sustain, service restoration addresses how much load can be serviced during an outage. This thesis also focused on practical service restoration in distribution networks. A comprehensive problem formulation was presented and practical algorithms yielding realizable, effective service restoration plans were developed. The algorithms have been implemented within a graphical user interface tailored towards all distribution personnel: planners, operators and linemen. Simulations results were presented on two large distribution networks.