Coordination of Distributed Energy Resources using Distributed Optimal Power Flow

Electrical power systems are deemed as one of the most complex systems mankind has built to date. Nonetheless, continuous changes on electricity generation and consumption are increasing the complexity of these systems. The uptake of behind-the-meter distributed energy resources (DER) (such as rooftop PV, battery systems, and flexible loads) requires a more precise coordination of all agents in the grid. This is particularly true for distribution systems, which are traditionally passive systems but require coordination under a high percentage of DER penetration. To perform such coordination, the centralized power system operation paradigm becomes incapable of handling the necessary information flow and computation burden required. Distributed coordination methods offer an alternative to this paradigm. In particular, distributed AC optimal power flow (DOPF) solved using a prosumer-based decomposition emerge as an effective tool for three main reasons: (i) they explicitly consider network constraints in their formulation, (ii) they permit a prosumer-based decomposition, which retains prosumer privacy and prerogative, and (iii) they computationally scalable. Distributed methods are further bolstered by the advances on technology systems towards an IoT, edge-computing scenario. The miniaturization of sensors and computing devices, and reliable and efficient telecommunication technologies, allow for deployment of coordination methods with progressively smaller investments. The present dissertation expands the frontiers of DER coordination, offering advances on both implementation aspects by presenting practical considerations on actual distributed, edge-computing hardware, as well as developing novel formulations for more advanced problems which assume realistic design considerations, a key to implementing DER coordination in real-world settings