Clustering Based Optimization and Automation of Utility Scale Solar Site Design

Abstract

EDFR has developed a series of methods for quickly drafting a set of utility scale photovoltaic plant layouts and choosing an optimized plant design from that set. The automated drafting methodology utilizes a standard clustering technique and a novel cluster equalizing post-processing algorithm to solve a problem in existing automated drafting software, which is that existing techniques cannot assign dc power in the form of trackers to inverters without human intervention. This step is crucial to create end to end automation of a PV plant layout. Without it, it is impossible to accurately determine the layout of dc wiring and associated electrical equipment. The work nearly eliminates the need for developer drafting of utility scale photovoltaic plant layouts and provides a foundation for reducing levelized cost of energy by allowing EDFR to select the most financially optimal project design without investing large amounts of time creating the feasibility space under which optimization can occur. (Abstract)