Concentrating photovoltaic solar power plants using dual-axis trackers are in increasing demand. In a utility-scale photovoltaic system, both capacity factor and ground coverage ratio are widely used to characterize systems in view of the land use efficiency. Current system modeling approaches lack accurate location-specific direct normal irradiance (DNI), miss a reliable electrical model for power optimization and conversion and are inadequate for optimizing the tracker array configuration. In this paper, a comprehensive modeling of a concentrating photovoltaic system is introduced. First, a more accurate estimation of hourly DNI is obtained by considering location-dependent DNI and air mass changes according to the sun's elevation. Second, mismatch effects of modules are factored in. Third, various power optimization and conversion levels are taken into account for optimization with self-shading in each module. The tracker array configuration has been optimized to maximize energy harvest by getting a maximum capacity factor for a given ground coverage ratio. Copyright © 2011 John Wiley & Sons, Ltd.