AN HEURISTIC APPROACH TO RENEWABLE ENERGY OPTIMIZATION

An analytical approach is often taken to predict the performance of renewable energy systems at a site, but an analytic approach requires detailed information on the system to be modeled that is better determined during schematic design than guessed-at during pre-design. This paper describes a heuristic approach to identify and prioritize renewable energy project opportunities before detailed system information is available. The method determines the combination of renewable energy technologies that minimize life-cycle cost at a facility, often with a specified goal regarding percent of energy use from renewable sources. Technologies include: photovoltaics (PV); wind; solar thermal heat and electric; solar ventilation air preheating; solar water heating; biomass heat and electric (combustion, gasification, pyrolysis, anaerobic digestion); and daylighting. The method rests upon the National Renewable Energy Laboratory's (NREL) capabilities in: characterizing of the empirical cost and performance of technologies; geographic information systems (GIS) resource assessment; and life-cycle cost analysis. For each technology, simple heuristic algorithms relate renewable energy resources at a site to annual energy delivery with coefficients that are determined empirically. Initial cost and operation and maintenance (O&M) cost also use empirical data. Economic performance is then calculated with a site's utility rates and incentives. The paper discusses how to account for the way candidate technologies interact with each other, and the solver routine used to determine the combination that minimizes life-cycle cost. Results include optimal sizes of each technology, initial cost, operating cost, and life-cycle cost, including incentives from utilities or governments. Results inform early planning to identify and prioritize projects at a site for subsequent engineering and economic feasibility study. Case studies include industrial sites, military bases, and civic buildings.