Two-stage simulation-based optimization for optimal development of wind farms considering wind uncertainty
Description
TitleTwo-stage simulation-based optimization for optimal development of wind farms considering wind uncertainty
Date Created2020
Other Date2020-05 (degree)
Extent1 online resource (xiii, 106 pages)
DescriptionAs one of the most promising renewable energy sources, wind power provides clean and carbon free energy and becomes more economically viable with significant environmental benefits. To feed the rapidly expanding energy market and to provide alternatives for the consumption of traditional non-renewable energy sources such as fossil fuel, wind farms have been developed to meet the ever-expanding growth of energy consumption. In the meantime, wind farm development and turbine manufacturing technology still need to address the challenges of high installation and operation costs, production stability, electricity capacity and economic efficiency. To enable economic feasibility of large-scale wind energy generation, optimal development of wind farms appears to be crucial for viable wind energy systems. This research presents stochastic models and optimization methods for optimal development of wind farms in different applications.
In this research, wind uncertainty is quantified using probabilistic models for stochastic wind speeds and directions. The two-stage optimization framework is developed to sequentially determine the optimal number of turbines and their most-productive placement. In the first stage of optimization, possible turbine locations are predefined at a number of candidate locations. A binary variable is associated with each location to determine whether a turbine is installed there. The first stage of global search optimization finds the optimal number of turbines needed and their corresponding locations. In the second stage of optimization, the solution from the previous stage is relaxed into a continuous solution space. The local search algorithm is then applied to further improve the locations of turbines identified by the optimization solution from the previous stage. With the two-stage optimization framework, the optimization procedure can determine the optimal number of turbines and refine the turbine placement for the most-productive layout design.
To minimize the objective function - Cost per Expected Power Production (CEPP), five different applications of wind farm models have been studied, in terms of geometric shape of wind farm, site selection and energy sources collaborations. First, the common rectangular wind farm model is studied with pre-defined cells, where the center of each cell represents a candidate turbine location. Next, more-realistic arbitrary-shaped wind farms are considered with engineering constraints, which fits flexibly in various surface conditions, applied to both onshore and offshore wind farm cases. Additionally, a wind farm model in Energy Storage Integrated Wind Energy System (ESIWES) is designed within a micro-grid. With energy storage functioning as backup supply, the wind farm generates electricity in order to meet the demand of the micro-grid community, and at the same time, maintain the minimal CEPP cost by leveraging storage of excess wind energy. The fifth application expands the renewable energy system to include biorefinery – the waste-to-energy recovery pipeline. With biorefinery, the Hybrid Wind, Biorefinery, Energy-storage based Renewable Energy System (HWBRES) is developed to generate more sustainable energy, and at the same time, tackle environmental risk problems caused by waste.
Last but not the least, an advanced scheduling and maintenance model is developed on top of the HWBRES system, in which the turbine operation scheduling and periodic inspection are both taken into consideration to save energy from excess production while maintaining the reliability of each turbine. Additionally, opportunistic maintenance is scheduled occasionally for the cluster of switched-off turbines, by implementing this model it ensures the reliable energy production with limited maintenance costs.
NotePh.D.
NoteIncludes bibliographical references
Genretheses, ETD doctoral
LanguageEnglish
CollectionSchool of Graduate Studies Electronic Theses and Dissertations
Organization NameRutgers, The State University of New Jersey
RightsThe author owns the copyright to this work.
Statistical Profile