Modelos de análisis para la inversión conjunta en centrales renovables y red de transporte

Resumen

The main purpose of this thesis is to develop procedures to assist a wind power producer and the system operator to decide on the optimal investment in wind power and in transmission facilities, respectively. Chapter 1 provides an introduction to the thesis work. First, an overview of electricity markets is provided. Second, the motivation of the work developed in this dissertation is stated. Third, the main features of the problems addressed in this thesis are described. Fourth, a literature review concerning the works related to this dissertation is carried out. Finally, the chapter concludes with a list of the objectives of this thesis. Chapter 2 provides models to characterize the different sources of uncertainty that affect the decision-making problems addressed in this thesis, including demand, wind power production, balancing market price, future demand growth, and future investment cost. Two different methods are proposed based on the load- and wind-duration curves, and the K-means clustering technique. A case study is provided to illustrate the modeling of these uncertain parameters. Chapter 3 provides a stochastic bilevel model for the wind power investment problem considering a static approach. The proposed model aims at maximizing the expected profit of a wind power investor. This chapter includes the reformulation of the proposed bilevel model as an MPEC that can be recast as a MILP problem. To tackle the computational burden of the resulting MILP problem, an approach based on Benders¿ decomposition is provided. Finally, we numerically justify that the minus expected profit of the wind power investor as a function of the investment decision variables has a convex enough envelope if a large number of operating conditions is considered, and thus, Benders¿ decomposition is effective. Chapter 4 proposes a stochastic bilevel model for the wind power investment problem considering a risk-constrained multi-stage approach. Instead of an investor that simply maximizes its expected profit as in Chapter 3, Chapter 4 considers a wind power investor that aims at maximizing its expected profit while minimizing its profit volatility. To do so, the CVaR metric is used. Uncertainties in demand, wind power production, future demand growth, and future investment cost are considered. A weighting parameter in the objective function is used to materialize different risk-aversion levels that lead to different investment strategies. The bilevel model is recast as a MILP problem that can be efficiently solved using Benders¿ decomposition. Chapter 5 provides a stochastic complementarity model for the optimal wind power investment of a strategic investor with market power. This investor participates and exercises market power in the day-ahead market, while it buys/sells its production deviations in the balancing market, in which it behaves as a deviator. First, a strategic offering for a wind power producer is developed to determine the optimal production level and price to be offered to the day-ahead market. Then, this model is extended to consider investment in wind power facilities. Both the strategic offering and the strategic investment models are recast as MILP problems. Chapter 6 proposes a stochastic bilevel model to determine the optimal transmission lines to be built by the TSO and the optimal wind power projects to be promoted among private profit-oriented wind power investors with the aim of minimizing the overall consumers¿ payments. Subsidies in wind power investment are considered as fixed percentages of the investment cost. The model is recast as a tractable MILP problem. Chapters 3-5 and 6 include clarifying examples based on a three-node system and the Garver¿s six-node system, respectively. Additionally, Chapters 3-6 include two realistic case studies based on the IEEE 24-node Reliability Test System (RTS) and the IEEE 118-node Test System (TS) to illustrate the features and applicability of the proposed models. Chapter 7 concludes this dissertation and provides a summary of the thesis contents, a number of relevant conclusions, the main contributions of the work carried out, and some suggestions for future research. Appendix A provides the wind power production and balancing market price scenario data used in the case studies of Chapters 2 and 5. Appendix B provides an overview of the mathematical tools used in this thesis, including the bilevel model used in Chapters 3-6, two linearization methods used in Chapters 3-6, and the Benders¿ decomposition algorithm used in Chapters 3-4. Appendices C and D provide the data of the IEEE 24-node RTS and the IEEE 118-node TS, respectively, used in the case studies of Chapters 3-6.