Bidding strategy and empirical analysis of bidding in electrical power market

Electric restructuring was started in the early 1990s as a way to increase electric power industry' efficiency and lower the energy cost. The traditional integrated system has now been separated in many parts of the country and some degree of competition introduced throughout the power industry. This thesis focused on how market participants (primarily generators) react under this new market operation mechanism. Specifically, this work contributed with the following three investigations: (1) Transmission system congestion influence on market clearing price and market participant bidding behavior in the framework of game theory was analyzed. The conclusion was drawn that deviation from idealized price-taker behavior is more serious when some market participants suffer disproportionately from the congestion. Due to the complexity of the calculations in the theoretical approach, this thesis suggests that a statistical analysis methodology is more appropriate. An intuitive probabilistic bidding methodology was proposed for the bidding problem to demonstrate feasibility. (2) A detailed statistical analysis has been carried out on the California real time imbalance energy market. A linear regression model was applied to a zonal energy price prediction process and a non-linear estimator based on a neural network was applied to predict bidding behavior. Sensitivity analysis was applied to understanding each factor's influence on market participant bidding behavior. (3) Statistical analysis results were applied to the optimal bidding strategy problem. The empirical conjecture approach was adopted using these results. Including risk as either an objective to be minimized or a constraint to be satisfied, a portfolio selection approach was applied. This method combines the statistical analysis technique with the optimal bidding problem. Although the results shown here are in the initial stage of development, it appears that this approach is more promising than an idealized game theoretic formulation.