Resumen de Robust generation scheduling in electricity markets
Generation scheduling is facing unprecedented levels of uncertainty due to the growing presence of intermittent and nondispatchable renewable-based generation, in particular of wind energy. Under such increased uncertainty, power system operation is seriously challenged and new effective generation scheduling methodologies are required. In this regard, the aim of this Ph.D. thesis is to provide practical generation scheduling models based on robust optimization to help system operators handle the unprecedented levels of uncertainty in current electricity markets. More specifically, this Ph.D. thesis addresses the following yet unresolved issues in robust generation scheduling under uncertainty: (i) the precise modeling of reserve offers in day-ahead co-optimized energy and reserve electricity markets under wind uncertainty and the loss of system components, (ii) the nonanticipativity of generation dispatch decisions, which is needed to be enforced in order to ensure full immunization against all possible uncertainty realizations within the prescribed uncertainty set, and (iii) the incorporation of flexible resources, namely storage devices and fast-acting generating units, in day-ahead co-optimized electricity markets. The resulting robust counterparts are formulated as instances of mixed-integer trilevel programming that are effectively solved by means of decomposition-based methods. Several case studies are examined to validate the effectiveness of the newly developed approaches and their operational advantages.
