Resumen
The deployment of new generation and storage technologies in modern power systems and their increase in interconnection capacities entails a paradigm change in operational trends. Accordingly, energy models require detailing the technical aspects of all the generation facilities placed in multi-area electricity markets with an hourly granularity to support decision-making processes.However, the inherent computational burden when addressing these concerns in the medium-term unit commitment problem at once has frequently forced the implementation of modeling simplifications to achieve affordable run times with reasonable computational resources. On the one hand, time series aggregation is a common practice to reduce complexity when dealing with medium- and long-term horizons but generally means a diversity loss in the data representation (intra-period details or extreme values) and the possible fracture of the chronological relationships. On the other hand, horizon-splitting methodologies also enable a complexity reduction while keeping chronological relationships inside each moving window representation. Nonetheless, they can not correctly manage medium-term decisions like hydro storage, fuel contracts, or third-party access. This paper proposes a flexible two-step optimization model that combines both approaches to overcome their limitations when representing real-size multi-area power systems in the medium term on an hourly basis. The performance and results of this model are compared to those of a whole horizon hourly case study to demonstrate its success.
A unit commitment model to represent real-size multi-area power systems in the medium term on an hourly basis: Energy storage, fuel contracts, and third-party access management
