Groupe d’études et de recherche en analyse des décisions

MIP in Practice: The Case of the California Energy Market

Guillermo Bautista Gestionnaire de l'analyse de marché, CAISO, États-Unis

The re-structuring of electric power systems in the last years has shift the paradigm of operating the power system based market-driven approaches, by finding the best operating point of a regional system while minimizing overall costs. The operation of a power system, including the commitment and dispatch of generation units, is cast naturally as a mixed-integer programming problem, which has been a challenging to solve mathematical problem due to its non convexities and non linearities. Its formulation includes binary variables used to model the commitment of the units, the disperse configuration of the transmission system, the highly nonlinear expression of power flows and the large size of real-life systems, which accounts for thousand of nodes, transmission lines and generation units. A typical technique to handle this difficult problem was until recently Lagrangian Relaxation. With the recent improvements in the mathematical formulations and computing power, MIP-based approaches are gaining ground to solve real-life problems in the power system area.

The California ISO is the clearing house of the electric energy market for the state of California. Its market clears supply and demand for power while enforcing all operational and market constraints.. It also determines the optimal commitment of units, the management of congestion of its transmission system and the optimal operating point of the resources participating in the market. Its market is modeled as a large and complex MIP problem and uses state of the art optimization MIP-based techniques to find a cost-minimization solution. This seminar will go over the market design in place for the California market and the practicalities of using a MIP-based technique to clear the market.