Best response-based individually look-ahead scheduling for natural gas and power systems

Abstract

The natural gas system and the electric power system are becoming increasingly coupled in consideration of more interdependent components, e.g., natural gas-fired units, in the systems. Energy supplies for these two systems need to be dispatched economically from their own perspectives when these two systems belong to different owners. This paper investigates the best response-based individually look-ahead scheduling for the natural gas system and the electric power system. The optimal scheduling strategies for these two systems are the best responses of the opposite sides, and the best response of each side is achieved by an iterative best-response-search approach. Each best-response-search process is established as a bilevel mixed integer linear programming model, in which the upper level aims to minimize each side’s operational cost constrained by on–off states of generators/lines/gas sources, and the lower level aims to minimize the total energy consumption cost constrained by the operational constraints in consideration of the scheduling of generators/lines/gas sources of two systems in the upper level. To deal with solving difficulty caused by binary variables in the lower level, the original bilevel mixed integer linear programming model is converted into a constrained single-level formulation, which is solved by a decomposition algorithm based on an iterative column-and-constraint generation method. Two cases validate the model and the algorithm, and the costs with the best response-based model and the joint optimization-based model are compared.