Economic Nonlinear Model Predictive Control for periodic optimal operation of gas pipeline networks

Abstract We study a Nonlinear Model Predictive Control (NMPC) formulation for optimizing the operational costs of gas pipeline networks. The major operating cost comes from running the compressors, which compensate for the frictional pressure loss as gas flows over long distances. We use an economic NMPC formulation, which directly considers the compressor operating cost as the controller objective. Due to diurnal gas demands, the optimal operation is a cyclic steady state. The controller objective and terminal constraints are suitably defined to ensure asymptotic convergence and closed-loop stability of the cyclic steady state. It is shown through simulations that the performance of the economic NMPC formulation is better than a tracking NMPC. The inherent robustness of the formulation also ensures convergence to a region around the cyclic steady state when demand forecasts are inaccurate. The large scale NLP is also solved within a reasonable CPU time making it practical for online application.