Nonlinear dynamic optimization for gas pipelines operation

Natural gas is one of the major energy resources employed in many sectors, and its transport is guaranteed by large-scale pipelines, which need to be properly managed in order to ensure an efficient operation. This paper proposes a Nonlinear Programming (NLP) algorithm to control the operation of gas networks and minimize the compressor energy consumption, as well as model the discretized dynamic gas transport equations in the pipelines, include detailed performance maps of compressors and their gas turbine drivers, and regulate control valves. This work proposes a nonlinear smoothing approach to model disjunctive operating configurations of the gas system, allowing to preserve the accuracy and exploit the computational speed of nonlinear algorithms. The algorithm is effectively applied to a test network featuring a complex branched topology, and the results are compared with a Mixed Integer Linear Programming (MILP) formulation, thus showing a significant reduction of the computational time, and an improvement in terms of accuracy of optimality of the solution.