Parametric Model Order Reduction for Gas Flow Models

Benner P, Grundel S, Himpe C

Abstract

Planning the dispatch of contracted gas denominations requires various simulations of the involved gas transport infrastructure. Furthermore, due to the growing interplay of traditional gas transport and fluctuating demands related to renewable energies, the number of necessary simulations vastly increases. Mathematically, a system of Euler equations, which are coupled according to the underlying gas network topology, embodies the associated nonlinear and hyperbolic model. Repeated simulation of large networks for varying supply and demand scenarios often necessitates model order reduction. Yet, beyond these variable boundary conditions, further attributes of the network may be uncertain or need to be kept variable throughout simulations, which motivates parametric model order reduction (pMOR).