Numerically efficient transient-state model for a pipeline based on the Fourier-Crump inverse Laplace transform

Abstract

The transient modelling of pipeline systems is typically performed by discrete nonlinear partial differential equation (PDE) solvers. Despite the sophistication of these methods, they are computationally expensive, and can suffer from stability and accuracy issues. An alternative approach to modelling pipeline systems is to use an approximate linear Laplace-domain model of the system coupled with an inverse transform. This approach retains the distributed characteristics of the original system and avoids the problems associated with the discrete methods. This paper explores the use of the Laplace-domain model combined with the Fourier-Crump inverse Laplace transform as an alternative time-domain simulator. A pipeline comprised of turbulent unsteady friction and viscoelastic pipe elements is studied and it is seen that the proposed method is faster than the standard discrete MOC for the cases considered, and it provides an accurate approximation to the original nonlinear dynamics.