A New Strategy for the Simulation of Gas Pipeline Network Based on System Topology Identification

Abstract

Abstract Pipeline network plays a very important role in chemical plant, refinery, oil and gas gathering. Design, optimization and scheduling of pipeline network are based on the simulation of pipeline network. With the increase of the number of pipe segments, pipeline network model becomes more difficult to be solved. Traditionally, nonlinear equations of pipeline network need to be calculated by equation oriented method. However, this method has large computational complexity and it highly depends on initial values. When there are more unknown variables, it is harder to assign appropriate initial values and also harder to get convergence. In this work, a new strategy for the simulation of gas pipeline network is introduced. A pipeline network model suitable for compressible gas is built. Pipeline network is identified and simplified by adjacent matrix searching. After adjacent matrix searching, the number of equations is reduced and pipeline network model becomes easier to solve. This solution is applied to a case study for verifying the proposed model and strategy.