MODELLING AND SIMULATION OF ISOTHERMAL UNSTEADY STATE FLOW IN NATURAL GAS PIPELINE NETWORK

Abstract

The consumption of natural gas is increasing day by day and has become the most commonly used primary source of energy all over the world due to its’ ease of use. It is costly and hazardous to store natural gas at high pressure or transport it by a vehicle because of its low density and high volatility. Natural gas deposits location are very far away from the places where it is consumed. Hence, the natural gas is transported from gas deposits to the market area by means of pipeline transport system and supplied to distribution system in which gas flows from gas station to a place where it is in demand such as households, industries etc. through a pipe line network. In general, steady state approach is used to analyze gas pipeline network. However, it is not uncommon that transient flow inside the network sets in due to rapid change in demands. Therefore, transient analysis study in these pipelines are important to find out the departure between simulation based on steady state and that based on transient. This report includes the modelling of given gas pipeline network. The present study attempts to investigate transient analysis phenomena inside a given pipeline network subjected to fluctuating demands with time. Many investigators have worked on this transient analysis problem. All of them have faced some common problems such as convergence problem, large computation time etc. After analysis all these problem, the present work for simulating gas pipeline network has been planned using COMSOL Multiphysics5.3a. Considered pipeline network consists of 9 pipes and 17 nodes modelled and solved on the basis of finite element method. The demand given in every outlet node got doubled after 30 min and then remains constant and accordingly pressure distribution along the gas pipeline network is calculated. The results obtained are compared with the traditional methods available in literature. And also, effect of different friction factor model on given pipeline network are also considered and it is calculated that Churchill friction factor model is best suited model with 8% minimum error comparable with other available models.