DESIGN AND OPTIMIZATION OF NATURAL GAS PIPELINE NETWORK

Abstract

As the world is looking towards cleaner and more environment friendly energy sources for meeting its energy demands and turning to non-conventional sources like solar, wind and many more; natural gas still remains as one of the most preferred conventional alternative because of the easy and extensive availability of its reserves and technologies for its distribution. Also, it is comparatively greener than its other conventional counterparts. For proper transport of natural gas, an elaborate transportation network is required. This transportation network consists of complex networks of pipelines. More and more pipeline networks are being built to facilitate the transport of natural gas to consumers to meet their rising demands. In any gas transportation system, major contributors to its fixed cost are the cost of pipeline network which includes the material cost, manufacturing cost and laying cost and the cost of compressor stations which includes compressor costs and installation costs. The focus of this study is on given natural gas pipeline networks which includes the work on modelling, simulation, costing, optimization and design of the gas pipeline networks based on steady state. In this study, firstly total cost of a given pipeline network is optimized using GAMS/CONOPT solver. The network is optimized such that the demands of natural gas at each node are fulfilled and pressures greater than minimum requirement of delivery pressure is achieved. The optimum values of design parameters (operating pressures at nodes, diameter of each pipe and thickness of each pipe) which gives the minimum cost of the given pipeline network are reported. The given network is optimized at different flow regimes (namely fully turbulent and partially turbulent). Different sub-models are generated to analyze the different flow regimes. These sub-models are generated as a result of changing friction factor equations within the main model for optimization. This was done to check the adaptability of the optimization model generated. It was concluded that the model generated can solve different types of flows. In the next section, a different network was generated for the same set of nodes and optimization of the new network was done. The cost of the new network was found to be lesser than that taken initially. The next section of this study analyzes the effect of total pipeline length on the pipeline costs. It was hence concluded that a layout which is of the shortest possible length may not always be the best option. This is because of the fact that the investment cost depends on many variables which are inter-dependent. The relations among these variables are nonlinear.