Gas Distribution Network Optimization by Genetic Algorithm

Natural gas is increasingly being used as a pure energy source. So it's valuable to reduce its total cost to be affordable for individual customer. In the past, interest was focused on efficient pipe network analysis method regardless of network cost; so many researches could be found for methods developed for network analysis. In contrast, very little researches could be developed to optimize the design of distribution networks. This, of course, reflects on software developed for the two purposes: the analysis and simulation of gas networks and the optimization purpose. The aim of this work is to develop a computer code that simulate and optimize gas distribution networks at all pressure ranges, i.e. low, medium and high pressure networks. The aim is to reduce the network diameter sizes to a minimum value while fulfilling the constraints of maximum link velocity and minimum node pressure. In this code, the analysis of gas distribution networks was based on the gradient algorithm which had never presented in gas networks before. In this study, the algorithm was presented and gave efficient analysis for any gas network, i.e. at all pressure ranges, at the least time any algorithm can record. Optimization of gas distribution networks was presented by the genetic algorithm. The code was applied on low pressure gas distribution networks and proved its efficiency and robustness. Therefore, this code was found to be useful at the stage of gas distribution networks design.