| Puguang gas field locates at Puguang town in Sichuan province,it is a typical high H2S content gas field,with H2S content up to 15.16% and CO2 content up to 8.64%.Puguang gas field has 39 wells which are all in 17 gas gathering stations,the production capacity is about 3120×104m3/d.All stations are connected by pipes in the type of tree.The gas and condensed liquid are transported together in the pipe,and are seperated in the station. The new gathering mode is introduced into the gas field in order to meet its own needs. Though the mode is innovative, it will also be influenced by man-made factors when it is designed. Therefore,some kind of optimal theories need to be introduced and used to evaluate the new gathering scheme on the basis of not interrupting the main gathering mode.During the study,a traditional optimization model was first used to find the best layout of tree-type pipelines in the first step,and optimize the diameters of pipes in the second step, during this step,the object function is the total cost of the gathering system,and kinds of constraint conditions were also considered. However,in the process of tradtional optimization design,the pipe layout and diameters were optimized step by step,which is dificult to get the global optimization scheme of pipe layout and diameters.To overcome this problem, An global optimization model is established, in which the minimal investment is taken as the objective function, and the pressure, flow rate and pipe diameters are taken as the constraint conditions. the dual coding combining binary coding with integer coding is adopted in the genetic algorithm to optimize the optimization model. Practical project test results show that the method can effectively get global optimization scheme of pipe layout and diameters,and reduce the network investment.In addition,in light of the high H2S content,the number of valves was also optimized.A software was developed to determine the optimal planning scheme, and for the accuracy and intuition of the scheme,three-dimensional terrain data of the gas field were gained and three-dimensional topographical map was also generated. |