| description abstract | Star-access-ring gathering pipeline network optimization (SARGPNO) is usually regarded as a kind of large-scale combination optimization problem. The structure is widely applied to square, circular, or elliptical oil and gas fields with many wells, stations, and large areas. Accordingly, it is of great practical significance to carry out research on the optimization of such a gathering and transportation network because the reasonable pipeline network structure can significantly reduce the pipeline network investment and improve the system operation reliability. This paper establishes a mixed-integer nonlinear programming (MINLP) model with the minimum total cost as the objective function. A step-by-step optimization strategy is proposed to classify SARGPNO into two subproblems: well group division optimization (WGDO) and ring structure optimization (RSO). These two subproblems are solved by hybrid genetic algorithm combined with local optimization of the mountain-climbing algorithm. In order to investigate the influence of algorithm parameters on the optimization results, such as crossover probability, mutation probability, the number of evolutionary generations, and initial population size, a real gas field was employed and analyzed. The results of pipeline network layout under different gathering radii and different well constraints were determined. Besides, considering the convenience of the construction site, the optimization design under discrete space was conducted. Then the optimized layout in discrete space was compared with the layout designed in continuous space. Furthermore, the optimization results of the pipeline network considering the obstacle constraints were also analyzed by using the gas field data from the literature. Finally, two examples were adopted to verify the validity and feasibility of the model and algorithm. | |
