基于代理辅助分层粒子群算法的页岩气藏压裂参数优化

Optimization of fracturing parameters for shale gas reservoir based on a surrogate-assisted hierarchical particle swarm optimization algorithm

针对现有页岩气藏压裂参数优化研究中单因素分析、非整体优化方法难以准确考虑井间干扰与缝间干扰、压裂参数之间关联性及优化时间花费过长等问题,提出基于代理模型辅助分层粒子群算法的井工厂模式整体优化方法。建立考虑多重介质、水平井摩阻、吸附解吸、非达西流动的页岩气藏流动数学模型,同时考虑井位置、井间距、裂缝条数、裂缝半长等参数对最终产能和经济效益的影响,基于拉丁超立方抽样生成初始种群,使用代理辅助分层粒子群算法以净现值为目标函数对压裂参数进行优化设计。结果表明,相比传统单因素分析,从整体角度优化压裂参数更合理有效,代理模型的辅助可以极大提高运行效率。

In terms of the existing single factor analysis and non-global optimization methods used in the optimization of fracturing parameters for shale gas reservoirs, it is difficult to accurately consider the interference between wells and fractures, the correlation between fracture parameters and the optimization time cost. A method based on surrogate model assisting particle swarm optimization was proposed to optimize the overall parameters from well to fractures in multi-well pads in shale gas reservoirs. In the model, multi-porosity, horizontal well friction, adsorption and desorption, and non-Darcy flow were considered. At the same time, parameters such as well position, well spacing, number of fractures and half-length of fractures were taken into account for overall gas production as well as the economic benefits. The initial population was generated based on Latin hypercube sampling, and the surrogate-assisted hierarchical particle swarm optimization algorithm was used to optimize the parameters with the net present value as the objective function. The results show that, compared with the traditional analysis, it is more reasonable and effective to optimize the fracturing parameters based on multi-well patterns, and the use of the surrogate model can greatly improve the optimization efficiency.