摘要
现有研究对四川盆地龙马溪组页岩压后各因素影响返排率的机理认识不清,返排率影响产能的规律不明朗,无法通过控制返排率控制产量,提高单井产量难度大。本研究基于地质数据、生产数据和施工数据,选取四川盆地龙马溪组WH区块的泊松比、黏土含量、有机碳含量、含气量、孔隙度、脆性指数、地层压力、层厚作为地质因素,水平段长、压裂水平段长、主压裂用液量、加砂强度、施工排量、支撑剂量作为工程因素,通过建立前馈神经网络模型,分析上述两类因素在预测页岩气井返排率时的影响权重,发现泊松比、孔隙度、黏土含量以及压裂水平段长、加砂强度、施工排量分别为地质和工程方面的主控因素。针对主控因素,应用多元非线性拟合,建立以工程指数为响应值、返排率与地质综合指数关系的预测图版;并进一步建立以综合指数为响应值的返排率与产能关系图版。图版分析表明,要注重综合因素对返排率和产能的影响。通过实际生产数据统计发现,四川盆地页岩气井存在最优返排率,达到最优返排率可以使页岩气井产气量最大化。本研究建立的返排率预测图版可有效预测返排率,并通过可控因素最优化返排率使其位于产量最大区间,进而助力提高页岩气井产量。
The mechanism of the factors affecting the flowback ratio after fracturing in the Longmaxi Formation in the Sichuan Basin is unclear.The mechanisms by which the flowback ratio affects the productive capacity are also unclear and increasing single well production is difficult.For this purpose,based on geological data,production data and construction data,this study selects Poisson's ratio,clay content,total organic carbon content(TOC),gas content,porosity,brittleness index,formation pressure,and layer thickness of the WH block of the Longmaxi Formation in the Sichuan Basin as the geological factors.The length of the horizontal section,the length of the fracturing section,the amount of liquid used for the main fracturing,the strength of the sand reinforcement,the construction displacement,and the supporting dose are selected as the engineering factors.By establishing a proper BP neural network model,the weights of these factors affecting the prediction of shale gas well flowback ratio are analyzed.It is found that Poisson's ratio,porosity,clay content,the length of the fracturing section,the strength of the sand reinforcement,and the construction displacement are the main controlling factors in geology and engineering.For the main controlling factors,a multivariate nonlinear fitting is applied to establish a prediction template with the engineering index as the response value based on the relationship between the flowback ratio and the geological comprehensive index,and a template of the relationship between the flowback ratio and the productive capacity with the comprehensive index as the response value is further established.The analysis shows that it is necessary to pay attention to the impact of comprehensive factors on the flowback ratio and productive capacity.According to the statistics of production data,it is found that the shale gas well in the Sichuan Basin has an optimum flowback ratio,and the optimum flowback ratio can maximize the gas production of the shale gas well.The flowback ratio prediction template established in this study can effectively predict and optimize the flowback ratio,thereby helping to improve the production of shale gas wells.
作者
郭建成
林伯韬
向建华
钟华
GUO Jiancheng;LIN Botao;XIANG Jianhua;ZHONG Hua(College of Petroleum Engineering,China University of Petroleum-Beijing,Beijing 102249,China;Engineering Technology Research Institute of PetroChina Southwest Oil and Gasfield Company,Guanghan 618300,Sichuan,China)
出处
《石油科学通报》
2019年第3期273-287,共15页
Petroleum Science Bulletin
基金
国家科技重大专项“大型油气田及煤层气开发”课题4“页岩气排采工艺技术与应用”(编号：2017ZX05037-004)资助
关键词
四川盆地
返排率
前馈神经网络法
最优返排率
多因素分析
定量预测
Sichuan Basin
flowback ratio
Back Propagation(BP)neural network
the optimum flowback ratio
multi-factor analysis
quantitative forecast
