页岩气藏水力压裂后大量压裂液滞留在基岩孔隙和压裂诱导裂缝内。目前页岩气藏压裂液微观赋存机理认识不清，导致难以准确认识返排过程中页岩气井压裂液返排率差异。建立页岩气藏多尺度孔缝介质压裂液微观赋存分析方法，揭示页岩气藏压裂液微观赋存机理。建立同时考虑岩石-流体作用力和气水界面毛管力的单孔隙气水赋存分析方法，进一步拓展至孔隙网络，采用侵入逾渗研究基岩压裂液赋存模式。基于水力压裂诱导裂缝CT扫描成像结果，采用水平集气、水界面追踪方法计算不同返排压力下气水分布，研究诱导裂缝压裂液赋存模式。研究结果表明：基岩系统返排率呈现缓慢上升—快速上升—趋于平缓的趋势。压裂液主要呈现水相饱和孔隙赋存、角落水相赋存、水膜状赋存3种赋存模式；压裂诱导裂缝系统返排率主要受诱导裂缝周围孔隙连通性影响，呈现快速上升后趋于平缓趋势。滞留压裂液主要赋存在水力压裂诱导裂缝周围盲端基质孔隙。

After hydraulic fracturing in shale gas reservoir, a significant volume of fracturing fluid retains in the matrix pores and induced fracture network. Currently, the pore scale fracturing fluid occurrence mechanisms are unclear. As a result, it is difficult to accurately understand the difference of fracturing fluid backflow rate of shale gas wells in the backflow process. In this work, the pore scale fracturing fluid occurrence mechanisms analysis method in shale multi-scale matrix-fracture system is developed and the fracturing fluid occurrence mechanisms in shale gas reservoir are elucidated in detail. To understand fracturing fluid occurrence pattern in shale matrix, singe pore gas-water occurrence method is established considering rock-fluid interaction and gas-water capillary pressure and is further extended into the pore network. Invasion percolation is applied to analyze the fracturing fluid occurrence pattern variation during different flowback stages. To understand fracturing fluid occurrence pattern in induced fracture network, the level-set gas-water interface tracking method is applied to calculate gas-water distribution at different flowback pressure based on induced fracture network CT imaging and the fracturing fluid occurrence pattern variation at different flowback stage is studied. Study results reveal that the fracturing fluid flowback rate in shale matrix first increases slowly and then increases fast. In the final stage, the fracturing fluid flowback rate in shale matrix reaches plateau. The fracturing fluid in shale matrix distributes in the forms of water saturated pores, corner water and water film. The fracturing fluid flowback rate in induced fracture network is influenced by pore connectivity around the induced fracture network. The fracturing fluid flowback rate first increases fast and then reaches plateau. The retained fracturing fluid distributes in the dead-end matrix pores around induced fracture network at the final stage.