It is difficult to accurately obtain the fracture height expansion information during the fractu-ring process,and the use of distributed optical fiber can accurately evaluate the fracture expansion infor-mation.Based on the finite element coupled cohesive force unit method,construct a forward model of fi-ber-optic strain induced in vertical adjacent wells due to the propagation of hydraulic fracturing fractures in horizontal wells,carry out numerical simulation of fiber-optic strain in vertical adjacent wells,and an-alyze the characteristics of fiber-optic strain evolution.Based on the fracturing construction parameters and the position of vertical adjacent well optical fiber deployment,it can be determined whether hydrau-lic fractures are within the effective monitoring range of the optical fiber.Utilizing large-scale true tri-axial fracturing experiments and distributed optical fiber sensing equipment to conduct real-time monito-ring of hydraulic fracturing physical experiments in horizontal wells using optical fibers.The results show that the fiber-optic strain evolution induced by the fracture height expansion is divided into four stages:strain-enhancing,tensile-strain-expanding,strain-linear-converging and strain-weakening,and the fiber-optic strain evolution is characterized by the appearance of tensile strain convergence bands in the middle part and compressive strain convergence bands on both sides.When the fracture is within the effective monitoring range of the deployed fiber optic,the vertical adjacent well fiber optic can effectively monitor the height expansion of fractures.The evolution results of fiber-optic strain in vertical adjacent wells and the verification of the correctness of the forward model by fiber-optic strain forward modeling calculation results.The characteristics of the strain evolution induced by fracture height expansion ob-served by an adjacent optical fiber can help to evaluate the status of fracture height expansion,providing reference and guidance for oilfield fracturing design.
维普
万方数据