Development and Application of Hydraulic Fracturing Simulation for Analysis of Fracture Interference and Distributed Acoustic Sensing Characterization

Abstract

Low-frequency distributed acoustic sensing (LF-DAS) is one of the promising diagnostic techniques to detect and characterize hydraulic fractures. LF-DAS signals can capture fracture hits and the strain field around the hydraulic fracture and provide continuous monitoring of fracture geometry and production at each stage of the wellbore. However, the interpretation of field LF-DAS data and the relationship between fluid allocation and production can be challenging due to the complexity of the underground conditions. This thesis develops a new workflow for coupling flow and geomechanical computations and simulating fracture propagation in the MATLAB Reservoir Simulation Toolbox (MRST) – an open-source reservoir simulation software. A new discrete fracture model (DFM) is implemented. Unstructured meshing is used to discretize the gridding domain. The matrix-to-matrix, fracture-to-fracture, and fracture-to-matrix fluxes are computed explicitly in the computational domain following the control volume formulation. The flow calculations and the geomechanical computations are solved sequentially: the governing equations for poroelasticity are incorporated, and the fixed stress splitting coupling methodology is employed. The hydraulic fracture is set to propagate along a prescribed path with a specific propagation or activation criterion. The accuracy of our model is also validated against the KGD analytical solutions for the leak-off-viscosity, storage-viscosity and leak-off-toughness dominated regimes. This study presents a comprehensive investigation of production performance and stress evolution in hydraulically fracturing formations using the proposed coupling strategy. This model considers the effects of poroelasticity and stress or strain variation in the fractured domains and factors influencing the behaviour of fracture interference, allowing for a more accurate representation of complex fracture interference behavior. Several case studies and sensitivity analyses demonstrate the approach's utility and examine fracture interference, closure, and stress shadowing effects. The modelling work facilitates interpreting field measurement data by investigating characteristics of fracture hits from adjacent wells. Both the matrix and fractures are discretized, enabling the investigation of how different rock properties impact strain variation. Additionally, this thesis presents several case studies utilizing the proposed model and raw DAS data analysis to explore the effects of cross-well fracture hits and completion designs on fracture propagation and production conductivity. The study aims to provide new and valuable insights into the implications of fracture interference on optimal designs during hydraulic fracturing stimulation, and completion design optimization and interpretation of DAS data, which will inform decision-making, ultimately leading to improved well productivity and efficiency.