A Lattice-Based Predictive Model for Interaction Mode of Hydraulic Fracture with Natural Fractures

Hydraulic fracturing (HF) is the prime technology for enhancing production from unconventional reservoirs. Ideally, the hydraulic fracture is expected to communicate the existing natural fracture networks, known as sweet spots in tight formations, to create a continuous path for the flow to the lateral section of the wellbore. However, the interaction of the hydraulic fracture with the natural fracture is complex and may result in different mechanisms (modes) known as crossing, arresting, and opening. In this study, we use the log–log plot of the bottom hole pressure versus time after fracture breakdown and during the fracture propagation stage to investigate the interaction modes. A qualitative indicator interaction pressurization rate index (IPRI) was proposed to predict interaction modes between hydraulic fracture and natural fracture. XSite, a lattice-based simulator, is used to model the interaction mechanism. A number of simulations were performed using the Bakken Shale data in North Dakota with a wide range of input parameters to study the impact of the input parameters on the interaction mode. The results formed a database used to develop a predictive model to determine the interaction mode. Several lab experiments are performed to validate the accuracy of the predictive model. Highlights A Lattice-based fracturing simulator was executed to simulate a large number of cases based on field data to build a database. A qualitative indicator Interaction pressurization Rate Index extracted from the pressure-time curve was proposed to analyze the hydraulic fracture propagation phase. A Support Vector Machine model was trained to predict interaction modes from the qualitative indicator. The trained model exhibited good performance in predicting the interaction modes of lab-scale experiments.