Inference of Induced Fracture Geometries Using Fiber-Optic Distributed Strain Sensing in Hydraulic Fracture Test Site 2
Stephen Bourne, K. Hindriks, Alexei A. Savitski, Gustavo Ugueto, Magdalena Wojtaszek
Abstract
Fiber-optic Distributed Strain Sensor observations in a vertical well allow rapid inference of the time evolution of nearby hydraulic fracture heights and their vertical aperture distributions during stimulation and depressurization. We describe a computationally efficient deconvolutional model based on an elastostatic kernel for fracture opening within an isotropic, homogeneous, linear elastic medium. Furthermore, an application of this model to observations from the Hydraulic Fracture Test Site 2 reveals that ultimate fracture heights are typically greater than those indicated by microseismic monitoring, and some hydraulic fractures continue their upward growth for several hours after high-pressure injections ceased.