Fiber Optic DAS and DTS in Multicluster, Multistage Horizontal Well Fracturing: Interpreting Hydraulic Fracture Initiation and Propagation Through Diagnostics

Abstract

Abstract The application of high-resolution fiber optic distributed acoustic sensing (DAS) and distributed temperature sensing (DTS) during multistage hydraulic fracturing has increased in recent years. However, fiber optic has been used for communication networks, surveillance, and monitoring and diagnosing wellbore integrity for quite some time. The technology has also been used to diagnose reservoir and well performance, e.g., production and injection profiling. Optical fibers were placed in a small stainless steel tube, which was clamped to the casing while running in the hole in a horizontal unconventional gas well. The casing was then cemented in place, and fracturing treatments were completed by limited-entry techniques using the "plug & perf" process, resulting in multicluster (each stage), multistage fractures. Perforations were placed on the topside of the casing, oriented to avoid damaging the optical fiber (zero degree phasing, at 5 shots per foot, one foot per cluster). This paper shares images and results from DAS and DTS data that allow for interpretation of fracturing treatments in multicluster, multistage horizontal wells. The images illustrate the dynamic nature of the spatial and temporal fluid distribution pertaining to initiation, propagation and arrest of hydraulic fractures during simultaneous stimulation of multiple clusters. First, fracture initiation appears not only during early time, as expected, but also late time initiation events may develop in previously dormant clusters. Second, propagation of multiple fractures can occur randomly within a cluster array, and these multiple fractures are sometimes in close proximity to one another. Third, dominant clusters are often observed during stimulation. Further, during the treatment cycle, the position of dominant cluster(s), within a given stage, may actually change. We leverage these observations with various other data sources in a modeling exercise to demonstrate the implications of unequal fracturing fluid distributions on propped fracture performance.