Simulating gas production from hydraulic fracture networks : a case study of the Barnett shale

Reese, Jennifer Lynn
Journal Title
Journal ISSN
Volume Title

The Barnett shale has become an extremely successful unconventional natural gas development, mainly due to the optimization of hydraulic fracturing treatments. The ideal stimulation job for the Barnett is a slickwater treatment with low proppant concentrations, because this type of waterfrac is believed to create longer and more complex fracture networks, contacting much greater surface areas of the reservoir while minimizing the fracture face damage through the use of low viscosity fluids with no gel solids. Fracture mapping research in the Barnett shale has shown that the hydraulic fracturing of vertical wells produces an extremely complex network of fractures, and the work presented here focuses on modeling these fracture networks to gain a better understanding of how hydraulic fractures perform in the Barnett. Over one hundred simulation runs were conducted with models of varying fairway sizes, aspect ratios, fracture spacings, total network lengths, and fracture conductivities in an effort to better understand the impact each parameter has on well performance. Results were analyzed according to the classic parameters of stimulated reservoir volume, fracture spacing and total fracture network length. Observed trends and production plateaus in the simulation data establish a way to optimize the stimulation treatment and production for a given well. Fracture conductivity is considered to be of secondary importance in the hydraulic fracturing of very low permeability formations, but the extensive network structures of the Barnett shale are so large that fracture conductivity becomes important again. Since small increases in fracture conductivity can yield significant production increases, operators in the Barnett shale can focus on fracture conductivity as a way to optimize stimulation jobs and yield efficient production wells. The simulation results were compared to field data gathered from a production database, and showed that the simulation model can duplicate both the shape and range of the cumulative production profiles observed in the field, thus validating the simulation modeling process. The fact that the simulation runs model the observed field production provides further evidence that Barnett shale wells actually produce from a complex fracture network and not from a single planar fracture.