Propagation of a Hydraulic Fracture into a Remote Observation Wellbore Results of C-Sand Experimentation at the GRI/DOE M-Site Project

Abstract

Propagation of a Hydraulic Fracture into a Remote Observation Wellbore Results of C-Sand Experimentation at the GRI/DOE M-Site Project P.T. Branagan, SPE, Branagan & Associates, Inc., R.E. Peterson, SPE, Branagan & Associates, Inc., N.R. Warpinski, SPE, Sandia National Laboratories, S.L. Wolhart, SPE, Gas Research Institute, R.E. Hill, Consultant Abstract A series of six hydraulic fracture and remote fracture diagnostics experiments were conducted in a thick Upper Mesa Verde group fluvial sandstone at the M-Site Project near Rifle, Colorado. A slanted openhole wellbore was emplaced 287 ft away from the frac well in the direct path of the propagating fractures. Pressure measurements were acquired in this observation well as the fractures grew and subsequently intersected the open bore hole. The approaching fracture tip effects were readily discernible from direct frac intersection and/or passing fractures. Borehole image and radioactive tracer logs provided information delineating the position of the borehole/fracture intersection. This unique data set attests to the complexities of fracture growth and geometry and the particular influence of fluid viscosity. Introduction Project Background. Hydraulic fracturing is a complex process frequently used to stimulate gas production from low permeability reservoirs. Analysis of the fracture treatments involves the application of simple 2D to complex 3D models that attempt to replicate the subsurface fracturing process. Model results are generally based upon history matching measured scalar parameters, primarily surface or bottomhole pressure, injection rates and viscosity. These fracture models also require additional input data regarding rock properties (e.g., in-situ stress, rock moduli, fracture toughness, etc). Rock properties are often estimated since actual measured data are typically unavailable. Thus, while fracture models may presently appear to provide our best approximation or prediction of fracture dimensions, the results are nevertheless subject to the absolute variability of the inputs, measured data and system heterogeneities as well as the individual subjectivity of the fracture modeler and/or user. Recent field-scale tests coupled with direct observations of far-field fractures provide obvious indications regarding the non-idealized complexities of the fracturing process. The questions must then be asked: "What are the critical fracturing parameters, and how is the actual fracturing process really unfolding?" On the basis of these questions, Gas Research Institute and the U.S. Department of Energy jointly conceived the idea of the Multi-Site Hydraulic Fracture Diagnostics Project (M-Site Project). The M-Site Project, located near Rifle, Colorado as shown in Figure 1, is a comprehensively instrumented field laboratory that includes the following principal elements:–two offset wells containing triaxial accelerometer arrays for detecting microseismic events associated with hydraulic fracture propagation,–a vertical inclinometer array buried deep in the subsurface and used to determine the earth's mechanical deformation resulting from the opening and closing of the hydraulic fracture,–inclined diagnostic wells designed to intersect the induced hydraulic fracture system in the "far-field", and–wellbores, surface infrastructure and data acquisition systems to conduct complex field experiments. Using these instrumentation systems and infrastructure, the project team focused on performing experiments and gathering independent diagnostic data to–develop a microseismic fracture mapping technique and verify that the imaged hydraulic fracture values (length, height, azimuth) represent the real "mechanical" hydraulic fracture geometry; and–directly observe and describe the far-field character of hydraulic fractures. P. 99^