| Abstract |
Hydraulic fracturing is planned to be performed in the SIGMA-V project at the EGS Collab Test site located in a drift (tunnel) at the Sanford Underground Research Facility (SURF) at a depth of 4,850 ft below ground surface. The anticipated stimulated sub-vertical fractures will be characterized using geophysical monitoring and hydrological flow tests. Four phases of the flow tests (i.e., step-rate pressure test, injection-withdrawal dipole tracer test, dipole thermal test, and a saline tracer test) will be employed to systematically characterize the fracture and matrix properties and fracture-matrix heat transfer area using a well-designed monitoring system. In this paper, we present the modeling results of multi-tracer tests. The matrix flow properties were estimated using the water-level surveys at the nearby kISMET wellbores, and the fracture geometry and aperture from stimulation modeling were used as the baseline properties of the dipole flow system. A comprehensive sensitivity analysis was conducted to investigate the importance of fluid uptake to the matrix (leak-off), diffusion into the matrix, advection in the fracture, and dispersion in the fracture. Two fracture geometries were considered: (1) a radial fracture of radius 20 m centered at the stimulation/injection well; and (2) a linear fracture 12 m × 10 m that is considered a possible geometry if the zero stress condition in the drift controls fracture development. The injection and production/observation wellbore intersections with the fracture are assumed to be 10 m apart, the injection rate is 10 ml/min, and the tracer-release duration is five hours. The comparison between the baseline and sensitivity cases shows that the tracer breakthrough curve is very sensitive to matrix diffusion coefficient, whose baseline value is 10-11 m2/s, but less sensitive to matrix advection (leak-off) which is controlled by matrix and fracture permeability. This indicates that multi-tracer tests with different diffusivity will provide critical monitoring data needed for model validation, e.g., to constrain fracture-matrix interaction processes. The choices of injection rate and tracer-release duration can be informed by value-of-information modeling that includes monitoring design to optimize the tracer test parameters and data collection strategy for the objective of model validation. |