Enhanced Geothermal Systems (EGS) are a promising concept for unlocking the great potential of Hot Dry Rock (HDR) resources for clean and sustainable energy production. It can be argued that three of the foremost unsolved challenges for EGS are: induced seismicity, uneconomically low flow rates, and premature cooling of the produced fluid. We propose that fracture caging could be a solution to these three challenges. Fracture caging is the placement of a ‘cage’ of boundary wells around injection wells before fluid stimulation or circulation begins. This fracture cage is intended to contain injected fluids and to thereby limit fracture growth. In the long term, this fracture cage permits sustained high-pressure fluid injection to hold fractures open using hydraulic pressure (i.e., ‘hydropropping’) instead of by using proppant particles or by shear asperity propping. In this study, we present an analytical model and laboratory experiments that quantitatively explore the limits of tensile hydraulic fracture caging. Discoveries from this work include: (1) the maximum flow rates that can be caged are limited by flow constrictions in the boundary wells but are not limited by the injection pressure, (2) a hydraulic fracture can be caged with as few as two boundary wells, and (3) tensile fractures can be hydropropped without growing larger during sustained high-pressure fluid injection into a cage.

Measuring hydro-mechanical properties of natural fractures is a prerequisite for optimizing hydraulic stimulation design and well placement. We completed experiments to characterize shear on natural fractures in schist, amphibolite, and rhyolite specimens drilled from EGS Collab Project’s field sites at the Sanford Underground Research Facility (SURF) in South Dakota. A triaxial direct shear method and coupled x-ray imaging were used to perform hydroshearing and mechanical shearing at the site’s in-situ stress conditions. This produced simultaneous measurements of fracture and matrix strength, permeability, stress-dependent aperture, dilation, and friction strength. Our results identified that only a subset of the natural fractures was weak enough for hydroshearing. Generally, hydroshearing increases fracture permeability by a factor of 10 or more and the enhancement is retainable over time. However, the shear slip does not always result in permeability enhancement. High content of phyllosilicates was found to associate with exceptionally weak fractures that also exhibited poor or even negative enhancement after stimulation. Combining our measurements with site data, we can predict that most observable fractures at the two EGS Collab sites do not meet the criteria for hydroshearing before tensile opening. In some cases, the visible fractures are low permeability and as strong as the adjacent rock. To induce hydroshearing before tensile opening, injection must target known weak and favorably oriented fractures with confirmed pre-existing permeability.