Enhanced/Engineered Geothermal Systems (EGS) have the potential to provide a significant amount of base load electricity and heat and to displace fossil fuel consumption globally. To determine the potential for the expansion of direct use geothermal energy, a detailed analysis of U.S. energy consumption was performed to estimate the amount of primary energy consumed as a function of its utilization temperature from 0 to 260 ? C. The analysis revealed that about 34 EJ annually, more than 30% of the U.S. annual energy demand is used for direct thermal use applications in the temperature range of 0 to 260 ? C. Both analytical and numerical models of discretely fractured reservoirs were developed to probe the thermal hydraulic behavior of model EGS reservoirs and quantify factors controlling performance. An analytical model for discrete, fixed aperture, rectangular fractures with specified uniform flow was used to illustrate the renew ability of EGS reservoirs with a ratio of production to renewal times of about 0.2 to 0.33. Fracture structure and connectivity were also shown to affect reservoir performance in modeling studies. In general, fracture connectivity is more important than aperture variations within the fractures. Flow channeling in fractures with spatially varying aperture fields were simulated using a developed numerical model. An ensemble of fracture realizations were used to illustrate how the magnitude of aperture variations lead to flow structures that often inhibit rather than enhance subsurface heat exchange. Finally, both conservative and reactive tracers were used to determine the spatially varying thermal field during heat extraction in a discrete fracture with variable aperture. Reduced order modeling of the fracture was used to create a tractable framework for inferring reservoir structure. Tracers revealed the capability to predict a reservoir's production temperature versus time, with reactive tracers providing better results. However, difficulties in accurately predicting the aperture field led to a non-unique outcome where more than one reservoir realization matched both the tracer curve and production temperature.