Numerical investigations of heat and mass transport in fractured porous rock masses

Abstract

Fluid flow processes in the subsurface are accompanied by heat and mass transport with several important feedbacks including reactive flow, and precipitation/dissolution processes. Heat and mass transport through fractured rock masses occurs in many natural systems such as the plumbing of volcanic systems, mesothermal ore deposits, and post-seismic fluid flow. Anthropogenically-driven systems, such as fluid-injection in Enhanced Geothermal Systems (EGS), and the injection of waste-water from hydrocarbon extraction also involve heat and mass transport through porous or fractured rocks. Understanding in detail how mass and heat transfer interact in natural or in industrial applications requires numerical models in combination with field and laboratory experiments to determine the dominating factors. This thesis examines the impact of heat and mass transport on high pressure fluid propagation in the subsurface, as well as different numerical approaches of transient heat flow in fractured porous media and the heat exchange between flowing fluid and host rock.<br /> Many fluid-triggered seismic events show a tendency for upward migration of the seismic cloud, generally assumed to reflect a fluid-pressure dependent permeability. In a numerical investigation that combines pressure-dependent permeability with thermal and salinity effects, it is found that over short timescales pressure-dependent permeability does indeed have the strongest influence on asymmetric diffusion. However, it is also demonstrated that over longer timescales, for example the lifetime of a geothermal reservoir, temperature and salinity effects play an increasingly important role. <br /> Assessing the thermal field of a geothermal resource or in a CO₂ sequestration project is essential for proper design and management. Typically, numerical simulations assume that the fluid and solid phases are in thermal equilibrium, an assumption that has to date not been investigated in detail. This assumption is examined in this work by simulating fluid and heat flow in a simple geometry to analyse the influence of site specific parameters on the simulation result. It is shown that the equilibrium model is not sensitive to porosity contrasts, while the non-equilibrium model shows a sensitivity to porosity contrasts, with simulation results diverging more strongly in less permeable zones. In a simulation of a hypothetical geothermal system, the equilibrium model shows higher production temperatures with a divergence of up to 7% between the approaches, which could impact the economic feasibility of a project.<br /> Finally, a new approach is introduced to determine the heat transfer coefficient h between rock walls and flowing fluid using the non-equilibrium model. Based on a numerical experimental setup with simple geometry and steady state scenario, a dynamic heat transfer coefficient is derived that depends on fracture aperture and flow velocity. This model is based on well-defined physical parameters, it is adaptable to complex geometries, and intrinsically adjusts to spatial heterogeneities and temporal changes in flow and temperature field. A possible extension of this dynamic approach is demonstrated in numerical simulations the reservoir scale.