Pore Scale Modeling of Mixing-Induced Carbonate Precipitation

Mixing of groundwaters of differing chemical composition can lead to precipitation of minerals, potentially modifying the transport and chemical properties of the subsurface materials. Carbonate minerals are particularly common secondary phases that form as a result of mixing, although in many instances their formation is also affected by a suite of complex dissolution and precipitation reactions that change the pH and alkalinity of groundwater. In the case of mixing, several distinct regimes are recognized, depending on the supersaturation generated by the mixing process. In the case where high degrees of supersaturation with respect to carbonate occur as a result of mixing (e.g., log Q/Keq > 1.5, where Q is the ion activity product and Keq is the equilibrium constant), homogeneous nucleation can generate reactive surface area for continued carbonate growth. In this case, no interaction between the mixing fluid and immobile solid phases is needed. In contrast, where supersaturation is more limited (log Q/Keq = 0.5 to 1.5), precipitation generally takes place via heterogeneous nucleation, in which case a templated mineral surface (normally carbonate) is required. Heterogeneous nucleation of carbonates is typically second order with respect to the supersaturation. At lower degrees of supersaturation (log Q/Keq < 0.5), precipitation takes place via crystal growth on discrete surface features of the carbonate mineral (e.g., via spiral growth) surface and shows a first order or quasi-first order dependence on supersaturation. Thus, the supersaturation induced by mixing largely controls the order of the reaction and the extent of interaction with pre-existing mineral surfaces in the subsurface. These in turn impact how the physical and chemical properties of the medium are modified by carbonate precipitation. We are investigating these carbonate precipitation regimes using pore scale reactive transport modeling based on Direct Numerical Simulation methods. Our computational approach relies on a new simulator based on operator splitting that combines a calculation of Navier-Stokes flow within discrete grain packs with the geochemical simulator CrunchFlow. Navier-Stokes flow and solute transport are handled by the Chombo software package, which implements Adaptive Mesh Refinement and embedded boundary methods for high resolution simulations. The pore scale computational results are validated by comparison with a variety of experimental studies in which carbonate precipitation results from mixing.