Dynamics of basaltic glass dissolution - Capturing microscopic effects in continuum scale models

The method of 'multiple interacting continua' (MINC) was applied to include microscopic rate-limiting processes in continuum scale reactive transport models of basaltic glass dissolution. The MINC method involves dividing the system up to ambient fluid and grains, using a specific surface area to describe the interface between the two. The various grains and regions within grains can then be described by dividing them into continua separated by dividing surfaces. Millions of grains can thus be considered within the method without the need to explicity discretizing them. Four continua were used for describing a dissolving basaltic glass grain; the first one describes the ambient fluid around the grain, while the second, third and fourth continuum refer to a diffusive leached layer, the dissolving part of the grain and the inert part of the grain, respectively. The model was validated using the TOUGHREACT simulator and data from column flow through experiments of basaltic glass dissolution at low, neutral and high pH values. Successful reactive transport simulations of the experiments and overall adequate agreement between measured and simulated values provides validation that the MINC approach can be applied for incorporating microscopic effects in continuum scale basaltic glass dissolution models. Equivalent models can be used when simulating dissolution and alteration of other minerals. The study provides an example of how numerical modeling and experimental work can be combined to enhance understanding of mechanisms associated with basaltic glass dissolution. Column outlet concentrations indicated basaltic glass to dissolve stoichiometrically at pH 3. Predictive simulations with the developed MINC model indicated significant precipitation of secondary minerals within the column at neutral and high pH, explaining observed non-stoichiometric outlet concentrations at these pH levels. Clay, zeolite and hydroxide precipitation was predicted to be most abundant within the column. (c) 2013 Elsevier Ltd. All rights reserved.