Substrate-explicit coupled modeling of organic matter degradation and sulfide production using a generalized kinetic...

In natural subsurface systems, Dissolved Organic Matter (DOM) is a potent reducing agent, as it sustains and catalyzes a number of microbially-mediated reactions. One of such reactions, namely the reduction of sulfate to sulfide mediated by sulfate-reducing bacteria (SRB) is particularly relevant for the evaluation of the long-term stability of deep geological repositories (DGRs) for spent nuclear fuel. Some of these proposed DGRs will be hosted at great depth in fractured crystalline rocks where sulfide, potentially produced by SRB, is practically the only oxidant of relevance for the copper canisters used to safely store and keep isolated the radionuclides from the natural environment. One important limitation to this reaction is substrate availability, as the DOM that typically infiltrates, e.g. from overlying surficial sediments, is typically bound in complex polymeric organic compounds (cellulose, starch, proteins, fats, etc.) that cannot be directly used by SRB. A further intermediate step is therefore needed which consists in the degradation of 'complex' organic matter into 'simpler' organic substrates. This intermediate step is performed by a different microbial group (typically acetogens and/or fermenters) that acts on a substrate characterized by significant heterogeneity. Previous studies have documented that the organic carbon substrate typically comprises thousands of species depending on the type of subsurface environment. In this study we argue that the evaluation of microbially mediated sulfate reduction by means of reactive transport modeling should explicitly account for the substrate heterogeneity. However, explicitly accounting for thousands of chemical species of complex organic matter presents two significant challenges. The first is how to parameterize each individual reaction. The second is how to include this large number of species in the mass balance calculations. The first challenge was addressed through the development of a Generalized Kinetic Model (GKM) that formulates all rates as a function of a universally accepted rate of electron transfer along with the underlying stoichiometry of each individual metabolic reaction. The computational challenge was overcome by using weight-based reaction rates. In practice, we solve transport and mass balance for the equivalent total concentration of complex organic carbon while retaining information on the underlying substrate and associated microbial population through the use of two weighting functions, which are updated on lagged time based on mass balance. The numerical framework is first verified and then applied to simulate sulfide production in a single fracture-matrix system. Important modeling outcomes, such as the dependence of SRB activity on the degradation of a complex organic substrate and the role of the matrix as a temporary storage term for substrates, are derived.