From conservative to reactive transport under diffusion‐controlled conditions

Abstract

International audienceWe assess the possibility to use conservative transport information, such as that contained in transit time distributions, breakthrough curves and tracer tests, to predict non-linear fluid-rock interactions in fracture/matrix or mobile/immobile conditions. Reference simulated data are given by conservative and reactive transport simulations in several diffusive porosity structures differing by their topological organization. Reactions includes non-linear kinetically-controlled dissolution and desorption. Effective Multi-Rate Mass Transfer models (MRMT) are calibrated solely on conservative transport information without pore topology information and provide concentration distributions on which effective reaction rates are estimated. Reference simulated reaction rates and effective reaction rates evaluated by MRMT are compared, as well as characteristic desorption and dissolution times. Although not exactly equal, these indicators remain very close whatever the porous structure, differing at most by 0.6% and 10% for desorption and dissolution. At shorter times, this close agreement arises from the fine characterization of the diffusive porosity close to the mobile zone that controls fast mobile-diffusive exchanges. At intermediate to larger times, concentration gradients are strongly reduced by diffusion, and reactivity can be captured by a very limited number of rates. We conclude that effective models calibrated solely on conservative transport information like MRMT can accurately estimate monocomponent kinetically-controlled non-linear fluid-rock interactions. Their relevance might extend to more advanced biogeochemical reactions because of the close representation of conservative concentration distributions, even by parsimonious models (e.g., MRMT with 3-5 rates). We propose a methodology to estimate reactive transport from conservative transport in mobile-immobile conditions

Reactive Transport Modeling in Porous Fractured Media: Application to Weathering Processes

“…In addition, the presence of damaged zones in the vicinity of fractures can enhance this contrast as they are more transmissive than the matrix (Laurent et al, 2019;Neretnieks, 1980;Wogelius et al, 2020). Such zones may be considered by sub-gridding the rock matrix, as done in the Structured INteracting Continua approach proposed by Babey et al (2016). Hereafter, rather than introducing a third subdomain in our multi-continua model, we merged the damaged zone area with the fracture region and used a D.P.…”

“…Such zones may be considered by sub‐gridding the rock matrix, as done in the Structured INteracting Continua approach proposed by Babey et al. (2016). Hereafter, rather than introducing a third subdomain in our multi‐continua model, we merged the damaged zone area with the fracture region and used a D.P.…”

Inherent relevance of MRMT models to concentration variance and mixing-induced reactivity

“…Among them, the mobile-immobile Multi-Rate Mass Transfer models (MRMT) [Carrera et al, 1998;Haggerty and Gorelick, 1995] does not only propose efficient characterization and upscaling methodologies [Willmann et al, 2008;Babey et al, 2015;Rapaport et al, 2017] but also a natural bridge to equivalent concentration distributions, which relevance can be assessed to model reactive transport [Donado et al, 2009;Henri and Fernandez-Garcia, 2015;Sanchez-Vila et al, 2010;Soler-Sagarra et al, 2016]. Synthetic experiments have shown that MRMT models provide close approximations of bulk reactivity even in non-linear equilibrium and kinetically-controlled cases [Babey et al, 2016]. This has been linked to the empirical observation that MRMT models do not only conserve mass by construction but also the porosity weighted integral of concentrations squared [de Dreuzy et al, 2013], which is directly linked to mixing-induced reactivity through the scalar dissipation rate [Le Borgne et al, 2010].…”

“…By reducing concentration variances, diffusive processes induced by mobile-immobile mass exchanges thus promote reactivity and, by consequence, reduce further mixing and reaction potentials.The scalar dissipation rate only differs from the reaction rate by a chemical term that derives from the nonlinearity of the reactivity in the reactant concentrations[De Simoni et al, 2005]. Numerical simulations have however shown that the influence of this chemical term remains limited for most immobile porosity structures and reaction types also extending to heterogeneous reactivity including sorption and precipitation/dissolution[Babey et al, 2016]. The relevance of MRMT to model reactive transport fundamentally comes from the representativity for the concentrationA C C E P T E D M A N U S C R I P TRelevance of MRMT models to reactive transport distribution of its mean and variance in diffusively-dominated conditions.…”