Geochemistry, Mineralogy, Volcanology Editors' Vox

Measurements and Models of Reactive Transport in Geological Media

Author Brian Berkowitz answers questions about his recently published article and the scientific and societal implications of his findings.


Last month, Reviews of Geophysics published “Measurements and Models of Reactive Transport in Geological Media” by Berkowitz et al. Here the author weighs in on new understandings and what they mean for future research in the field.

Reactions between rocks and groundwater both alter rocks and determine the chemistry of water, including of contaminants as it flows through the ground. This process has been studied for some time. What recent advances in particular are leading to a new understanding or synthesis?

Reactive chemical transport plays a key role in geological media across scales from pores to an aquifer. The combination of fluid and chemical transport with chemical reactions leads to a rich spectrum of complex dynamics. The principal challenge in modeling reactive transport is to account for the subtle effects of fluctuations in the flow field and species concentrations; spatial or temporal averaging generally suppresses these effects. It is critical, too, to ground model conceptualizations and test model outputs against laboratory experiments and field measurements.

This review emphasizes the integration of these aspects in the context of development and solution of reactive transport models based on continuum-scale and particle tracking approaches. The advent of new experimental methods and powerful computational resources enable model-data synthesis and more interactive interrogation of key processes governing the emergence of larger-scale patterns of transport and reaction.

As a consequence, four major features of reactive transport are recognized as controlling measurement and model interpretation from core to aquifer scales. First, chemical reactions always occur within small volumes and narrow zones of mixing and reaction fronts, at the pore scale. Second, reaction dynamics are fundamentally dependent upon the initial distribution of reactants and how other reactants are subsequently introduced into the system. Third, preferential pathways are ubiquitous, and their occurrence leads to nonuniform flow fields and reaction fronts. While reactions must be quantified at small scales, flow and transport must be identified at small scales but then quantified as they propagate over much larger scales. Fourth, probabilistic considerations are required to account for the small-scale fluctuations in chemical species velocity (spreading), concentration, and mixing that ultimately control the spatiotemporal occurrence of chemical reactions.

What are the societal implications of the new understanding?

The modeling and experimental methods, results and insights, and discussion of what can one can realistically expect to measure and quantify are relevant to virtually any “real-world” problem—whether induced naturally or anthropogenically—involving biogeochemical reaction scenarios in settings with flowing solutions through porous or fractured materials. The review offers a comprehensive state-of-the-art analysis of what has and has not been measured, and what can and cannot be effectively modeled, in such scenarios. The analysis has direct application to assessment of soil and groundwater quality, the fate of anthropogenic pollutants (e.g., contamination from landfill leachates, septic tanks, and leakage from industrial sites), environmental management and remediation strategies (e.g., irrigation, agrochemical application, occurrence of arsenic in groundwater), as well as subsurface CO2 and nuclear waste sequestration that also account for chain reactions that include precipitation and dissolution, and even evolutionary phenomena in geological formations.  Careful integration of measurements to calibrate a realistic model can enable, for example, prediction of where heavy metals leaking from an industrial site might be expected to precipitate, or where injected CO2 might precipitate as calcium carbonate, and how patterns of water flow might be affected from the resulting change in aquifer heterogeneity.

What are the major unsolved or unresolved questions, and where are additional data or modeling efforts needed?

A range of distinct factors influence reactive transport dynamics, both spatially and temporally: host medium fluid flow field and biogeochemical properties, type of chemical species, relevant chemical reaction, and type of chemical release. Reaction and transport processes are intimately coupled, creating a complex web of feedback loops propagating through various spatial and temporal scales. For example, introduction of one chemical species into a subsurface environment can initiate a series of “chain reactions”, mobilizing other species which are then mobilized and can precipitate, affecting hydraulic conductivity, porosity and flow field patterns at different scales.

Multispecies reaction systems generally involve an extensive series of chemical reactions which can occur, either reversibly or irreversibly, in parallel. The reaction dynamics are such that one or more “bottleneck” reaction(s) tend(s) to control the system behavior; these bottlenecks are the rate-determining steps, and different reactions can become bottlenecks as conditions change over time and space. As a consequence, more sophisticated modeling tools must be invoked, and corresponding experimental design, data collection, and measurement methods must be incorporated, to quantify and predict desired hydrological and geochemical parameters.

These realities—and the intimate coupling between transport and reaction— pose complex and subtle considerations for model development, which can involve both continuum and particle tracking formulations. Such coupling has been demonstrated in various continuum formulations, but solution of these equations remains largely elusive and further model development and analysis is required. On the other hand, numerical models based on particle tracking approaches are attractive, largely because they can account implicitly for the coupling between transport and reaction and for the effects of spatiotemporal fluctuations in velocity and chemical concentrations. However, both continuum and particle tracking methods are limited in terms of capturing chemical reaction dynamics, particularly at the field scale. The inescapable reality is that all of the resulting numerical models require detailed information that characterizes the structural heterogeneity of the domain, as well as initial and boundary conditions, to first enable determination of the flow field at as high a resolution as possible. For chemical reactions, however, the real enigma is how to incorporate pore-scale velocity and concentration fluctuations, and the degree of mixing and reaction, to remain computationally feasible. Continuum methods account for reactions by averaging concentrations over domain elements, which suppresses fluctuations, while particle tracking methods account for reactions by defining a reaction radius or collocation probabilities, which can become unrealistically large.

With the understanding that there are no “all-purpose”, “magic bullet” models or experiments that are representative of and capture all such dynamics, we recognize the need to develop scenario-based analyses that explicitly consider these complexities.

Dr. Brian Berkowitz, Weizmann Institute of Science, Department of Earth and Planetary Sciences, Rehovot, Israel; [email protected]