The manner in which air moves through soils depends on its velocity, flow direction, amount of resident water in the soil, and soil characteristics such as grain shapes and void (pore) connectivity. In turn, the flow patterns and distribution directly affect our capabilities to describe, quantify, predict, and evaluate this flow.
A recent study in Reviews of Geophysics explores the physics governing air flow from the micron to centimeter scale, and different approaches and limitations to modeling this behavior. We asked the lead author to give an overview of airflow through soil, describe how scientists study and model it, and outline what questions remain.
In simple terms, why is it important to understand how air moves through soil?
The importance of air (and other gases) flow through porous media is drawn from its role in many natural processes and artificial practices, such as industrial filters and reactors, petroleum production, terrestrial gas emission, gas sequestration in confined aquifers, aeration of cultivated soils, remediation of contaminated geological formations, and medical applications. For example, even the ability to respire through a moist face mask is an air flow through a partially-saturated porous medium problem.
Understanding the biogeochemical processes involved, and engineering and operating the systems of the different practices, hinges on quantifying the different processes at different scales and modeling gas flow in wet porous media.
What factors influence how air flows through soil?
In terms of governing forces, the flow of air is subjected mostly to capillary (surface tension), viscous (drag), and buoyance (gravitational) forces. At the pore scale, the capillary to viscous forces ratio dominates the phase distribution, that is, the air/water-filled pore size distribution. At the larger (Darcy) scale, for upward air flow, the interplay (ratio) between buoyancy and viscosity is the most significant factor in determining the air flow pattern.
The interplay between the different forces is determined by: media characteristics (e.g., wettability, porosity, grain size, heterogeneity, connectivity, angularity, roughness), fluid characteristics (e.g., viscosity, density, miscibility, volatility, compressibility), and fluid dynamics (e.g., velocity, gravity/direction, pressure, air/water content).
For illustration, if we inject (sparge) air into otherwise water-saturated porous media, the airflow pattern, at the macroscale, will evolve from small bubbles to large ganglia and to air channels, with increasing injection rates for which the transition (injection rate) between the different flow patterns, depends on the media characteristics (e.g., grain size).
What are “flow mechanisms” and how do they differ from “flow patterns”?
The microscale flow mechanism refers to how the two phases (e.g., air and water) migrate from one pore to another (Fig. 1a). The macroscale flow pattern refers to how air is distributed and flows through a practical (centimeter to meter) scale (Fig. 1b). The same factors and forces affect both microscale mechanism and macroscale flow patterns and, in some cases, also share some morphological resemblance.
Moreover, the microscale and macroscale affect each other. For example, the flow mechanism is strongly affected by wettability. In hydrophobic media, air will flow near the grain surface; in hydrophilic media, air will invade through the pore center. In turn, this will affect the connectivity and trapping of the air phase and its larger, macroscale flow pattern and content. However, the relationship between the mechanisms and patterns is not straightforward. For example, a stable displacement mechanism may prevail on the moving fronts of an air channel or an air bubble. While micro and macro scale flows are interconnected, quantifying this linkage remains one of the major open challenges in multiphase flow.
Why is it beneficial for scientists to study flow at both the microscale and macroscale?
Different processes are related to different scales. Thus, evaluating complex phenomena requires an across-scale perspective. For illustration, if one wants to design a bioventing remediation system that improves soil aeration status to enhance the biological degradation of hydrocarbons, the macroscale flow (if it can be quantified) can be used to evaluate some design parameters such as the distance between injection wells, discharge rates, and so on.
However, some processes, such as chemical and biological reactions, occur on much smaller (pore) scales. This means we cannot estimate the efficiency of the aeration and remediation duration without accounting for the microscale processes. Upscaling these processes, even in stable flow, is a challenging task, more so in unstable displacement in which the flow is not well described.
What are some of the limitations of different modeling approaches?
Pore-scale models are mostly limited by the computational resources demands, such as running time and memory, needed to solve the air/water displacement, even in a small flow domain. Pore network models usually have very limited predictive capabilities and are also hindered by the ability to capture real media heterogeneity. In turn, Darcy-scale models are limited by the continuum approach limitations, which rely on the flowing phase/s to be continuous so that each phase is propelled by its potential gradient, which is usually not the case for unstable and non-coherent flow. Another limitation of the Darcy-scale models is that many processes that occur on the pore scale cannot be upscaled to the representative elementary volume (REV) without averaging some key (critical to the process) information.
What are some of the knowledge gaps where additional research, data, or modeling are needed?
One of the main limitations in modeling air/water systems lies in the relevant length scale of the physical process being much smaller than the scale of interest for most practical applications. Thus, while comprehensive simulations of realistic multiphase flow problems in porous media hinge on upscaling or coupling between scales, the physical heterogeneity and the unknown nature of porous media on all scales hinder most efforts. Modeling is even more challenging in the context of unstable air flow, and developing an upscaling method that accounts for the expected flow patterns is a major undertaking. To address this challenge, we first need to resolve many open questions, such as:
- Can we quantify the bidirectional effect of the flow dynamics and non-equilibrium phase configuration?
- How do the interactions between wettability and pore geometry affect the air-water displacement mechanisms? Moreover, how do the flow dynamics affect these interactions?
- How do the microscale air-water displacement mechanisms and phase configuration affect the macroscale flow patterns and phase distributions?
- How can the effect of microscale phase configuration on the hydraulic properties of the media be evaluated from the flow dynamics and media characteristics?
- How do air flow dynamics affect the medium permeability to air?
- Even if obtained, can the predictability of the flow pattern (from the governing factors) be used implicitly to solve the unstable flow problem?
Editor’s Note: It is the policy of AGU Publications to invite the authors of articles published in Reviews of Geophysics to write a summary for Eos Editors’ Vox.