Reactive Random Walks - Embracing Complexity, Incomplete Mixing, and Heterogeneity to Predict Reactions in Environmentally Relevant Porous Media Systems
Abstract
Chemical reactions are ubiquitous in the natural environment, influencing behaviors and outcomes of many systems of practical interest. However, predicting complex reactions in realistic environmental settings, characterized by high heterogeneity and processes occurring over multiple spatial and temporal scales, still remains a demanding challenge. Traditional, and even state of the art, models often fail at reproducing observations, even for relatively simple systems. This is exacerbated as systems become more complex, which is the norm in the natural environment. Agreement between traditional models and field observations is often obtained through unphysical calibration of model parameters, which works for hind-casting and fitting, but highlights the unphysical basis of these models and reveals problems with their use in a predictive sense. For many processes, the challenge is as follows. On one hand, controlled laboratory experiments allow the study of mechanisms by which chemicals move, degrade and transform in environments, where multiple phases and interfaces (e.g., soil/air/water) are present. On the other hand, the idealized nature and high degree of control is not representative of the range of scales and conditions in the natural environment. Consequently, translating high quality laboratory-acquired information to field scales is immensely challenging, particularly due to ubiquitous heterogeneity and non-ideal conditions that exist at all scales in real environmental systems. While conventional approaches to modeling reactive transport are built on a deterministic framework, it should be noted that models that aim to successfully work at scales of environmental relevance must embrace the complex, uncertain, and quasi-stochastic nature of environmental systems. Novel Lagrangian reactive random walks, rooted in stochastic theory, are a clear alternative. They are suited to environmental multi-scale problems and already, for simple reactive systems, have accurately predicted reactions in heterogeneous flows without parameter tweaking. Their success arises because particles explicitly represent concentration and transport fluctuations at all scales. The physics of mixing and reaction emerge naturally at the smallest scales, and macroscopic behavior evolves without unphysical external impositions or parameter tweaking. However, the theory on how to apply them in the context of multi-phase systems with complex reaction chains is still in its infancy. This work proposes to extend Lagrangian approaches by deriving and testing accurate, physically-based model equations and simulation techniques for transport and reactions in complex environmental systems, particularly relating to redox reactions and mobility of metals in subsurface waters. Specific objectives include: (i) Finalize the theory for Lagrangian random walks for arbitrarily complex reactions; (ii) Implement the Lagrangian reactive random walk framework in a general and flexible coding environment so that it can be used by the scientific community at large; (iii) Couple with a suite of available geochemical libraries to take advantage of state of art geochemical knowledge and (iv) Apply to complex reactive systems of environmental relevance including metal cycling in soils, metal mobilization by natural organic matter and uranium mobilization at the well known Rifle site in Colorado. The focus will be on geochemical reactions in porous media, motivated by abundant examples, where mismatches between predictions and observations arise. However, success will not be measured in this field alone. The ultimate goal is to develop a fundamentally novel and transformative framework for modeling reactions that embraces complexity and heterogeneity at all scales. The outcome will be sufficiently general to be broadly applicable, with impacts in predicting reactions in many environmental systems such as the atmosphere, lakes, rivers, oceans and beyond.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Feb 14, 2019
- Source ID
- W911NF1810338
Entities
People
- Diogo Bolster
Organizations
- Army Contracting Command
- United States Army
- University of Notre Dame