Abstract
When modeling reactive transport, reaction rates are often overestimated
due to the improper assumption of perfect mixing at the support scale of
the transport model. In reality, fronts tend to form between
participants in thermodynamically favorable reactions, leading to
segregation of reactants into islands or fingers. When such a
configuration arises, reactions are limited to the interface between the
reactive solutes. Closure methods for estimating
control-volume-effective reaction rates in terms of quantities defined
at the control volume scale do not presently exist, but their
development is crucial for effective field-scale modeling. We attack
this problem through a combination of analytical and numerical means.
Specifically, we numerically study reactive transport through an
ensemble of realizations of two-dimensional heterogeneous porous media.
We then employ regression analysis to calibrate an analytically-derived
relationship between reaction rate and various dimensionless quantities
representing conductivity-field heterogeneity and the respective
strengths of diffusion, reaction and advection.
Original language | English GB |
---|---|
Journal | Geophysical Research Abstracts |
Volume | 21 |
State | Published - 1 Dec 2017 |
Externally published | Yes |
Keywords
- 1816 Estimation and forecasting
- HYDROLOGY
- 1869 Stochastic hydrology
- 1873 Uncertainty assessment
- 1942 Machine learning
- INFORMATICS