Premium
Ensemble‐averaged equations for reactive transport in porous media under unsteady flow conditions
Author(s) -
Wood Brian D.,
Kawas M. Levent
Publication year - 1999
Publication title -
water resources research
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.863
H-Index - 217
eISSN - 1944-7973
pISSN - 0043-1397
DOI - 10.1029/1999wr900113
Subject(s) - covariance , mathematics , flow (mathematics) , operator (biology) , mathematical analysis , exponential function , divergence (linguistics) , statistical physics , physics , geometry , biochemistry , statistics , chemistry , linguistics , philosophy , repressor , transcription factor , gene
We present a method for deriving the ensemble‐averaged reactive solute transport equation for unsteady, non‐divergence‐free flow field conditions. Our approach uses a cumulant expansion, Lie group theory, and time‐ordered exponentials to develop the ensemble‐averaged transport equation. The cumulant expansion is in powers of a αт c , where α measures the magnitude of the perturbations of the transport and reaction operators and т c is the correlation time of these perturbations. Because the cumulant expansion avoids secular terms (terms in powers of time), the problem can be closed by rationally truncating the expansion. The truncated terms can be shown to be of lower order than those terms that are kept, provided that a particular constraint (in terms of the Kubo number) is met. The use of Lie group theory allows one to automatically combine the Eulerian and Lagrangian approaches. A particular time‐ordered exponential that arises in the analysis can be interpreted as a translation operator that possesses a well‐defined algebra. These translation operators appear in the second‐order (covariance) terms of the cumulant expansion, and their effect is to shift one of the terms of the covariance functions relative to the other along the trajectory formed by the ensemble‐averaged velocity field. This approach does not require neglecting the local dispersion tensor and has the advantage that no integral transformations are conducted; therefore all results are expressed in terms of real space variables.