PMFCT-2D - A TRANSPORT SIMULATOR FOR VARIOUS GRID PECLET NUMBERS

Authors
Citation
Nj. Aimo et M. Oostrom, PMFCT-2D - A TRANSPORT SIMULATOR FOR VARIOUS GRID PECLET NUMBERS, Ground water, 35(1), 1997, pp. 30-38
Citations number
20
Categorie Soggetti
Geosciences, Interdisciplinary
Journal title
ISSN journal
0017467X
Volume
35
Issue
1
Year of publication
1997
Pages
30 - 38
Database
ISI
SICI code
0017-467X(1997)35:1<30:P-ATSF>2.0.ZU;2-R
Abstract
Solute transport simulation using numerical models is an important and widespread tool for evaluation of clean-up strategies as well as for prediction of future transport. Classical simulation algorithms for ad vective-dispersive transport usually introduce large numerical errors where concentrations are lowest. In general, numerical errors tend to spread (disperse) the solute more than physical processes alone. For s imulations where the Peclet number (Re) is greater than about 2, numer ical dispersion can be very significant and could lead to erroneous co nclusions. Recent numerical techniques for simulating advective transp ort minimize numerical errors and provide much better solutions. One s uch technique, Flux-Corrected Transport (FCT), can preserve sharp conc entration fronts by virtually eliminating numerical dispersion. In gen eral, it has been observed that as more detailed knowledge of subsurfa ce flow fields is obtained, smaller dispersivity values are needed to match observed and simulated data. However, for many numerical codes t he use of small dispersivities is not practical, because it requires f ine grids to keep the grid Peclet number limited. A general purpose tr ansport code, PMFCT-2D, has been developed, including a fast and effic ient FCT algorithm, to simulate advective-dispersive transport in vari ably saturated, heterogeneous porous media, with nonuniform aquifer th ickness in the third dimension. PMFCT-2D can be used to accurately sim ulate high Peclet number transport, including purely advective transpo rt (Pe = infinity), resulting from transient or steady-state flow cond itions. The code is easily coupled to any flow simulator via generated velocity, saturation, and cell thickness fields.