PORE-SCALE MODELING OF NONWETTING-PHASE RESIDUAL IN POROUS-MEDIA

Citation
Mi. Lowry et Ct. Miller, PORE-SCALE MODELING OF NONWETTING-PHASE RESIDUAL IN POROUS-MEDIA, Water resources research, 31(3), 1995, pp. 455-473
Citations number
76
Categorie Soggetti
Limnology,"Environmental Sciences","Water Resources
Journal title
ISSN journal
00431397
Volume
31
Issue
3
Year of publication
1995
Pages
455 - 473
Database
ISI
SICI code
0043-1397(1995)31:3<455:PMONRI>2.0.ZU;2-S
Abstract
The complexity of the formation and removal of nonwetting-phase residu al in porous media provides a significant challenge to the development of theories needed to advance more effective aquifer remediation sche mes. Constitutive theories are required to overcome theoretical defici encies in describing behaviors such as the formation, dissolution, and mobilization of residual. Current macroscopic approaches are often un able to capture adequately the complexity of such processes, in these cases, pore-scale research should provide the necessary conceptual and quantitative basis for constitutive theory development. This work con tributes to these efforts by investigating factors affecting residual in strongly wet, capillary-dominated systems. Porous media are modeled as three-dimensional networks of pore bodies and throats. A random ne twork approach is developed, which provides stochastic geometry and va riable connectivity. Immiscible displacement occurs via piston-displac ement and film-flow mechanisms. Residual ganglia become trapped during imbibition. Wetting-phase continuity is maintained during drainage, d isallowing the entrapment of wetting-phase ganglia. Results show that pore-scale geometric parameters significantly affect residual saturati on, the ganglia-volume distribution, and nonwetting-wetting-phase inte rfacial area. Thus constitutive theory of residual should incorporate pore-scale knowledge to be valid for a wide range of conditions and po rous media types. Comparison with quantitative experimental ganglia-vo lume distributions yields good agreement with the data trends but poor quantitative agreement. These results indicate the need for further e xperimental comparisons with theoretical results, linking macroscale a nd microscale theory and experiment.