A two-site filtration model for silica nanoaggregate mobility in porous media under high salinity conditions
2018
A variable-viscosity colloid transport simulator is developed to model the mobility behavior of surface-engineered nanosilica aggregates (nSiO2) under high salinity conditions. A two-site (2S) filtration approach was incorporated to account for heterogeneous particle-collector surface interactions. 2S model was then implemented along with the conventional clean bed filtration (CFT) and maximum retention capacity (MRC) particle filtration models to simulate the results of a series of column tests conducted in brine (8% wt. NaCl and 2% wt. CaCl2)-saturated Ottawa sand columns at various pore velocities (7 to 71 m/day). Simulation results reveal the superiority of the MRC and 2S model classes over CFT model with respect to numerical performance criteria; a general decrease of normalized sum of squared residuals (ca. 20–90% reduction) and an enhanced degree of normality of model residuals were detected for 2S and MRC over CFT in all simulated experiments. Based on our findings, conformance with theories underpinning colloid deposition in porous media was the ultimate factor that set 2S and MRC model classes apart in terms of explaining the observed mobility trends. MRC and 2S models were evaluated based on the scaling of the fitted maximum retention capacity parameter with variation of experimental conditions. Two subclasses of 2S that consider a mix of favorable and unfavorable attachment sites with irreversible attachment to favorable sites (with and without physical straining effects) were found most consistent with filtration theory and shadow zone predictions, yielding theoretical conformity indices of 0.6 and higher, the highest among all implemented models. An explanation for such irreversible favorable deposition sites on the surface of silica nanoaggregates might be a partial depletion of stabilizing steric forces that had led to the formation of these aggregates.
Keywords:
- Correction
- Source
- Cite
- Save
- Machine Reading By IdeaReader
48
References
6
Citations
NaN
KQI