Modeling Escherichia coli and Rhodococcus erythropolis transport through wettable and water repellent porous media.

Abstract

Water protection and bioremediation strategies in the vadose zone require understanding the factors controlling bacterial transport for different hydraulic conditions. Breakthrough experiments were made in two different flow conditions: i) an initial bacteria pulse under ponded infiltration into dry sand (-15,000 cm); ii) a second bacteria pulse into the same columns during subsequent infiltration in constant water content and steady-state flow. Escherichia coli (E. coli) and Rhodococcus erythropolis (R. erythropolis) were used to represent hydrophilic and hydrophobic bacteria, respectively. Equilibrium and attachment/detachment models were tested to fit bromide (Br-) and bacteria transport data using HYDRUS-1D. Derjaguin-Landau-Verwey-Overbeek (DLVO) and extended DVLO (XDLVO) interaction energy profiles were calculated to predict bacteria sorption at particles. Adsorption of bacteria at air-water interfaces was estimated by a hydrophobic force approach. Results suggested greater retention of bacteria in water repellent sand compared with wettable sand. Inverse parameter optimization suggested that physico-chemical attachment of both E. coli and R. erythropolis was thousands of times lower in wettable than repellant sand and straining was 10-fold lower in E. coli for wettable vs repellant sand compared to the exact opposite by orders of magnitude with R. erythropolis. HYDRUS did not provide a clear priority of importance of solid-water or air-water interfaces in bacteria retention. Optimized model parameters did not show a clear relation to the (X)DLVO adsorption energies. This illustrated the ambivalence of (X)DLVO to predict bacterial attachment at solid soil particles of different wetting properties. Simultaneous analysis of mass recovery, numerical modeling, and interaction energy profiles thus suggested irreversible straining due to bacteria sizing as dominant compared to attachment to liquid-solid or liquid-air interfaces. Further studies are needed to distinguish straining mechanisms (i.e. pore structure or film straining) in different hydraulic conditions.