Hydrodynamic surface association and chemotactic dispersion affect bacterial transport in porous media

The transport of bacteria in porous media is poorly understood. The focus of this dissertation is on two considerations that can significantly influence bacterial fluxes in subsurface environments. The first is a phenomenon called surface association, a non-sorptive hydrodynamic interaction. Motile bacteria are particles that actively execute runs and tumbles using flagellar motors. Brownian-based transport models fail to adequately model transport in porous media due to the inelastic nature of surface interactions. In the absence of an internally-mediated switch to tumble, the cell continues to swim in its previous direction until such time that a tumble re-orients its body appropriately. This dissertation develops a model to predict the bacterial residence times due to this mechanism, and the pore size effects. Analysis of experiments conducted by others with columns where the porous media was pre-treated with unlabelled bacteria with a surface association model provided a mechanistic explanation for the observations by these investigators of low apparent diffusion coefficients. The results of the surface association model, in the form of a ratio of kinetic parameters, k1/k -1 was consistent with theoretical predictions. The k1/k -1 value for chemotaxis experiments was larger than those obtained from random motility experiments, consistent with the increased run lengths due to the presence of a gradient. Also, the k1/k-1 value for P. putida was larger than the comparable value for E. coli, consistent with the larger Knudsen number in experiments with P. putida, and with experimental observations of higher tortuosity in these experiments.; The second consideration is the influence of chemotaxis on the macroscopic transport of bacteria. The approach was to start with a microscopic description of porous media, and to use volume averaging method to develop closed forms of the macroscale description in terms of volume-averaged quantities. This dissertation shows that the effect of a chemotactic driving force is an enhancement of the bacterial dispersion coefficient. Simulations indicated that the dispersion coefficient could be modeled in the form DB gamma =DBgammaI+( alpha+betasigma)·vgamma where DBgamma is the bulk diffusion coefficient, alpha is the hydrodynamic dispersivity vector, beta is a chemotactic dispersivity tensor, sigma is the dimensionless Chemotaxis number and vgamma is the fluid velocity.