THE ROLE OF SORPTION IN THE TRANSPORT OF KLEBSIELLA-OXYTOCA THROUGH SATURATED SILICA SAND

Experiments were conducted to investigate the role of sorption during the transport of the bacterium Klebsiella oxytoca through saturated si lica sand. The sorption process was visualized at the pore scale in a minicell (3 mm deep X 6.0 mm wide X 7.0 cm long) using scanning confoc al laser microscopy. The sorption process was also studied by conducti ng column experiments at three scales (3.8, 10, and 40 cm long), Resul ts of image analyses of the sorbed and unattached cells in pore throat s and the bacterial breakthrough data from the column experiments exhi bited similar trends. Breakthrough peaks were attenuated with respect to the input concentrations and well-defined tailing was observed. Vis ualization suggested that the sorption process was dominated by revers ible and irreversible sorption (k(irr)). In the case of reversible sor ption, the rate of forward sorption (k(f)) was different from the rate for reversible sorption (k(r)). Visualization also showed that the ba cterial coverage on the sand grains, although extensive, covered < 0.5 % of the available surface area. A 1D solution for advective-dispersiv e transport was used to estimate k(irr), k(f), and k(r) with appropria te values for the coefficient of hydrodynamic dispersion and average l inear pore-water velocity (determined from Cl data). Simulated best fi ts to the bacterial peaks were good for the 3.8 cm columns but underes timated peak heights in the 10 and 40 cm columns by one order of magni tude. Best-fit k(irr) values decreased with increasing scale (0.6, 0.1 3, and 0.062 hr(-1) for the 3.8, 10, and 40 cm columns, respectively) and showed that a k(irr) value determined at one scale cannot be used to determine concentrations of K. oxytoca with time at another scale. These results suggested that k(irr) was a function of t(o) (length of column over velocity). The equivalent irreversible sorption parameter (A, where A = t(o) . k(irr)) was a constant (mean value of 3.36) for t he three scales investigated. This observation suggested that the use of the value A, determined at one scale of investigation, may prove ef fective in approximating the value of k(irr) predicting bacterial tran sport at other scales. Best-fit determinations yielded the same k(f) a nd k(r) values at all three scales (0.1 and 0.02 hr(-1)). This suggest ed that reversible sorption may be independent of column length. This study emphasized the need for more comprehensive investigations of the role of sorption in the transport of microorganisms in the subsurface .