Effect of fluid velocity on model-estimated rates of radial solute diffusion in a cylindrical macropore column

Solute transfer between regions of high and low permeability will limit rates of aquifer remediation in the field and can also affect solute transport through laboratory columns. Transport in systems with such rate limitations is often modeled by assuming that advective transport occurs only in the high-permeability region (mobile domain) and that solute movement within the lower-permeability region (immobile domain) is via diffusion only. The validity of such models was explored by experimentally observing the influence of pore water velocity on the apparent rates of transfer within a geometrically well defined, two-domain system. Tritiated water tracer (³H₂O) experiments were conducted in a 7-cm soil column with a central cylindrical zone of highly permeable material surrounded by an annulus of relatively impermeable finer-grained sediment. Experimental pore water velocities spanned over two orders of magnitude and included ́long" and "short" solute input pulse lengths. The breakthrough data were interpreted using a first-order mass transfer model (FOM) and a Fickian pore-diffusion model (PDM). As expected, the rate parameters fitted to the PDM were less sensitive to pore water velocity than those estimated using the FOM. Nonetheless, parameters for the PDM exhibited a negative correlation with pore water velocity. Reestimated parameters from experiments of Young and Ball (1998) showed similar scaling. The velocity dependency indicates deficiencies in inherent model assumptions of both the FOM and the PDM. For the PDM, reasons for the velocity dependency may include mass transfer limitations in the mobile domain and/or heterogeneity of diffusion rates within the immobile domain.