Potential artifacts in interpretation of differential breakthrough of colloids and dissolved tracers in the context of transport in a zero-valent iron permeable reactive barrier

Abstract

Many published studies have used visual comparison of the timing of peak breakthrough of colloids versus conservative dissolved tracers (hereafter referred to as dissolved tracers or tracers) in subsurface media to determine whether they are advected differently, and to elucidate the mechanisms of differential advection. This purely visual approach of determining differential advection may have artifacts, however, due to the attachment of colloids to subsurface media. The attachment of colloids to subsurface media may shift the colloidal peak breakthrough to earlier times, causing an apparent "faster" peak breakthrough of colloids relative to dissolve tracers even though the transport velocities for the colloids and the dissolved tracers may actually be equivalent. In this paper, a peak shift analysis was presented to illustrate the artifacts associated with the purely visual approach in determining differential advection, and to quantify the peak shift due to colloid attachment. This peak shift analysis was described within the context of microsphere and bromide transport within a zero-valent iron (ZVI) permeable reactive barrier (PRB) located in Fry Canyon, Utah. Application of the peak shift analysis to the field microsphere and bromide breakthrough data indicated that differential advection of the microspheres relative to the bromide occurred in the monitoring wells closest to the injection well in the PRB. It was hypothesized that the physical heterogeneity at the grain scale, presumably arising from differences in inter- versus intra-particle porosity, contributed to the differential advection of the microspheres versus the bromide in the PRB. The relative breakthrough (RB) of microspheres at different wells was inversely related to the ionic strength of ground water at these wells, in agreement with numerous studies showing that colloid attachment is directly related to solution ionic strength.