Modeling Colloid‐Facilitated Contaminant Transport in the Vadose Zone

Abstract

Subsurface colloids can enhance the movement of strongly sorbing contaminants, a phenomenon called colloid‐facilitated contaminant transport. In the presence of mobile subsurface colloids, contaminants may move faster and farther than in the absence of colloids, thereby bypassing the filter and buffer capacity of soils and sediments. Fate and transport models neglecting colloid‐facilitated transport therefore often underpredict contaminant movement. Long‐term predictions of contaminant fate and transport as well as risk assessment rely on an accurate representation of subsurface processes, and in the case of strongly sorbing contaminants, need to consider mobile colloids as potential contaminant carriers. The purpose of this review is to discuss the current knowledge and recent developments of modeling colloid‐facilitated contaminant transport in the vadose zone. The main part of this review is devoted to the discussion of conceptual models used to describe colloid‐facilitated contaminant transport in the vadose zone and their mathematical implementation. Modeling of colloid‐facilitated contaminant transport involves various interactions, including colloid attachment to and detachment from the solid matrix and the air–water interface, contaminant adsorption to and desorption from colloids and transport with mobile colloids, and contaminant adsorption to and desorption from the solid matrix. Most of these processes in colloid‐facilitated contaminant transport models have been described by first‐ or second‐order kinetics. The unique feature of the vadose zone is the presence of an air phase, which affects colloid and contaminant transport in several ways. Colloids can be trapped in immobile water, strained in thin water films and in the smallest regions of the pore space, or attached to the air–water interface itself. The modeling of colloid‐facilitated contaminant transport in the vadose zone has mostly been theoretical, and tested only with column experiments; field applications are still lacking.