Applicability of the Linear Sorption Isotherm Model to Represent Contaminant Transport Processes in Site Wide Performance Assessments

Abstract

The estimation of flux of contaminants through the vadose zone to the groundwater under varying geologic, hydrologic, and chemical conditions is key to making technically credible and sound decisions regarding soil site characterization and remediation, single-shell tank retrieval, and waste site closures (DOE 2000). One of the principal needs identified in the science and technology roadmap (DOE 2000) is the need to improve the conceptual and numerical models that describe the location of contaminants today, and to provide the basis for forecasting future movement of contaminants on both site-specific and site-wide scales. The State of Knowledge (DOE 1999) and Preliminary Concepts documents describe the importance of geochemical processes on the transport of contaminants through the Vadose Zone. These processes have been identified in the international list of Features, Events, and Processes (FEPs) (NEA 2000) and included in the list of FEPS currently being developed for Hanford Site assessments (Soler et al. 2001). The current vision for Hanford site-wide cumulative risk assessments as performed using the System Assessment Capability (SAC) is to represent contaminant adsorption using the linear isotherm (empirical distribution coefficient, K{sub d}) sorption model. Integration Project Expert Panel (PEP) comments indicate that work is required to adequately justify the applicability of the linear sorption model, and to identify and defend the range of K{sub d} values that are adopted for assessments. The work plans developed for the Science and Technology (S&T) efforts, SAC, and the Core Projects must answer directly the question of ''Is there a scientific basis for the application of the linear sorption isotherm model to the complex wastes of the Hanford Site?'' This paper is intended to address these issues. The reason that well documented justification is required for using the linear sorption (K{sub d}) model is that this approach is strictly empirical and is often applicable only under a limited range of physical-chemical conditions. As a result, K{sub d} values can be applied with confidence only to conditions under which the linear adsorption isotherm has been demonstrated to be applicable. If the sediment/soil mineralogy or physical properties, solution chemistry, or contaminant loading/concentration of the system to be modeled is significantly different from that for which the K{sub d} values were determined, significant error could potentially occur in the estimated transport rates. This is because many factors can affect the degree to which a particular contaminant adsorbs to a particular sediment or soil. These factors include: sediment mineralogy and surface area, major ion concentration (complexation and competitive adsorption), pH of the solution, and concentration of the adsorbate in solution and on the adsorbent. Another important variable that can affect adsorption is kinetics. If the contact time between the contaminant in solution and the sediment is limited by hydrologic factors, equilibrium may not be attained. In this case, modeling equilibrium K{sub d} values will overestimate the degree of adsorption. In some cases careful application of currently available geochemical knowledge can often significantly reduce the number of variables that must be considered for evaluating K{sub d} values for each particular contaminant.