You are here:
DEVELOPMENT AND EVALUATION OF AN INTEGRATED MODEL TO FACILITATE RISK-BASED CORRECTIVE ACTION AT SUPERFUND SITES
Impact/Purpose:
Risk-based corrective action (RBCA) involves decisions related to the need for, and extent of, clean-up at Superfund sites. Such decisions may begin with prescriptions of acceptable human health risks, which are then used to determine maximum allowable dose. Source and exposure models can then be used to determine clean-up requirements, i.e., how rapidly and to what extent a source must be reduced so that a maximum dose is not exceeded. However, there are currently no models which integrate knowledge related to chemical migration from soil to building environments, volatilization of chemicals from contaminated water to indoor air, human activity patterns and exposure, health risks, and corrective action requirements. The objectives of this research are to (1) develop a fully integrated risk model that can be used for RBCA at Superfund sites, and (2) evaluate the model based on unique data sets associated with field studies and a series of controlled laboratory experiments.
The overall objective of the proposed research is to develop a state-of-the-art model which can be used for advanced studies of human exposure to hazardous chemicals near Superfund sites. As RBCA is often driven by indoor exposures to volatile contaminants, the research proposed here will focus on such conditions. The fully integrated source and exposure model (FISEM) will be designed with several features that are intended to assist industrial staff and consultants who are involved with advanced modeling and monitoring studies of human exposures at Superfund sites, and policy-makers involved with important decisions regarding remediation requirements at such sites. In particular, FISEM will be both forward and backward executable, the latter allowing for the use of prescribed dose thresholds to estimate the need for, or extent, of soil and/or groundwater remediation. An important feature of the
Risk-based corrective action (RBCA) involves decisions related to the need for, and extent of, clean-up at Superfund sites. Such decisions may begin with prescriptions of acceptable human health risks, which are then used to determine maximum allowable dose. Source and exposure models can then be used to determine clean-up requirements, i.e., how rapidly and to what extent a source must be reduced so that a maximum dose is not exceeded. However, there are currently no models which integrate knowledge related to chemical migration from soil to building environments, volatilization of chemicals from contaminated water to indoor air, human activity patterns and exposure, health risks, and corrective action requirements. The objectives of this research are to (1) develop a fully integrated risk model that can be used for RBCA at Superfund sites, and (2) evaluate the model based on unique data sets associated with field studies and a series of controlled laboratory experiments.
The overall objective of the proposed research is to develop a state-of-the-art model which can be used for advanced studies of human exposure to hazardous chemicals near Superfund sites. As RBCA is often driven by indoor exposures to volatile contaminants, the research proposed here will focus on such conditions. The fully integrated source and exposure model (FISEM) will be designed with several features that are intended to assist industrial staff and consultants who are involved with advanced modeling and monitoring studies of human exposures at Superfund sites, and policy-makers involved with important decisions regarding remediation requirements at such sites. In particular, FISEM will be both forward and backward executable, the latter allowing for the use of prescribed dose thresholds to estimate the need for, or extent, of soil and/or groundwater remediation. An important feature of the
Description:
We developed a numerical model to predict chemical concentrations in indoor environments resulting from soil vapor intrusion and volatilization from groundwater. The model, which integrates new and existing algorithms for chemical fate and transport, was originally coded in FORTRAN and this version of the model was tested. An algorithm to account for human activity and water usage patterns is currently available as a user-prescribed input.Based on a previous progress report, a reviewer noted that it may be desirable to use the same programming language, i.e., C++, for all tasks. We agreed with this recommendation and converted the IAM-UT source code from FORTRAN to C++. Both codes were evaluated with identical input files to ensure consistency of results. The model was also evaluated with respect to several simple cases. These cases were chosen in order to yield analytical solutions, e.g., cases with two or three zones and a limited number of emission sources. The IAM-UT model was also tested for robustness of solution technique, i.e., solution stability and time-step sensitivity.We completed development of a general, analytical framework to calibrate the fate and transport model (Task 1) with field and laboratory data. This framework, named the Second Moment Bayesian Method (SMBM), allowed for efficient implementation of both experimental design and data analysis. Development included evaluating the accuracy of analytical approximations within the framework using numerical solutions, and coding the framework into a C++ program that can be linked with the numerical fate and transport model. We completed development of the source codes calculating first and second derivatives of all model variables. We linked these codes with a general likelihood function developed previously. The integrated IAM-UT/SMBM model allowed for analysis of data collected in Task 3. Two residential test houses near a site characterized by sub-surface contamination in Paulsboro, New Jersey were selected for this study. Two major components of the soil gas to indoor air exposure pathway were studied: (1) chemical migration through the basement foundation and (2) ventilation rate from basement to the remainder of the house. A total of 12 experiments were completed (7 basement foundation experiments and 5 basement ventilation experiments). All experiments were completed using sulfur hexafluoride (SF6) as an inert tracer gas. In addition, several VOC samples were taken during the basement foundation experiments using sorbent tubes with subsequent GC analysis. The following parameters were also measured on a continuous basis during each experiment: carbon dioxide concentration, temperature, relative humidity, and differential pressure.Soil gas entry rates into the basement were location-dependent but generally increased with increased soil-basement pressure differences. Soil-basement pressure differences for these experiments ranged from -0.2 to -6.2 Pa (relative to basement). Basement ventilation rates ranged from 0.17 to 0.75 ACH, with basement-ambient pressure differences ranging from -1.1 to -7.6 Pa (relative to ambient).The data collected for this Task was analyzed within a probabilistic framework using the integrated IAM-UT/SMBM model described in Task 2. The updated means and variances for all model parameters were determined based on the maximum likelihood described in Task 2.