A science team led by researchers at Rutgers University discovered a new tool for removing contaminants from water. Tiny glowing crystals designed...
In a project to develop a risk-management model for the cleanup of contaminated groundwater sites, researchers at the University of Illinois at Urbana-Champaign have heightened the role of uncertainty in risk-based remediation.
"Many states now allow site-specific, risk-based criteria to be used in lieu of fixed drinking-water standards," said Barbara Minsker, a UI professor of civil and environmental engineering. "As we move toward using risk as the criteria for how we clean up contaminated sites, it is very important that we consider uncertainty. By evaluating tradeoffs among cost and risk under conditionsof uncertainty, our model will help decision makers make better decisions."
Minsker's risk-management model combines a genetic algorithm with a fate and transport simulation model and a risk assessment module to identify potention remediation designs. The generic algorithm searches the decision space for remediation designs that best meet a specified management objective, such as minimizing cost and risk. The simulation model and risk-assessment module are used to predict the risk associated with candidate remediation designs.
Currently, the model can evaluate tradeoffs between cost and risk without considering uncertainty, or it can minimize cost given a specified maximum risk level nder conditions of uncertainty.Ultimately, the model will allow tradeoffs among cost, risk and cleanup time to be considered under uncertainty during the remediation design process.
"With such information readily available, negotiations can focus on design issues that have the most impact on the cost and effectiveness of the remediation."
Minsker is presenting the model during a symposium on scientific uncertainty and risk management at the American Chemical Society meeting held Aug. 23-24 in Washington, D.C. The National Science Foundation and the U.S. Army Research Office are supporting development of the model.