Skip to main content
U.S. Department of Energy
Office of Scientific and Technical Information

Doing a conservative risk assessment by doing your best probabilistic assessment

Conference ·
OSTI ID:354292
 [1];  [2];  [3]
  1. Gradient Corp., Houston, TX (United States)
  2. Sandia National Labs., Albuquerque, NM (United States)
  3. Intera-Albuquerque, NM (United States)

Traditional human health risk assessments combine average, conservative, and worst-case values to derive a point estimate of risk that is presumed to be protective of public health. This approach is based on the belief that intentionally large estimates of risk can adequately account for uncertainties in risk parameters and that decisions based on such estimates will protect the public despite the uncertainty. Computer-based risk assessment models provide a way to derive realistic risk estimates while still maintaining conservatism. This is especially true if the models contain two features: (1) multiple pathway analysis, wherein contaminant mass is conserved and the distribution of mass between pathways is internally consistent, and (2) probabilistic methods for translating uncertainty in model input parameters into uncertainty in model output. Probabilistic risk assessments have been conducted with the Precis model developed by Sandia National Laboratories/New Mexico (SNL/NM). A probabilistic approach can use the best data available, including their uncertainty, to produce objective estimates of cleanup goals and their associated uncertainty. This information can be used to identify the major sources of risk, to focus remediation efforts, and to support risk management decisions.

OSTI ID:
354292
Report Number(s):
CONF-970677--
Country of Publication:
United States
Language:
English