Uncertainty in Multi-Pathway Risk Assessment for Combustion Facilities

摘要

Multi-pathway risk assessments (MPRAs) of contaminant emissions to the atmosphere consider both direct exposures, via ambient air, and indirect exposures, via deposition to land and water. MPRAs embody numerous interconnected models and parameters. Concatenation of many multiplicative and incompletely defined assumptions and inputs can result in risk estimates with considerable uncertainties, which are difficult to quantify and elucidate. Here, three MPRA case-studies approach uncertainties in ways that better inform context-specific judgments of risk. In the first case, default values predicted implausibly large impacts; substitution of site-specific data within conservative methods resulted in reasonable and intuitive worst-case estimates. In the second, a simpler, clearly worst-case water quality model sufficed to demonstrate acceptable risks. In the third case, exposures were intentionally and transparently overestimated. Choices made within particular MPRAs depend on availability of data as suitable replacements for default assumptions, regulatory requirements, and thoughtful consideration of the concerns of interested stakeholders. Explicit consideration of the biases inherent in each risk assessment lends greater credibility to the assessment results, and can form the bases for evidence-based decision-making.