Secondary Logo

Journal Logo

Institutional members access full text with Ovid®

An Approach for Addressing Hard-to-Detect Hot Spots

Abelquist, Eric W.*; King, David A.*; Miller, Laurence F.; Viars, James A.*

doi: 10.1097/HP.0b013e3182812867

The Multi-Agency Radiation Survey and Site Investigation Manual (MARSSIM) survey approach is comprised of systematic random sampling coupled with radiation scanning to assess acceptability of potential hot spots. Hot spot identification for some radionuclides may not be possible due to the very weak gamma or x-ray radiation they emit—these hard-to-detect nuclides are unlikely to be identified by field scans. Similarly, scanning technology is not yet available for chemical contamination. For both hard-to-detect nuclides and chemical contamination, hot spots are only identified via volumetric sampling. The remedial investigation and cleanup of sites under the Comprehensive Environmental Response, Compensation, and Liability Act typically includes the collection of samples over relatively large exposure units, and concentration limits are applied assuming the contamination is more or less uniformly distributed. However, data collected from contaminated sites demonstrate contamination is often highly localized. These highly localized areas, or hot spots, will only be identified if sample densities are high or if the environmental characterization program happens to sample directly from the hot spot footprint. This paper describes a Bayesian approach for addressing hard-to-detect nuclides and chemical hot spots. The approach begins using available data (e.g., as collected using the standard approach) to predict the probability that an unacceptable hot spot is present somewhere in the exposure unit. This Bayesian approach may even be coupled with the graded sampling approach to optimize hot spot characterization. Once the investigator concludes that the presence of hot spots is likely, then the surveyor should use the data quality objectives process to generate an appropriate sample campaign that optimizes the identification of risk-relevant hot spots.

A Bayesian statistical approach, coupled with a graded sampling campaign, offers an effective solution for addressing risk-relevant, hard-to-detect hot spots.

*Oak Ridge Associated Universities, P.O. Box 117, MS-22, Oak Ridge, TN 37831; †The University of Tennessee, Knoxville, TN.

The authors declare no conflicts of interest.

©2013Health Physics Society