Modifying the MDA Calculation to Include Anticipated Residual Contamination
A.E. Desrosiers (Bartlett Services, Inc.)
The equation commonly used to calculate a priori estimates of the minimum detectable activity (MDA) assumes that the sample count rate is equal to the background count rate. This assumption produces an accurate estimate of the quantity of radioactivity that can be distinguished from cosmic, terrestrial, and instrument background. The expected frequency of false positive and false negative results are based on Poisson statistics. Typical applications for using the MDA equation include verification of adequate sensitivity for free release surveys, design of environmental surveillance programs, and development of final status survey programs in support of decommissioning. However there is oftentimes a fundamental difference in the objectives of free release surveys and decommissioning surveys. For many licensed users of radioactive materials, there is no allowable limit of contamination for released materials. In contrast to environmental release specifications, which allow trivial, but measurable, amounts of radioactive materials in effluents, and therefore in environmental media, current NRC regulations restrict certain licensees to from releasing any non-zero levels of contamination in cleared solid materials. The decommissioning regulations, nevertheless allow measurable contamination in soil and on building surfaces, after satisfying ALARA and dose assessment analyses. The limits for these residual limits in decommissioning criteria are determined according to risk based dose scenario calculations. Logically then, the presence of residual contamination in soil or on building surfaces will result in measurements with higher variances than would be observed if the soil and surfaces were at background levels. The higher variance effectively lowers the sensitivity and increases the minimum detectable activity (MDA) of the measurements, all other factors being equal. In the Multi-Agency Radiation Survey and Site Investigation Manual (MARRSIM), this situation is addressed by developing a procedure for discounting the MDA. The MARRSIM process requires a prior estimate of the variance of the residual activity. If the final status survey will be performed after remediation, the variance of the residual activity must be estimated. This paper explains a method for calculating an a priori estimate of the variance that would be experienced during the release survey. This method is similar to the current generally accepted method for calculating MDAs, but an additional term is added to account for the contribution of the residual activity. The use of an a priori method simplifies the process of planning final status surveys and clearance surveys. An additional innovative feature of the proposed method is a comparison the actual variance of the distribution of measurements in each item to the a priori estimates of variance that are used to calculate the minimum counting time. If the actual variance significantly exceeds the a priori estimate such that the observed mean concentration of residual contamination cannot be distinguished from the desired limits with sufficient reliability, the entire item can be recounted with longer counting times.