Answer to Question #6175 Submitted to "Ask the Experts"

Thank you for your question concerning the differences between decommissioning survey guidance found in NUREG/CR-5849 (1992) and the MultiAgency Radiation Survey and Site Investigation Manual (MARSSIM 2000). First, it is important to recognize that the MARSSIM has superseded NUREG/CR-5849, so MARSSIM would most likely be the proper survey guidance to use. That being said, a number of ongoing decommissioning and decontamination (D&D) projects are using NUREG/CR-5849. One possible explanation for this exception is that their decommissioning plan may have been approved several or more years ago, so they are in effect "grandfathered" to use NUREG/CR-5849.

There are a number of differences between NUREG/CR-5849 and MARSSIM. I'll mention just a few differences here, and note that Decommissioning Health Physics (Abelquist 2001) discusses this matter over the course of several pages in Chapter 12.
                   
In general, I would sum up the differences by saying the NUREG/CR-5849 approach is prescriptive, while the MARSSIM approach is flexible, given that it is based on the DQO (data quality objectives) Process for data collection. For example, NUREG/CR-5849 provides a prescriptive survey approach to determining the sample size for both direct measurements and soil samples. 

The number of samples required in a particular survey unit is only based on its classification—either affected or unaffected. The MARSSIM approach to sample size is based on the contaminant derived concentration guideline level (DCGL), the expected standard deviation of the contaminant in background and in the survey unit, and the acceptable probability of making Type I and Type II decision errors. 

And perhaps most discomforting to the MARSSIM user is that the sample size may be ratcheted higher depending on the scanning instrument's ability (or rather inability) to detect hot spots that could be significant relative to regulatory limits.

A couple of related differences include:

1. How the site is divided into survey units based on contamination potential—The NUREG/CR-5849 guidance recommends that the site be divided into affected and unaffected areas for purposes of designing the final status survey; MARSSIM uses three classifications of contamination potential (Class 1, 2, and 3). In NUREG/CR-5849 affected area survey units are further divided into grid blocks (1 m² for structure surfaces and 100 m² for land areas). MARSSIM recommends suggested survey unit sizes that vary according to the classification of the survey unit (there is no further division into grid blocks . . . big plus for MARSSIM).
   
   
2. Statistical test used to demonstrate compliance with release criteria—NUREG/CR-5849 recommends using the 95 percent upper confidence level via the t-test to demonstrate that the survey data comply with release criteria. MARSSIM favors nonparametric statistical tests to demonstrate compliance; specifically MARSSIM recommends using hypothesis tests on the survey data, using either the Sign or Wilcoxon Rank Sum test.

Hope this helps.

Eric W. Abelquist, CHP

Reference
Abelquist EW. Decommissioning health physics—A handbook for MARSSIM users. Philadelphia: Institute of Physics Publishing; 2001.