Re: STATISTICAL FREE RELEASE RAD SURVEYING TECHNIQUES.

[David_B_Dave_Ottley@rl.gov]

     Westinghouse Hanford Company's (WHC) material and equipment release 
     procedure currently allows statistical release as a time saving 
     technique.  It is not used very often because the technicians believe 
     it is more complicated than just doing a 100% survey.  The procedure 
     is based on Durham, J. S. and D. L. Gardner, "Contamination Surveys 
     for Release of Material" which was presented at the 1994 HPS midyear 
     meeting in Albany, NY.  Dr. Durham can be reached at (509) 373-0886 
     for more information.  The statistics of this paper were based on 
     Eger, E. J., "The Use of One-sided Tolerance Tests for Surveys During 
     Decontamination and Decommissioning."
     
     WHC is also removing radiological postings from outdoor contamination 
     areas based on statistical surveys.  This is not an unrestricted 
     release because the land is in a controlled area but once unposted, 
     the land could eventually be released.  This procedure is based on the 
     above mention documents and the DOE Draft Environmental Implementation 
     Guide for Radiological Survey Procedures, November 1992.  This 
     document is very similar to the NRC NUREG/CR-5849, "Manual for 
     Conducting Radiological Surveys in Support of License Termination," 
     June 1992.  James Berger at (615) 531-3669 is the author of the NRC 
     manual.  Federal agencies are interested in issuing one manual to 
     cover release.  It will be called the Multiagency Manual for 
     Environmental Radiological Surveys (MMERS).  My latest draft is
     May 1994.  The NRC has just issued NUREG-1505, "A Nonparametric 
     Statistical Methodology for the Design and Analysis of Final Status 
     Decommissioning Surveys," August 1995.  It seems to represent the 
     direction the agencies are going and is based on EPA 230-R-94-004, 
     "Statistical Methods for Evaluating the Attainment of Cleanup 
     Standards," Volume 3, December 1992.  This is based on the statistical 
     work of Dr. Richard Gilbert of PNL.
     
     We believe that our procedure is good for finding big problems like 
     spills or burial sites.  It is not good for finding small isolated 
     contamination like animal feces that have been brought into the area.  
     The procedure has been scrutinized by many reviewers.  The reviewers 
     have a hard time with the statistics because it goes against their 
     tradition of 100% survey for small items and I have a hard time 
     getting them to take the time to read the above referenced documents 
     and coming up to speed.
     
     I believe that the use of statistics is the only way to reasonably 
     release large items/areas.  I am very interested in this subject and 
     can be reached at (509) 376-8124 or David_B_Ottley@rl.gov.