Methods for Reducing Exposure to Personnel Leak-testing Sealed Radium Sources

Abstract

In a previous report, we proposed a group test procedure that would minimize exposure to personnel leak-testing sealed radium sources by reducing the required number of tests. The procedure was based on two assumptions: (a) there will be at most a single leaking source in any given test group, and (b) the resulting exposure is directly proportional to the number of leak tests performed. Whereas the validity of the first assumption finds its basis in leak test data available in the literature, the requirement of the second assumption may not be readily satisfied.In the present work, this requirement is lifted and consideration is given to the more realistic situation where the exposure is directly proportional to the activity of the group being tested. This leads to a group leak-testing procedure which in many instances results in a considerable reduction in exposure.Consideration is also given to storage arrangements and it is shown that, given an inventory of sealed radium sources, minimum exposure during leak-testing results when the distribution of sources among drawers is as alike as possible with respect to the number, arrangement, and activity of the sources per drawer-assuming use of our recommended leak-testing procedure

Group-testing, halving procedures, and binary search∗

“…is given by (a)), but did not explicitly examine the numerical value of this uncertainty or its implications; this was done subsequently by Thomas et al (1975) and Pasternack et al (1976).…”

Group-Sequential Leak-Testing of Sealed Radium Sources

“…Furthermore, this application includes the concept of permitting the preassignment of a minimum probability to the correct classification of all units by group-testing methods, a notion which first appeared implicitly in [18] and subsequently explicitly in [14,15,19]. This probability can be regarded as a joint confidence level.…”

“…Thomas et al [19] call the overall strategy, group plus sequential testing when the screen test turns up positive, the RGS procedure. By assuming an underlying conditional binomial model they are able to evaluate and compare the RGS procedure with other test methods.…”

“…(It is of interest to note that an alternative interpretation for P* is given in [19]. There the authors indicate that the expected exposure for any test procedure conditioned on there being at most one leaker in the group, based on the conditional binomial model with d = 1 as above, can be regarded as an estimate of the posterior expectation given an upper bound for the actual number of leakers present.…”

“…The primary purpose of this paper is to reevaluate the R GS and Rs procedures using an unconditional binomial model (d = n) and to compare the results with that obtained in [19] where the conditional binomial model (d = 1) is used.…”