Double sampling as a cost-effective method to estimate mean or total amounts of radioactivity

In many field studies to determine the quantities of radioactivity at commercial low-level radioactive waste sites, preliminary appraisals are made with field radiation detectors, or other relatively inaccurate devices. More accurate determinations are subsequently made with procedures requiring chemical separations or other expensive analyses. Costs of these laboratory determinations are often large, so that adequate sampling may not be achieved due to budget limitations. In this paper we propose double sampling as a way to combine the expensive and inexpensive approaches to substantially reduce overall costs. The underlying theory was developed for human and agricultural surveys, and is partially based on assumptions that are not appropriate for commercial low-level waste sites. Consequently, extensive computer simulations must be conducted to determine whether the results can be applied in circumstances of importance to the Nuclear Regulatory Commission. Our initial simulation results lead us to believe that double sampling may be useful for some surveys at commercial low-level waste sites. Additional simulations are currently being evaluated using additional models and the range of parameter values likely to be encountered in field studies