RECOMMENDATIONS FOR ASSESSING THE UNCERTAINTY IN TANK 18-F WALL SAMPLES
Tank 18-F in the F-Area Tank Farm at the Savannah River Site (SRS) has had measurements taken from its inner vertical sides in order to determine the level of radionuclide and other analyte concentrations attached to the tank walls. In all, three samples have been obtained by drilling shallow holes into the carbon steel walls and consolidating the material. An Upper Wall Sample (Sample ID: Tk 18-1) was formed by combining two drill samples taken at a height of 17 ft above the tank floor, and a Lower Wall Sample (Sample ID: SPD4) was formed by combining two drill samples taken between 10 and 12 ft above the tank floor. A Scale Sample (Sample ID: Tk 18-2) was formed by combining 5 drill samples obtained between 6 and 7 ft above the tank floor. Photographs of the sampled material and a more detailed description of the samples and the concentration results are presented by Hay and others [2009]. The objective of this report is to determine a method and use it to place an upper confidence bound on the concentrations in the wall samples using only the currently available sample information. None of the three wall locations (tank heights) has been measured more than once. For radionuclides, only the variation among the concentrations per unit mass (g) of the wall samples, ignoring locations, or the variation among the concentrations of the floor samples are possibilities for establishing an upper confidence bound. The wall samples and floor samples were examined for comparability by (a) observing whether the wall sample concentrations fell inside the footprints created by prediction intervals for floor sample radionuclide concentrations and (b) whether the variation among the wall samples was approximately the same as the variation among floor samples. Most of the radionuclide concentrations satisfied (a) but the variation among radionuclide concentrations (b) was smaller for the floor samples. Consequently, upper 95% confidence bounds were established separately for radionuclide concentrations at each of the sampled tank heights using the conservatively estimated variation among the wall samples. A final step to convert concentrations by unit mass (g) to concentrations by sq ft was performed for the Upper Wall Sample and the Lower Wall Sample regions of the tank wall. The Upper Wall Sample and the Lower Wall Sample were not measured for elemental constituents. Consequently, the only possibility for establishing an upper bound for nonradionuclide concentrations for the Scale Sample was using the concentrations from floor samples. However, most non-radionuclide wall concentrations failed to fall within the footprint generated prediction intervals based on the non-radionuclide concentrations for the floor samples. The report concludes that there is no way to establish upper confidence bounds for elemental constituents attached to the inner liner of Tank 18-F based on currently available data.
- Research Organization:
- Savannah River Site (SRS), Aiken, SC (United States)
- Sponsoring Organization:
- USDOE
- DOE Contract Number:
- DE-AC09-08SR22470
- OSTI ID:
- 1001174
- Report Number(s):
- SRNL-STI-2010-00525; TRN: US201102%%412
- Country of Publication:
- United States
- Language:
- English
Similar Records
RECOMMENDATIONS FOR SAMPLING OF TANK 18 IN F TANK FARM
Hydrology of the low-level radioactive-solid-waste burial site and vicinity near Barnwell, South Carolina