Saxton: Soil Survey I
SMCM Soil Survey
March 2003
The Saxton Nuclear Experimental Corporation (SNEC) Facility located in Saxton, Pennsylvania accumulated approximately 11,200 tons of building debris (primarily concrete) and approximately 5,000 tons of soils that would be suitable for use as fill for excavations that will remain following d econtamination and decommissioning (D&D). GPU Nuclear (GPU) determined that the concentration of man-made radionuclides in the materials would be measured prior to using the soil and debris for fill. Materials with concentrations above a fraction of the DCGLw stated in the License Termination Plan (LTP) (5.6 pCi/g) were separated from the material used for fill.
GPU used a system of conveyors and radiation monitors, along with sampling and laboratory analysis, to document that the material is substantially below limits proscribed in the facility LTP filed with the Nuclear Regulatory Commission (NRC). Shonka Research Associates, Inc (SRA) was contracted to build, operate, and summarize data from the radiation monitors.
The conveyor system was retrofit with the SMCM which is a complement of four sodium iodide radiation detectors mounted in an enclosure that was placed above a section of the conveyor. The detectors were operated as energy spectrometers. The system measured the concentration of cesium (Cs-137) as well as natural potassium (K-40), uranium (via Bi-214), and thorium (via Tl-208).
The GPU survey request number 55 (SR-55) comprised 38 batches that included over 11,200 tons of building debris. Based on the data recorded the system operated with a minimum detectable contamination (MDC) of 0.36 pCi/g Cs-137 on a batch basis assuming a uniformly distributed contamination. The system operated with an alarm setpoint of 2.91 pCi/g Cs-137 on a 5-acquisition (25-second) basis assuming a uniformly distributed contamination within 7.3 cubic feet of material. No alarms occurred during the survey. The SMCM net potassium, uranium, and thorium showed good agreement with the soil samples, which were collected and counted independently by GPU. Thirty-one ten-hour days were required to survey the 38 each SR-55 batches.
The GPU SR-62 survey comprised 18 batches that included over 5,000 tons of soil. Twenty-eight alarms occurred during the survey. The system operated with an alarm setpoint of 2.91 pCi/g Cs-137 on a 5-acquisition (25-second) basis assuming a uniformly distributed contamination. The SMCM net potassium, uranium, and thorium showed good agreement with the soil samples, which were collected and counted independently by GPU. Twenty-six 10-hour days were required to survey the 18 SR-62 batches.
Total costs for this survey were an order of magnitude lower than DOE’s experience with competing conveyorized soil counting systems.
