Identification of artificial gamma-emitting nuclides using a scintillator-based gamma-ray spectral logging system

The standard spectral gamma ray logging system used at the Nevada Test Site (NTS) contains a large sodium iodide detector which provides high efficiency but low energy resolution. To enhance the capabilities of this system for identifying artificial gamma-emitting nuclides, I developed and implemented a simple procedure for extracting artificial components from low-resolution gamma-ray spectra. This procedures uses three bias spectra, developed by a consultant using numerical modeling, representing the spectral response of the downhole instrument to naturally occurring potassium, uranium and thorium family gamma rays in a large-diameter air-filled borehole. To extract the artificial spectral components, the three basis spectra are first scaled to the recorded field spectrum using the usual spectral windows for K, U and Th; these windows bracket the photopeaks at 1.46, 1.76 and 2.61 MeV. Since most of the contribution from artificial nuclides will fall below 1.26 MeV (the lower limit of the potassium window) this scaling process should be insensitive to the presence of artificial nuclides. The scaled basis spectra are then subtracted from the field data, leaving a residual spectrum consisting of noise plus the contribution of any artificial gamma-emitting nuclides. This process is repeated for each spectrum in the log, or the spectra can be accumulated over any desired depth range for better statistics. Rather than inspect each spectrum visually, a parameter can be computed which indicates the presence of artificial nuclides; this parameter can be plotted along with the usual K, U and Th concentration estimates as a function of depth. These techniques have been used successfully on field data and provide us with an inexpensive screening tool to detect artificial nuclides along boreholes. 11 refs., 6 figs