Detector resolution required for accurate identification in common gamma-ray masking situations

Abstract

Accurate nuclide identification depends on the ability to determine if specific peaks are present in the spectrum. Several current handheld nuclide identifiers and portal monitors use a variant of a peak quality value for this. The peak quality is usually calculated as the peak area divided by the uncertainty of the peak area and when this quotient is above a threshold value, the peak is said to be present. Other works [Terracol et al. In: 2004 IEEE Nuclear Science Symposium Conference Record, Rome, Italy, 2004, Ryder In: Scanning Electron Microscopy/1977 V. 1, Proceedings of the Workshop on Analytical Electron Microscopy, Chicago, 1977] have developed a formalism to calculate the peak uncertainty for interfering peaks based on the detector resolution, background, individual peak areas, and peak separation. The threshold on peak uncertainty determines the minimum activity that will be identified or detected. Care must be used in the selection of the threshold in order to comply with the false positive and false negative requirements of the detection system regime, or “concept of operations”. The performance standards for the handheld identifiers and portal monitors specify the nuclides required to be identified. From this list and other commonly expected nuclides, the energies of the expected gamma rays can be tallied, yielding a table of the separations of adjacent peaks possible in the collected spectrum. Using the formalism, the peak quality value can be determined as a function of the detector resolution, peak area and background for the energy separations in the table determined above. Results are shown for the cases of HEU and plutonium with the masking nuclides of NORM, 133Ba, or 57Co for both germanium and sodium iodide detectors. Typical resolutions, efficiencies and counting times were used.