Evaluation of Monte Carlo simulation of photon counting efficiency for germanium detectors.

The reliability of calculating the full-energy peak counting efficiency of germanium detectors by Monte Carlo simulation was evaluated by comparing MCNP-4 code results with measurements over a range of conditions. The measurements were performed for two detectors of 20% and 110% nominal efficiencies with a point source at two distances and with four different volume sources, including a reentrant beaker, placed on the end caps. The radionuclides were National Institute of Standards and Technology standard sources that emit photons between energies of 42.8 and 1,596.4 keV. Each detector was modeled in detail with respect to the dimensions of the detection volume and attenuation material in the can, dead layer, and crystal holder foil. The measurements of the 42.8-keV photon were used to check dimensions and to change slightly the value of the dead-layer thickness so that the simulation agreed with measurements at this energy. After this adjustment, the average ratio of simulation results to measurements for 13 photon energies above 42.8 keV was between 0.97 and 1.03 for all sets of point and volume source comparisons. Ratios at individual energies were between 0.92 and 1.06 for the point source and between 0.94 and 1.09 for volume sources. The observed differences were consistent with the estimated standard deviations of simulation and of measurement, which ranged from 1 to 6% and typically were 2 to 3% except for larger uncertainties at low energies. On the basis of this comparison, simulation with the MCNP-4 code is equivalent to measurement with photon standards if the detector and source configuration can be modeled within a tenth of a millimeter.