The examination of source distribution in a large sample by Monte Carlo simulation Daniela Gurau a,n , Octavian Sima b a Horia Hulubei National Institute for Physics and Nuclear Engineering, P.O.B. MG-6, 077125 Magurele, Romania b Physics Department, University of Bucharest, P.O.B. MG-11, 077125 Magurele, Romania article info Available online 27 February 2012 Keywords: Gamma-ray spectrometry Waste drums Monte Carlo simulation Spatial distribution abstract In this work, realistic Monte Carlo simulations were carried out for several distributions of activity in a waste drum to observe the dependence of the efficiency on the source distribution and to test whether the efficiency can be correlated with the shape of the spectra. & 2012 Published by Elsevier Ltd. 1. Introduction A particular problem in gamma-ray spectrometry measure- ment of large samples, such as 220l waste drums, is the source distribution variability in the volume. In some cases the actual source can be approximated by a source uniformly distributed in the volume (Radu and Sima, 2010), in other cases the source can be localized in a smaller domain within the volume. Due to the strong variation of the efficiency in function of the source position, the estimated activity may have a large error if the actual source distribution differs strongly from the assumed distribution. Therefore, any information concerning the source distribution can be useful in the assessment of the activity of such type of samples. In this study, we test by the Monte Carlo simulation whether the shape of a spectrum is correlated with the location and distribution of the source and furthermore whether the shape can provide information on the efficiency. 2. Methodology In order to study the dependence of the spectra on the distribution of the source in a waste drum of length L ¼ 880 mm and radius R ¼ 285 mm, we considered the drum volume divided into a number n of smaller volumes that were each independently submitted to the Monte Carlo simulations. The volume elements were defined (Fig. 1) by applying two divisions of the drum volume (Gurau and Sima, 2012). The Monte Carlo simulations were carried out using GEANT 3.21 code (Brun et al., 1987). The details of the detector geometry (p-type high-purity germanium with 25% relative efficiency), of the source (dimensions, composition) and measurement config- uration (source position, collimator characteristics) as well as all the component materials were implemented in the geometry and the material routines of the code. The source matrix considered was concrete with standard composition encased in an iron container with 1.25 mm wall thickness. The distance from the center of the coordinate system associated with the detector to the center of the drum was 500 mm (Gurau and Sima, 2012). In the first stage of the simulations, a random emission point was sampled in one of the intersections of the S1,y,S8 and the V1,y,V5 domains selected for the current computation, i.e. a uniform distribution was assumed in 1 of the 40 domains. For each location of the source the simulations were carried out for the gamma rays of 60 Co (1173.2 and 1332.5 keV) (one of the nuclides of interest in radioactive waste assay) and for the main photons emitted by 152 Eu (which is used for drum calibration using the shell sources approach (Bruggeman et al., 1999)). For each energy 1.2  10 10 photons were traced in total for the 40 volume elements. In the second stage of the simulations, point sources were considered. One set of simulations included point sources placed on the detector axis (that is perpendicular to the axis of the waste drum) at several distances from the detector, such that the photon path lengths in the drum were 0.12, 5.12, 10.12, 15.12, 20.12 and 25.12 cm. Another case was for a point source located off axis of the detector, on a line inclined with 301 with respect to the axis; the position of the source was chosen such that the path length through the drum was also 25.12 cm, but the distance to the detector was larger. In each case the simulations were carried Contents lists available at SciVerse ScienceDirect journal homepage: www.elsevier.com/locate/apradiso Applied Radiation and Isotopes 0969-8043/$ - see front matter & 2012 Published by Elsevier Ltd. doi:10.1016/j.apradiso.2012.02.075 n Corresponding author. E-mail address: daniela.gurau@ymail.com (D. Gurau). Applied Radiation and Isotopes 70 (2012) 2141–2143