Methods and programmes of γ spectrum analysis. Intercomparison results

Laboratoire de Metrologie des Rayonnements Ionisants organised a nation-wide interlaboratory test - P.T.I. R 6/1974 - on gamma spectrum analysis. This intercomparison dealt with the investigation of total absorption peaks in a complex spectrum and the calculation of their main parameters: centroids and surface area. Eleven laboratories participated in this intercomparison, proposing one or more methods of analysis. The results sent in show that the peak investigation is the result of a compromise between elimination of 'pseudo-peaks' and elimination of some actual peaks; a programmed automatic localisation is all the better as the rejection criteria are elaborated. The calculated centroids are generally credible as long as a developed analysis procedure is used, for example least squares adjustment of a mathematical model. On the other hand whatever the method adopted the calculated surface areas do not always coincide with the reference values. A special effort must be made to estimate surface errors, particularly an underestimation observed in the case of interferences or of peaks superimposed on a strong background