Evaluation of the prediction uncertainty in tritium production based on results from experiments on an Li₂O annular blanket surrounding a 14 MeV simulated line source

Abstract

The experimental analysis of phase III experiments of the USDOE/JAERI collaborative Program on Fusion Neutronics is described. The annular Li₂O test assembly totally surrounds a simulated 14 MeV line source of length 2 m. This simulation is achieved by cyclic movement of the assembly relative to the stationary 14 MeV point source. Three phases of experiments were performed and analyzed. In phase IIIA, a stainless steel first wall 1.5 cm thick was used. An additional carbon layer 2.45 cm thick was added in phase IIIB, and a large opening (42.55 cm × 37.6 cm) was made at one side of the annular assembly in phase IIIC. Calculations were performed independently by the USDOE and JAERI for many measured items that included the tritium production rate (TPR) from ⁶Li(T₆), ⁷Li(T₇); in-system spectrum measurements; and various activation measurements. In this paper, the calculated-to-measured values (C/E) for the TPR were used to derive qualitative estimates to the mean values of the prediction uncertainty in the line-integrated TPR, and the possible spread around these values. Distinction was made between results based on the discrete ordinates and on the Monte Carlo method. Furthermore, a comparison was made with the pertaining results when all the experiments conducted during the USDOE/JAERI collaborative program were considered.