Neutron Spallation Studies for an Accelerator Driven Subcritical Reactor

Nuclear power production can benefit from the development of more comprehensive alternatives for dealing with long-term radioactive waste. One such alternative is an accelerator-driven subcritical reactor (ADSR) which has been proposed for both energy production and for burning radioactive waste. Here we investigate the effects of the size of the ADSR spallation target on the total neutron yield integrated over the neutron energy and emission angle. The contribution to the total neutron yield from the (n, xn) neutron interactions is evaluated at proton beam energies between 0.4 and 2 GeV. Calculations have been carried out with the GEANT4 simulation code using the Liege intranuclear cascade model and the results are compared to the the LAHET/MCNP code package predictions. INTRODUCTION Due to their inherent safety features and waste transmutation potential, accelerator driven subcritical reactors are the subject of research and development in many countries around the world. The ADSR consists of three parts: the accelerator, spallation neutron target and sub-critical reactor core. The spallation target is at the heart of any accelerator driven system. Because the ADSR is operated in a subcritical state, the target system has to provide the neutrons needed to sustain fission. These are generated by the spallation process resulting from high energy protons impacting the spallation target installed at the centre of the core. Therefore the target materials must have high neutron production efficiency. One of the best candidate target material is lead or a lead/bismuth eutectic. In the present work, the GEANT4 simulation code [1] has been used to calculate the total neutron yield for spallation reactions in Pb for proton energies between 0.4 and 2 GeV. The results of this work provide also a benchmark of the GEANT4 simulation results against the MCNP/LAHET (RAL model) code output published in reference [2]. COMPUTATION DETAILS GEANT4 provides an extensive set of hadronic physics models for energies up to 10 15 GeV, both for the intranuclear cascade region and for modelling of evaporation. In MCNP/MCNPX codes, the Bertini model is used by default for nucleons and pions, while the ISABEL model is used for other particle types [3]. The Bertini model does not take into account the nuclear structure effects in the inelastic interactions during the intranuclear cascade and therefore the code modelling of interactions at energies much below 100 MeV is questionable [2]. This becomes an important issue when dealing with thick targets, as although the primary neutrons are produced by the high energy proton beam, these are relatively low energy neutrons which can produce further spallation processes inside the target leading to secondary neutrons. For the Geant4 simulations, we selected the Liege intranuclear cascade model together with the independent evaporation/fission code ABLA. This model has been added recently to the Geant4 code and has been validated against experimental data for spallation processes in many different heavy elements [4]. This model is valid for proton, neutron, pion, deuteron and triton projectiles of energies up to 3 GeV and heavy target materials (Carbon Uranium). It models the Woods-Saxon nuclear potential, Coulomb barrier, non-uniform time-step, pion and delta decay cross sections, delta decay, Pauli blocking and utility functions, making it an independent code. The Liege model is largely free of parameters and is preferred by validation and, compared to the other theoretical models available in Geant4 (Binary and Bertini being currently the most widely used), it is more data driven [5]. The MCNPX data with which our GEANT4 simulations are compared have been taken from reference [2]. In all the simulations described below, a 60 cm long Pb target was considered. RESULTS Figure 1 shows the dependence of the number of neutrons on the target radius, both for neutrons produced and leaving the target volume. This is shown separately for the primary neutrons generated by the initial protons, secondary neutrons generated by the primary neutrons inside the target and the total number of neutron induced. It can be seen that the main “source” of neutrons inside the target is not the initial proton, but the primary neutron. The big difference in the number of produced neutrons and the number of neutrons leaving the target volume is due to the absorption of low energy generated neutrons. While the Bertini model does describe accurately the proton induced spallation processes, there is concern over its validity in describing the low-energy neutrons induced spallation. This model is the default MCNPX intra-nuclear cascade model for protons and neutrons, and has been chosen by the authors in reference [2]. The spallation neutrons enProceedings of PAC09, Vancouver, BC, Canada TU6PFP029 Applications of Accelerators U03 Transmutation and Power Generation 1351 TARGET DIAMETER (cm) 0 50 100 150 200 250 300 Y IE L D ( n /p ) 0 20 40 60 80 100 120 primary neutrons produced primary neutrons out of target secondary neutrons produced secondary neutrons out of target total neutrons produced total neutrons out of target Neutron spallation 1GeV p Pb target Figure 1: The variation of the calculated total neutron yield with the target radius (GEANT4). htemp Entries 148647 Mean 21.74 RMS 60.9 (MeV) n E 0 200 400 600 80