• pebble-bed
• PBMR
• pebble
• HTR
• MCNP
• PANTHERMIX

The work described in this thesis has been performed mainly in the frame of the PUMA project. The aim of that project is to investigate HTGR manly Gen-IV nuclear reactors physics in order to improve their potential as Pu/MA transmuters. One of the main goals of this work is to find a better model for the PBMR core design by one of the codes used to study the problem of Pu/MA trasmuters: PANTHERMIX. This code couples PANTHER (a neutronic diffusion code in two energy groups) and DIREKT/THERMIX (a thermo-hydraulics code developed mainly for Helium heat transfer). Particularly this research has the target to obtain a solution for the so-called “recirculation problem” for the PANTHEMIX recirculation system: in PBMR-400 there is a recirculation system that reinserts the not-exhausted pebbles on the top of the core. The presence of deep burnt pebbles (up to 600 GWd/t) leads to a nonlinear behaviour for cross-sections and local k-infinitive. For this reason PANTHERMIX needs a new model for describing macroscopic cross-sections vs. burn-up in order to correct the older values.
The preparatory work (literature study of HTR and its fuel, familiarization with MCNP code, etc.) was mainly performed at DIMNP of the University of Pisa and is briefly documented in the first chapters; the research activity was mainly developed in the Netherlands, at NRG (Petten) and TUD (Delft), obviously in continuous and strict collaboration with the DIMNP Group.
The first chapter gives a general description of the power plant (particularly of the reactor vessel and the recirculation system), with a strong emphasis on the intrinsic safety due to the neutronic stability and to the long duration in case of LOFC (loss of forced circulation). The second chapter is focused on the fuel and on the fuel cycle, one of the main targets of the PUMA project. Generally speaking, the use of some different fuel compositions is allowed by the flexibility of this reactor. Specifically for the present research we have considered 3 different types of pebbles. After the preliminary analyses performed to setup the new mixing model (Chapter 3), Chapter 4 gives a brief description of the codes used in the analyses: PANTHERMIX and MCNP5®. The first is the main object of the present analyses while the second is used for support calculations in order to obtain the correlations between the single pebble neutronic characteristics and a Mesh (or Mixing Cell) with different pebbles inside. Finally, in Chapter 5 the detailed description of the work is reported; particularly the original evolution of the Mixing Model is highlighted, having a look to the goal to obtain at least an acceptable grade of accuracy.
In conclusion, with the new model, PANTHERMIX can properly predict the behaviour of a pebble-bed core loaded with mixtures of pebbles at different burn-up values (even at burn-up higher than 600 GWd/t). The basic idea is to apply the superimposition effect method to create a sort of “mosaic” and to obtain a solution for all the different configurations. This mixing model works efficiently and it is able to predict the behaviour of various mixtures.
The work gives new ideas for setting-up further models and more generally for interesting future developments.