Abstract:

TINTE is a well established reactor analysis code which models the transient behaviour of pebble bed reactor cores but does not include the capabilities to model the direct cycle power conversion unit (PCU). This raises the issue that TINTE cannot model full system transients. One way to overcome this problem is to supply TINTE with time-dependant thermal-hydraulic boundary conditions which are obtained from PCU simulations.

This study investigates a method to provide boundary conditions for the nuclear code TINTE during full system transients. This was accomplished by creating a high level interface between the systems CFD code Flownex and TINTE. An indirect coupling method is explored whereby characteristics of the PCU are matched to characteristics of the nuclear core. This method eliminates the need to iterate between the two codes. The program Flownex-TINTE Interface (FTI) was created to perform data transfer and program synchronization.

All the simulations were performed on a 268MW PBMR core which is connected to a three-shaft Breyton-cycle plant. The validation study shows that the coupling method introduces errors when reproducing mass flow, temperature and pressure. These errors are relatively small in steady state analysis, but become more pronounced when dealing with fast thermal-hydraulic transients. Decreasing the maximum time step length in TINTE reduces this problem, but increases the computational time dramatically.

A number of transients are simulated using FTI and then compared against the stand-alone Flownex simulation. Load following is simulated to determine if the thermal-hydraulic response of the stand-alone Flownex model and the coupled code coincide. Positive results were obtained, but there were still some discrepancies due to differences in the thermal-hydraulic models.

A slow control rod withdrawal is also performed. The purpose of this simulation is to analyse the neutronic behaviour of the two models. It is shown that the behaviour of the Flownex V502 model is not exactly the same as that of the TINTE 268MW model. This is due to assumptions made in the Flownex point-kinetic model.

The main conclusion that can be drawn from this study is that the indirect coupling method can provide rough boundary conditions during transients when interfacing TINTE and Flownex. This was implemented in FTI which is now a practical tool to compare TINTE diffusion reactor models with Flownex point-kinetic reactor models during full system transients. Typical running times increase when using the coupled code, but this can be minimized by choosing appropriate variable time steps.