The modular High-Temperature Gas-cooled Reactor (HTGR) is one of the six generation IV advanced nuclear reactors. With the final purpose of operator training and licensing, the engineering simulation system (ESS) has been studied to model the pebble-bed type reactor core and has been successfully implemented into the full scope simulator of HTR-PM.
As stated in corresponding industrial standards, one important feature of the nuclear power plant simulator is real-time calculation, and the other one is simulation results with high fidelity (compared to design parameters or operational data in different stages). In ESS, each macro cross-section was in the form of polynomial by function of several variables (like burn-up, buckling, temperatures), the expression of which was finalized by multivariate regression analysis from large scattered database generated by the VSOP. Since the polynomial is explicit and prepared in advance, the macro cross-sections are quickly calculated in running ESS. However, some variables (such as temperature) in HTGR are in larger scope so that the polynomial is not easy to meet full range accuracy. One normal idea is to optimize the expression of polynomial, while another means was proposed and tested in present paper.
Other than focusing on the polynomials, a new method, called the fast searching, was described to significantly improve the accuracy of macro cross-section calculation while it was also fast to maintain the real-time feature. Instead of setting up a regression polynomial from the large cross-section database, the fast searching method treated the database as scatted points in the multi-dimension space, and aimed to locate the target position of unknown macro cross-section by fast searching and interpolating. Searching was to find the neighbouring database points around the target point in the multi-dimension space, which naturally improved the accuracy. While interpolating was to predict the macro cross-section of target point based on those neighbouring database points. To keep the searching and interpolating fast, the original database of macro cross-sections was analysed. A series of searching and interpolating methods have been described, programmed, tested and compared to find appropriate methods to calculate all the macro cross-sections in limited time cost. Finally, the fast searching method and its program was implemented into ESS to show better performances.