Status of the OECD/NEA TVA Watts Bar 1 benchmark and Monte Carlo modeling of the 'Zero power physics test' exercise

This paper describes the status of the OECD/NEA Tennessee Valley Authority (TVA) Watts Bar 1 (WB1) Multi-Physics Multi-Cycle depletion benchmark and the results associated with the First Exercise - Zero Power Physics Tests (ZPPT) at Hot Zero Power (HZP) conditions. The OECD/NEA TVA WB1 benchmark is being developed for verification and validation of traditional and novel high-fidelity multi-physics codes for Pressurized Water Reactors (PWRs), Modeling and Simulation of different steady states and cycle depletion scenarios. The benchmark relies on real plant data of the first three cycles of the TVA WB1, a Westinghouse 17 x 17 PWR, and covers the distinct phases of typical operational conditions over 7 benchmark exercises. Exercise 1 corresponds to the start-up fresh-fuel only ZPPT stand-alone three-dimensional neutronics at HZP conditions; Exercise 2 covers the Hot Full Power (HFP) Beginning of Cycle (BOC) operations; Exercise 3 focuses on the WB1 Cycle 1 depletion; Exercise 4 covers the WB1 fuel shuffle and decay of Cycle 2 BOC ZPPT; Exercise 5 models the WB1 depletion for Cycle 2, Exercise 6 and Exercise 7 are similar to Exercise 4 and Exercise 5 but applied to core reload for Cycle 3. Validating cycle depletion modeling and simulation tools over the first 3 cycles (which are 18-month cycles associated with the current state-of-the-art PWR fuel management practices) is very important because it covers the transition of start-up cycles to near (pseudo) equilibrium cycle of operation. The paper provides the results for the first exercise obtained with the Monte Carlo code Serpent 2.131 and the comparison with the Monte Carlo code CE-KENO-VI. The Serpent calculations are performed with two nuclear cross-section libraries ENDF/B-VII.0 and ENDF/B-VII.1 to measure the sensitivity of the model upon the nuclear data library upgrade. Ten critical configurations are presented along with nine rod worth cases to evaluate the control bank worth for each control bank. Serpent and KENO have a good agreement with 56 pcm in absolute value as the highest difference. Serpent results are improved with a factor of ∼22 pcm as compared to measured data when upgrading from ENDF/B-VII.0 to ENDF/B-VII.1 cross-section library. (authors)