Accurate simulation of nuclear power plant behavior is necessary for both engineering and training applications. An engineering grade simulator, used for design, safety analysis and operations, is characterized by high fidelity, computational power, lack of real-time capability, and user non-interactive environment. By contrast, a training grade simulator, used for operator training and education, is characterized by lower fidelity, less computational power, real- time capability, and user interactive environment (e.g. freeze, interrupt, restart, retrace, redirect and visual display capabilities). This projects objective is to take the most overall desirable characteristics of both nuclear power plant engineering and training grade simulators, while increasing their respective fidelities, to create a simulator with a more productive environment for engineering applications and facilitate commonality across engineering and training grade simulators. This will result in increased engineering productivity, economic improvements via enhanced designs and operations via margin utilization, and more realistic training and education. Given the limited time and budget associated with Phase 1, focus of work to be done under Phase 1 will be on improving fidelity of one component of the nuclear power plant simulator, that being the neutronics portion of the core simulator. This will be done by employing a higher fidelity core simulator than that of currently used simulators to inform such simulators so that their fidelity can be improved. The higher fidelity core simulator that will be employed is the Consortium for Advanced Simulation of Light Water Reactors VERA-CS code. This simulator employs transport theory (MOC and SPN), many energy groups, and fine spatial mesh, with many isotopes tracked. In-core thermal-hydraulic (T-H) solution employs a subchannel two-fluid, three-field methodology, and finite difference solution for fuel rod heat conduction. Currently employed simulators utilize the few-group nodal diffusion method with limited isotope tracking and coupled in-core T-H model most often employing a closed channel, homogenous equilibrium mixture, drift flux, or a six equation, two-fluid model (e.g. RELAP5-3D). For this project the NESTLE nodal simulator will be utilized. All these models utilize a coarse spatial mesh, so homogenization techniques are important to obtaining reasonable fidelity. For the radiation transport solution, this is today done by completing lattice physics calculations. Lattice physics outputs include homogenized nuclear data (nodal cross-sections, discontinuity factors and pin form factors) as functions of thermal-hydraulics conditions and history effects. Lattice calculations are two dimensional (radial plane) and assume zero current boundary condition, and spatially and time constant temperatures and densities; hence, core-wide spatial and history effects are not correctly captured, requiring various ad hoc corrections to be introduced at the core-wide solution level using information from lattice branch depletions. Ideally, one could utilize VERA-CS in place of currently employed simulators to overcome these limitations, since VERA-CS introduces none of them. Unfortunately, VERA-CS computer resource requirements and execution times currently do not make this practical, given the volume of engineering grade simulations and need for real-time capability for training grade simulations. Executing VERA-CS a limited number of times to inform a currently employed simulator presents a practical path forward. For neutronics this would be done by completing core depletion and instantaneous branch cases using VERA-CS, to provide information that would be used, along with 3-D one-node nodal solutions, to generate consistently (with respect to preserving nuclear interaction rates, leakage, and pin powers) collapsed (with respect to space, energy and methodology) nuclear data for each spatial node, thereby correctly capturing core-wide and history effects. Research, development and verification of this consistent collapse approach, using VERA-CS and NESTLE, would be completed during Phase 1 with a focus on steady-state conditions. Phase 2 would then further characterize the nuclear data to enable transient conditions simulations, extend the consistent collapse methodology to in-core thermal-hydraulics to better capture subchannel effects, incorporate the core simulator into the overall nuclear power plant simulator, and refine the interactive environment characteristic of training grade simulators to better support engineering applications.
