Modeling and validation of a mechanistic tool (MEFISTO) for the prediction of critical power in BWR fuel assemblies

Westinghouse is currently developing the MEFISTO code with the main goal to achieve fast, robust, practical and reliable prediction of steady-state dryout Critical Power in Boiling Water Reactor (BWR) fuel bundle based on a mechanistic approach. A computationally efficient simulation scheme was used to achieve this goal, where the code resolves all relevant field (drop, steam and multi-film) mass balance equations, within the annular flow region, at the sub-channel level while relying on a fast and robust two-phase (liquid/steam) sub-channel solution to provide the cross-flow information. The MEFISTO code can hence provide highly detailed solution of the multi-film flow in BWR fuel bundle while enhancing flexibility and reducing the computer time by an order of magnitude as compared to a standard three-field sub-channel analysis approach.Models for the numerical computation of the one-dimensional field flowrate distributions in an open channel (e.g. a sub-channel), including the numerical treatment of field cross-flows, part-length rods, spacers grids and post-dryout conditions are presented in this paper. The MEFISTO code is then applied to dryout prediction in BWR fuel bundle using VIPRE-W as a fast and robust two-phase sub-channel driver code. The dryout power is numerically predicted by iterating on the bundle power so that the minimum film flowrate in the bundle reaches the dryout criteria. Predicted dryout powers (including trends with flow, pressure, inlet subcooling and power distribution) and predicted dryout locations (both axial and radial) are compared to experimental results, using the entire Westinghouse SVEA-96 Optima3 dryout database, and are shown to yield excellent results.

► The MEFISTO code efficiently and accurately predicts the dryout event in a BWR fuel bundle, using a mechanistic model.
► A hybrid approach between a fast and robust sub-channel analysis and a three-field two-phase analysis is adopted.
► MEFISTO modeling approach, calibration, CPU usage, sensitivity, trend analysis and performance evaluation are presented.
► The calibration parameters and process were carefully selected to preserve the mechanistic nature of the code.
► The code dryout prediction performance is near the level of fuel-specific empirical dryout correlations.