The NEPTUNE project constitutes the thermal-hydraulics part of a long-term joint development program for the next generation of nuclear reactor simulation tools. This project is being carried through by EDF (Electricite´ de France) and CEA (Commissariat a` l’Energie Atomique), with the co-sponsorship of IRSN (Institut de Radioprotection et de Suˆrete´ Nucle´aire) and AREVA NP. NEPTUNE is a multi-phase flow software platform that includes advanced physical models and numerical methods for each simulation scale (CFD, component, system). NEPTUNE also provides new multi-scale and multi-disciplinary coupling functionalities. This new generation of two-phase flow simulation tools aims at meeting major industrial needs. DNB (Departure from Nucleate Boiling) prediction in PWRs is one of the high priority needs, and this paper focuses on its anticipated improvement by means of a so-called “Local Predictive Approach” using the NEPTUNE CFD code. We firstly present the ambitious “Local Predictive Approach” anticipated for a better prediction of DNB, i.e. an approach that intends to result in CHF correlations based on relevant local parameters as provided by the CFD modeling. The associated requirements for the two-phase flow modeling are underlined as well as those for the good level of performance of the NEPTUNE CFD code; hence, the code validation strategy based on different experimental data base types (including separated effect and integral-type tests data) is depicted. Secondly, we present comparisons between low pressure adiabatic bubbly flow experimental data obtained on the DEDALE experiment and the associated numerical simulation results. This study anew shows the high potential of NEPTUNE CFD code, even if, with respect to the aforementioned DNB-related aim, there is still a need for some modeling improvements involving new validation data obtained in thermal-hydraulics conditions representative of PWR ones. Finally, we deal with one of these new experimental data needs and present a scaling method for the design of the associated experimentation devoted to the analysis of the dynamics-related modeling of a bubbly flow in PWR representative conditions.

Hypothetical Small Break Loss Of Coolant Accident is identified as one of the most severe transients leading to a potential huge Pressurized Thermal Shock on the Reactor Pressure Vessel (RPV). This may result in two-phase flow configurations in the cold legs, according to the operating conditions, and to reliably assess the RPV wall integrity, advanced two-phase flow simulations are required. Related needs in development and/or validation of these advanced models are important, and the ongoing TOPFLOW-PTS experimental program was designed to provide a well documented data base to meet these needs. This paper focuses on pre-test NEPTUNE_CFD simulations of TOPFLOW-PTS experiments; these simulations were performed to (i) help in the definition of the test matrix and test procedure, and (ii) check the presence of the different key physical phenomena at the mock-up scale.

Thermal-Hydraulic (T/H) core code prediction of the existence and localization of boiling zones is crucial in the framework of axial offset anomaly (AOA) risk assessment of PWR cores. In this prospect, an experimental program — NESTOR — has been completed by Commissariat a` l’Energie Atomique (CEA, France), Electricite´ de France (EDF, France) and Electric Power Research Institute (EPRI, USA). The aim of the NESTOR program has been to develop an accurate prediction model for the onset of nucleate boiling (ONB) boundary in a nuclear fuel bundle based on an ONB wall superheat criterion associated with a dedicated single-phase heat transfer model. The experimental scope of NESTOR program involved using two loops to measure axial velocities in sub-channels and heater rod surface temperatures, respectively, in identical 5×5 rod bundles. The first set of experimental measurements were devoted to a bundle configuration containing only simple support grids (SSG) in order to resemble, as closely as possible, a bare rod bundle. Test data analyses in this configuration have since been jointly carried out by the three NESTOR partners, each using its own T/H core code (FLICA IV for CEA, THYC-COEUR for EDF and VIPRE-I for US Penn State University on behalf of EPRI). This paper describes the analyses results and conclusions based on SSG configuration data. The data analyses methodology consisted of three successive stages: (i) T/H core code calibration — determination of specific input data related to the bundle configuration and required by further core code simulations of the tests; (ii) Single-phase heat transfer analysis — development of dedicated single-phase heat transfer models using single-phase test data along with sub-channel-averaged temperatures and velocities obtained from T/H code simulations. The heat transfer models were unique for each of the three codes and included both a heat transfer correlation and grid enhancement correction factor; (iii) ONB test analysis — assessment of an ONB wall superheat criterion based on stage (ii) models and, if necessary, development of a new ONB wall superheat criterion. The analyses showed that while the heat transfer models could correctly represent the single-phase test data, the ONB wall superheat is over-predicted by 1–3.5 K when compared to experimental values and open literature wall superheat correlations. However, the actual impact of this inconsistency on prediction of ONB boundary localization in this experimental SSG bundle configuration is low. Similar concurrent data analyses for tests on a mixing vane grid bundle configuration are under progress and results should be available by the end of 2010.