Michele Andreani

Pretest Calculations of Phase A of ISP-42 (PANDA) Using the GOTHIC Containment Code and Comparison with the Experimental Results

Abstract The pretest calculations of phase A of the International Standard Problem 42 (ISP-42) using the GOTHIC containment code are presented in this paper, together with the comparison with the experimental results. The focus of the analyses presented is on the mixing process in the drywells (DWs), initially filled with air, during the initial steam purging transient. Consequently, a large effort has been made to capture the flow pattern produced by the jet created by the steam injection, including in the model a large number of nodes for the three-dimensional (3-D) representation of the two vessels. The influence of the nodalization of the DWs on the calculation was investigated by means of two additional models using one volume for each of the DWs and a 3-D calculation using a much coarser mesh, respectively. Since the fluid in the DWs was well mixed and stratification occurred only below the injection level, all the models could predict very accurately the global variables such as pressure and temperature. The 3-D simulation also reproduced the wall and gas temperature distributions fairly well. The only (inferred) discrepancy with the test was the overprediction in the upward deflection of the buoyant steam jet.

SIMULATION OF MIXING INDUCED BY A HOT PAR EXHAUST PLUME

Passive Autocatalytic Recombiners (PARs) are installed in various reactor containment designs to mitigate the hydrogen risk. For the evaluation of the effectiveness of these devices, validated computational tools are needed. To build confidence in the codes, their capability must also be assessed against separate effect tests addressing specific phenomena. Within the OECD SETH 2 project three experiments have been performed in the large-scale PANDA facility, where the thermal effect of a PAR was simulated by means of a heater and the plume generated by the heat source interacted with an initially stratified ambient. In these tests, helium was used instead of hydrogen. The position of the heater and the presence of simultaneous injection of steam were varied in these tests. These experiments have been analyzed with the GOTHIC and the ANSYS CFX codes. This paper reports only the results obtained with the GOTHIC code. In general, the GOTHIC code in conjunction with a coarse mesh could predict the mixing process reasonably well. The only substantial discrepancy with the experiments was the overprediction of the velocity at the inlet of the heater case, but this had little effect on the simulation of the overall mixing.Copyright

Simulations of Basic Gas Mixing Tests With Condensation in the PANDA Facility Using the GOTHIC Code

In the framework of the OECD SETH project, a number of experiments related to safety issues in the containment of a nuclear reactor have been performed in the large-scale facility PANDA. The tests have been designed to provide an adequate database for basic assessment of CFD and advanced lumped parameter (LP) codes. The test geometry consists of two interconnected vessels (compartments) with fluid injected in one vessel. The gas distribution in the injection vessel and the distribution of gases and the propagation of the stratification in the adjacent vessel are measured. Four of these tests were performed with initial and boundary conditions that resulted in substantial condensation rates. Three of these experiments featured vertical injection (with production of a plume), and in one, the transient response due to a high-momentum horizontal injection (jet) was investigated. The injected fluid was either saturated steam or a superheated mixture of steam and helium, and the fluid initially present in the vessels was pure air. These experiments have been analysed with the advanced containment code GOTHIC, and the main results are presented here. In general, the results obtained with the code and the standard mesh were in good agreement with the data. Limitations in modeling local phenomena controlled by complex flow patterns (e.g. heat transfer in the region of an impinging jet) and the need for refined meshes to reproduce certain aspects of the transients (e.g. erosion of the interface between layers of different gas composition) were also identified.Copyright

Evaluation of the PAR Mitigation System in Swiss PWR Containment Using the GOTHIC Code

Abstract The installation of passive autocatalytic recombiners (PARs) in the containment of operating nuclear power plants (NPPs) is increasingly based on three-dimensional studies of severe accidents that accurately predict the hydrogen pathways and local accumulation regions in containment and examine the mitigation effects of the PARs on the hydrogen risk. The GOTHIC (Generation Of Thermal-Hydraulic Information for Containments) code is applied in this paper to study the effectiveness of the PARs installed in the Gösgen NPP in Switzerland. A fast release of a mixture of hydrogen and steam from the hot leg during a total station blackout is chosen as the limiting scenario. The PAR modeling approach is qualified simulating two experiments performed in the frame of the OECD/NEA (Organisation for Economic Co-operation and Development/Nuclear Energy Agency) THAI (Thermal-hydraulics, Hydrogen, Aerosols and Iodine) project. The results of the plant analyses show that the recombiners cannot prevent the formation of a stratified cloud of hydrogen (10% molar concentration), but they can mitigate the hydrogen accumulation once formed. In the case of the analyzed fast release scenario, which is characterized by increasing loads with large initial flow rate and high hydrogen concentration values, it is shown that, when a large number of recombiners are installed, the global outcome in relation to the combustion risk does not depend on the details of the single PAR behavior. The hydrogen ignition risk can be fully mitigated in a timeframe ranging from 15 to 30 min after the fast release, according to the dependence of the PAR efficiency model on the adopted parameters.