![CAnes logo[96]](https://canes-mit-edu.ezproxyberklee.flo.org/hs-fs/hubfs/CAnes%20logo%5B96%5D.png?width=300&height=91&name=CAnes%20logo%5B96%5D.png "CAnes logo[96]")
Request Copy
Report Date: March 2004
Appendices: No
Abstract
The objective of this report was to investigate the key safety features of the pebble bed reactor under challenging conditions. The first part of the report explored the "no meltdown" claim of the proponents of the technology without the use of any active emergency core cooling systems after a loss of coolant accident. Using a conservative HEATING-7 analysis, it was shown that the peak fuel temperature was approximately 1640ºC after the initial loss of coolant which is about 1500ºC below the UO2 fuel melting temperature. Sensitivity studies also showed that the peak fuel temperature was insensitive to thermal properties, such as the soil conductivity, emissivities of the concrete wall and the pressure vessel. It was established that although the fuel would not melt, the temperature of the reactor vessel and the reactor cavity concrete exceeded design limits. A separate study by Professor Hee Cheon No using new code developed for this application (PEB-SIM) confirmed the HEATING-7 peak temperature results without convection cooling in the reactor cavity. His analysis was extended to perform a sensitivity study to determine how much air would have to be circulated in the reactor cavity to bring the temperatures of the reactor vessel and reactor cavity within design limits. This study showed that approximately 6 m/s of air-flow would be required. This study also showed that the peak fuel temperature was unaffected by the reactor cavity cooling system. The conclusion was that although the core will not melt even under these very conservative conditions by a large margin, some form of reactor cavity cooling system will be required to keep the reactor vessel and reactor cavity concrete within design limits. Future work in this area will be the design of a passive reactor cavity core cooling system based on the design inferences identified in this study.
The second part of this report was to develop a better understanding of the details of air ingress accidents in pebble bed and prismatic reactors. A theoretical study of an open cylinder of pebbles to better understand the key processes involved in air ingress was performed using the previous results from the LOCA analysis as initial conditions. The HEATING-7 model with side calculations to model the buoyancy and resistance to flow in a pebble was used to predict the peak fuel temperatures and air ingress velocity. The results of this simple analysis showed that heat source contribution from the chemical reaction was relatively low and confined to the lower reflector region. The peak temperature increase from the non-chemical LOCA analysis was about 20ºC (a maximum of 1660ºC at 92 hours). Dr. No’s analysis using PEB-SIM with chemical reactions showed similar results (peak temperature of 1617ºC at 92 hours). The other interesting result was that the air ingress velocity decreased after about 350ºC. This negative feedback could be a significant factor in air ingress accidents in real reactors since the average post LOCA temperature in the reactor is on the order of above 1300ºC. Using these fundamental insights, attention focused on developing a benchmarked computational fluid dynamics modeling capability for air ingress events. Two series of tests were used to benchmark the CFD code selected for this analysis - FLUENT 6.0. The first series of tests were performed at the Japan Atomic Energy Research Institute (JAERI). These tests were aimed at understanding the fundamental processes of air ingress accidents in separate effects tests. There were initial diffusion, natural circulation and the chemical reactions with heated graphite in a prismatic reactor configuration. Particularly, A multi-component diffusion model, surface reaction model and volume reaction model were developed. The FLUENT model and methodology developed was able to predict quite accurately each of these tests. The second experimental benchmark was the Julich Research Center test performed at the NACOK facility. This series of tests was to model natural circulation in a pebble bed reactor under varying hot and cold leg temperatures to assess air mass flow rate. The FLUENT methodology was able to predict the mass flow rates for the forty experiments with very good results. This work will be used to benchmark the NACOK chemical corrosion tests in the future. An outline for a work plan for continuation of this work has been prepared for development of a benchmarked CFD capability to analyze the details of air ingress accidents.
Program: ANP : Advanced Nuclear Power Program
Type: TR
RPT. No.: 102