Sophistication of burnup analysis system for fast reactor (2)

Improvement on prediction accuracy for neutronics characteristics of fast reactor cores is one of the most important study domains in terms of both achievement of high economical plant efficiency based on reasonably advanced designs and increased reliability and safety margins. In former study, considerable improvement on prediction accuracy in neutronics design has been achieved in the development of the unified cross-section set as a fruit of a series of critical experiments such as JUPITER in application of the reactor constant adjustments. For design of fast reactor cores improvement of not only static characteristics but also burnup characteristics is very important. For such purpose, it is necessary to improve the prediction accuracy on burnup characteristics using actual burnup data of 'JOYO' and 'MONJU', experimental and prototype fast reactors. Recently, study on effective burnup method for minor actinides becomes important theme. However, there is a problem that analysis work tends to become inefficient for lack of functionality suitable for analysis of composition change due to burnup since the conventional analysis system is targeted to critical assembly systems. Therefore development of burnup analysis system for fast reactors with modularity and flexibility is being done that would contribute to actual core design work and improvement of prediction accuracy. In the previous study, we have developed a prototype system which has functions of performing core and burnup calculations using given constant files (PDS files) and information based on simple and easy user input data. It has also functions of fuel shuffling which is indispensable for power reactor analysis systems. In the present study, by extending the prototype system, features for handling of control rods and energy collapse of group constants have been designed and implemented. Computational results from the present analysis system are stored into restart files which can be accessible by users to retrieve detailed information. However, there have been difficulties in actual data management because of complex manner in data access; it was quite hard for anyone other than experts to retrieve data. In the present study, as a remedy for the difficulty, a mechanism for database management has been developed by extending the idea on restart files in order to help user easily access arbitrary data of the results. (author)