Abstract
The gamma radiation field outside of a nuclear reactor carries information about the coolant inventory and the nuclear fuel distribution inside the reactor pressure vessel. Hence, it may serve as an indicator for changes in the reactor internal structures, e.g. in the course of a severe accident. To study the feasibility of using external gamma radiation measurements for the detection of reactor state changes, three-dimensional Monte-Carlo simulations were performed to evaluate the vertical gamma flux distribution outside of a generic pressurized water reactor pressure vessel. The gamma flux was calculated for a reactor with different decreased coolant levels and different core melt states. The results indicate that the gamma flux is very sensitive to the reactor states. The shape and magnitude of the gamma flux distribution are unequivocally subject to the coolant levels and to the relocation of corium into the reactor lower head. Therefore, the results strongly suggest that a measurement of the external gamma radiation distribution, especially if coupled with sophisticated algorithms for state detection, should be sufficient to infer the state of the reactor during a severe accident. A simple state detection algorithm was tested to infer predefined reactor states. The results of the test showed that upon the measured gamma flux distribution the reactor state can be inferred with an approximate accuracy of 0.983.
Original language | English |
---|---|
Pages (from-to) | 175-180 |
Number of pages | 6 |
Journal | Annals of Nuclear Energy |
Volume | 114 |
DOIs | |
State | Published - 1 Apr 2018 |
Externally published | Yes |
Keywords
- Bayesian inference
- Gamma radiation
- Monte Carlo
- Severe accident
ASJC Scopus subject areas
- Nuclear Energy and Engineering