LASSO reconstruction scheme to predict radioactive source distributions inside reactor building rooms; Evaluation activities

Aoki, Yuto; Ito, Rintaro; Kitamura, Akihiro   ; Machida, Masahiko  ; Suzuki, Masahiro ; Omori, Takazumi ; Taniguchi, Tatsuro ; Ide, Hiroshi  

Prior to the full-scale implementation of fuel debris removal operations in decommissioning of Fukushima Daiichi Nuclear Power Plants (1F), it is crucial to improve the on-site environment in order to safely and efficiently construct access routes inside reactor buildings. Among various improvement operations, it is one of the most essential activities to identify distributions of scattered radioactive sources based on the on-site structural data and air dose rate measurement data. In order to meet the demands, we now develop a prototype system enabling to inversely predict radioactive source distributions from air dose rate measurement data and structural data. In order to evaluate this system, a blind test was conducted in a real reactor building, JMTR in JAEA Oarai area. The test results reveal that if we cooperatively use two structural datasets, i.e., detailed and simplified structural datasets, then we can successfully reproduce radioactive source distributions and air dose rate fields inside the target room. These results suggest that the prototype system is promising in the future decommissioning in 1F.