Tests at three different scales were conducted in order to understand the mechanisms of pool scrubbing under a wide range of two-phase fluid dynamic conditions with a special focus on rapid depressurization caused by venting. In the small-scale separate effect test, high resolution two-phase flow measurement techniques utilizing equipment such as a high-speed digital camera, wire mesh sensor and PIV were applied to capture the behaviors of a single bubble (deformation, rising velocity and aerosol tracks) and the evolution of two-phase flow structures (bubble globule breakup, bubble sizes, bubble shapes and swarm rising). In the large-scale integral effect test, the dependence of the aerosol removal efficiency on submergence and pool temperature was measured under constant pressure and depressurized conditions. Attention was also paid to aerosol materials with different particle size distributions, hydrophilicity and solubility from the viewpoint of their importance for aerosol removal efficiency. To clarify relationships between individual phenomena and combined phenomena observed in two tests, the mid-scale integral effect test was undertaken. In this paper, we overviewed existing analytical models of pool scrubbing, their deficiencies to be improved, the purpose of this experimental project, an outline of experimental approaches and findings obtained so far.

　When excess pressure and temperature are added to a containment vessel of a nuclear power plant during a severe accident, damage of the containment vessel and the release of radioactive materials into the environment are expected. Filtered containment venting systems (FCVS) installed in exhaust systems reduce the release of radioactive materials by the use of multistage filters. In actual FCVS, several types of alkaline solution are used in the scrubbing stage to capture the radioactive iodine (I₂). Therefore, it is difficult to remove CH₃I by the alkaline solution, and molecular sieves such as silver zeolites installed downstream of the scrubbing stage are used to capture the organic iodide (CH₃I). Silver nitrate aqueous solution (AgNO₃ aq.) is highly reactive against iodine; hence, by using AgNO₃ aq., the removal of organic iodide in the scrubbing stage can be expected. We have conducted basic evaluations of the removal properties of iodine and organic iodide by AgNO₃ aq. using small-scale and pool-scrubbing (inner diameter: 300 mm) test sections at ambient temperature and pressure. The small-scale test results show that AgNO₃ aq. has the same I₂ removal performance as sodium hydroxide aqueous solution (NaOH aq.). Moreover, the pool-scrubbing test results show that a CH₃I decontamination factor (DF) of over 50 can be expected under the conditions of a AgNO₃ concentration ≥ 10 wt％ and submergence ≥ 1.14 m.

　New Zr alloys for fuel cladding with different compositions from conventional ones have been developed to increase the safety of nuclear power plants and to utilize existing nuclear power plants more effectively. Since the irradiation growth of fuel cladding is one of the most important parameters regarding the dimensional stability of a fuel rod and/or fuel assembly during irradiation, the irradiation growth behavior of the improved Zr alloys for light-water reactor fuel cladding was investigated. Coupon specimens, which were prepared from fuel cladding tubes with improved Zr alloys, were irradiated in the Halden reactor in Norway at temperatures of 300 and 320℃ under a typical water chemistry condition of a PWR and at 240℃ under the coolant condition of the Halden reactor up to a fast neutron fluence of ～8×10²⁵ (1/m², E＞1 MeV). During and after the irradiation test, the amount of irradiation growth of each specimen was evaluated. The effect of the difference in alloy composition on irradiation growth behavior seemed insignificant if the other conditions, such as the final heat treatment condition at fabrication, the irradiation temperature and the amount of hydrogen precharged in the specimen, were the same.

　The objective of this study is to develop a source term evaluation method applicable to a dynamic probabilistic risk assessment (PRA) for a level 2 PRA. The dynamic PRA for a level 2 PRA needs a method that can calculate the behaviors of fission products (FPs) in nuclear power plants (NPPs) during a severe accident (SA) appropriately with much a smaller calculation load than that of existing SA analysis methods. In this study, a new method called the source term phenomenological relationship diagram (PRD) is developed by applying the PRD method. The source term PRD consists of the major phenomena related to FP behaviors and their quantification processes. The trial evaluation of the method is conducted for a long-term station blackout scenario. The FP behaviors evaluated using the source term PRD are consistent with those obtained by the conventional SA analysis with a much smaller computational load than that of the conventional SA analysis methods. These facts verify that the source term PRD is applicable to a level 2 dynamic PRA. In the future, the analysis of the source term PRD coupled with thermal hydraulic analysis will be conducted for the establishment of the level 2 dynamic PRA.