In the foreseen decommissioning and debris removal plans of the damaged Fukushima Daiichi reactors, the fuel debris will be broken into small pieces using laser or mechanical cutting techniques prior to removing them from the reactor buildings. Regardless of the technique to be employed, submicron radioactive aerosol particles will be generated and dispersed in the gas space of primary containment vessel during cutting operations. The water spray system has been proven to be an applicable method in removing aerosol particles. However, it cannot remove Greenfield-gap aerosol particles (with diameters between 0.1–1 μm) so effectively. To solve this problem, a new agglomeration method by addition of water mist before spray injection was developed. With preexisting water mist, aerosol particles were expected to aggregate with water mist and form larger-sized agglomerated aerosol-mist particles, which increased the effect of inertial impaction mechanism leading to higher scavenging efficiency. The new method has been verified to be capable of improving the spray scavenging efficiency for the Greenfield gap particles by conducting aerosol scavenging experiments without and with mist in the newly built UTARTS facility in the University of Tokyo. The experiment results showed that the aerosol removal rate increased along with the increasing of mist concentration level. To verify the new agglomeration method in different experiment facilities and to investigate the effects of vessel’s size scale on aerosol collection efficiency, similar experiments were repeated in the TOSQAN facility of IRSN, France. Though the cylindrical vessel in two facilities have same internal diameter, the vessel’s height of TOSQAN facility is 4.8 m, which is larger than the one in the UTARTS facility (2.5 m). The experiment results in TOSQAN facility also showed that water mist has potential to improve aerosol spray scavenging efficiency. The corresponding numerical simulations about aerosol removal by spray droplets without mist in both UTARTS and TOSQAN facilities were conducted to better understand the aerosol removal process, including time evolution of aerosol mass fraction and flow field of the gas phase.

Control rod blades are comprised of a stainless steel sheath, which contains neutron absorber tubes (filled with boron carbide powder). During decommissioning, the first stage of size reduction consists of cutting the connector (bottom portion) of the control rod, while the second stage consists of separating the blades of the control rod by cutting through the tie rod. The last stage consists of segmenting the control rod blades by cutting through absorber tubes. In this study, the control rod blades segmentation (last stage of size reduction) is investigated using an actual control rod (unused). During the experiments, we used a forming press on the cut locations followed by a plasma arc cutting underwater. The purpose of this cutting technique is to minimize the scattering of boron carbides into water by using the stainless sheath melt to seal the absorber tubes. After the segmentation, we confirmed the sealing of the absorber tubes by visually examining the cut cross-sections. The water analysis showed that the boron carbide scattering was relatively low (only 0.07% of the total boron carbides was scattered). Finally, we confirmed that the off-gas emission is considerably reduced by using Argon plasma instead of Argon-Hydrogen plasma.

In decommissioning sodium-cooled fast reactors, the operators can be exposed to radiation during dismantling of cold trap equipment (C/T). The C/T is higher dose equipment because the C/T trapped tritium of fission products during the operation to purify the sodium coolant. In this study, thermal decomposition temperature and rate of sodium hydride (NaH) were measured as a fundamental research for development of “thermolysis” process prior to the dismantling. We measured the thermal decomposition temperature and rate using NaH powder (95.3%, Sigma-Aldrich) in alumina pan with ThermoGravimetry-Differential Thermal Analysis (TG-DTA) instrument (STA2500 Regulus, NETZSCH Japan). The heating rates of TG-DTA were set to β = 2.0, 5.0, 10.0 and 20.0 K/min. The DTA showed endothermic reaction and the TG showed two-steps mass-loss over 580K. This first-step mass-loss was consistent with change of chemical composition of the NaH with heating (NaH → Na+1/2H 2 ). The thermal decomposition temperature and rate were obtained from the onset temperature of the mass-loss and the simplified Kissinger plots, respectively. Furthermore, we set to the thermal decomposition temperature of around 590K, and the mass-loss rates were measured. As a result, over 590K, the thermal decomposition occurred actively, and showed good agreement with the estimation curves obtained by the simplified Kissinger plots. The thermal decomposition rate strongly depended on the heating temperature.

Decommissioning of Nuclear Power Plant is a long-term project during which generations are expected to change. Therefore, it is necessary to appropriately transfer knowledge, technology and skills to the next generation. In recent years, in the world of decommissioning, attempts have been made to apply advanced technologies such as utilization of knowledge management and digital technology. This study describes adaptation in decommissioning from viewpoint of utilizing IT technology called digital twin and aspect of knowledge management.

To support the 1F decommissioning project, the Institute of Applied Energy has been analyzing the course of the accident using the SAMPSON code with an aim to grasp radionuclide distribution in 1F NPP. The present study has mainly focused on molybdenum release from a fuel pellet and chemical forms of iodine during the early phase of the accident because a release rate of semi-volatile FPs has various uncertainties. The present simulation set the detrimental factor of molybdenum used in FP release model as a sensitive parameter. The simulation results indicate that more molybdenum release causes more generation of hydrogen iodide and more deposition of iodine inside Reactor Pressure Vessel (RPV) because the adsorption velocity of hydrogen iodide is comparatively higher than deposition velocities of another FPs.

A long-reach articulated manipulator that can deploy various sensors in a large but confined workspace is urgently needed in decommissioning tasks at Fukushima Daiichi Nuclear Power Stations. The most critical problem involves managing the large gravitational torques acting on the manipulator’s joints. In previous studies, several prototype models were developed; however, they were extremely heavy and bulky. This paper presents fundamental ideas for creating a lightweight super long-reach articulated manipulator. First, a mechanical structure could be utilized to support gravity. Second, a vertical force could be generated to compensate for gravity. Third, a tendon-driven mechanism could be utilized because it permits the installation of heavy actuators on the base. The tendons can transmit the actuator power to each joint. Thus, the weight of the arm is significantly reduced. We have discussed the advantages and drawbacks of each recommendation based on the hardware prototyping.

In January 2012, Toshiba Energy Systems & Solutions Corporation inserted a camera into the damaged Primary Containment Vessel (PCV) of Fukushima Daiichi Nuclear Power Station unit 2 for the first time since the Great East Japan Earthquake. Since then, we, as a member of International Research Institute for Nuclear Decommissioning (IRID), has developed several devices and has investigated inside of PCV. The investigation devices for unit 2 are roughly categorized into three types. For the investigations in early stages, just a camera and sensors with simple support tools were inserted through a small hole. Then small crawler type robots were developed to expand the investigation area to inside of Pedestal which exists the central part of PCV. Since the devices found that there were not enough areas to run around by the robots in the Pedestal, a long telescopic guiding pipe device was chosen to deliver cameras, sensors, and tools into the Pedestal. This type of device took movie of wide area of inside of the Pedestal. An image processing software was also developed to get a high quality and extensive picture from the movie taken by a low resolution camera and insufficient lighting. In February 2019, the other telescopic guiding pipe device finally delivered a camera and fingers like tong to grab deposits fallen on the bottom of pedestal for the first time after the earthquake, and found that some of deposits can be lift up. Some of projects were implemented as subsidized projects of “Decommissioning and Contaminated Water Management” founded by the Ministry of Economy, Trade and Industry.

Decommissioning of the Fukushima Daiichi nuclear power station is challenging due to industrial and chemical hazards as well as radiological ones. The decommissioning workers in these sites are indicated to wear Personal Protective Equipment (PPE) appropriately for radiation protection. In response to the difficulties of on-site PPE management in decommissioning site of Fukushima Daiichi NPS, this paper proposes the combination of deep learning-based object detection and individual detection using geometry relationships analysis to automatically facilitate the safety monitoring task of decommissioning workers to ensure hardhats, full-face masks and dust masks are correctly used for each zone. The experimental results demonstrate that the high precision, high recall and fast speed of the proposed approach can effectively detect decommissioning workers’ incorrect use of PPE.

Boron carbide is used as a neutron-absorbing material in Fukushima-Daiichi Nuclear Power Station (1F), producing borides that are twice as hard as oxides (such as UO 2 and ZrO 2 ). The high neutron absorption of boron affects the evaluation of re-criticality during the process of debris retrieval. Therefore, it is important not only to determine the presence of boron but also to investigate the distribution of boron inside the material in a non-destructive manner during decommissioning. To address the uncertainties in the core material relocation behavior of boiling water reactor (BWR) during a severe accident (SA), solidified melt specimens of a simulated fuel assembly were prepared by plasma heating. If core material melting and relocation (CMMR) specimens can be used to estimate the B distribution in 1F Unit-3, that will provide valuable information in the decommissioning of 1F. To address this, the authors focused on the energy-resolved neutron-imaging system, RADEN, which utilizes a wide energy range, from meV to keV. This is an innovative three-dimensional analysis technology for boride distribution that affects the evaluation of hardness and re-criticality. In the calibration standard samples (Zr x B 1-x and Fe x B 1-x ), there was a good correlation between boron concentration and the energy-dependence of the cross sections of cold and epithermal neutrons. In the CMMR specimens, boron distribution was confirmed from the contrast difference between cold and epithermal neutrons. In the future, the results of calibration standard samples will be applied to the results of CMMR specimens. With this method, three-dimensional boron distribution will be measured, and the understanding of boride distribution 1F Unit-3 will be improved, which may be reflected in an improved SA code.

Decontamination is a major activity in decommissioning of Nuclear Power Plants. In efforts to reduce the overall volume of nuclear waste, retrieve reusable materials, and reduce the environmental impact, many different technologies have been developed/used in prior decommissioning projects and many more are being developed. However due to the amount of technologies available and the specific use cases for each, the ability to choose an appropriate and optimal technology is a challenge. An approach was adopted to develop a tool to assist in selection of decontamination technologies appropriate for Canadian Applications. The first step is the creation of a database to compile information of the different decontamination methods currently available in one location. The next step was the development of a software program to provide a search optimization for the database based on a set of initial user conditions. The program considers a radio-isotopic breakdown of a component as identified by the user and compares its concentration (Bq/g) to regulation limits set by the Canadian Nuclear Safety Commission (CSNC) for Unconditional Clearance Levels. Then, by using the CNSC guidelines, it will determine if the component is under Unconditional Levels or not. If the component is not, the code will calculate the minimum cumulative Decontamination Factor ( DF R ) required to make the component compliant with unconditional requirements. The software allows for users to plan their decontamination roadmap at a present state as well as a future state where natural decay opens up the ability for a wider range of decontamination technologies and for a combination of multiple components to use a given decontamination technique.

Understanding the spreading behavior of the molten core is important for predicting the progress of severe accidents and for smooth decommissioning work. In this study, we performed molten metal drop experiments using Sn and Cu to obtain fundamental data and its uncertainties that can be used to verify the analysis model and to evaluate the effect of scaling parameters on the spreading and deposition behaviors. We obtained spreading area and deposition thickness of the metals after floor impingement. The standard deviation on average were 6.0% and 6.7% for the thickness and the area, respectively. In addition, the correlation with each dimensionless number was confirmed, and strong correlations were obtained between the deposition behavior and the Re number, and between the spreading behavior and the Pe number. We constructed a correlation equation using our experimental data as a function of Re number, We number, Fr number, Pe number, and dimensionless length ( L/d ). The errors for our experimental data were within 30%.

The laser 3D scanning model reconstruction is an effective technical means for the reconstruction of 3D virtual scenes of old nuclear facilities. However, the existing technology have the following problems: First, the scanning equipment needs to be installed on the triangular fixed bracket, which requires personnel to be installed on site, thereby bringing radiation safety risks to personnel; second, it requires the person to reconstruct and splice the model one face by one face, which is inefficient. Aiming to deal with these problems, this paper proposes a mobile laser 3D scanning method and a rapid model reconstruction technology. These methods include: (1) In order to realize fast scanning of large-scale scenes, and at the same time ensure accurate scanning of local important objects, a laser three-dimensional composite scanning method is proposed. (2) Design and development of a remote control automatic walking and fixing device, which can transport the laser 3D scanner to the position of measurement which is needed. The device has a certain ability to overcome obstacles, and can be reliably fixed to ensure the stability and scanning accuracy of the laser scanner at high speed working station. You can remotely control the laser scanner to start or stop working. The difficulty of this technology is how to make sure the device has the flexibility in motion and the stability on working; (3) An automatic model reconstruction and feature identification method based on a large number of point cloud data is proposed, and the corresponding software is developed. This method can realize the automatic recognition of the standard and common models, and the automatic model reconstruction of the point cloud data in the scene. To achieve this goal, the neural network algorithm is used. These technologies can effectively reduce the radiation safety risk of personnel during laser scanning in high-radioactive places, reduce the intensity of personnel operations, and improve work efficiency.

Assessing the production and dispersion of aerosols carrying contamination during Fukushima fuel debris retrieval is IRSN’s contribution to a research project undertaken by a French consortium jointly with ONET Technologies and CEA on behalf of METI and managed by the Mitsubishi Research Institute (Georges et al. 2017). The objective is to obtain quantified data for evaluating the risk of disseminating contamination when implementing cutting tools such as laser or others such as mechanical one, over the next few years, in the process of decommissioning the damaged reactors at the Fukushima-Daiichi plant.

Operation and decommissioning of nuclear facilities will produce radioactive waste, and different radionuclides in the waste will bring different hazards to the public and the environment. The waste would be sorted more reasonably by distinguishing different radionuclides. Yet it is still very difficult to measure directly the pure Beta radioactive waste in situ, though in situ Gamma-analytical and Alpha-waste-barrel measurement techniques have become more sophisticated. The aim is to propose a scientific technique to sort the radioactive waste in situ. This study focused on the 90Sr-contaminated material in China Institute of Atomic Energy and optimized the design of the existing solid waste disposal facilities. A novel technique to measure the radioactive waste 90Sr-90Y online was proposed, trying to sort the radioactive waste as optimally as possible to realize further separation of exemption waste. Theoretically, the exemption waste can be further sorted, and it can guide the design of radioactive waste disposal system.

Authors are developing an experimental technology to realize experiments simulating Severe Accident (SA) conditions using simulant fuel material (ZrO 2 with slight addition of MgO for stabilization) that would contribute not only to Fukushima Daiichi (1F) decommissioning but also to enhance the safety of worldwide existing and future nuclear power plants through clarification of the accident progression behavior. Based on the results of the prototype test, improvement of plasma heating technology was conducted. The Core Material Melting and Relocation (CMMR)-1/-2 experiments were carried out in 2017 with the large-scale simulated fuel assembly (1 m × 0.3 mϕ) applying the improved technology (higher heating power and controlled oxygen concentration). In these two tests, heating history was different resulting basically in similar physical responses with more pronounced material melting and relocation in the CMMR-2 experiment. The CMMR-2 experiment is selected here from the viewpoint of establishing an experimental technology. The CMMR-2 experiment adopted 30-min heating period, the power was increased up to a level so that a large temperature gradient (> 2,000 K/m) expected at the lower part of the core in the actual 1F accident conditions. Most of the control blade and the channel box migrated from the original position. After the heating, the simulated fuel assembly was measured by the X-ray Computed Tomography (CT) technology and by Electron Probe Micro Analyzer (EPMA). CT pictures and elemental mapping demonstrated its excellent performance with rather good precision. Based on these results, an excellent perspective in terms of applicability of the non-transfer type plasma heating technology to the SA experimental study was obtained.

The phase–out of all Nuclear Power Plants (NPP) until 2022 in Germany offer the opportunity to apply new and efficient decommissioning technologies, which allow further reduction of decommissioning costs and minimization of the collective dose for personnel. One challenging task in this process is the decontamination of protective paints containing polychlorinated biphenyl (PCB). PCB-containing surfaces demand consecutive radiological decontamination and removal of the PCB-containing paints before the demolition is possible. Laser technology, commonly used in many different industrial sectors, presents an advanced approach to this problem. Successful thermal decomposition of PCB-containing paints using high power diode lasers has been reported by our group in former publications [1]. Ongoing investigations focus on the technology transfer from static laboratory state to industrial application. This includes the verification of safety and efficiency issues for the complete laser supported process. The current presentation covers all aspects of the laser process: concrete sample preparation as well as an experimental set-up utilizing a mobile laser system that includes an innovative laser head. Particular focus will be on the phenomena of the generated concrete surfaces and the arising by-products during laser processing. A 10 kW high power diode laser, Laserline LDF 1500-10000 was used for the ablation experiments. Concrete samples were coated with typical decontamination epoxy-based paints, which match the commonly used paint systems in German NPPs in terms of composition, structure and thickness. The influence of significant processing parameters (laser power, feed rate, type of concrete and composition of paint) on the laser ablation are shown in relation to the energy input per length. The release of particles is subjected to these parameters as well. Therefore the size and the shape of the arising particles have been measured online in accordance with these processing parameters by an Engine Exhaust Particle Sizer (EEPS, Model RP-3090, TSI) in a range between 5.6 to 560 nm. The quantity of particles has also been counted by a Condensation Particle Counter (CPC, Model 3022, TSI). Overall up to 6 million particles per cubic decimeter are detected with a dilution factor of 1:100 throughout the experiments. Both measuring systems are connected to the exhaust air pipe downstream, next to the laser head. The shift of the particle size in accordance with the above mentioned parameters is crucial for a comprehensive understanding of the laser decontamination. The mobile laser supported decontamination technology will be tested in 2018 in a German nuclear facility. This test will contain the decontamination process with actual surface configurations like floors, walls, ceilings and corners, and will act as an effective proof-of-concept for the developed laser system prototype.

Retrieval of fuel debris is one of the most urgent issues regarding the decommissioning of Fukushima Daiichi Nuclear Power Plant (NPP). A retrieval of fuel debris requires a measurement system which can be applied in the severe environment to investigate inside the Reactor Pressure Vessel (RPV) and the Control Rod Drive (CRD) housing. In this study, an air-coupled ultrasonic measurement is proposed because of the feasibility of its application in the severe environment such as the high-radiation atmosphere. However, compared with using it in water or solids, there are some difficulties with applying the air-coupled ultrasonic technique. Therefore, we employed a point-focused transducer which can converge ultrasonic energy in a small area. As a fundamental study, this paper focuses on the experimental assessment of the potential feasibility and property of the developed air-coupled ultrasonic imaging. Firstly, to obtain the property and characteristics of the point-focused transducer, the sound field of the transducer was measured. The sound field showed the convergence rate, acoustic intensity and the beam width of the transducer. Secondly, in order to improve Signal to Noise ratio (SN ratio), the ultrasonic imaging method using a frequency correlation was designed taking the feeble ultrasonic echoes from complex shape object into consideration. Finally, the shape of a test specimen, which imitates a fuel debris, was measured with the developed imaging system. The ultrasonic image of the specimen showed a good agreement with a photo. In conclusion, the imaging using air-coupled point-focused ultrasound was developed to successfully visualize a mock fuel debris from a relatively short distance of 40 mm – 100 mm.

As for the decommissioning of Fukushima Dai-ichi nuclear power plant (1F), a long-term waste storage container with high safety is requested to store radioactive materials such as fuel debris for a long period of time. Since hydrogen is generated by radioactive decomposition of water, it is important to keep the concentration of hydrogen gas below the explosion limit in order to ensure the safety of the container. Then, use of passive autocatalytic recombiner (PAR) was proposed to reduce the hydrogen concentration. PAR is installed in the container. In order to experimentally confirm the reduction of hydrogen concentration by PAR and hydrogen behavior in the container, an experimental apparatus consisting of a small-scale modeled container and a hydrogen supply system was provided. Preliminary experiments were begun for confirming fundamental performance of the experimental apparatus under the conditions that PAR and simulated fuel debris are not installed in the container. Moreover, the hydrogen behavior in the container was analyzed numerically. In addition, the steam behavior generated by the chemical reaction of hydrogen and oxygen by PAR was also predicted. This paper describes both results of the preliminary experiments and numerical simulations. The experimental results showed that the hydrogen behavior can be predicted using the temperature distributions in the container. The analysis results clarified the controlling factors on the hydrogen behavior and the steam distribution in the container by PAR.

Decommissioning of nuclear systems and consolidation of resulting waste remains a major hurdle for the industry. The radioactive nature of nuclear waste makes manual handling unfeasible, while teleoperation and inflexible automation also have drawbacks. This effort utilizes virtual fixtures to augment path creation for non-contact material reduction tasks. It builds on previous Nuclear and Applied Robotics Group work with Variable Normal Surface Virtual Fixtures, which generate surfaces offset from task surfaces. Offset surfaces can then provide poses at a set orientation to task surfaces. Once the Variable Normal Surface Virtual Fixture is generated from task information, operators build a list of desired path poses. The Robot Operating System Descartes package takes the pose list and plans a smooth trajectory for task execution. Planar and cylindrical demonstrations based on experimental studies at the United Kingdom’s Sellafield site were performed. This methodology augments waste reduction by allowing flexible laser cutting routes.

When a severe accident of a nuclear reactor occurs, decommissioning work becomes important task. In the decommissioning work of a boiling water reactor after the severe accident, estimation of the sedimentation place of the molten debris is important. However, the technique to estimate exactly the sedimentation place has not been enough developed. Therefore, the detailed and phenomenological numerical simulation code named “JUPITER” is under development for predicting the molten core behavior in JAEA (Japan Atomic Energy Agency). The comparison between experimental and numerical results is necessary to clarify the validity of the numerical analysis code. The study provides the experimental data to examine the numerical simulation code. As a basic study to examine the numerical simulation code, a liquid film flowing in a modeling flow channel was studied by using water. The flow was visualized, and the flow data were obtained by image processing.