Numerical evaluation of criticality in debris beds formed during severe accidents in light water reactors

Abstract

After the Fukushima accident, the interest of the scientific community in severe accident (SA) research has been renewed. Great efforts are being made internationally to reassess and strengthen the safety of nuclear power plants. The recriticality potential in debris beds formed after the core meltdown is one of the SA research issues that needs further attention, and it is also the focus of this work. An inadvertent criticality event may cause the release of nuclear radiation and have severe consequences. Thus, the criticality in debris beds must be evaluated to predict possible risks and establish the appropriate control measures if necessary. The available criticality data for debris beds are still very scarce. Thus, the Japan Atomic Energy Agency has begun the ambitious task of building a criticality map for debris beds. That is an arduous enterprise, which requires the investigation of appropriate debris bed models and numerous computations under a broad range of possible conditions. A global effort and international cooperation are essential. The present work aims to contribute to this common endeavor by improving debris bed models, extending the criticality database, and facilitating future analyses. Alternatives for modeling the debris bed characteristics with a potential influence on the criticality are discussed in this thesis, from the most conservative assumptions to more realistic approaches. Among other things, it was found that debris beds can be modeled with high accuracy as spheres regularly arranged in a water matrix if an adequate equivalent diameter is chosen. Besides, coupled neutronic-thermohydraulic calculations were proven to be not necessary for assessing the criticality of Fukushima debris beds. This work also investigates the criticality characteristics of UO2-concrete systems. The calculation results prove the good moderation capacities of concrete, which has a significant positive reactivity effect at very low porosities. Not only the bound water is capable of thermalizing neutrons but also the SiO2, a major component of concrete. Consequently, MCCI products should be treated carefully in the criticality analyses. A preliminary conservative criticality assessment of Fukushima debris beds has revealed safety parameter ranges, i.e., conditions for which recriticality can be excluded. On the one hand, dry debris beds cannot become critical under any conditions due to the lack of sufficient moderator. On the other hand, debris beds submerged in water will remain subcritical if the porosity is sufficiently low (< 0.24 for debris beds without concrete, < 0.1 if concrete is mixed with fuel), the mass is sufficiently small (< 124 kg), or the cooling water is sufficiently borated (> 2600 ppm B). Finally, a statistical method is proposed as an alternative and more realistic way to evaluate the criticality in debris beds. A first exploratory analysis of the debris bed at Fukushima Unit 1 reveals that the probability of recriticality is extremely low. Additionally, the sensitivity analysis has concluded that the amount of control rod material (B4C) mixed with fuel is by far the most relevant parameter. Other parameters with a strong correlation with keff are the percentage of fuel in the corium, the amount of debris in particulate form, and the debris bed spreading. Based on them, future areas of research and improvement are proposed.