Conference Agenda

During a severe accident in a light water nuclear reactor, large amounts of hydrogen could be generated and released into the containment during reactor core degradation. Additional burnable gases (H₂ and CO) may be released into the containment in case of molten corium/concrete interaction. As observed during the Fukushima accidents, H₂ and CO combustion could cause high pressure peaks that could challenge the reactor containments.To prevent this risk, most of the mitigation strategies adopted in European countries are based on the implementation of Passive Autocatalytic Recombiners (PARs). Nevertheless, studies indicate that, despite the installation of PARs, it is difficult to prevent, at all times, the formation of a combustible mixture potentially leading to local flame acceleration. To better understand the phenomena associated with the combustion hazard and to address the issues highlighted after the Fukushima events, such as the explosion hazard inside the venting systems. The AMHYCO project aims to propose innovative enhancements in the way combustible gases are managed in case of a severe accident in operating reactors.

As first step, a critical review of the available literature had been performed with the objective to form the basis for the project regarding (1) PAR efficiency under ex-vessel conditions, (2) existing PWR Emergency Operating Procedures (EOPs) and SAMGs regarding containment risk management (3) H2/CO combustion and the available engineering correlations for combustion risk estimation, (4) equipment and instrumentation surveillance under severe accident conditions. This paper provides a survey on the available literature related to the four topics mentioned above.

Most countries with nuclear power plants have implemented measurement systems to assess hydrogen concentration by extracting air from the containment building during severe accidents. This method samples the atmosphere to determine hydrogen concentration, rather than depending on sensors within the containment, to preserve sensor integrity and ensure accurate readings. However, uncertainties may emerge. Firstly, steam condensation during sampling can change the gas composition ratio. Secondly, the recorded time for hydrogen concentration includes a delay from sampling, which can be compared to the time taken for direct pressure and temperature measurements inside the containment. This time lag may influence flammability predictions based on thermal-hydraulic conditions.
We are evaluating the uncertainties associated with hydrogen concentration measurements. The sampling-type measurement system consists of an accident environmental control module, a hydrogen concentration sensing module, and a data processing module. The control module extracts mixed gases from an experimental setup that simulates an accident environment, filters them through a water tank, and returns them to the experimental vessel. The pool scrubbing method eliminates steam and aerosols. Part of the mixed gases passing through the water tank is entering the sensing module with a consistent sampling flow rate. The data processing module employs an algorithm to adjust the output values from the hydrogen sensor, accurately reflecting the true hydrogen concentration in the accident environment. The accuracy of hydrogen concentration monitoring in a severe accident environment is assessed based on the test results obtained from the experimental matrix.

The generation of combustible gases (H₂ and CO) during a severe accident (SA) and their potential accumulation in the containment atmosphere could threaten the containment integrity and/or safety components in case of uncontrolled combustion. The AMHYCO project (2020-2025), funded by the European Commission, aims to enhance the understanding of H₂/CO combustion risk within the containment of a nuclear power plant, particularly in the late phase of a severe accident, to revise the management of combustible gas risk.

This work, performed in the frame of AMHYCO, explores the impact of passive autocatalytic recombiners (PARs) performance on SA progression, and particularly their interaction with other safety systems (i.e., sprays and fan-coolers). Two Western PWR scenarios (a double-ended guillotine LOCA and an SBO) were simulated with the MELCOR 2.2 code. In the LOCA scenario, steam concentration is strongly reduced shortly after the initiating event by the automatic spray actuation. The suppression of steam promotes the formation of flammable gas mixtures in the ex-vessel phase. Parametric cases showed that cooling systems' unavailability or deactivation could reduce combustion risk. Contrarily, the SBO accident initially evolves at high pressure with a high steam content in the containment. In this sequence, a late spray operation significantly affects the gas mixture's flammability. In both sequences, the oxygen depletion by the PARs operation leads to containment inertization in the late phase of the accident. As a future step, CIEMAT will launch a calculation campaign to assess how uncertainties may impact the insights gained through best-estimate analyses.

The mitigation of the hydrogen risk with passive auto-catalytic recombiners (PARs) is state-of-the-art in nuclear power plants with water-cooled reactors. In the ex-vessel phase of a severe accident, the operation of PARs faces several challenges. While hydrogen is continuously released from the interaction between molten corium and concrete, carbon monoxide is also produced, along with other gases. Inside the PAR, hydrogen and carbon monoxide compete for the available oxygen, which is continuously consumed. As a consequence, the performance of the PAR in terms of recombination rates and overall efficiency decreases. In order to enable a realistic assessment of the availability and performance of the measures to control combustible gases, numerical models developed for PAR operation during the in-vessel phase need to be enhanced towards these boundary conditions.
The REKO-DIREKT code was developed to describe the operational behavior of PARs with a mechanistic approach. This model describes heat and mass transfer processes inside the catalyst section in detail in order to simulate PAR performance and conversion efficiency. Based on an experimental program which was part of the European AMHYCO project, REKO-DIREKT was further developed to cope with the ex-vessel conditions of a severe accident.
The paper describes aspects of the code development work and gives examples for the code validation based on the existing database.

The Japan Nuclear Regulation Authority recognizes that the hydrogen explosion at Fukushima Daiichi Nuclear Power Plant Unit No. 3 originated not on the operating floor but on the lower level of the reactor building. This study aims to obtain knowledge on hydrogen behavior by examining analytical conditions under severe accident scenarios to develop an evaluation method for the retention and diffusion behavior of hydrogen leaked into the lower level of the reactor building, which represents a typical BWR plant in Japan. Based on plant walk-down data and the results of safety analysis evaluation of the actual plant, the dimensional shape and heat transfer characteristics of the floor area where hydrogen may leak, and the fluid characteristics of the leaking gas, were organized. An analytical model was developed using representative parameters as basic conditions. Sensitivity analysis of various parameters showed that the height from the leak point to the ceiling and the horizontal distance to the ceiling cavity were highly sensitive to the hydrogen concentration in the ceiling cavity. The hydrogen concentration increased as the vertical distance from the leak location to the ceiling decreased, and as the horizontal distance to the ceiling cavity decreased. By contrast, other parameters, such as the temperature of the leaking gas, had little effect on the hydrogen concentration in the ceiling cavity. The results of the sensitivity analysis indicate that these are the three main factors that increase the hydrogen concentration in the ceiling cavity.

During a severe accident, hydrogen distribution in a containment building and characteristics of hydrogen depletion by PARs differ depending on the thermal-hydraulic behaviors occurring in the containment. Various pressure control systems are installed in the containment building to prevent overpressure in severe accident conditions. Representative systems include a spray, a fan-cooler (RCFC: reactor containment fan-cooler), a filtered containment venting system (FCVS), and a passive containment cooling system (PCCS). The containment pressure control system ensures the integrity of the containment building by maintaining the containment pressure lower than the design pressure in a severe accident condition. However, during the operation of this pressure control system, the effectiveness of the hydrogen control system and the hydrogen safety in the containment building must be ensured. This study intends to experimentally evaluate the hydrogen removal characteristics of a PAR when pressure control systems such as RCFC, and PCCS are operating. The following were obtained from the experiment. In the PAR-PCCS experiments, the hydrogen removal rates of the PAR show a similar value to the correlation even during the PCCS operation, so it seems that the PCCS has little effect on the PAR operations. It is judged that the operation of the RCFC does not hurt the removal of hydrogen from the PAR through the evaluation experiment of the PAR performance according to the operation of the fan cooler.