Conference Agenda

Following the TMI accident, the concept of ‘risk’ was introduced to comprehensively evaluate the safety of complex systems in NPPs. Risk means potential losses that may occur in the future due to certain factors and is defined as the probability of an event occurring multiplied by the consequence of the event. Current methodologies such as Probabilistic Safety Assessment (PSA) are used for risk evaluation. However, these methodologies have some limitations, making them inefficient for assessing the risks of individual accident sequences. To address this issue, this study proposes a methodology to develop risk profiles for each accident sequence through source-term analysis. In this study, a source term analysis uncertainty assessment was conducted specifically for core damage accident sequences of Loss of Feedwater (LOFW) and Small Loss of Coolant Accident (SLOCA) based on the OPR-1000 level 1 PSA model. Based on the result, we quantified the consequences for each sequence and developed risk profiles by visualizing the risk through frequency-consequence curves (F-C curve). This approach can efficiently evaluate the risks of each accident sequence efficiently and present them in a clear visualization. These results contribute valuable information to the risk communication process.

As part of ASNR's development of Level 1 and 2 Probabilistic Safety Assessments (PSA), various support studies are conducted for internal events (IE) PSAs and internal and external hazards PSAs (fire, internal and external flooding, internal explosion, seismic, heat wave…) for operating French nuclear power plants and for some other nuclear facilities. Among these, several thermohydraulic studies are performed using tools developed by ASNR such as the SOFIA simulator (Simulator for Observation of Incident and Accident Scenarios) for the Level 1 PSA, and the ASTEC integral code for the Level 2 PSA. Those tools can simulate a wide range of operational conditions, from full power to shutdown states for 900, 1300, and 1450 MWe PWRs, as well as for the EPR. These thermohydraulic simulations play a crucial role in assessing the kinetic and the consequences of accidental scenarios, to determine whether core or fuel damage occurs, to identify the mitigation systems success criteria and to understand when and how the core uncover. They also contribute to the human reliability assessment by providing available time for diagnosis, decision-making and operator actions. Furthermore, these studies allow the examination of uncertainties inherent to key parameters, such as the size and location break in primary circuit, and their impact on the progression of the accident.

The paper presents the status and perspectives of these studies for PWRs or other facilities and introduces some expectations for possible other reactor designs (e.g. Small Modular Reactors - SMRs).

Passive safety systems employing physical processes and phenomena, such as natural circulation, have been widely applied to the contemporary design of Light Water (LW) Small Modular Reactors (SMRs). The demonstration of passive system reliability requires mechanistic analysis of the system performance in all possible accident scenarios. During the assessment, identification of the “failure domains” i.e. the domains of scenario parameters where the passive system fails to fulfil its mission, and associated “failure modes” of the system is challenging due to a wide range of operational conditions that need to be assessed. The brute-force search is computationally impractical due to the high-dimensional nature of the input space and the significant computational cost associated with Full Model (FM) evaluations.

The goal of this work is to demonstrate the feasibility of using advanced search methods, i.e. genetic algorithm, for the identification of the “failure domain” and its boundaries in the multidimensional space of accident scenario parameters. The primary objective is to improve the search efficiency by reducing a Figure of Merit (FOM) defined as the total number of FM evaluations by the number of identified boundary points. Three frameworks are developed, tested and compared on a benchmark case. The method that integrates GA and Adaptive Triangulation Sampling (ATS) demonstrates a good performance.

The Organisation for Economic Co-operation and Development (OECD) Nuclear Energy Agency (NEA) Committee on the Safety of Nuclear Installations (CSNI) has directed the NEA Expert Group on Small Modular Reactors (EGSMR) to identify and prioritize knowledge gaps where cooperative research would facilitate the safe deployment of small modular reactors (SMRs). EGSMR is executing a pilot project to demonstrate a method to generate research recommendations for advanced technology (AT-), non-water cooled, designs. High-temperature Gas Cooled Reactors (HTGRs) were selected to pilot this process; however, it is meant to be generally applicable with future application to other AT-SMR technologies. The identification and prioritization of phenomenological knowledge gaps has been built into a procedure to be completed by a task team in coordination with subject matter experts and NEA Working Groups. Phenomena Identification and Ranking Tables (PIRTs) are tools used to identify phenomena important to reactor safety by numerical ranking of importance and knowledge level. In the current work, PIRT information is collected by the task team and sorted into phenomenological groupings before being ranked based on PIRT knowledge gaps, safety significance and suitability for international collaborative experimental research. Consulting with subject matter experts, the task team will refine the prioritization list into a final set of high priority research subjects. In a later phase of the AT-SMR pilot project, these subjects will be linked to experimental facilities and compiled into a final set of detailed research activity recommendations.