We propose a given-data approach to estimate reliability importance measures by post-processing a single dataset of stochastic simulation runs. Modern discrete-event and hybrid simulators produce rich outputs, with information on component behavior and system failure times over many scenarios. Our method exploits this information to estimate, at no additional cost, any importance measure based on conditioning. We propose estimators for six classical indices, the Birnbaum, Barlow–Proschan, Fussell–Vesely, RAW, RRW, and Differential importance measures. We prove strong consistency of the resulting estimates under standard assumptions. We quantify estimation uncertainty through bootstrap resampling, which also avoids additional model evaluations. We present numerical experiments for a simple system with analytically known values and a water supply system representative of a nuclear power plant, with a more complex logic that includes dependent (common-cause) failures. Both systems are modeled using the Event Modeling Risk Assessment using Linked Diagrams (EMRALD) platform. For the system, post-processing the resulting output dataset allows us to compare numerical estimates with analytical benchmarks. The experiments show convergence for all estimated importance measures as the sample size increases. For the nuclear system, the simultaneous estimation of multiple importance measures provides complementary insights into their risk-significance, thus supporting a rigorous prioritization and providing information useful in categorizing components borderline for Fussell–Vesely and Risk Achievement Worth but important according to other measures.

Several advanced reactor designs are now under active development in the United States and elsewhere, promising sustainable solutions to the growing world energy needs. The designs currently being considered are quite diverse and different from the more established light water reactor technology that has dominated the operating commercial nuclear landscape. While advanced reactor concepts were first explored in the dawn of the nuclear age, they are now being reconsidered under the light of modern needs, and specifically, for their flexible operating conditions and inherent safety characteristics. In response, the U.S. Nuclear Regulatory Commission staff is moving forward with development of 10 CFR Part 53 rulemaking, which is a more risk-informed, technology-agnostic framework for licensing and regulating such new designs. The nuclear industry response to this regulatory initiative resulted in the technical report by the Nuclear Energy Institute, NEI 18-04 Revision 1, which provides an implementation roadmap of the risk-informed approach when defining the safety case for a new plant design. The implementation of this safety case may be a nontrivial exercise for an actual reactor design. This paper provides a demonstration of performing such an analysis for a representative advanced reactor. Public information from the General Atomics high-temperature gas reactor design was considered in this demonstration. The analysis workflow was facilitated with the FPoliSolutions' proprietary Risk-Informed System Engineering (RISE) digital platform, a product that was presented in previous publications. RISE is one application of FPoli's enterprise digital platform, which was created to facilitate orchestration of complex workflows leveraging recent technologies developed at national laboratories, such as Idaho National Laboratory's RAVEN and EMRALD frameworks. The analysis described in the paper includes the selection and classifications of events, the integration of probabilistic risk analysis artifacts, and event modeling simulations for consequence evaluations. The results are then used for system, structures, and components safety classification and a synthesis of the safety case for the design in line with the frequency-consequence targets presented in NEI 18-04. The purpose of the analysis, as framed in RISE, is to readily produce outputs and views that can aid users and regulators in making risk-informed decisions to demonstrate their plant safety case.

The U.S. Nuclear Regulatory Commission has developed regulations regarding the siting and design of nuclear power plants (NPPs) that are aimed at addressing various natural hazards, including flooding. Flood barriers are designed to prevent water from entering NPP areas containing structures, systems, and components (SSCs) important to safety. The barriers are used at NPPs along with drains, sumps, pumps, valves, plugs, and site grading as part of the plant flood protection features that protect SSCs from experiencing external or internal flooding and mitigate the effects of flooding on NPP operations. The performance of flood protection features, including flood barriers at NPPs, has been an ongoing concern. Domestic and international operational experience provides clear indications that flood barrier performance has significant safety implications, especially for aging NPPs. The observed deficiencies show that flood barriers should be designed and installed properly, then adequately tested, inspected, and maintained in order to ensure that they perform their intended functions during flooding events.This paper reviews available information related to flood barriers employed at U.S. NPPs and provides an overview and categorization of NPP flood barriers. It identifies potential domestic and international flood barrier testing facilities, including operating and decommissioned U.S. NPPs. Finally, this paper presents the technical and logistical considerations that should be made when developing specific testing strategies and protocols for flood barriers, such as the selection of flood barriers, test locations, testing approach, performance criteria, and testing parameters.

Probabilistic Safety Assessment (PSA) of complex facilities is performed to arrive at the risk posed by them. PSA also accounts for the contribution of the human errors towards the overall risk through Human Reliability Analysis (HRA) in terms of Human Error Probability (HEP). Human operators are part of the system and do not work in isolation. Their performance is influenced by the context in which the actions are performed. As a result, quantification of HEP requires operator performance data under the given context. Some good sources of operator performance data are plant's operation data, simulator data and expert judgement. The plant operation data pertaining to HRA is generally sparse. In this situation, a full scope plant simulator provides a good alternative for operator performance data generation. Many of the currently practised HRA methods have been developed by combining the empirical evidence with expert judgement and contain a lot of uncertainty in their estimates. Bayesian inference is suitable for updating the prior HRA estimates with the simulator evidence to obtain the posterior HEP. In this study, posterior HEP has been calculated for postulated accident scenarios in advanced reactor (first of its kind) at design stage, using plant simulator.

We present the Data and Risk-Informed Chemical Assessment Technique (DRICAT), a quantitative/qualitative risk analysis technique for assessing the risk of a potential chemical release incident that may lead to a mass casualty event in United States communities. Risk assessment is a comprehensive, structured, and logical analysis approach aimed at identifying and assessing risks in “systems” for the purpose of improving management of these systems. DRICAT leads to better understanding and effective management of risks from chemical incidents through risk and scenario identification and ranking by severity by helping community planning to minimize morbidity and mortality during and after potential events. As such, DRICAT is designed to be reproducible, evidence-based, practical, and scalable for different types of communities and the possible chemical hazards present in that community. Recognizing that many communities have assessment protocols and response mechanisms already in place, we believe these DRICAT characteristics will enhance both existing chemical incident awareness and readiness activities while providing a baseline approach for communities lacking chemical release risk analysis techniques. To understand where potential hazards might exist within a community, it is useful to consider drivers for hazards (i.e., those factors that influence the presence of the hazards and the uncertainty). DRICAT uses essential elements of information (EEI) to identify chemical initiating events (IE) which are a part of potential hazardous scenarios. Both formal and informal approaches can be used to identify initiators arising from chemical hazards. DRICAT focuses on precursor events and a deductive approach using a hazard identification diagram. EEI data sources for community chemical hazards range from informal (e.g., social media, and local news outlets) to formal vetted databases. EEIs include systematic identification of hazards, community factors that affect hazards (e.g., population density, weather, and commodity flows), associated IEs, and grouping of individual causes into like categories. IE characteristics may vary among communities and include IEs that may lead directly to a chemical release or may require additional mitigative failures. DRICAT leverages EEIs to identify IEs to build and rank chemical accident scenarios from IE to the potential outcome. The likelihood of this release and the consequence of the release determine the overall risk.