Dynamic Probabilistic Risk Analysis (PRA) methods couple stochastic methods (e.g., RAVEN) with safety analysis codes (e.g., RELAP5-3D) to determine risk associated to complex systems such as nuclear plants. Compared to classical PRA methods, which are based on static logic structures (e.g., Event-Trees, Fault-Trees), they can evaluate with higher resolution the safety impact of timing and sequencing of events on the accident progression. Recently, special attention has been given to nuclear plant sites which consist of multiple units and, in particular, on the safety impact of system dependencies, shared systems and common resources on core damage frequencies. In the literature, classical PRA methods have been employed to model multi-unit sites in a limited number of cases while Dynamic PRA methods have never been applied to analyze a full multi-unit model. This paper presents a PRA analysis of a multi-unit plant using Dynamic PRA methods. We employ RAVEN as stochastic tool coupled with RELAP5-3D. The site under consideration consists of three units (each unit is composed by a reactor and its associated spent fuel pool) while the considered initiating event is a seismic induced station blackout event. This paper describes in detail how the multi-unit site has been constructed and, in particular, how unit dependencies and shared resources are modeled from both a deterministic and stochastic point of view.

Catalysis informatics is a distinct subfield that lies at the intersection of cheminformatics and materials informatics but with distinctive challenges arising from the dynamic, surface-sensitive, and multiscale nature of heterogeneous catalysis. The ideas behind catalysis informatics can be traced back decades, but the field is only recently emerging due to advances in data infrastructure, statistics, machine learning, and computational methods. In this work, we review the field from early works on expert systems and knowledge engines to more recent approaches utilizing machine-learning and uncertainty quantification. The data information knowledge hierarchy is introduced and used to classify various developments. The chemical master equation and microkinetic models are proposed as a quantitative representation of catalysis knowledge, which can be used to generate explanative and predictive hypotheses for the understanding and discovery of catalytic materials. We discuss future prospects for the field, including improved quantitative coupling of experiment/theory, advanced microkinetic models, and the development of open-source software tools. Ultimately, integration of existing chemical and physical models with emerging statistical and computational tools presents a promising route toward the automated design, discovery, and optimization of heterogeneous catalytic processes.

Computation-based human reliability analysis (CoBHRA) provides the opportunity for dynamic modeling of human actions and their impacts on the state of the nuclear power plant. Central to this dynamic HRA approach is a representation of the human operator comprised of actions and the time course over which those actions are performed. The success or failure of tasks is time dependent, and therefore modeling different times at which the operator completes actions helps predict how timing differences affect the human error potential for a given task. To model the operators' timing variability, Goals, Operators, Methods, and Selection rules (GOMS) task level primitives were developed based on simulator logs of operators completing multiple scenarios. The logs have sufficient detail to determine the timing information for procedure steps and to map the procedure steps into the task level primitives. The task level primitives can then be applied to other procedures that were not evaluated, since they represent generic task level actions applicable to all procedure steps. With these generic task level primitives, untested scenarios can be dynamically modeled in CoBHRA, which expands the usefulness of the approach considerably. An example is provided of a station blackout scenario, which demonstrates how the operator timing of task level primitives can enhance our understanding of human error in nuclear process control.

This paper introduces the virtual human reliability analyst model (VHRAM). The VHRAM is an approach that automates the HRA process to enable HRA elements to be included in simulations in general and simulation based risk analysis in particular. Inspirations from clinical AI and game development are discussed as well as the possibilities for a VHRAM to be used outside of a simulated virtual twin of a nuclear power plant.

The classification of nuclear power plant procedures at the sub-task level can be accomplished via text mining. This method can inform dynamic human reliability calculations without manual coding. Several approaches to text classification are considered with results provided. When a discrete discriminant analysis is applied to the text, this results in clear identification procedure primitive greater than 88% of the time. Other analysis methods considered are Euclidian difference, principal component analysis, and single value decomposition. The text mining approach automatically decomposes procedure steps as Procedure Level Primitives, which are mapped to task level primitives in the Goals, Operation, Methods, and Section Rules (GOMS) human reliability analysis (HRA) method. The GOMS-HRA method is used as the basis for estimating operator timing and error probability. This approach also provides a tool that may be incorporated in dynamic HRA methods such as the Human Unimodel for Nuclear Technology to Enhance Reliability (HUNTER) framework.

In this paper, we describe the development of behavioral primitives for use in human reliability analysis (HRA). Previously, in the GOMS-HRA method, we described the development of task level primitives, which model basic human cognition and actions. Like generic task types found in some HRA methods, the task level primitives provide a generic or nominal human error probability. These generic task types are often modeled at the task level-grouped according to a high-level goal that includes many activities. In contrast, task level primitives represent a finer level of task decomposition, corresponding not to a group of actions that comprise an overall task but rather individual steps toward that task. In this paper, we further elaborate on the task level primitives by grouping task level primitives into procedure level primitives. This terminology reflects standard groupings of activities that are performed by reactor operators when following operating procedures. For the purposes of HRA, it is desirable to model operator actions according to these prescribed procedure categories. We present mappings of the procedure level to the task level primitives found in the GOMS-HRA method. We provide examples and conclude that procedure level primitives are a useful tool to streamline HRA modeling and quantification, especially for dynamic HRA applications.