Functional reasoning support for nuclear power plant field operators

Transient operations at Nuclear Power Plants (NPPs), involving system startups or shutdowns, are critical yet challenging phases. These operations are characterized by rapid changes in system conditions and an increased potential for unexpected equipment behaviors. In such settings, the mental models that operators rely on become essential. Mental models enable operators to diagnose the current state of the plant and anticipate future events. However, these mental models are vulnerable to flaws during transient operations. Such flaws can significantly misalign an operator's perception with the actual state of the plant, potentially leading to operational errors that compromise safety. Therefore, this study investigates the flawed mental models of field operators during transient operations at NPPs. Chapter 2 aimed to understand the mechanisms behind flawed mental models among field operators at NPPs. This chapter utilized document analysis to identify that field operations consist of six task stages: pre-job briefing, procedure walk-down, place-keeping, operator transit, response planning, and response implementation. An expert panel survey was conducted to explore and classify the components of mental models within NPP operations into three critical areas: workspace dynamics, workflow prognostics, and hazard identification. This analysis revealed that operators needed to continually activate and update their mental models to meet the specific demands of each task stage and to maintain alignment with the actual state of the plant. The chapter also identified that flaws in mental models frequently arose from a lack of comprehensive system dynamics information and inadequate guidance provided by existing operational procedures. Chapter 3 aimed to develop a formal framework based on functional reasoning to enhance the capture and utilization of system dynamics information within NPP operations. The chapter began by developing an ontology that formalized the connectivity, states associated with components, and the mapping relationships between these states and components' capabilities. This ontology served as a foundation, enabling the structured interpretation of component function types and behaviors. Building on the ontology, Chapter 3 then developed algorithms for functional analysis and behavioral analysis that translated observable sensor readings into detailed descriptions of component functions and behaviors. Lastly, this chapter utilized a Batch Plant to demonstrate the implementation outcomes of the developed framework. Chapter 4 assessed the impact of the generated system dynamics information on operator decision-making. This chapter detailed the design and execution of human-subject experiments involving industrial professionals from the nuclear power sector. These experiments aimed to measure how the provision of system dynamics information affected operators' decision-making across different operational scenarios. Findings from the study revealed that while providing system dynamics information improved decision-making accuracy and efficiency in handling complex tasks, it also increased the perceived workload among operators. The study noted that operators required a period of adaptation to effectively integrate and utilize the increased volume of information provided. The culmination of this research underscored the critical importance of mental models in ensuring operational safety. One of the key practical implications of this study is the potential integration of advanced decision-support systems into NPP operations. These systems aim to deepen operators' understanding of plant dynamics, thereby enhancing operational safety. Looking forward, future research should investigate innovative methods of information delivery that could reduce cognitive load and improve the assimilation of complex information. This exploration might also extend these concepts to other high-stakes industries, broadening the impact of this work and offering new avenues for enhancing safety and decision-making across various fields.