{"id":{"repo_id":"uiuc","oai_identifier":"oai:www.ideals.illinois.edu:2142/130127"},"canonical_url":"https://search.dev.ndltd.org/etd/uiuc/oai:www.ideals.illinois.edu:2142/130127","repository":{"repo_id":"uiuc","name":"University of Illinois - Urbana-Champaign","base_url":"https://www.ideals.illinois.edu/oai-pmh"},"display":{"title":"Temporal coupling of maintenance human performance, physical degradation, and digital twin models for probabilistic risk assessment in nuclear power plants","abstract":"Submission published under a 24 month embargo labeled 'Closed Access', the embargo will last until 2027-08-01","abstract_html":"Submission published under a 24 month embargo labeled 'Closed Access', the embargo will last until 2027-08-01","abstract_has_math":false,"creators":["Beal, John"],"institution":"University of Illinois Urbana-Champaign","degree_name":"Ph.D.","degree_level":"Dissertation","degree_discipline":"Nuclear, Plasma, Radiolgc Engr","degree_department":null,"school":null,"contributors":["Mohaghegh, Zahra","Sakurahara, Tatsuya","Bui, Ha","Sriver, Ryan","Alam, Syed","Vergari, Lorenzo"],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2025,"date_issued":"2025-07-07","date_published":"2025-07-07","updated_at":"2026-07-22T22:25:06Z","subjects":["Probabilistic Risk Assessment","Nuclear Power Plants","Uncertainty Analysis","Human Reliability Analysis","Maintenance"],"languages":["en","eng"],"rights":["Copyright 2025 John Beal"],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"https://hdl.handle.net/2142/130127","outbound_label":"Handle","outbound_source":"dc:identifier"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Mohaghegh, Zahra","Sakurahara, Tatsuya","Bui, Ha","Sriver, Ryan","Alam, Syed","Vergari, Lorenzo"]},{"key":"dc:creator","label":"Author","values":["Beal, John"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2025-07-07","2025-08"]},{"key":"dc:type","label":"Dc Type","values":["text"]},{"key":"thesis:degree_discipline","label":"Discipline","values":["Nuclear, Plasma, Radiolgc Engr"]},{"key":"thesis:degree_level","label":"Degree Level","values":["Dissertation"]},{"key":"thesis:degree_name","label":"Degree Name","values":["Ph.D."]},{"key":"thesis:institution_name","label":"Thesis Institution Name","values":["University of Illinois Urbana-Champaign"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Probabilistic Risk Assessment","Nuclear Power Plants","Uncertainty Analysis","Human Reliability Analysis","Maintenance"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language","label":"Dc Language","values":["en","eng"]},{"key":"dc:rights","label":"Dc Rights","values":["Copyright 2025 John Beal"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["https://hdl.handle.net/2142/130127"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["Submission published under a 24 month embargo labeled 'Closed Access', the embargo will last until 2027-08-01","The student, John Beal, accepted the attached license on 2025-06-18 at 07:04.","The student, John Beal, submitted this Dissertation for approval on 2025-06-18 at 10:06.","This Dissertation was approved for publication on 2025-07-07 at 12:45.","DSpace SAF Submission Ingestion Package generated from Vireo submission #22343 on 2025-10-25 at 15:52:33","Probabilistic Risk Assessment (PRA) has played a critical role in improving the safety and performance of existing Nuclear Power Plants (NPPs) and is a key element of the U.S. Nuclear Regulatory Commission (NRC)’s Risk-Informed Performance-Based Regulation (RIPB). PRA methods have evolved over the years, and their continued advancement is increasingly important in light of the Accelerating Deployment of Versatile, Advanced Nuclear for Clean Energy Act of 2024 (or ADVANCE Act of 2024). Such advancements ensure that the safety assessment of next-generation NPPs will move beyond solely relying on prescriptive, deterministic, or subjective approaches, such as Maximum Credible Accident (MCA), which have demonstrated limitations in terms of safety and efficiency. To support the quantification of probabilistic inputs for the PRA of NPPs, this research utilizes modeling and simulation (M&S) to estimate repairable component reliability, especially in cases where empirical data is lacking. This is particularly relevant for new reactor designs or aging plants, where historical data may no longer be applicable due to design changes or evolving operational approaches, such as the integration of digital twins (DTs) into operations. Furthermore, the use of M&S enhances risk management strategies by incorporating the underlying failure mechanisms directly into the PRA framework. Although the methodological contributions of this thesis are demonstrated for NPP reactor coolant system piping components subject to long-term stress corrosion cracking (SCC) using the Extremely Low Probability of Rupture (xLPR) Probabilistic Fracture Mechanics (PFM) code, they can be applied (with some adjustments) to other types of components and physical degradation mechanisms. Three key scientific contributions of this Ph.D. thesis include: Contribution I: Modeling component reliability by developing and coupling human reliability analysis (HRA)-based maintenance models with physical degradation models. The analysis of repairable components is a multi-faceted problem because the underlying process involves time-dependent interactions between physical phenomena and social mechanisms. The physical process can encompass random shocks and degradation mechanisms that can deteriorate the functional performance of a component. The social mechanisms are related to maintenance activities and involve human performance and organizational factors. A review of the related literature identified critical gaps related to (a) simulating the underlying maintenance performance model and (b) capturing the effects of interactions between maintenance performance models and physical degradation models. This Ph.D. research addressed the gaps by: (i) developing a human reliability analysis (HRA)-based maintenance performance model to quantify maintenance outcomes under multiple types of maintenance programs, and (ii) developing a methodological and computational platform to couple the HRA-based maintenance performance model with physics-of-failure models. Event sequences are generated using dynamic PRA methods to capture temporal interactions between HRA-based maintenance models and physics-of-failure models. Bi-directional, time-dependent interactions are explicitly modeled in this approach, enhancing the resolution of the coupling. Contribution II: Uncertainty-based validation of coupled maintenance-physical degradation models in the absence of maintenance performance data. This research investigates the key challenges associated with the validation of coupled maintenance-physical degradation simulation models, especially when little to no applicable empirical validation data is available. The Probabilistic Validation (PV) methodology, an uncertainty-based validation method, was leveraged and advanced. The original PV methodology required advancements for this research because (a) the nature of human performance models differs significantly from the physics-based simulation models for which the original PV methodology was demonstrated, (b) the original PV methodology could not validate simulation models with time-dependent, bi-directional feedback between models within the same hierarchical level, and (c) in the original PV methodology, it was assumed that limited output-level validation data would be available to estimate model-form uncertainty; however, this research creates an advanced approach to quantify model-form uncertainty in the absence of any output-level validation data. The literature review revealed inconsistent consideration of uncertain parameters in maintenance models and limited focus on human performance, preventing its uncertainty from being reflected in system-level risk assessments. Contribution II of this thesis addresses these gaps through two key steps: (i) developing a novel methodology to systematically identify, characterize, and quantify uncertainty in HRA-based maintenance models, even in the absence of output-level validation data and (ii) developing a computational platform to propagate uncertainty from the HRA-based maintenance model to the system response level using a nested Monte Carlo-based approach. The new methodology guides the construction of a graphical causal model that captures the sources of uncertainty and their interactions, followed by quantification using a Bayesian Belief Network (BBN)-based approach. This enables the coverage of a broader range of epistemic uncertainties, including model-form uncertainty, than existing methods. Contribution III: Development of bi-directional temporal coupling among human performance, physical twin, and digital twin (DT) models, and incorporating the coupled system into the PRA of NPPs. The nuclear industry is investigating the applications of DTs in NPPs, where real-time sensor data from a Physical Twin (PT) (i.e., a physical element) is utilized in the DT (i.e., a time-synchronized, virtual representation of the physical element) to enhance operations and maintenance (O&M) decision-making and predictive maintenance. Contribution III of this thesis focuses on developing an Integrated Probabilistic Risk Assessment (I-PRA) methodological framework for NPPs that are equipped with DTs. In this framework, a fully simulation-based approach is used, including a simulated high-fidelity PT model (i.e., Replicated PT), a physical degradation-based DT model, and a human model (containing simulation of O&M decision-making and an HRA-based maintenance human performance model). The I-PRA framework incorporates the simulation of the coupled human-PT-DT system into PRA, while characterizing and propagating epistemic and aleatory uncertainties associated with their inputs. This research offers a theoretical foundation, grounded on highly relevant research articles and regulatory documents, for the interconnections within the coupled human-PT-DT system in the I-PRA framework. To establish the methodological and computational approach, existing methodologies supported in the literature (e.g., for interactions between the DT and PT) are adapted into the I-PRA framework where possible. The novel methodological and computational developments of the I-PRA introduced in this thesis, which are not found in the existing literature, include: (i) the bidirectional and dynamic interactions between the replicated PT model and the human model, incorporating simulations of maintenance activities (e.g., repair and replacement) and their effects on the state of the replicated PT in conjunction with the DT model; and (ii) the bi-directional integration of the coupled human-PT-DT system with PRA. Within the I-PRA framework, futuristic plant-level risk estimates are based on projected component performance from DTs, providing forward-looking insights to inform both short-term actions (e.g., immediate maintenance) and long-term O&M decisions (e.g., supporting RIPB regulation and applications, such as risk-managed technical specifications (RMTS)). Component reliabilities are estimated from the simulated output of the coupled human-PT-DT system (explicitly incorporating human-in-the-loop decision-making and maintenance performance), serving as inputs to the PRA and supporting short-term risk estimation (e.g., during plant mission time). The case study implementation for NPP piping reliability analysis showed an 80% decrease in the mean cumulative leak probability at the end of the component lifetime, demonstrating the potential impact of DT-based predictive maintenance on component performance."]},{"key":"dc:format","label":"Dc Format","values":["application/pdf"]},{"key":"dc:title","label":"Title","values":["Temporal coupling of maintenance human performance, physical degradation, and digital twin models for probabilistic risk assessment in nuclear power plants"]}]}],"canonical_facts":{"dc:contributor":["Mohaghegh, Zahra","Sakurahara, Tatsuya","Bui, Ha","Sriver, Ryan","Alam, Syed","Vergari, Lorenzo"],"dc:creator":["Beal, John"],"dc:date":["2025-07-07","2025-08"],"dc:description":["Submission published under a 24 month embargo labeled 'Closed Access', the embargo will last until 2027-08-01","The student, John Beal, accepted the attached license on 2025-06-18 at 07:04.","The student, John Beal, submitted this Dissertation for approval on 2025-06-18 at 10:06.","This Dissertation was approved for publication on 2025-07-07 at 12:45.","DSpace SAF Submission Ingestion Package generated from Vireo submission #22343 on 2025-10-25 at 15:52:33","Probabilistic Risk Assessment (PRA) has played a critical role in improving the safety and performance of existing Nuclear Power Plants (NPPs) and is a key element of the U.S. Nuclear Regulatory Commission (NRC)’s Risk-Informed Performance-Based Regulation (RIPB). PRA methods have evolved over the years, and their continued advancement is increasingly important in light of the Accelerating Deployment of Versatile, Advanced Nuclear for Clean Energy Act of 2024 (or ADVANCE Act of 2024). Such advancements ensure that the safety assessment of next-generation NPPs will move beyond solely relying on prescriptive, deterministic, or subjective approaches, such as Maximum Credible Accident (MCA), which have demonstrated limitations in terms of safety and efficiency. To support the quantification of probabilistic inputs for the PRA of NPPs, this research utilizes modeling and simulation (M&S) to estimate repairable component reliability, especially in cases where empirical data is lacking. This is particularly relevant for new reactor designs or aging plants, where historical data may no longer be applicable due to design changes or evolving operational approaches, such as the integration of digital twins (DTs) into operations. Furthermore, the use of M&S enhances risk management strategies by incorporating the underlying failure mechanisms directly into the PRA framework. Although the methodological contributions of this thesis are demonstrated for NPP reactor coolant system piping components subject to long-term stress corrosion cracking (SCC) using the Extremely Low Probability of Rupture (xLPR) Probabilistic Fracture Mechanics (PFM) code, they can be applied (with some adjustments) to other types of components and physical degradation mechanisms. Three key scientific contributions of this Ph.D. thesis include: Contribution I: Modeling component reliability by developing and coupling human reliability analysis (HRA)-based maintenance models with physical degradation models. The analysis of repairable components is a multi-faceted problem because the underlying process involves time-dependent interactions between physical phenomena and social mechanisms. The physical process can encompass random shocks and degradation mechanisms that can deteriorate the functional performance of a component. The social mechanisms are related to maintenance activities and involve human performance and organizational factors. A review of the related literature identified critical gaps related to (a) simulating the underlying maintenance performance model and (b) capturing the effects of interactions between maintenance performance models and physical degradation models. This Ph.D. research addressed the gaps by: (i) developing a human reliability analysis (HRA)-based maintenance performance model to quantify maintenance outcomes under multiple types of maintenance programs, and (ii) developing a methodological and computational platform to couple the HRA-based maintenance performance model with physics-of-failure models. Event sequences are generated using dynamic PRA methods to capture temporal interactions between HRA-based maintenance models and physics-of-failure models. Bi-directional, time-dependent interactions are explicitly modeled in this approach, enhancing the resolution of the coupling. Contribution II: Uncertainty-based validation of coupled maintenance-physical degradation models in the absence of maintenance performance data. This research investigates the key challenges associated with the validation of coupled maintenance-physical degradation simulation models, especially when little to no applicable empirical validation data is available. The Probabilistic Validation (PV) methodology, an uncertainty-based validation method, was leveraged and advanced. The original PV methodology required advancements for this research because (a) the nature of human performance models differs significantly from the physics-based simulation models for which the original PV methodology was demonstrated, (b) the original PV methodology could not validate simulation models with time-dependent, bi-directional feedback between models within the same hierarchical level, and (c) in the original PV methodology, it was assumed that limited output-level validation data would be available to estimate model-form uncertainty; however, this research creates an advanced approach to quantify model-form uncertainty in the absence of any output-level validation data. The literature review revealed inconsistent consideration of uncertain parameters in maintenance models and limited focus on human performance, preventing its uncertainty from being reflected in system-level risk assessments. Contribution II of this thesis addresses these gaps through two key steps: (i) developing a novel methodology to systematically identify, characterize, and quantify uncertainty in HRA-based maintenance models, even in the absence of output-level validation data and (ii) developing a computational platform to propagate uncertainty from the HRA-based maintenance model to the system response level using a nested Monte Carlo-based approach. The new methodology guides the construction of a graphical causal model that captures the sources of uncertainty and their interactions, followed by quantification using a Bayesian Belief Network (BBN)-based approach. This enables the coverage of a broader range of epistemic uncertainties, including model-form uncertainty, than existing methods. Contribution III: Development of bi-directional temporal coupling among human performance, physical twin, and digital twin (DT) models, and incorporating the coupled system into the PRA of NPPs. The nuclear industry is investigating the applications of DTs in NPPs, where real-time sensor data from a Physical Twin (PT) (i.e., a physical element) is utilized in the DT (i.e., a time-synchronized, virtual representation of the physical element) to enhance operations and maintenance (O&M) decision-making and predictive maintenance. Contribution III of this thesis focuses on developing an Integrated Probabilistic Risk Assessment (I-PRA) methodological framework for NPPs that are equipped with DTs. In this framework, a fully simulation-based approach is used, including a simulated high-fidelity PT model (i.e., Replicated PT), a physical degradation-based DT model, and a human model (containing simulation of O&M decision-making and an HRA-based maintenance human performance model). The I-PRA framework incorporates the simulation of the coupled human-PT-DT system into PRA, while characterizing and propagating epistemic and aleatory uncertainties associated with their inputs. This research offers a theoretical foundation, grounded on highly relevant research articles and regulatory documents, for the interconnections within the coupled human-PT-DT system in the I-PRA framework. To establish the methodological and computational approach, existing methodologies supported in the literature (e.g., for interactions between the DT and PT) are adapted into the I-PRA framework where possible. The novel methodological and computational developments of the I-PRA introduced in this thesis, which are not found in the existing literature, include: (i) the bidirectional and dynamic interactions between the replicated PT model and the human model, incorporating simulations of maintenance activities (e.g., repair and replacement) and their effects on the state of the replicated PT in conjunction with the DT model; and (ii) the bi-directional integration of the coupled human-PT-DT system with PRA. Within the I-PRA framework, futuristic plant-level risk estimates are based on projected component performance from DTs, providing forward-looking insights to inform both short-term actions (e.g., immediate maintenance) and long-term O&M decisions (e.g., supporting RIPB regulation and applications, such as risk-managed technical specifications (RMTS)). Component reliabilities are estimated from the simulated output of the coupled human-PT-DT system (explicitly incorporating human-in-the-loop decision-making and maintenance performance), serving as inputs to the PRA and supporting short-term risk estimation (e.g., during plant mission time). The case study implementation for NPP piping reliability analysis showed an 80% decrease in the mean cumulative leak probability at the end of the component lifetime, demonstrating the potential impact of DT-based predictive maintenance on component performance."],"dc:format":["application/pdf"],"dc:identifier":["https://hdl.handle.net/2142/130127"],"dc:language":["en","eng"],"dc:rights":["Copyright 2025 John Beal"],"dc:subject":["Probabilistic Risk Assessment","Nuclear Power Plants","Uncertainty Analysis","Human Reliability Analysis","Maintenance"],"dc:title":["Temporal coupling of maintenance human performance, physical degradation, and digital twin models for probabilistic risk assessment in nuclear power plants"],"dc:type":["text"],"thesis:degree_discipline":["Nuclear, Plasma, Radiolgc Engr"],"thesis:degree_level":["Dissertation"],"thesis:degree_name":["Ph.D."],"thesis:institution_name":["University of Illinois Urbana-Champaign"]},"updated_at":"2026-07-22T22:25:06Z"}