Results for “failure prediction” — Page 5

DOE OSTI2023

Simulations of biomass compression-screw feeding using a compressible non-Newtonian constitutive model

There is global interest in the conversion of biomass into sustainable low-carbon-footprint fuels and chemicals as an alternative to non-renewable fossil feedstocks. Feeding biomass solids into pressurized reactors is one of the key steps in biomass conversion. Predicting mechanical failure and energy requirements for this step helps avoid upstream processing bottlenecks and enables efficient operation of a biorefinery. Here, in this work, we developed a predictive computational model for biomass screw feeders that capture the highly viscous, non-Newtonian and compressible behavior of biomass slurries. Biomass compressible behavior is formulated by an equation of state and the non-Newtonian rheology is represented by a density-dependent viscosity model. Experimental data from two compression screw-feeder systems are presented as a validation for our model. Our model successfully predicted the location of the compressed biomass “plug”, biomass flow rate, and the required torque at different operating conditions for the experimental conditions studied in this work.

09 BIOMASS FUELS↗

DOE OSTI2020

The Corrective Maintenance Paradigm Shift at Hanford's Tank Farms - 20077

Hanford's Tank Farms facilities have been used to safely store waste for over 70 years, with the first single-shell tanks being constructed in 1943. Tank Farm facilities consist of 149 single-shell tanks, 28 double-shell tanks, an evaporator facility, and wastewater treatment facilities. Tank Farm facilities are aging, with a tremendous corrective maintenance burden on the Tank Farm contractor. The mission of Tank Farm facilities is soon changing from waste storage to waste staging for the Hanford Waste Treatment and Immobilization Plant (WTP). WTP operations will demand a significant increase in Tank Farm facility operations, in which corrective maintenance outage windows will shrink drastically. This realization has forced the Tank Farm contractor to consider a paradigm shift in Tank Farm facilities Maintenance planning, and the use of reliability Engineering tools. The Tank Farm Production Operations Engineering Cognizant System Engineering (CSE) organization has led the way in motivating this paradigm shift. This shift has been realized through the use of: 1) technical exchange with other Department of Energy (DOE) contractors to develop improvements in the CSE program, 2) a shift from the use of lagging to leading system health indicators, and 3) a Plant Health Committee to unite Engineering, Operations, and Maintenance personnel toward a productive maintenance strategy. The CSE organization has held several technical exchanges with other DOE contractors to discuss CSE concepts, and how to better maintain aging infrastructure. The technical exchange with other contractors has greatly reduced the time required to make improvements in the Tank Farm CSE program. Other DOE contractors have already faced issues surrounding aging infrastructure, and have vast experience in improving the reliability and usable life of structures and components in nuclear facilities. The past CSE program used lagging health indicators to determine the health of systems. The key lagging indicator used to determine system health was availability, which is the percentage of time that a facility was ready for operation compared to the time the facility was demanded for operation. Availability was a good indicator of health in the waste storage mission of Tank Farms, where safe storage was the most important function of the facility, and where maintenance outage windows were typically long-duration. In current and future operations, outage windows are reducing, resulting in the need for much more reliable systems. Systems that have had high availability may suddenly become inoperable due to a failed component or sub-system. In several instances, the use of availability as an indicator of system health failed to predict system/equipment failure before its occurrence. In discussions with other DOE contractors, a set of reliability tools, including leading indicators of health, has been implemented in the CSE program. This primarily involves the use of failure modes and effects analysis and the study of equipment failure to develop system monitoring plans that focus on trending data to detect oncoming equipment failure ahead of time. In addition, the use of a Plant Health Committee has added significantly to the paradigm shift from a corrective maintenance philosophy to the use of predictive and preventive maintenance. The Plant Health Committee is a chartered team consisting of Engineering, Operations, and Maintenance personnel. CSEs use this forum to present the results of their performance monitoring, including the presentation of health via leading health indicators. The most positive aspect of this committee is the communication that it creates within these critical organizations. The Operations and Maintenance organization benefit from focusing maintenance on the reliability-centered focus provided by Engineering. Engineering benefits from the operational experience of the Operations organization and from the failure data that can be provided by Maintenance personnel. The continued use of the Plant Health Committee is expected to further decrease maintenance outage times, in better support of oncoming 24/7 operations. (authors)

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

DOE OSTI2020

Thermodynamic Consistent Neural Networks for Learning Material Interfacial Mechanics

For multilayer materials in thin substrate systems, interfacial failure is one of the most challenges. The traction-separation relations (TSR) quantitatively describe the mechanical behavior of a material interface undergoing openings, which is critical to understand and predict interfacial failures under complex loadings. However, existing theoretical models have limitations on enough complexity and flexibility to well learn the real-world TSR from experimental observations. A neural network can fit well along with the loading paths but often fails to obey the laws of physics, due to a lack of experimental data and understanding of the hidden physical mechanism. In this paper, we propose a thermodynamic consistent neural network (TCNN) approach to build a data-driven model of the TSR with sparse experimental data. The TCNN leverages recent advances in physics-informed neural networks (PINN) that encode prior physical information into the loss function and efficiently train the neural networks using automatic differentiation. We investigate three thermodynamic consistent principles, i.e., positive energy dissipation, steepest energy dissipation gradient, and energy conservative loading path. All of them are mathematically formulated and embedded into a neural network model with a novel defined loss function. A real-world experiment demonstrates the superior performance of TCNN, and we find that TCNN provides an accurate prediction of the whole TSR surface and significantly reduces the violated prediction against the laws of physics.

Zhang, Jiaxin↗

DOE OSTI2024

Machine Learning and Data Science to Advance Laboratory Earthquake Prediction and Illuminate the Mechanics of Precursors to Failure

Earthquakes represent one of our greatest natural hazards and in recent years human induced seismicity is adding to the threat. Even a modest improvement in the ability to forecast devastating large earthquakes or smaller shallow events associated with fluid injection could save thousands of lives and billions of dollars. Current efforts to forecast earthquakes are limited by knowledge of earthquake physics and hampered by a lack of reliable lab or field observations. However, recent work has provided a critical opportunity for advancement. We have found: 1) clear and consistent precursors prior to earthquake-like failure in the laboratory and 2) that lab earthquakes can be predicted using machine learning (ML). These works show that stick-slip failure events –the lab equivalent of earthquakes– are preceded by a cascade of micro-failure events that radiate elastic energy in a manner that foretells catastrophic failure. Remarkably, ML predicts the fault zone stress state, the failure time and in some cases the magnitude of lab earthquakes. In addition, the observations include clear precursors to failure in the form of changes in fault zone properties prior to lab earthquakes. Precursors have been observed in previous laboratory studies but their origin is poorly understood and their possible connection to ML based earthquake prediction is unknown. The work conducted under our project has dramatically expanded these efforts. We have developed an integrated data science approach to illuminate the physics of earthquake precursors and lab earthquake prediction. Our work has accelerated the development of ML, artificial intelligence (AI), and related data science approaches by providing massive data sets that are tightly connected to critical scientific problems and by bringing together leading subject matter experts and data scientists. Earthquake physics involves phenomena that are far from equilibrium. Our work has leveraged data science methods to illuminate these phenomena and investigate how they relate to earthquake prediction. In addition to a large database with many types of labeled events that is available to everyone, our work has advanced the fundamental understanding of seismic forecasting, earthquake physics, and fault rheology

58 GEOSCIENCES↗

DOE OSTI2021

Machine Learning for Improved Availability of the SNS Klystron High Voltage Converter Modulators

Beam availability has increased at the SNS, however, the targeted availability is greater than 95 %, while the SNS has failed to meet lower targets in the past. The HVCM used to power the linac klystrons have been one source of lost beam time and was chosen to explore using AI/ML techniques to improve reliability. Among the possibilities being explored are automating the tuning of HVCMs and predicting component failures such as capacitor aging, rectifier assemblies containing hundreds of diodes, and insulating oil degradation. The methodology pursued includes data cleaning, de-noising, post-analysis data labeling, and machine learning model development. We explore using Long Short-Term Memory and autoencoders for anomaly detection and prognostication used to schedule maintenance. We evaluate the use of model regularizers and constraints to improve the performance of the model and investigate methods to estimate the uncertainty of the models to provide a robust prediction with statistical interoperability. This paper describes the operational experience and known failures of the HVCMs and the proposed ML methodology and the preliminary results of training the AI/ML algorithms.

Pappas, G. C.↗

DOE OSTI2021

Machine Learning for Improved Availability of the SNS Klystron High Voltage Converter Modulators

Beam availability has increased at the SNS, however, the targeted availability is greater than 95 %, while the SNS has failed to meet lower targets in the past. The HVCM used to power the linac klystrons have been one source of lost beam time and was chosen to explore using AI/ML techniques to improve reliability. Among the possibilities being explored are automating the tuning of HVCMs and predicting component failures such as capacitor aging, rectifier assemblies containing hundreds of diodes, and insulating oil degradation. The methodology pursued includes data cleaning, de-noising, post-analysis data labeling, and machine learning model development. We explore using Long Short-Term Memory and autoencoders for anomaly detection and prognostication used to schedule maintenance. We evaluate the use of model regularizers and constraints to improve the performance of the model and investigate methods to estimate the uncertainty of the models to provide a robust prediction with statistical interoperability. This paper describes the operational experience and known failures of the HVCMs and the proposed ML methodology and the preliminary results of training the AI/ML algorithms.

Pappas, Chris↗

DOE OSTI2021

Machine Learning for Improved Availability of the SNS Klystron High Voltage Converter Modulators

Beam availability has increased at the SNS, however, the targeted availability is greater than 95 %, while the SNS has failed to meet lower targets in the past. The HVCM used to power the linac klystrons have been one source of lost beam time and was chosen to explore using AI/ML techniques to improve reliability. Among the possibilities being explored are automating the tuning of HVCMs and predicting component failures such as capacitor aging, rectifier assemblies containing hundreds of diodes, and insulating oil degradation. The methodology pursued includes data cleaning, de-noising, post-analysis data labeling, and machine learning model development. We explore using Long Short-Term Memory and autoencoders for anomaly detection and prognostication used to schedule maintenance. We evaluate the use of model regularizers and constraints to improve the performance of the model and investigate methods to estimate the uncertainty of the models to provide a robust prediction with statistical interoperability. This paper describes the operational experience and known failures of the HVCMs and the proposed ML methodology and the preliminary results of training the AI/ML algorithms.

Pappas, Chris↗

DOE OSTI2025

Quantifying precursors to void nucleation and coalescence in aluminum

Ductile rupture is a common failure mode for engineering alloys. It is generally comprised of three mechanisms: void nucleation at secondary particles, void growth, and coalescence of voids. Conventional models for ductile rupture have limited precision for predicting macroscopic failure. This is partially because they have been corroborated using ex-situ observations of these three mechanisms and calibrated by macroscopic strain or stress metrics. In this study, in-situ high-energy X-ray characterization was conducted to correlate sites of void nucleation via particle cracking and sites of void growth with grain-scale metrics. Here it is shown that particle cracking is not predicated by elevated stress metrics, which disagrees with classical models. Instead, particle cracking tended to occur for the largest, least spherical particles. This trend persisted on the aggregate-scale and within individual neighborhoods around particles. Furthermore, an Eshelby analysis illuminated a statistically significant increase in maximum principal stress within a cracked particle, compared to an uncracked particle. In-situ observations also revealed two separate examples of intragranular void growth comprised of flat, crack-like features that grew showing alignment with slip systems of the highest resolved shear stress. Post-mortem fractography revealed a secondary population of dimples on these crack-like features, implying that void sheeting may be occurring during this crack-like void growth. Furthermore, these in-situ observations imply that classical models for void growth, that assume a homogenous medium, should be extended to account for the anisotropic grain-level behavior to correctly capture distinct features of void nucleation and growth.

Al-2219↗

DOE OSTI2022

High Accuracy Indicators of Androgen Suppression Therapy Failure for Prostate Cancer—A Modeling Study

Prostate cancer is a serious public health concern in the United States. The primary obstacle to effective long-term management for prostate cancer patients is the eventual development of treatment resistance. Due to the uniquely chaotic nature of the neoplastic genome, it is difficult to determine the evolution of tumor composition over the course of treatment. Hence, a drug is often applied continuously past the point of effectiveness, thereby losing any potential treatment combination with that drug permanently to resistance. If a clinician is aware of the timing of resistance to a particular drug, then they may have a crucial opportunity to adjust the treatment to retain the drug’s usefulness in a potential treatment combination or strategy. In this study, we investigate new methods of predicting treatment failure due to treatment resistance using a novel mechanistic model built on an evolutionary interpretation of Droop cell quota theory. We analyze our proposed methods using patient PSA and androgen data from a clinical trial of intermittent treatment with androgen deprivation therapy. Our results produce two indicators of treatment failure. The first indicator, proposed from the evolutionary nature of the cancer population, is calculated using our mathematical model with a predictive accuracy of 87.3% (sensitivity: 96.1%, specificity: 65%). The second indicator, conjectured from the implication of the first indicator, is calculated directly from serum androgen and PSA data with a predictive accuracy of 88.7% (sensitivity: 90.2%, specificity: 85%). Our results demonstrate the potential and feasibility of using an evolutionary tumor dynamics model in combination with the appropriate data to aid in the adaptive management of prostate cancer.

59 BASIC BIOLOGICAL SCIENCES↗

DOE OSTI2021

wtDigiTwin (Wind Turbine Digital Twin) [SWR-21-15]

This wind turbine digital twin software (wtDigiTwin) provides a digital twin solution for wind turbine applications. The focus of wtDigiTwin is to estimate loads, motions and environmental conditions for an operating wind turbine. The program uses supervisory control and data acquisition (SCADA) measurements as inputs, together with a wind turbine model. The wind industry is currently challenged by the high cost of operation and maintenance. These costs could be mitigated if component failures are predicted, but such predictions are difficult unless the turbines are equipped with expensive measuring devices. The alternative is to use a digital twin such as wtDigiTwin to estimate the necessary signals. wtDigiTwin can perform online prediction of signals that are otherwise not measured, using a limited set of reliable measurements and a physics-based model. The predicted signals can be used in applications that have direct cost benefits: 1) real-time estimation of the fatigue consumption of key components of the wind turbine; 2) root cause analyses and failure detections ; 3) lifetime reassessments ; 4) improvements to follow-on designs. The current version provides examples to estimate wind speed, thrust, torque, tower-top position, and tower loads on an onshore wind turbine using the following measurements tower top acceleration, generator torque, pitch, and rotational speed. The model combines a linear state-space model, a wind speed estimator, and a Kalman filter algorithm that integrates measurements with the state model to perform state estimations. The state space model is obtained either using OpenFAST linearizations, or using the yams package provided with the software.

Branlard, Emmanuel↗

DOE OSTI2023

Determining Stress in Metallic Conducting Layers of Microelectronics Devices Using High Resolution Electron Backscatter Diffraction and Finite Element Analysis

Delayed failure due to stress voiding is a concern with some aging microelectronics, as these voids can grow large enough to cause an open circuit. Local measurements of stress in the metallic layers are crucial to understanding and predicting this failure, but such measurements are complicated by the fact that exposing the aluminum conducting lines will relieve most of their stress. In this study, we instead mechanically thin the device substrate and measure distortions on the thinned surface using high resolution electron backscatter diffraction (HREBSD). These measurements are then related to the stresses in the metallic layers through elastic simulations. This study found that in legacy components that had no obvious voids, the stresses were comparable to the theoretical stresses at the time of manufacture (≈300 MPa). Distortion fields in the substrate were also determined around known voids, which may be directly compared to stress voiding models. In conclusion, the technique presented here for stress determination, HREBSD coupled with finite element analysis to infer subsurface stresses, is a valuable tool for assessing failure in layered microelectronics devices.

HREBSD↗

DOE OSTI2025

Toward Drilling the Perfect Geothermal Well: An International Research Coordination Network for Geothermal Drilling Optimization Supported by Deep Machine Learning and Cloud Based Data Aggregation

The EDGE project, supported by the U.S. Department of Energy Geothermal Technologies Office under award DE-EE0008793, established a data-driven framework for improving the efficiency, cost-effectiveness, and reliability of geothermal well drilling. The project focused on developing scalable data infrastructure, advanced machine learning and probabilistic models, and integrated analytics tools to support continuous drilling optimization. A central objective was to reduce geothermal drilling costs by up to seventy percent while minimizing the risk of well failure through predictive diagnostics and adaptive planning. Over the project period, a comprehensive data repository was designed and deployed, incorporating records from over one hundred geothermal wells across varied geological settings. This repository supported both structured and unstructured data and adhered to FAIR data principles, enabling provenance tracking, quality control, and standardized metadata. The project introduced automated ingestion pipelines and a cloud-hosted platform that facilitated access to raw, processed, and derived datasets. This infrastructure served as the foundation for model development and analysis. Machine learning workflows were developed to predict key drilling metrics including rate of penetration, non-productive time, and total drilling costs. Self-organizing maps and dimensionality reduction methods were used to uncover operational patterns and outliers, while supervised learning algorithms such as random forests and deep neural networks were applied to forecast performance outcomes. The models were validated on heterogeneous datasets from both U.S. and Icelandic fields, demonstrating variable but significant predictive accuracy. The results indicated that finer temporal resolution, inclusion of lithological data, and consistency in operational annotations could substantially improve model performance. The project also implemented process mining techniques to reconstruct state-transition models from drilling event logs. These models enabled the identification of deviations from optimal workflows and provided insights into recurring failure modes. Analysis of non-productive time highlighted the impact of equipment failures, geological challenges, and human factors, offering opportunities for targeted mitigation strategies. The EDGE Dashboard was developed as a web-based expert system integrating data visualization, model outputs, and user-driven queries. It provided an accessible interface for operators to explore historical data, evaluate predicted outcomes, and compare drilling scenarios. Initial feedback from project partners suggested that the dashboard could serve as a foundation for more advanced advisory and optimization tools. Overall, the EDGE project demonstrated the feasibility and value of applying modern data science techniques to geothermal drilling. It delivered a set of interoperable tools and models that can support more efficient, lower-risk well development. The findings point toward a viable path for transitioning from advisory analytics to semi-autonomous drilling systems, contingent on continued collaboration, expanded datasets, and field validation. The project results have immediate relevance for drilling operations, data management practices, and future geothermal R&D efforts aimed at achieving reliable, cost-competitive geothermal energy at scale.

15 GEOTHERMAL ENERGY↗

DOE OSTI2022

LCOE reduction through proactively optimized monitoring of PV Systems (Final Technical Report)

The project demonstrates the value proposition for a high-resolution monitoring system (HRMS) with diagnostic-prognostic capability and determine its impact on LCOE. The HRMS differs from conventional monitoring systems in a number of ways. First it will include the capability to automatically measure IV curves at the string and module levels. This provides a much richer view into the DC performance of the PV system and allows classification of many typical failure and degradation modes. Second, it will incorporate software capable of quantifying power and energy losses in the field as well as define the location and mechanism of the power and energy loss. Moreover, we aimed to deliver a prognostic system that is capable of predicting certain failures before they occur and giving system operators the opportunity to more efficiently plan operations and maintenance (O&M) activity in order to lower costs and increase yield over the life of the system. The research identifies cost targets required for different monitoring stages, including at the string combiner, at the individual string, and at the module level, to lower the levelized cost of energy (LCOE). Finally, a comprehensive guide determines the value PV monitoring brings to PV field operations. The guide assists plant operators in maximizing value from existing plants and identify the trade-offs of different monitoring solutions for future plants depending on the size, location, and expected system lifetime.

14 SOLAR ENERGY↗

DOE OSTI2019

Ground-Truth Prediction to Accelerate Soft-Error Impact Analysis for Iterative Methods

Understanding the impact of soft errors on applications can be expensive. Often, it requires an extensive error injection campaign involving numerous runs of the full application in the presence of errors. In this paper, we present a novel approach to arriving at the ground truth--the true impact of an error on the final output--for iterative methods by observing a small number of iterations to learn deviations between normal and error-impacted execution. We develop a machine learning based predictor for three iterative methods to generate ground-truth results without running them to completion for every error injected. We demonstrate that this approach achieves greater accuracy than alternative prediction strategies, including three existing soft error detection strategies. We demonstrate the effectiveness of the ground truth prediction model in evaluating vulnerability and the effectiveness of soft error detection strategies in the context of iterative methods.

iterative methods, failure prediction, Soft error ↗

DOE OSTI2023

A Multicontinuum-Theory-Based Approach to the Analysis of Fiber-Reinforced Polymer Composites with Degraded Stiffness and Strength Properties Due to Moisture Absorption

Marine energy generation technologies such as wave and tidal power have great potential in meeting the need for renewable energy in the years ahead. Yet, many challenges remain associated with marine-based systems because of the corrosive environment. Conventional materials like metals are subject to rapid corrosive breakdown, crippling the lifespan of structures in such environments. Fiber-reinforced polymer composites offer an appealing alternative in their strength and corrosion resistance, but can experience degradation of mechanical properties as a result of moisture absorption. An investigation is conducted to test the application of a technique for micromechanical analysis of composites, known as multicontinuum theory and demonstrated in past works, as a mechanism for predicting the effects of prolonged moisture absorption on the performance of fiber-reinforced composites. Experimental tensile tests are performed on composite coupons with and without prolonged exposure to a salt water solution to obtain stiffness and strength properties. Multicontinuum theory is applied in conjunction with micromechanical modeling to deduce the effects of moisture absorption on the behavior of constituent materials within the composites. The results are consistent with experimental observations when guided by known mechanisms and trends from previous studies, indicating multicontinuum theory as a potentially effective tool in predicting the long-term performance of composites in marine environments.

36 MATERIALS SCIENCE↗

DOE OSTI2023

Reactivity-initiated accidents in two pressurized water reactor high burnup core designs

Here, this paper presents a safety analysis of two proposed core loadings for 24-month Pressurized Water Reactor (PWR) fuel cycles. This analysis focuses on reactivity-initiated accidents (RIAs) and evaluates core safety performance impacts of rod-averaged burnup limits up to 75 GWD/MTU and less than 7 % enriched UO 2 . The capabilities of Polaris, PARCS, and RELAP5-3D are leveraged to evaluate the core neutronic and thermal–hydraulic behavior for normal-operation, uncontrolled control rod withdrawal (CRW) transients, and control rod ejection (CRE) accidents. The two core designs are compared to identify features of realistic high burnup/extended enrichment core design approaches which have significant safety impact, identify experimental data needs for high-fidelity predictive modeling, and provide recommendations for future high burnup core designs. The first core design evaluated in this study was developed by Southern Nuclear Company and used an ZrB 2 Integral Fuel Burnable Absorber (IFBA) and B 4 C Wet-Annular Burnable Absorber (WABA)-based burnable poison strategy. The second core design assessed in this work used a Gd 2 O 3 -doped UO 2 burnable poison, similar to that used in boiling water reactors or French PWRs. Results indicate that fuel thermal limits are maintained for limiting CRW and hot full power (HFP) CRE transients. Cladding failure is predicted for the highest energy deposition rods in each core during limiting hot zero power (HZP) CRE accidents (where maximum radially averaged enthalpy exceeds 120 cal/g), though licensing may be permissible with a limited number of failed rods. While concerns exist regarding high critical boron concentration during steady state for the IFBA core and large plenum pressures for the gadolinia core design, the analysis demonstrates adequate safety performance during limiting RIA accident scenarios for two representative high burnup core designs. Design changes limiting plenum pressures and implementation of accident tolerant fuel (ATF) cladding features which minimize hydriding and susceptibility to pellet-cladding mechanical interaction (PCMI) are recommended for future high burnup fuel concepts. To support the technical basis for burnup limit increases, high-fidelity fuel performance models are needed to address physical effects not considered in this analysis, and high burnup irradiated fuel tests are required to extend applicability of the fuel failure limits and validate existing and future models.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

DOE OSTI2025

Risk-informed Graded Approach for Reliability and Performance Assessment of Machine Learning and Artificial Intelligence for Advanced Condition Monitoring Techniques

With the shift away from time-based maintenance and toward condition-based maintenance, and to reduce overall maintenance costs, there has been an upsurge in the usage and development of advanced condition monitoring (ACM) techniques for real-time monitoring of nuclear power plant (NPP) components. ACM is particularly useful in the development of digital twins, which are designed to predict the failure or degradation of plant components. Successful implementation of ACM requires an assessment to inform the development of a risk-informed approach to evaluate the use of ACM to meet Nuclear Regulatory Committee (NRC) regulations for in-service testing (IST) programs. This includes the monitoring and diagnostics of reactor components and systems in current, new, and advanced reactors. A key component in ACM is the usage of machine learning (ML) and artificial intelligence (AI) algorithms that can employ real-time data from instrumentation and sensors to detect and predict reactor component degradations. Such predictive capabilities enable early detection of component degradation so as to help plant personnel plan and execute necessary maintenance. For successful implementation of ML/AI in ACM such that regulatory requirements are met, a risk-informed graded approach is needed to assess the reliability and performance of ML/AI for ACM. The American Society for Mechanical Engineers (ASME) developed their Operations and Maintenance (O&M) Code to provide guidance on safe, reliable O&M of NPPs. The IST section of the O&M Code specifically establishes requirements for IST and examination to gauge operational readiness of components in water-cooled NPPs. This paper presents a state-of-the-art review of how reliability and risk assessment can be integrated with ACM to assess component performance by non-nuclear industries. This is followed by different methodologies and approaches for conducting performance and reliability assessments so as to meet IST requirements for NPP components.

97 - MATHEMATICS AND COMPUTING↗

DOE OSTI2025

Risk-informed Graded Approach for Reliability and Performance Assessment for Advanced Condition Monitoring Techniques

With the shift away from time-based maintenance and toward condition-based maintenance, and to reduce overall maintenance costs, there has been an upsurge in the usage and development of advanced condition monitoring (ACM) techniques for real-time monitoring of nuclear power plant (NPP) components. ACM is particularly useful in the development of digital twins, which are designed to predict the failure or degradation of plant components. Successful implementation of ACM requires an assessment to inform the development of a risk-informed approach to evaluate the use of ACM to meet Nuclear Regulatory Committee (NRC) regulations for in-service testing (IST) programs. This includes the monitoring and diagnostics of reactor components and systems in current, new, and advanced reactors. A key component in ACM is the usage of machine learning (ML) and artificial intelligence (AI) algorithms that can employ real-time data from instrumentation and sensors to detect and predict reactor component degradations. Such predictive capabilities enable early detection of component degradation so as to help plant personnel plan and execute necessary maintenance. For successful implementation of ML/AI in ACM such that regulatory requirements are met, a risk-informed graded approach is needed to assess the reliability and performance of ML/AI for ACM. The American Society for Mechanical Engineers (ASME) developed their Operations and Maintenance (O&M) Code to provide guidance on safe, reliable O&M of NPPs. The IST section of the O&M Code specifically establishes requirements for IST and examination to gauge operational readiness of components in water-cooled NPPs. This paper presents a state-of-the-art review of how reliability and risk assessment can be integrated with ACM to assess component performance by non-nuclear industries. This is followed by different methodologies and approaches for conducting performance and reliability assessments so as to meet IST requirements for NPP components.

99 - GENERAL AND MISCELLANEOUS↗