Ensuring Interpretability of ML Technologies for Predictive Maintenance on Operating Nuclear Plants
Poster to explain app that we are presenting at the conference
46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
SEARCH · Engineering Papers
Results for “Operations and Maintenance”
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
Nuclear Power Fault Diagnostics and Preventative Maintenance Optimization
Operation and maintenance costs for nuclear power plants are very large. Reactors are starting to shut down even after their operating licenses have been extended, because they are not price competitive compared to other energy sources. The nuclear industry is witnessing early closure of nuclear power plants due to economic reasons despite excellent safety records. Therefore, it is imperative to reduce costs to prevent these early closures. This paper showcases recent research into advanced fault diagnostics techniques and preventative maintenance optimization to reduce these maintenance costs. This report focuses on the condensate and feedwater system for both pressurized and boiling water reactor systems. The computerized maintenance management system, which contains the plant’s digital record of all the corrective- and preventative-maintenance work orders, was used as a ground truth to locate potential faults and label the process data as healthy or faulty. Various feature extraction techniques were utilized to further differentiate the faults from the healthy data. Support vectors machines were used to categorize other test sets of process data as healthy or faulty through a cross validation procedure. Similar faults were not found within this system leading to preventative maintenance optimization. Unnecessary amounts of preventative maintenance lead to inflated maintenance costs. This paper summarizes the steps for preventative maintenance optimization from component health determination to recommendation for action. This optimization was completed for condensate pumps, condensate booster pumps, and the respective motors that drive them.
21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗
Demonstration of the Human and Technology Integration Guidance for the Design of Plant-Specific Advanced Automation and Data Visualization Techniques
Nuclear power continues to be a safe, reliable, and carbon-free electricity generating source for the United States, though the cost of operating and maintaining the current United States nuclear power plant fleet has become uncompetitive with other sources. This gap is attributed to the advent of new digital instrumentation and control technologies that other electricity generating industries are currently leveraging to streamline work and greatly reduce operating, maintenance, and support costs. Digital instrumentation and control systems and control room modernization offers significant opportunities to reduce operating and maintenance costs to ensure the continued operation of the existing United States light-water reactors.
22 GENERAL STUDIES OF NUCLEAR REACTORS↗
Abstract for CRADA between NETL and Electric Power Research Institute (EPRI) (AGMT-1113)
The National Energy Technology Laboratory (NETL) and EPRI (Participant) will collaborate in the investigation of flashovers in electric power distribution maintenance operations. Live line and barehand maintenance techniques are used on overhead transmission lines around the world. While flashovers during this type of work are rare, several unexplained flashovers have occurred in North America. In each case, available evidence suggests these occurred at steady state (60 Hz) system voltage and at voltage stress levels well below that which is known to cause flashover. Understanding the cause of these flashovers is essential so they can be avoided in the future and enhance line worker safety. Some utilities are considering adding sheds to their live line tools (e.g. hot sticks) as a possible mitigation measure. The effect of adding sheds to the hot sticks was modelled and published by Chalmers University. A strong case can be made for sheds if the model can be validated through laboratory testing. This research is envisioned to be a multiple-year project. The work proposed under this agreement represents the initial research steps which will involve reproducing, verifying, and extending the modeling research previously published by Chalmers University.
24 POWER TRANSMISSION AND DISTRIBUTION↗
Technoeconomic Analysis of the Miba Segmented Journal Bearing
This presentation describes a technoeconomic analysis using NLR's Windfarm Operations & Maintenance cost-Benefit Analysis Tool (WOMBAT) to compare operational expenses and levelized cost of energy for wind turbines using standard spherical roller bearings with a novel segmented journal bearing developed by Miba.
17 WIND ENERGY↗
Wind Turbine Maintenance Costs: Assessing the Potential of Gear Oil Improvements
Wind turbine operations & maintenance (O&M) costs constitute a sizable portion of total energy cost for wind power. There are many components in a utility-scale wind turbine that need to be lubricated with either oil or grease. This study uses gearbox oil as an example and assesses how lubricant technology improvements may impact wind turbine power production and levelized cost of energy. Using the modeling tools (i.e., WOMBAT, reV, and SAM) developed at National Renewable Energy Laboratory and lubrication oil technology scenarios defined based on inputs provided by ExxonMobil and industry stakeholders, we quantify the potential for increasing energy production and reducing maintenance costs across the current and future U.S. fleet of wind turbines as a result of improvements in lubrication technologies. The modeled improvements reduce the median levelized cost of energy by 1-2% across the U.S. fleet. Cumulative saving from gear oil technology improvements in 2050 for U.S. fleet is estimated to be approximately $6 billion. Extending the lubricant replacement interval has a larger impact on total cost and energy production than reducing lubricant cost.
costs↗
Artificial Linear Brush Abrasion of Coatings for Photovoltaic Module First Surfaces
Natural soiling and the subsequent requisite cleaning of photovoltaic (PV) modules result in abrasion damage to the cover glass. The durability of the front glass has important economic consequences, including determining the use of anti-reflective and/or anti-soiling coatings as well as the method and frequency of operational maintenance (cleaning). Artificial linear brush abrasion using Nylon 6/12 bristles was therefore examined to explore the durability of representative PV first-surfaces, i.e., the surface of a module incident to direct solar radiation. Specimens examined include silane surface functionalized-, roughened (etched)-, porous silica-coated-, fluoropolymer-coated-, and ceramic (TiO2 or ZrO2/SiO2/ZrO2/SiO2)-coated-glass, which are compared to monolithic-poly(methyl methacrylate) and -glass coupons. Characterization methods used in this study include: optical microscopy, ultraviolet-visible-near-infrared (UV-VIS-NIR) spectroscopy, sessile drop goniometry, white-light interferometry, atomic force microscopy (AFM), and depth-profiling X-ray photoelectron spectroscopy (XPS). The corresponding characteristics examined include: surface morphology, transmittance (i.e., optical performance), surface energy (water contact angle), surface roughness, scratch width and depth, and chemical composition, respectively. The study here was performed to determine coating failure modes; identify characterization methods that can detect nascent failures; compare the durability of popular contemporary coating materials; identify their corresponding damage characteristics; and compare slurry and dry-dust abrasion. This study will also aid in developing an abrasion standard for the PV industry.
41 EE - Solar Energy Technologies Office (EE-4S)↗
Survey of prospective techniques for molten salt reactor feed monitoring
Safeguards verification measurements of nuclear material content in fresh fuel salt for liquid-fueled molten salt reactors (MSRs) are likely to be required as part of nuclear material accountancy for International Atomic Energy Agency safeguards. Here, this paper presents a comprehensive review and evaluation of 18 potential candidate techniques to quantify total uranium and 235 U for input accountancy measurements for liquid-fueled MSRs. As part of an overall screening and down-selection effort to identify the most promising techniques for further development for an MSR feed monitoring system, this paper defines eight figures of merit (FOMs): reasonably achievable measurement uncertainty, measurement time required, capital cost, burden upon the facility operator, maintenance intensity, technological maturity, human capital requirements for operation, and whether the technique introduces a path for potential material removal. Each candidate technique is then evaluated across these FOMs to identify the techniques with the highest potential for future development for fresh fuel accountancy measurements in MSRs. Our findings indicate that no single technique or combination thereof currently has the requisite technological maturity for immediate implementation in nuclear material accountancy at a liquid-fueled MSR facility. While several promising techniques are identified, there is a critical lack of experimental data for most systems in the context of molten salt applications.
11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗
FFTF Acceptance and Startup Testing for GAIN
The Fast Flux Test Facility (FFTF) is the most recent liquid metal reactor (LMR) to be designed, constructed, and operated by the U.S. Department of Energy (DOE). The 400-MWt sodium-cooled, fast-neutron flux reactor plant was designed for irradiation testing of nuclear reactor fuels and materials for liquid metal fast breeder reactors. Following the demise of the breeder reactor program in the United States, FFTF continued to play a key role in providing a test bed for demonstrating performance of advanced fuel designs and demonstrating operation, maintenance, and safety of advanced liquid metal reactors. FFTF operations ceased in April 1992 after a determination by DOE that no combination of proposed missions was financially feasible over a ten-year period. The reactor is currently deactivated and in a long-term surveillance and maintenance (S&M) mode. This report provides information on the extensive and rigorous process that was used to conduct turnover from construction followed by acceptance and startup testing of the FFTF. This paper is in support of the Gateway for Accelerated Innovation in Nuclear (GAIN), which provides the nuclear energy community with access to the technical, regulatory, and financial support necessary to move new or advanced nuclear reactor designs toward commercialization while ensuring the continued safe, reliable, and economic operation of the existing nuclear fleet. The information obtained from the design, startup, and operation of the FFTF provides valuable insight for follow-on reactor projects, such as the Versatile Test Reactor (VTR), in the areas of plant system and component design, component fabrication, fuel design and performance, prototype testing, site construction, reactor startup and operations, and reactor deactivation and decommissioning (D&D). The focus of this report is on the process used to startup the FFTF and to ensure that operations could be conducted efficiently and safely. A reference section is provided of documents detailing the successful turnover and testing process implemented for startup of the reactor and its supporting systems. The documents listed can be retrieved upon request and are believed useful for future reactor startup endeavors.
21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗
RASA Field Performance at US International Monitoring Stations
This report attempts to collate, analyze, and summarize RASA performance over time at each of 11 US operated IMS stations. The performance over time was compared to routine maintenance operations. The data, in general, were summarized into a set analyzed almost exclusively in Excel. The data set allows for further probing into RASA performance over time. The data set would permit data to be routinely added for continued analysis. In general, all RASA systems ran within the CTBTO mandated the MDA requirements over their current lifetime of operations. The largest impacts to performance appeared to arise from variabilities of Germanium gamma detectors as they were replaced. It is noted that the air volumes of samples were not provided and had to be assumed to be constant over the period analyzed.
61 RADIATION PROTECTION AND DOSIMETRY↗
A reinforcement learning approach to long-horizon operations, health, and maintenance supervisory control of advanced energy systems
In this work, we develop a Reinforcement Learning (RL) approach to the supervisory control problem for advanced energy systems, such as novel nuclear reactors and other demand-driven, mission-critical, and component-health-sensitive energy plants. The inclusive problem landscape considered captures the stochastic confluence of plant performance, component health evolution, power demand from the grid, diverse maintenance actions, and operator-defined goals and constraints, all considered over meaningfully long-enough reasoning horizons. Key aspects of the proposed approach are a receding horizon control-inspired technique dictating time- or event-triggered supervisory policy (re-)constructions, as well as additional capability-enabling contributions such as timescale compression, to handle long reasoning horizons and uncertainty in parts of the problem, and practical yet demonstrably-effective handling of hybrid action spaces with continuous and discrete decision variables. The resulting algorithm consists of a simulation-based RL agent constructing stochastic supervisory control policies over nontrivial action spaces and for long horizons, applying the learned policy to the system for a much shorter interval, and perpetually repeating, to construct the next long-horizon policy. That next policy will only be applied, again, for a short interval, yet originally far-in-time events move progressively closer, their associated uncertainty decreases, and new events and aspects enter the reasoning horizon. The proposed methodology bridges fundamental receding horizon concepts with the unequivocally stronger and more scalable reasoning of contemporary RL. Numerical examples using Soft Actor–Critic Deep RL illustrate the operation and efficacy of the proposed technique for a power plant tasked with health-aware load following missions in a dynamic electricity market landscape.
97 MATHEMATICS AND COMPUTING↗
Artificial Intelligence/Machine Learning Technologies for Advanced Reactors (Workshop Summary Report)
A workshop on artificial intelligence and machine learning (AI/ML) for advanced reactors (AR) was held October 5-6, 2021. The workshop was to be attended in-person at ANL but COVID restrictions forced the workshop to go virtual. The objectives of the workshop were to identify the most promising AI/ML opportunities for improving advanced reactor design, optimizing plant performance, and enhancing economic competitiveness and to develop an understanding of the scientific, engineering and licensing challenges facing their application. The workshop planning committee included GAIN, EPRI and NEI and members of three national laboratories (ANL, INL, and ORNL). The workshop was attended by more than 200 individuals representing academic and scientific institutions and the nuclear power industry. The definition put forth for an AI/ML system was one that perceives its environment and takes actions that maximize its chance of achieving its goals. In this report AI/ML refers to next generation algorithms that include deep learning, statistical analysis and data analytics and associated scientific computing and their potential application to the design, licensing, operation and maintenance of ARs. These methods typically incorporate models built from process data and may also include data generated by simulations that represent the behavior of a system. The workshop was organized in response to the growing interest in application of AI/ML for improving the economic competitiveness of nuclear energy. Increasingly more resources are being allocated to investigating the benefits of AI/ML methods. The DOE created the Artificial Intelligence & Technology Office to promote their development. And within the Office of Nuclear Energy, resources have been allocated to explore and understand the potential benefits of AI/ML. Additionally, the national laboratories are strategically positioned with DOE computing facilities such as Summit, Perlmutter, Aurora and Frontier that support large-scale simulations, hybrid HPC models with AI surrogates, and the exploration of new types of generative models emerging from multi-model data streams and sources. The workshop was organized with members of the AR community to understand the effort and to identify the level of interest and progress in this emerging technology. The workshop discussions focused on identifying opportunities for AI/ML across diverse areas of the nuclear industry and identifying current scientific and engineering challenges for advanced reactors that might be addressed through transformational uses of AI/ML. Discussion panels focused on four high-interest technical domains for advanced reactors: design, maintenance and operations, energy storage, and materials. The results of those discussions are summarized in this report. This includes opportunities that were identified for exploiting AI techniques and methods to improve the efficacy and efficiency of reactor analysis and to improve the operation and optimization of advanced reactors. Advanced reactor developers expressed an interest in learning more about AI/ML methods and their application. This included understanding whether ML methods can provide an advantage over existing nonlinear data regression methods for collapsing high-fidelity simulation results into faster running models. A consensus emerged that AR advances planned for the next decade will benefit from the use of AI/ML tools. The need exists to understand and model complex systems across length scales and modalities. AI/ML is a tool for discovery that can yield a set of engineering principles for use by nuclear engineers, licensing bodies, and operators to solve problems in plant design, safety analyses, autonomous operation, and predictive maintenance. While AI/ML represents a new set of tools, an awareness by the nuclear community of the full potential is still in the early stages so there is a need to increase awareness. It appears that the wide-spread adoption of AI/ML tools for ARs would be facilitated by future educational workshops that describe foundational methods and capabilities and describe successful applications.
21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗
Co-design optimization of combined heat and power-based microgrids
With the emergent need for clean and reliable energy resources, hybrid energy systems, such as the microgrid, are widely adopted in the United States. A microgrid can consist of various distributed energy resources, for instance, combined heat and power (CHP) systems. Here, the CHP module is a distributed cogeneration technology that produces electricity and recaptures heat generated as a by-product. It is an energy-efficient technology converting heat that would otherwise be wasted to valuable thermal energy. For an optimal system configuration, this study develops a novel co-design optimization framework for CHP-based cogeneration microgrids. The framework provides the stakeholder with a method to optimize investments and attain resilient operations. The proposed co-design framework has a mixed integer programming (MIP) model that outputs decisions for both plant designs and operating controls. The microgrid considered in this study contains six components: the CHP, boiler, heat recovery unit, thermal storage system, power storage system, and photovoltaic plant. After solving the MIP model, the optimal design parameters of each component can be found to minimize the total installation cost of all components in the microgrid. Furthermore, the online costs from energy production, operation, maintenance, machine startup, and disruption-induced unsatisfied loads are minimized by solving the optimal control decisions for operations. Case studies based on designing a CHP-based microgrid with empirical data are conducted. Moreover, we consider both nominal and disruptive operational scenarios to validate the performance of the proposed co-design framework in terms of a cost-effective, resilient system.
42 ENGINEERING↗
ARM FY2024 Radar Plan
The U.S. Department of Energy (DOE) Atmospheric Radiation Measurement (ARM) user facility’s radar capability aims to provide high-quality radar observations to the scientific community, advancing the understanding of clouds and precipitation processes and ultimately improving climate models. ARM operates radars across a spectrum of frequencies with diverse scanning capacities (Table 1) to meet these scientific goals. The maintenance, operations, and management of these distributed radars at sites in various climate regimes require a substantial staffing commitment. However, the existing staffing resources are constrained, making it impractical to sustain the consistent high-quality operations that are expected by the ARM user community. In response, ARM needs to develop priorities for the radar team each year. This report lays out the forthcoming priorities for Financial Year 2024 (FY24), aiming to establish a structure for the ongoing management of radar operations and data processing.
54 ENVIRONMENTAL SCIENCES↗
Early-Stage Radiation Safety Analysis for the Spallation Neutron Source Second Target Station Bunker Operations
The Second Target Station project at Oak Ridge National Laboratory will develop a cold neutron source to meet growing experimental needs. This paper describes calculations of the residual dose rates associated with the monolith shield plug and the beamline bunker, two key conventional operations and radiation safety features. While neutron production is active, the instrument hall outside the bunker must be generally accessible with dose rates of less than 0.25 mrem/h. When neutron production is halted, the bunker must be accessible for hands-on maintenance operations. Further, these two requirements form the cause for the assessments reported herein of residual dose rates caused by the monolith shield plug and residual dose rates in the bunker. The monolith shield plug was shown to not produce significant dose rates inside the bunker after a 20-year lifetime, and the residual dose rates inside the bunker for the case of an operating beamline were shown to reasonably allow for hands-on maintenance. These calculations are based on preliminary design models of the relevant systems. Additionally, an example showing the significance of considering neutron supermirror physics in transport calculations that track nuclide production and destruction rates to produce gamma sources for residual dose rate calculations is included. The example shows that if neutron supermirror physics is not considered, dose rate fields may be significantly underpredicted.
61 RADIATION PROTECTION AND DOSIMETRY↗
Analytics-at-scale of Sensor Data for Digital Monitoring in Nuclear Plants (3 rd Annual Report)
Nuclear power plants collect and store large volumes of heterogeneous data from various components and systems. With recent advances in machine learning (ML) techniques, these data can be leveraged to develop diagnostic and short-term forecasting models to better predict future equipment condition. Maintenance operations can then be planned in advance whenever degraded performance is predicted, thus resulting in fewer unplanned outages and the optimization of maintenance activities. This enables lower maintenance costs and improves the overall economics of nuclear power. This report primarily focuses on developing a short-term forecasting process that leverages a feature selection process to distill large volumes of heterogeneous data and predict specific equipment parameters. A variety of feature selection methods, including Shapley Additive Explanations (SHAP) and variance inflation factor (VIF), were used to select the optimal features as inputs for three ML methods: long short-term memory (LSTM) networks, support vector regression (SVR), and random forest (RF). Each combination of model and input features was used to predict a pump bearing temperature both 1 and 24 hours in advance, based on actual plant system data. The optimal inputs for the LSTM and SVR were selected using the SHAP values, while the optimal input for the RF consisted solely of the response variable itself. Each model produced similar 1-hour-ahead predictions, with root mean square errors (RMSEs) of roughly 0.006. For the 24-hour-ahead predictions, differences could be seen between LSTM, SVR, and RF, as reflected by model performances of 0.036 ± 0.014, 0.0026 ± 0, and 0.063 ± 0.004 RMSE, respectively. As big data and continuous online monitoring become more widely available, the proposed feature selection process can be used for many applications beyond the prediction of process parameters within nuclear infrastructure. This report summarizes the Fiscal Year 2021 research progress encompassing the (1) data cleaning and feature selection necessary for ML applications; (2) development of short-term forecasting models to predict future plant process parameters for both single and multiple time steps ahead; and (3) validation of the feature selection methods and short-term forecasting models given new data from different systems.
46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
Development of Short-Term Forecasting Models Using Plant Asset Data and Feature Selection
Nuclear power plants collect and store large volumes of heterogeneous data from various components and systems. With recent advances in machine learning (ML) techniques, these data can be leveraged to develop diagnostic and short-term forecasting models to better predict future equipment condition. Maintenance operations can then be planned in advance whenever degraded performance is predicted, thus resulting in fewer unplanned outages and the optimization of maintenance activities. This enables lower maintenance costs and improves the overall economics of nuclear power. This paper focuses on developing a short-term forecasting process that leverages a feature selection process to distill large volumes of heterogeneous data and predict specific equipment parameters. A variety of feature selection methods, including Shapley Additive Explanations (SHAP) and variance inflation factor (VIF), were used to select the optimal features as inputs for three ML methods: long short-term memory (LSTM) networks, support vector regression (SVR), and random forest (RF). Each combination of model and input features was used to predict a pump bearing temperature both 1 and 24 hours in advance, based on actual plant system data. The optimal inputs for the LSTM and SVR were selected using the SHAP values, while the optimal input for the RF consisted solely of the response variable itself. Each model produced similar 1-hour-ahead predictions, with root mean square errors (RMSEs) of roughly 0.006. For the 24-hour-ahead predictions, differences could be seen between LSTM, SVR, and RF, as reflected by model performances of 0.036 +- 0.014, 0.0026 +- 0, and 0.063 +- 0.004 RMSE, respectively. As big data and continuous online monitoring become more widely available, the proposed feature selection process can be used for many applications beyond the prediction of process parameters within nuclear infrastructure.
46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
Equipment Self-Assessment Guide Checklist
This Equipment Self-Assessment Checklist is designed for asset owners and operators (AOOs) responsible for the deployment, operation, maintenance, or cybersecurity oversight of grid systems and digital energy technologies. It provides a structured inspection checklist for evaluating the security, integrity, and operational trustworthiness of equipment across substations, generation sites, distributed energy resources (DERs), and control environments.
32 - ENERGY CONSERVATION, CONSUMPTION, AND UTILIZA↗