PSAM 16 - Session Th15 Overview

PSAM 16 Conference Session Th15 Overview

Session Chair: Ronald Boring (ronald.boring@inl.gov)

Paper 1 ME315

Lead Author: Tom Ulrich Co-author(s): Stephen Hancock, stephen.hancock@inl.gov Roger Lew, rogerlew@uidaho.edu Tyler Westover, tyler.westover@inl.gov Richard Boardman, richard.boardman@inl.gov

Simulators and Operating Concepts for Hydrogen Production
Operating concepts for close-coupled nuclear and hydrogen plants require simulator development to address operator human factors and performance issues. In addition, this research supports the development of plant-to-plant control systems interfaces that may involve mixed analog and digital distributed control system logic. This work supports operational safety while enabling nuclear plants to dispatch electricity to the hydrogen plant or the grid within minutes of a request from the grid operators. The Human Systems Simulation Lab at the Idaho National Laboratory was used to evaluate the ability of nuclear power plant operators to respond to normal and potentially off-normal events when switching between the traditional fully electric generation and hybrid electric and thermal energy operation modes to support hydrogen production with a nuclear power plant.

Simulators and Operating Concepts for Hydrogen Production

Operating concepts for close-coupled nuclear and hydrogen plants require simulator development to address operator human factors and performance issues. In addition, this research supports the development of plant-to-plant control systems interfaces that may involve mixed analog and digital distributed control system logic. This work supports operational safety while enabling nuclear plants to dispatch electricity to the hydrogen plant or the grid within minutes of a request from the grid operators. The Human Systems Simulation Lab at the Idaho National Laboratory was used to evaluate the ability of nuclear power plant operators to respond to normal and potentially off-normal events when switching between the traditional fully electric generation and hybrid electric and thermal energy operation modes to support hydrogen production with a nuclear power plant.

Paper ME315 | Download the paper file. | Download the presentation PowerPoint file.

Name: Tom Ulrich (thomas.ulrich@inl.gov)

Bio: Dr. Ulrich is a human factors and reliability research scientist at the Idaho National Laboratory. He has led and participated in several full-scope, full-scale simulator studies using the Human Systems Simulation Laboratory (HSSL) to investigate a range of nuclear control room topics. Dr. Ulrich possesses expertise in human performance assessment methodology with an emphasis on situation awareness and attention assessment via eye-tracking techniques. He is an expert in simulation and interface prototyping and has helped develop the advanced computerized operator support system (COSS), the Rancor Microworld, the ATRC digital control board, and numerous digital Turbine Control Systems. Dr. Ulrich’s active research includes dynamic human reliability analysis methodology development for nuclear power plant FLEX activities and development of prognostic human-machine interfaces (HMIs) for existing main control rooms and advanced reactor HMIs. Dr. Ulrich has also led an interdisciplinary res

Country: USA
Company: Idaho National Laboratory
Job Title: Human Factors and Reliability Research Scientist

Paper 2 NI96

Lead Author: Kanoko Nishiono Co-author(s): Marilia Ramos, marilia.ramos@ucla.edu Yoshikane Hamaguchi, hamaguchi_yoshikane_b3n@nra.go.jp Ali Mosleh, mosleh@ucla.edu

Dependency Analysis within Human Failure Events for Nuclear Power Plant: Comparison between Phoenix and SPAR-H
Robust and realistic Human Error Probability (HEP) estimation within Human Reliability Analysis (HRA) relies upon, among other factors, appropriate consideration of the dependency between human failure events (HFEs). The approach for assessing dependency varies throughout HRA methods. The reasoning and cognitive basis behind the different approaches for dependency, their quantitative rationale, and their impact on the HEP are still subject to investigation from the HRA community. This paper aims to discuss the characteristics of HRA methodology considering dependency through a comparison between two approaches: Phoenix, developed by the University of California, Los Angeles, and SPAR-H, developed by Idaho National Laboratories. The comparison of their qualitative frameworks will be performed through three elements: HRA variables, environmental factors considered, and causal modeling methods. Additionally, the following two elements will be discussed for comparing the quantitative analysis: dependency value estimation method and HEP estimation method considering dependency. SPAR-H and Phoenix include dependencies between HFEs. In SPAR-H, there are four factors to assess a dependency condition: Crew (same or different), Time (close in time or not close in time), location (same or different), and cues (additional or no additional). It determines the dependency level on five grades using a Dependency Condition Table. In Phoenix, the analyst selects several factors which may serve to influence the crew performance among eight factors (Performance Impact Factors - PIFs): Human System Interface, Procedures, Resources, Knowledge/Abilities, Team Effectiveness, Bias, Stress, and Task Load. Bayesian Belief Network models the causal model. The discussions of this paper aim to establish a foundation for a complete comparison between the approaches for dependency throughout an application within a Probabilistic Risk Assessment (PRA) scenario. Furthermore, our findings could be exploited for comparison between other HRA methods.

Dependency Analysis within Human Failure Events for Nuclear Power Plant: Comparison between Phoenix and SPAR-H

Robust and realistic Human Error Probability (HEP) estimation within Human Reliability Analysis (HRA) relies upon, among other factors, appropriate consideration of the dependency between human failure events (HFEs). The approach for assessing dependency varies throughout HRA methods. The reasoning and cognitive basis behind the different approaches for dependency, their quantitative rationale, and their impact on the HEP are still subject to investigation from the HRA community. This paper aims to discuss the characteristics of HRA methodology considering dependency through a comparison between two approaches: Phoenix, developed by the University of California, Los Angeles, and SPAR-H, developed by Idaho National Laboratories. The comparison of their qualitative frameworks will be performed through three elements: HRA variables, environmental factors considered, and causal modeling methods. Additionally, the following two elements will be discussed for comparing the quantitative analysis: dependency value estimation method and HEP estimation method considering dependency. SPAR-H and Phoenix include dependencies between HFEs. In SPAR-H, there are four factors to assess a dependency condition: Crew (same or different), Time (close in time or not close in time), location (same or different), and cues (additional or no additional). It determines the dependency level on five grades using a Dependency Condition Table. In Phoenix, the analyst selects several factors which may serve to influence the crew performance among eight factors (Performance Impact Factors - PIFs): Human System Interface, Procedures, Resources, Knowledge/Abilities, Team Effectiveness, Bias, Stress, and Task Load. Bayesian Belief Network models the causal model. The discussions of this paper aim to establish a foundation for a complete comparison between the approaches for dependency throughout an application within a Probabilistic Risk Assessment (PRA) scenario. Furthermore, our findings could be exploited for comparison between other HRA methods.

Paper NI96 | Download the paper file. | Download the presentation PowerPoint file.

Name: Kanoko Nishiono (nishiono_kanoko_yg4@nra.go.jp)

Bio: Kanoko NISHIONO research about Human Reliability Analysis and Dynamic HRA at Japan Nuclear Regulation Authority. In bachelor and master course, I worked on the environmental dynamics of Cs-137 and developing the internal dose estimation methods. I aim to be a researcher who can provide reliable risk assessment information to the public by utilizing the off-site knowledge I learned as a student and the on-site knowledge I developed in my work.

Country: JPN
Company: J-NRA
Job Title: Technical Official

Paper 3 RO162

Lead Author: Roger Lew Co-author(s): Ronald L. Boring Thomas A. Ulrich

Applications of the Gamified Rancor Microworld Simulator Model for Dynamic Human Reliability Simulation
Significant effort has been put into the development of high-fidelity thermohydraulic modeling for nuclear power and process control generally. Plants utilize full-scope simulators for engineering and training purposes. While not perfect, they tend to represent the physical configuration and control systems of plants with a high degree of fidelity. The disadvantages of such models are that they are complex, difficult to modify, and difficult to couple with other models. Full-scope simulators are also not optimized for speed, and even with modern computers conducting Monte Carlo simulations with 10s of thousands or 100s of thousands of runs is not logistically feasible. Reduced order models (ROMs) are simplified engineering models that validate particular aspects (e.g. steady state performance) against physical systems or higher fidelity models. They can then be utilized within their validated envelopes for gaining insights into engineered systems. ROMs address the complexity and coupling disadvantages of full-scope models due to their simplified nature. In the human factors domain an analogous problem exists with full-scope simulators. The simulators represent all of the sub-systems and components in physical plants and licensed operators go through years of training to operate these plants competently. Licensed operators are a finite and expensive commodity for laboratory human reliability studies to the extent that obtaining statistically useful error probability rates is not feasible. An alternative approach is to utilize microworld simulators such as DURESS/II or the Rancor Nuclear Power Plant Microworld Simulator. The Rancor Microworld Simulator was jointly developed by Idaho National Laboratory and the University of Idaho to investigate attention and situation awareness with novice operator. It has subsequently been used to design and validate the concept of operations, procedures, and interface design of an integrated energy system for nuclear power. A shortcoming of the Rancor Microworld was that the simulation model was only accessible through a graphical user interface. Here we describe how the simulation model of Rancor has been extracted and made accessible to a variety of platforms and applications by modularizing the model to a .NET core library that can be utilized from .NET compatible environments including Visual Studio and Unity3d. The model has also been ported to Python with the ability to load initial conditions, and trigger faults. Integrating it with the Human Unimodel for Nuclear Technology to Enhance Reliability (HUNTER) will enable dynamic human reliability simulations. The fidelity of the Rancor microworld is limited compared to full-scope simulators or even simplified educational/training simulators. However, the tradeoff is performance, with yet-to-be-optimized code running 100x the speed of full-scope simulators. Furthermore, the simplicity and flexibility of Rancor is favorable to proof-of-concept testing for HUNTER. Dynamic human reliability simulation fundamentally requires a model with a deterministic fault tree, the ability to specify the probability of faults and accept human control actions, and the ability to conduct enough simulation runs to capture the fault tree. The Rancor python model meets these specifications. In this manner the Rancor model can capture theoretical principles of dynamic human reliability analysis (DHRA) ahead of more lengthy and complex integration with higher fidelity models that more precisely capture the temporal and failure dynamics of nuclear power systems.

Applications of the Gamified Rancor Microworld Simulator Model for Dynamic Human Reliability Simulation

Significant effort has been put into the development of high-fidelity thermohydraulic modeling for nuclear power and process control generally. Plants utilize full-scope simulators for engineering and training purposes. While not perfect, they tend to represent the physical configuration and control systems of plants with a high degree of fidelity. The disadvantages of such models are that they are complex, difficult to modify, and difficult to couple with other models. Full-scope simulators are also not optimized for speed, and even with modern computers conducting Monte Carlo simulations with 10s of thousands or 100s of thousands of runs is not logistically feasible. Reduced order models (ROMs) are simplified engineering models that validate particular aspects (e.g. steady state performance) against physical systems or higher fidelity models. They can then be utilized within their validated envelopes for gaining insights into engineered systems. ROMs address the complexity and coupling disadvantages of full-scope models due to their simplified nature. In the human factors domain an analogous problem exists with full-scope simulators. The simulators represent all of the sub-systems and components in physical plants and licensed operators go through years of training to operate these plants competently. Licensed operators are a finite and expensive commodity for laboratory human reliability studies to the extent that obtaining statistically useful error probability rates is not feasible. An alternative approach is to utilize microworld simulators such as DURESS/II or the Rancor Nuclear Power Plant Microworld Simulator. The Rancor Microworld Simulator was jointly developed by Idaho National Laboratory and the University of Idaho to investigate attention and situation awareness with novice operator. It has subsequently been used to design and validate the concept of operations, procedures, and interface design of an integrated energy system for nuclear power. A shortcoming of the Rancor Microworld was that the simulation model was only accessible through a graphical user interface. Here we describe how the simulation model of Rancor has been extracted and made accessible to a variety of platforms and applications by modularizing the model to a .NET core library that can be utilized from .NET compatible environments including Visual Studio and Unity3d. The model has also been ported to Python with the ability to load initial conditions, and trigger faults. Integrating it with the Human Unimodel for Nuclear Technology to Enhance Reliability (HUNTER) will enable dynamic human reliability simulations. The fidelity of the Rancor microworld is limited compared to full-scope simulators or even simplified educational/training simulators. However, the tradeoff is performance, with yet-to-be-optimized code running 100x the speed of full-scope simulators. Furthermore, the simplicity and flexibility of Rancor is favorable to proof-of-concept testing for HUNTER. Dynamic human reliability simulation fundamentally requires a model with a deterministic fault tree, the ability to specify the probability of faults and accept human control actions, and the ability to conduct enough simulation runs to capture the fault tree. The Rancor python model meets these specifications. In this manner the Rancor model can capture theoretical principles of dynamic human reliability analysis (DHRA) ahead of more lengthy and complex integration with higher fidelity models that more precisely capture the temporal and failure dynamics of nuclear power systems.

Paper RO162 | Download the paper file. |

Name: Roger Lew (rogerlew@uidaho.edu)

Bio: Roger is a associate research professor at the University of Idaho and holds a Master's in Human Factors Psychology and a Ph.D. in Neuroscience. His research interests include human factors for nuclear power, transportation safety, visual perception, and decision-support for pre- and post-wildfire modeling.

Country: USA
Company: University of Idaho
Job Title: Associate Research Professor