Abstract
After years of stagnation, interest in the development and construction of new nuclear power facilities is accelerating. Some of the proposed designs, such as small modular reactors and microreactors, are significantly different from the plants currently in operation. Therefore, much of the nuclear human factors knowledge acquired over the last 40 years may not be easily transferrable to these plants. At the same time, other kinds of human factors research, such as work on high-levels of automation and remote operation, become increasingly relevant. This panel aims to provide guidance for human factors practitioners in this shifting landscape by gathering experts with backgrounds in research, regulation, and operations.
Keywords
Introduction
Nuclear power generation is a significant source of reliable, low-emission electricity. There are over 90 reactors in operation in the United States (U.S.), and over 400 worldwide. The nuclear industry has experienced a number of ebbs and flows in terms of acceptance and viability on a global level throughout its history. Likewise, the nature and indeed existence of human factors research for the nuclear domain has experienced a similar ebb and flow.
Human Factors Engineering (HFE) has been a central area of concern in the design and operation of nuclear facilities, largely precipitated following the accidents at Three Mile Island (1979) and Chernobyl (1986), where issues including inadequate control room instrumentation, lack of procedures, risk awareness, normalization of risky behaviors, and poor teamwork and communication were found to be contributing factors. These events led to a strong emphasis on the important role of HFE for the design, operation, and oversight of nuclear facilities.
These historic events triggered important developments for HFE research and guidance for the nuclear domain and risk assessment methods (e.g., Human Reliability Analysis) that have since expanded to many other application domains.
These methods and guidance have since been tried-and-true in supporting the licensing and oversight of safe, traditional operations. Around the early 2000s, the nuclear industry was experiencing a “nuclear renaissance” with countries seeking to expand their safe, clean energy profiles through the commissioning of nuclear power plants. It was during this period that new approaches to the design of large light water reactors (LLWR) and operations using more passive safety systems and increased number of digital interfaces in the control room (e.g., the AP1000 reactor design by Westinghouse) were being developed by industry and eventually certified by regulatory bodies.
However, following the Fukushima accident (2011), which resulted in a changed global landscape and world-wide public perception of nuclear power, in coincidence with unanticipated construction costs for new LLWRs, construction of new reactors has come almost to a standstill. For example, only two new reactors have been built in the U.S. in decades, with several projects cancelled during the design or construction phase. Some countries even shut down all their reactors in the wake of Fukushima.
Regarding human factors research, after an initial spike in activities to address post-Fukushima concerns (e.g., response to extreme events and severe accidents), less emphasis was given to more future looking research and focus shifted to operating reactors. With many decades of operational experience from the current fleet of reactors, there is an abundance of knowledge to support the safety analysis for existing plants. Some interesting challenges arise from the lifetime extension projects needed to operate the current reactors beyond their original licensing period. As the majority of operating plants are operated using analog instrumentation and controls, these projects often involve significant digital upgrades to the plants and sometimes a complete upgrade of the main control room to screen-based interfaces with soft controls.
In the last few years, this landscape has started to change drastically with regard to interest in new and advanced reactor technology. In the U.S., a new regulatory framework is under development to accommodate these anticipated advancements. Driven by concerns around energy security and climate change, the demand for reliable, low-emission energy sources has intensified. While this has resulted in accelerated construction planning for current-generation reactors in some countries, it also highlighted the potential of new reactor technology design concepts in supporting the achievement of safe, reliable energy production. The U.S. Department of Energy (DOE) is granting funding towards the demonstration of new nuclear technologies as part of President Biden’s Clean Energy Act, which reinforces development efforts of these important technologies and novel designs. One such design concept includes Small Modular Reactors (SMR) which are designed to radically simplify the construction and operation of nuclear facilities while making them more scalable and easily deployable. If these benefits can be achieved, they could significantly contribute to a fast decarbonization of the energy sector. These designs have therefore received much public attention, with several companies proposing designs, and several governments worldwide considering pilot installations. Indeed, in 2022 the U.S. Nuclear Regulatory Commission (NRC) certified the NuScale Power SMR design.
In addition to SMRs, there are even smaller reactor designs such as microreactors that are being considered. Microreactors are on a scale similar to diesel generators and hold promise for use in remote or industrial settings that may not warrant larger generating capabilities but still benefit from the reliability afforded by nuclear power. Because microreactors do not require frequent refueling, they are especially attractive in settings where maintaining the supply chain for fossil fuel based electricity generation is challenging.
The rapidly rising interest in new nuclear installations using advanced technologies, digital interfaces, and concepts of operations implies that nuclear human factors is facing a period of new challenges where knowledge about the role of the operator in a significantly different operational concept will be required. Many of the factors that support the economic viability of these new technologies, such as increased use of automation and the resulting fewer human actions, introduce potentially new human factors concerns not able to be addressed by operating experience. The accumulated body of knowledge from current reactors will be of limited use. Unlike conventional plants, in SMRs and microreactors, several smaller reactor vessels run in parallel. Current proposals are for designs with between 1-12 units. Monitoring and operating multiple reactor units is a significantly different task from single-unit operation as seen in current reactors, in particular in case of emergencies.
At the same time, SMRs and microreactors will have higher levels of automation, potentially creating known human factors issues with transparency, workload and deskilling. SMRs will likely also have a much smaller workforce in the order of dozens rather than hundreds of on-site staff. Microreactors may have even fewer personnel.
In addition to SMRs and microreactors, other nuclear power concepts are being considered for applications other than just adding power to the grid. For instance, there is growing interest in the use of nuclear reactors to power large shipping vessels, to provide heating in remote communities, or for hydrogen production. Some of these facilities may have no on-site support staff but might be operated remotely from a control center. The human factors challenges of such installations would be different from both conventional plants and advanced, smaller nuclear power plants. Another dimension is added by potential threats from physical and cyber-attacks. Emergency or incident response for such events demands thorough consideration of human capabilities and limitations.
It is this evolving landscape that points to an era of new opportunities for the nuclear human factors domain. However, with these new opportunities comes uncertainty. Nuclear human factors may need to incorporate research areas that were less relevant in the past, such as increased use of high-level automation and remote operation. As we gain operational experience with real world application of newer reactor design and digital control rooms such as AP1000 and NuScale, we can begin to inform future research priorities.
The panel is an attempt to help a general human factors audience understand how their own research might support the nuclear industry in the future, and what skills and experience they might need in order to get involved in future nuclear human factors research by sharing some of this operational experience and presenting the associated challenges.
To this end, we have assembled a panel with a wealth of knowledge and experience in nuclear human factors, and with involvement in new reactor designs. To provide a balanced perspective, the panelists come from diverse professional background including research, regulation, and operation. With panelists from Europe, China and the U.S., we hope to give a comprehensive account of the current state of the art in nuclear human factors and challenges to be addressed in future research.
Panel Summary
Human-Automation Research Horizon for Multi-Unit Operations, Dr. Claire Blackett, Institute for Energy Technology (IFE), Norway
A key technological development in the design of advanced reactors such as SMRs or microreactors is that they will be very highly automated, and even capable of remote and/or autonomous operation. This design goal, in conjunction with increased use of passive safety systems, is frequently cited as a justification for reducing the numbers of staff that will be required to monitor and operate an advanced reactor plant. A reduction in staff is expected to deliver significant cost savings by comparison to a conventional large nuclear power plant, and so the idea of highly automated or autonomous plants is very attractive to power companies and plant operators. What is less clear, however, is the degree to which such high levels of automation or autonomy are actually achievable in practice, or what the effects may be on human performance for those staff who are still tasked with maintaining an overview of plant operations and safety. Research within the Halden HTO project suggests that higher levels of automation can actually increase cognitive complexity for control room operators. This may be compounded in a multi-unit control room, which is often proposed for advanced reactor plants, where operators have to oversee several highly automated or autonomous plants operating in parallel.
Recent high-profile events involving automation and autonomous systems in the automotive and aviation industries have highlighted the faults with many levels of automation (LOA) models that are used in industry today. These events suggest that simple like-for-like replacement of manually performed tasks with automated sequences can result in hidden automation failures going undetected, lack of understanding of human-automation interaction resulting in poor decision making by humans, and even human misunderstanding of the role, capabilities and limitations of automation in the system.
Automation has the potential to substantially augment human capabilities to reduce risk, enhance safety and enable more efficient operation of plants. However, switching from highly manual to highly automated, or even autonomous, operation of power plants requires significant investigation and evaluation. We need better understanding of the potential human factors issues, as well as human-automation interaction to avoid potential human error traps, such as those seen in the transport industries.
Regulatory Achievements in Operating, New, and Future Reactor Regulation in the United States, Dr. Brian Green, U.S. Nuclear Regulatory Commission (NRC)
Demands for carbon free energy and energy independence are rising, sparking heightened interest in, and accelerated development of, new nuclear reactor technologies. The human factors staff at the NRC have had many recent achievements licensing and constructing SMRs and new light water reactors, and has been developing draft regulation for advanced, non-light water reactors. This discussion will describe these recent successes, challenges, and the work that still needs to be done to ensure that human factors principles continue to be applied at nuclear plants.
In recent years, the NRC has completed design certification reviews for large light water reactors, including the AP1000. As of early 2023, two AP1000 units are nearly complete and preparing for operation in the United States. When these units are complete, they will be the first new nuclear plants to go online in the United States in decades. In 2022, the NRC completed the final inspections, tests, analyses, and acceptance criteria (ITAAC) inspections at Vogtle Unit 3, authorizing the facility to startup which occurred in March 2023. The ITAAC is nearly complete for Vogtle Unit 4.
Among these ITAAC inspections are a number of human factors activities including observing integrated system validation tests confirming that operators can use the digital controls to safely operate the plant under normal, abnormal, and accident conditions; observing validation tests confirming that operators can perform certain actions outside of the control room; and confirming that the environment in the final, as-constructed plant is suitable for use by operators.
The NRC also completed the design certification rulemaking for the NuScale Power design in 2022. This is the first SMR design certified in the U.S. The design uses a unique concept of operations, which includes control of twelve units by a single crew, which is different from the operating fleet, which uses a single crew for each reactor. The NRC staff was successful at reducing the time needed to conduct the licensing review, in part, due to an innovative approach to the review of NuScale’s human factors integrated system validation tests.
The NRC is conducting licensing reviews of multiple operating reactor control rooms that are being modernized by replacing analog components with new digital alarms, controls, and displays. Certain unique challenges were experienced due to limitations and availability of control room simulators for use in integrated system validation. Therefore, licensees and the NRC staff are currently gaining experience using glass top simulators during a multi-stage validation process, which helps provide additional flexibility in the licensing process, and which helps develop regulatory confidence in the modified design earlier in the design process.
Recent legislation has accelerated the need for performance-based, technology independent regulations for new advanced reactor designs. NRC human factors staff has incorporated the lessons learned from the projects described above and has developed a new draft regulation to be used with advanced non-light water reactors, which are generally smaller and rely on inherent safety systems rather than active systems monitored and operated by humans. As such, the new draft rule and supporting guidance consider the changing role of the operator in safety and proposes new regulations that are more flexible while still ensuring that operators have the human-system interfaces, training, procedures, and tools needed to safely operate the plant. The staff are continuing to develop guidance related to remote operation, autonomous operation, unique concept of operations, and other human factors topics.
The Urgency for Human Factors Research for Advanced Reactors, Dr. Ronald Boring, Idaho National Laboratory (INL)
Idaho National Laboratory (INL) has served as the birthplace of nuclear power in the U.S., from the first electricity generated by nuclear energy on December 20, 1951, through the prototypes for boiling water and pressurized water reactors that serve as the backbone of the U.S. commercial nuclear fleet, to early sodium-cooled fast reactor prototypes that underlie many next generation nuclear plant designs. Joining the 52 legacy reactors built at INL, there comes a host of new reactor construction, including near-term plans to build three microreactors as proofs of concept for advanced reactors, and siting for the first commercial SMR.
Although the Nuclear Renaissance of the early 2000s did not fully materialize as predicted, a new form of nuclear power is emerging with an even greater promise to change the energy landscape. Dozens of new reactor designs have emerged, often backed by startup companies exploring smaller scale reactors. The economy of scale for nuclear power may not be in large, baseload reactors such as the current fleet. Rather, smaller designs bring with them the opportunity to achieve construction savings by prefabricating modular systems and transporting them to reactor sites rather than constructing reactors piece-by-piece onsite. The scale of the reactors ranges from fission batteries (< 2MWe) to microreactors (2-10MWe) to SMRs (50- 300MWe), but individual units can be combined for greater plant output. These smaller reactors fill carbon-neutral gaps by replacing smaller scale fossil generating capabilities such as diesel or coal, providing local energy and price security for municipalities and remote industries.
Because many of the vendors for these new reactors represent smaller enterprises than the historically large reactor vendors, the focus of their research and development is on plant hardware such as innovative cooling systems, reactor control systems, and reactor fuels. The development timelines for new plants is aggressive, ranging from five years for the smallest reactor types to a dozen years for the more advanced plants such as those that move beyond light water reactor technology. Therein lies the challenge. A singular focus on advancing the reactor technology often comes in the absence of parallel developments in the human-system interfaces.
An accelerated timeline must also prioritize the development of control systems for advanced reactors to ensure timely licensing. We are setting the stage for a mismatch between the Technology Readiness Level and Human Readiness Level of new reactors. Such a mismatch can result in significant delays or even failure to deploy otherwise technologically and economically viable designs.
There is a need for active research to support advanced reactor vendors’ needs for human factors solutions. The current fleet of U.S. reactors predates the digital revolution and is gradually migrating its control rooms from analog to digital. As a lagging technology with high staffing requirements, current control rooms do not serve as a template for advanced reactors. Even the latest reactor—the Westinghouse AP1000—has main control rooms that, while mostly digital, represent a nearly quarter-century old concept of operations, foregoing many recent automation, monitoring, and prognostics developments in sister industries.
Research is necessary in a number of areas, not limited to: (1) advanced visualization technologies to assist in monitoring, (2) intelligent root cause alarm systems, (3) computerized operator support systems, (4) prognostics and predictive maintenance, (5) human-automation collaboration, (6) remote operations, (7) staffing reductions and ensuing crew communications, (8) load following and alternate uses of thermal power beyond electricity generation, (9) span of control by operators over multiple units in parallel, and (10) novel human error traps resulting from advanced human-systems interactions. This research should not be the purview of a single vendor, which creates a disproportionate practical and regulatory burden on the first-in-line reactor. It is important that industry-wide initiatives be put in place to pool resources and answer human factors problems systematically. Additionally, with the significant advances coming into play for nuclear control systems, there is the potential for nuclear power to find itself on the forefront of human-system interfaces. The time is ripe for sharing lessons learned across safety-critical industries that are moving into similar realms of human-automation collaboration.
A failure to conduct human factors research for advanced reactors risks propagating outdated control solutions or dampening the momentum and viability of new reactors. Slow-moving legacy nuclear funding must adapt to the pressing needs of advanced reactors. The timelines are immediate, and human factors is needed now.
Transitioning from Analog to Digital-Based Control Rooms: Experience with the AP1000 in the U.S. and China, R. Scott Egli, U.S. Nuclear Regulatory Commission (NRC)
Since the 1960s nuclear power generation has grown to be a very important part of the U.S. non-carbon emitting energy mix. All the currently operating U.S. nuclear facilities utilize Main Control Rooms (MCRs) that are primarily analog in nature. The forthcoming Westinghouse AP1000 units will be nearly entirely digital and utilize a completely different MCR design.
The NRC staff of the Technical Training Center (TTC) in Chattanooga, Tennessee have gained a wealth of knowledge and understanding from former experience as licensed operators in traditional analog control rooms and current experience with training and operations with the Westinghouse AP1000 design. TTC staff led the development of AP1000 training for NRC including the commissioning of an AP1000-like simulator and the certification of instructors on the platform. Additional operating experience in the U.S. includes training on the AP1000 MCR simulator at the Westinghouse headquarters and observations of operating training conducted on Vogtle’s AP1000 MCR simulators. The Vogtle units are located just outside Waynesboro, Georgia. The NRC has also observed MCR operations during startup testing activities in 2018 at the first implementation of the AP100 Sanmen Nuclear Power Station in China.
This operational experience has resulted in a good perspective about the pros and cons of the digital MCR environment in comparison to traditional operation in analog environment. The NRC has also conducted human factors research experiments examining operator performance in both the analog and digital MCR environments including operator use of computer-based procedure (CBP) systems. The introduction of CBPs, a digital feature of the MCR, changed the conduct of operations (ConOps) from a reader-doer where the senior reactor operator (SRO) reads and instructs the Reactor Operator (RO) and Balance of Plant (BOP) according to the traditionally paper-based procedures to a situation where the RO now owns the procedure and implements the actions while the SRO and BOP can follow along on their own desktop CBP display. This adjustment to the ConOps results in a quieter control room, because there is not the same need for out-loud communications, but perhaps enhanced situation awareness and ability for the SRO to maintain good overall perspective of the operational scenario as they are no longer in charge of place-keeping in a paper-based procedure.
New features, New Challenges and New Solutions for New Reactors, Shuhui Zhang, Shanghai Nuclear Engineering R&D Institute (SNERDI)
There are 54 units in operation and 16 units under construction in mainland of China at the end of year 2022. The 3rd generation reactors have become the main force of the nuclear energy industry in China, and at the same time, 4th generation reactors have achieved demonstrative breakthroughs, such as the grid connection of the high-temperature gas-cooled reactor in 2021. All these newly built reactors have advanced digital main control rooms. New features from digital technology and cogeneration bring advantages together with new challenges to the HFE and operation of reactors. Designers and utilities are working in a collaborative manner to generate more integrated solutions, collect deeper operation experiences, and progress new improvements.
Digital instrumentation and control (I&C) systems and human system interfaces (HSIs) play a vital role in the normal operation and emergency conditions. In advanced digital control rooms, the number of analog HSIs are limited and most of the hard controls are on system or function level. On the contrary, digital HSIs are much more sophisticated compared with traditional analog ones. Degraded HSIs and I&C conditions arise as a new challenge for HFE. Degradations vary from minor errors, which are hard for operators to identify, to the loss of the whole HSI resource which could impact the performance of the crew.
In the CAP1400 reactor, performance-based HFE tests with full scope simulators and operators were conducted to investigate different potential kinds of degradations in order to clarify the influence on the monitoring, analyzing, decision making, action, response process for an operator, and communication, task allocation manner for a crew. Examples of investigated degradations include: loss of half of visual display unit (VDU)-based workstations (available VDUs reduce from 4 to 2 for each operator); failure of certain soft controls; alarm system updating failure during accident; degradation of computerized procedure system (wrong suggestion of the adopted procedure or step); failure of computerized procedure system in the middle of accident mitigation. Findings are not only useful to improve current designs and refine the conduct of operations, but also helpful when new digital functions or systems are introduced in the future.
Cogeneration is the expansion of nuclear power generation technology in non-electric applications, such as heating, process heat, cooling, desalination and hydrogen. CAP1400 incorporates the cogeneration feature for city residents’ winter heating during the design stage. A separate HFE program was developed to ensure the integration of cogeneration function into the main control room shall not introduce HFE problems. All potential impacted HSI resources such as VDU-based displays, large panel displays, alarms, procedures were identified at first. Then, an analysis from a functions and tasks perspective was conducted separately to gather different HFE insights of the design. Finally, two scenarios were designed to support the performance-based validation tests to validate the new concept of operation. In the future, to achieve higher economic competitiveness or community value, the goal of a nuclear reactor facility would be diversified and complexed, and the integration level of main control room would be increased. HFE should play an important role in the change.
Footnotes
Acknowledgements
This document is a work of authorship of multiple institutions. The views expressed herein are those of the authors and do not necessarily represent an official position of the U.S. Nuclear Regulatory Commission or the U.S. Department of Energy. This material is declared as a work of the U.S. Government and is not subject to copyright protection in the United States.