Diablo Canyon will host a commercial installation of the first on-site generative artificial intelligence deployment at a U.S. nuclear plant.

Pacific Gas & Electric is deploying Atomic Canyon’s Neutron Enterprise to assist the utility’s management of datasets associated with operations of Diablo Canyon. The software, which runs on Nvidia’s full-stack AI platform, enables intelligent document processing, computation of semantic embeddings, and generative capabilities. Its infrastructure allows nuclear facilities to process and analyze vast amounts of complex documentation with unprecedented speed and accuracy, according to the company.

The Federal Energy Regulatory Commission has denied plans for Talen Energy to supply additional on-site power to an Amazon Web Services’ data center campus from the neighboring Susquehanna nuclear plant in Pennsylvania.

Washington, D.C. — Craig Piercy, CEO of the American Nuclear Society (ANS), issued the following statement:

"The American Nuclear Society applauds the announced partnerships between Google and Kairos Power and by Amazon and X-energy. Together, these deals will add at least 820 megawatts of zero carbon electricity to the U.S. energy supply. This is a major step toward securing the commercial deployment of advanced nuclear technologies that will make the world a cleaner and more prosperous place."

Atomic Canyon is developing a generative AI search for the nuclear energy sector and is working with the Department of Energy’s Oak Ridge National Laboratory to get it done. On September 26, Atomic Canyon announced its initial results about six months after the collaboration was first announced in March.

Generative artificial intelligence paired with advanced diagnostic tools could detect potential problems in nuclear power plants and deliver a straightforward explanation to operators in real time. That’s the premise of research out of the Department of Energy’s Argonne National Laboratory, and just one example of the DOE’s increasing exploration of AI applications in nuclear science and technology research. Training and restraining novel AI systems take expertise and data, and the DOE has access to both. According to a flurry of reports and announcements in recent months, the DOE is setting out its plans to ensure the United States can use AI to its advantage to enhance energy security and national security.

While no development details have been released, Constellation is asking to rezone 658.8 acres of land it owns around the Byron nuclear plant in Illinois for possible long-term use.

The U.S. Department of Energy plans to invest up to $900 million to support the initial deployment of small modular reactor technology.

The DOE issued a notice of intent to fund projects from President Biden’s infrastructure law with the goal of accelerating advanced nuclear projects to support energy infrastructure. The department estimates the country will need up to 950 gigawatts of reliable and clean energy to help reach the goal of net-zero emissions by 2050. Nuclear currently generates 18.6 percent of U.S. electricity.

The Nuclear Regulatory Commission hosted a public meeting yesterday to gather comments on its web-based ADAMS (WBA) system—a public document search tool introduced in 2010. It’s a tool that novice users find daunting and frequent users find frustrating, whether they’re searching for a single document or for thousands of documents on a single topic.

Ken Petersen
president@ans.org

Nuclear has been on a good roll lately and it is getting better. The 2022 Inflation Reduction Act (IRA) provides a nuclear power production tax credit. This has stopped the early retirement of deregulated units. The IRA also provides a benefit for the clean production of hydrogen. Many utilities have committed to a net-zero goal by 2050. Duke and other utilities have plans to transition coal plants to nuclear with small modular reactors.

And now, nuclear has a new supporter—tech companies.

The big U.S. utility companies (like Exelon, Duke, Dominion, Southern, and Entergy) are all projecting growth in electricity demand—primarily in the commercial sector but some residential growth is also expected. Commercial growth is being driven by new factories (thank you, IRA and CHIPS, that is, the Creating Helpful Incentives to Produce Semiconductors and Science Act). It is also being driven by data centers.

A team of engineers, physicists, and data scientists from Princeton University and the Princeton Plasma Physics Laboratory (PPPL) have used artificial intelligence (AI) to predict—and then avoid—the formation of a specific type of plasma instability in magnetic confinement fusion tokamaks. The researchers built and trained a model using past experimental data from operations at the DIII-D National Fusion Facility in San Diego, Calif., before proving through real-time experiments that their model could forecast so-called tearing mode instabilities up to 300 milliseconds in advance—enough time for an AI controller to adjust operating parameters and avoid a tear in the plasma that could potentially end the fusion reaction.

Researchers at Savannah River National Laboratory (SRNL), in concert with Lawrence Berkeley National Laboratory, Massachusetts Institute of Technology, Pacific Northwest National Laboratory, and Florida International University, are leading the Advanced Long-Term Environmental Monitoring Systems (ALTEMIS) project to move groundwater cleanup from a reactive process to a proactive process, while also reducing the cost of long-term monitoring and accelerating site closure.

The Department of Energy announced $29 million in funding for seven team awards for research in machine learning, artificial intelligence, and data resources for fusion energy sciences on August 31. In all, 19 institutions will build algorithms to address high-priority research opportunities in fusion and plasma sciences using interdisciplinary collaborations of fusion and plasma researchers teamed with data and computational scientists.

Imagine if your employees had a Virtual Assistant that could create condition reports, work requests, or plan work orders simply by asking for it. Or better yet, what if the Virtual Assistant could create work packages and plan your employees’ day automatically without even having to ask? How much more productive would we be if we all had a Virtual Assistant to help perform our work and capture and share our activities as they occurred?

To help plan and prepare for new technologies involving artificial intelligence, the Nuclear Regulatory Commission has released its Artificial Intelligence Strategic Plan (NUREG-2261) for fiscal years 2023–2027.

The NRC said that it expects license applications that include the use of AI technologies to be submitted to the agency for review and approval within the next few years. The strategic plan is meant to help ensure that NRC staff are prepared to review and evaluate such applications.

In the foreword, the NRC Office of Nuclear Regulatory Research director Raymond Furstenau introduces the strategic plan, writing, “We recognize that interest in AI is growing rapidly in both the public and private sectors. As such, I think [it] is important to lay the groundwork needed to ensure the safe and secure use of AI in NRC-regulated activities.”

Scientists are searching for new materials to advance the next generation of nuclear power plants. In a recent study, researchers at the Department of Energy’s Argonne National Laboratory showed how artificial intelligence could help pinpoint the right types of molten salts, a key component for advanced nuclear reactors.

The Nuclear Regulatory Commission has issued a request for comments as it develops a strategic plan for evaluating artificial intelligence in its regulations. Specifically, the NRC is asking for input on the agency’s overall AI strategy, as well as the strategic goals presented in the NRC’s draft report Artificial Intelligence Strategic Plan: Fiscal Year 2023–2027 (NUREG-2261).

The request for comments on the NRC’s AI Strategic Plan was issued in the July 5 Federal Register with a deadline of August 19. The NRC also plans to hold a public webinar on August 3 from 1–3 p.m. eastern time to receive comments on the draft plan.

Researchers from the Department of Energy’s Argonne National Laboratory are developing a “tool kit” based on artificial intelligence that will help better determine the properties of materials used in building a nuclear reactor.

Clifford

Lackey

Two student interns at Pacific Northwest National Laboratory looking for an easier way to monitor the acidity and phosphate concentrations of a process fluid like dissolved nuclear fuel have published research on a monitoring method that provides real-time data without the need for physical sampling of the substance. Their story was published on October 27 on PNNL’s website.

Student leaders: Hope Lackey conducted pH measurement and chemical analysis research during her Science Undergraduate Laboratory Internships (SULI) experience at PNNL in 2018 while she was working toward her undergraduate degree in environmental studies at the College of Idaho. Andrew Clifford, also a SULI intern and a student at the College of Idaho, partnered with Lackey between his junior and senior year, while studying for a dual bachelor’s in chemistry and math/physics.

Within the National Nuclear Security Administration, the Office of Defense Nuclear Nonproliferation Research and Development (DNN R&D) is leading efforts to drive advances in artificial intelligence and accelerate the adoption of AI-enabled technologies to solve nuclear nonproliferation and national security challenges.

The goal is to incorporate AI into advanced techniques for detecting nuclear weapons and materials. According to the NNSA, these detection capabilities support the nuclear nonproliferation and arms control goals of the United States, while also driving the development of new capabilities.

PPPL physicist Dan Boyer. (Photo: Amber Boyer/Kiran Sudarsanan)

Researchers at the Department of Energy’s Princeton Plasma Physics Laboratory are using machine learning to predict electron density and pressure profile shapes on the National Spherical Torus Experiment-Upgrade (NSTX-U), the flagship fusion facility at PPPL that is currently under repair.

The hope is that such predictions, generated by artificial neural networks, could improve the ability of NSTX-U researchers to optimize the components of experiments that heat and shape the fusion plasma.

“This is a step toward what we should do to optimize the actuators,” said PPPL physicist Dan Boyer, author of the paper, “Prediction of electron density and pressure profile shapes on NSTX-U using neural networks,” published by Nuclear Fusion, a journal of the International Atomic Energy Agency. “Machine learning can turn historical data into a simple model that we can evaluate quickly enough to make decisions in the control room or even in real time during an experiment.”