Seeing more, repairing smarter: How advanced NDE and repair technologies are changing nuclear asset management

The achievement represents more than a successful technology deployment. It reflects a broader shift taking place across the nuclear industry. Inspection and repair activities are becoming more digital, more intelligent, and more integrated. Utilities are looking for ways to improve outage performance, support long-term operation, preserve critical expertise, and maintain the highest standards of safety and quality. At the same time, they must manage aging assets, increasing data volumes, and complex inspection and repair challenges while maintaining compliance with regulatory codes and standards.

Recent advances in nondestructive evaluation (NDE), digital work execution, AI-assisted analysis, laser-based surface preparation, nonmetallic repair technologies, and adaptive welding are helping utilities meet those challenges. While the technologies differ in application, they share a common goal: enabling better decisions through more effective inspection and repair.

Better decisions through better inspection

Inspection has always been about more than just finding flaws. At its best, NDE provides the information needed to assess the structural integrity of an in-service component, whether repair is necessary, and what actions should be taken to maintain long-term reliability.

The same principle applies to repair technologies. A promising repair method becomes valuable only when it can be qualified, deployed, inspected, and accepted within the industry's rigorous regulatory and code framework.

Historically, the challenge wasn't collecting inspection data, it was turning that data into actionable information. Today's technologies are helping experts focus their attention where it adds the greatest value.

It follows that today's most important innovations are focused not only on new tools but also on improving the quality of decisions that utilities make about critical assets.

Increasingly, the industry's objective is to help experts focus their attention where it provides the greatest value. That trend is evident in advances ranging from ultrasonic testing and AI-assisted analysis to mixed-reality applications and remote welding technologies.

Replacing disruption with better information

One of the clearest examples of the industry's shift toward more-informed inspection and repair decisions is the potential for using ultrasonic testing (UT) in lieu of radiographic testing (RT).

For decades, RT has been the standard approach for volumetric weld inspection in nuclear construction and repair. The technique earned widespread acceptance because of its ability to produce a permanent visual record and identify changes in material density. However, decades of construction and operating experience have also revealed important limitations. RT requires exclusion zones, interrupts nearby work activities, and relies on radioactive sources that introduce additional safety and logistical considerations. On large projects and outage work, those impacts can create significant scheduling challenges.

The limitations are not solely logistical. RT produces a two-dimensional image that can reveal flaw length and width but cannot directly characterize flaw depth or proximity to critical surfaces. It is also less effective at identifying planar flaws such as cracks or lack of fusion, defect types that are often the most important from a structural integrity perspective. In some cases, flaws that are not detected during RT are later identified during UT performed for preservice inspection.

Modern UT offers a different approach. Advanced UT techniques can provide three-dimensional characterization of flaw size, orientation, and through-wall location. Rather than relying on changes in material density, UT evaluates reflected sound waves, making it particularly effective at identifying planar flaws and providing information that supports engineering evaluation. This allows utilities to move beyond simple workmanship-based acceptance criteria and make more-informed decisions about the significance of inspection findings.

The Electric Power Research Institute's recent technical basis work for new construction examines how qualified UT could serve as a practical alternative to RT while maintaining rigorous quality assurance standards. At the same time, EPRI has conducted benefit-cost analyses evaluating factors such as radiation protection, boundary protection, outage duration, parallel work impacts, inspection costs, and weld repairs. The work indicates that significant benefits may be achieved when exclusion zones are eliminated and inspection results provide more actionable engineering information.

In addition, structurally insignificant welding flaws identified and repaired during fabrication can introduce residual tensile stresses that later initiate in-service flaws. UT can help teams disposition such unnecessary weld repairs during fabrication.

Industry efforts are also focused on qualification. Demonstration programs are developing specimens for stainless and ferritic piping ranging from 2-inch (50.8mm) to 24-inch (609.6mm) diameters, supporting qualification programs based on the American Society of Mechanical Engineers (ASME) Code Case N-831-1 and related requirements. These efforts are intended to ensure that UT implementation is supported by the same level of rigor expected for nuclear applications.

Ronnie Swain, senior technical executive of plant support at EPRI, said that utilities are looking for ways to obtain better inspection information while minimizing the operational impacts associated with traditional radiography. "Ultrasonic testing provides an opportunity to improve both inspection quality and outage execution," he said. "The goal is not simply to replace radiography. The goal is to provide better information with less operational disruption."

The broader story is not that UT simply replaces RT. It is that the industry is seeking inspection methods that improve decision-making while reducing disruption. As utilities prepare for long-term operation and a new generation of nuclear construction projects, better information may prove just as valuable as more efficient inspections.

AI changes the inspection workflow

If UT demonstrates how inspection methods are evolving, AI-assisted analysis shows how inspection workflows themselves are changing.

The challenge facing many inspection programs is not in the collection of data-it's in the review. Modern ultrasonic inspections generate enormous amounts of information, much of which contains no actionable indications. Highly trained analysts often spend significant time reviewing data that ultimately confirm acceptable conditions.

To address this challenge, EPRI, in collaboration with TrueFlaw, developed an AI/ML-assisted ultrasonic analysis tool that automatically filters irrelevant signals and highlights areas requiring expert attention. Rather than replacing human expertise, the technology enables analysts to focus on potential indications and engineering evaluation.

Sweden's Vattenfall became the first utility to qualify for and deploy the technology for reactor vessel upper head penetration inspections at its Ringhals-3 nuclear power plant. Using existing mock-ups and advanced defect simulation techniques, the utility completed a full ENIQ qualification through the independent Swedish Qualification Centre. (ENIQ stands for European Network for Inspection Qualification.) The qualification demonstrated that the AI-assisted process could satisfy demanding nuclear inspection compliance requirements.

The operational benefits were immediate. Review workload was reduced by up to 95 percent, allowing analysts to focus on the most relevant data. During a subsequent deployment at Ringhals-4, the utility reported approximately one week of schedule savings associated with the inspection activity.

"Qualifying this technology was a major milestone for Vattenfall and the industry. It allowed us to reduce data review by up to 95 percent, focus on critical areas, and complete inspections more efficiently without compromising safety," said Ola Johansson, a senior NDE specialist at Vattenfall.

Perhaps most importantly, the project established confidence that qualified AI applications can support safety-critical nuclear inspections while preserving human oversight and engineering judgment. As the volume and complexity of inspection data continue to grow, that capability will become increasingly valuable across the industry. Reducing data review time by up to 95 percent allows experts to spend less time searching for information and more time evaluating what matters.

Digitizing the workface

Another important trend is the use of mixed-reality technologies to improve inspection execution and data traceability.

Inspection quality depends on more than the inspection technique itself. Technicians must locate the correct component, collect data at the correct locations, follow the appropriate procedures, and ensure results are properly documented. In large and complex facilities, those seemingly simple requirements can create significant challenges.

Mixed-reality applications are designed to make field execution more consistent and repeatable. Using virtual reality (VR) headset technologies like Microsoft HoloLens and Meta Quest 3, technicians can access digital work packages, view guided inspection locations, connect to inspection equipment, and record results directly within a digital workflow.

Field demonstrations have shown that virtual gridding can significantly reduce the time required for flow-accelerated corrosion inspections while maintaining inspection quality. By projecting a virtual grid onto piping through a VR headset, inspectors can avoid physically marking the pipe. The software, developed by Quasset, also records the grid location and ultrasonic signature, helping ensure future inspections provide consistent comparisons.

Beyond efficiency improvements, these tools provide better traceability, improved data management, and enhanced opportunities for workforce training and knowledge transfer. They also create new possibilities for remote expert support, helping experienced personnel guide activities without being physically present at the work location.

"The value of mixed reality is not simply speed," said Malek Abedrabouh, an engineer on the plant support team at EPRI. "It is consistency. Every technician sees the same guidance, follows the same workflow, and captures data in a repeatable way."

Cleaning without collateral damage

Surface preparation and decontamination represent another area of active innovation.

Laser ablation is the use of laser energy to remove corrosion products, coatings, and contaminants from component surfaces. The technology has attracted industry interest because it could potentially replace or supplement conventional methods such as grinding, blasting, and needle guns while reducing manual effort and improving efficiency.

However, adoption requires answering important technical questions. Utilities must understand how laser ablation affects material properties, inspection capability, worker safety, and code compliance.

Research activities, therefore, have focused on material effects, reflected energy, combustion behavior, microparticle capture, and potential ASME code implications. Initial results have been encouraging and suggest laser ablation may gain broader adoption and eventually provide utilities with more options for cleaning and decontamination activities.

Expanding repair options

Inspection advances are only part of the story. Utilities also need repair technologies capable of addressing difficult-to-access assets and challenging operating environments.

Nonmetallic repair technologies are being evaluated for applications involving fuel pools, transfer canals, reactor cavities, piping, vessels, and tanks. The work focuses on understanding available repair options, capturing operating experience, identifying technology gaps, and supporting future implementation.

Current efforts include evaluation of geopolymers, modified silane polymers, patch technologies, ribbon technologies, and robotic delivery systems. Planned demonstrations are expected to further explore how these technologies can be applied in nuclear environments where traditional repair methods may be difficult or impractical.

Fully adaptive welding becomes reality

One of the strongest examples of repair innovation moving into practice comes from the Department of Energy's Hanford Site.

Project teams needed a fully adaptive, remote, code-compliant approach to seal 18 dry storage systems containing high-activity cesium and strontium capsules. High radiation fields, restricted access, and limited availability of qualified personnel made conventional welding approaches impractical.

To address the challenge, Fluor Enterprises, Central Plateau Cleanup Company, Liburdi Dimetrics, and EPRI deployed a fully autonomous gas tungsten arc welding system incorporating adaptive controls and real-time parameter adjustment. EPRI contributed research in adaptive welding, welding optimization, materials performance, and NDE, while project partners translated the technology into a field-ready system and integrated it into operations.

The resulting process produced repeatable, leak-tight, ASME code-compliant closure welds while eliminating worker exposure in high-radiation environments. The project is expected to save approximately $12 million over a two-year campaign through reduced delays, fewer weld repairs, and reduced retraining requirements.

"For Hanford, success meant a closure weld we could execute remotely, repeatedly, and to code without putting people in harm's way," said Fluor Enterprises senior fellow Gary Cannell. "This technology gives the site schedule and cost certainty while meeting the rigor expected for a mission-critical cleanup milestone."

The broader significance is that adaptive welding has moved beyond research and into practical deployment for high-consequence nuclear applications.

"Adaptive welding demonstrates what becomes possible when robotics, real-time sensing, welding science, and qualification programs are brought together to solve a real operational challenge," said Jon Tatman, senior principal team lead for plant support at EPRI. "Hanford proved that fully remote, code-compliant welding can move from concept to deployment in one of the industry's most demanding environments."

The common thread

These technologies differ in purpose and maturity, but they all share a common path to adoption.

Nuclear innovation succeeds only when supported by technical bases; qualification programs; field demonstrations; operating experience; code alignment; and collaboration among utilities, research organizations, vendors, and regulators. That collaborative model is evident throughout these efforts, from UT qualification programs and AI deployment at Vattenfall to mixed-reality demonstrations, laser ablation research, nonmetallic repair development, and adaptive welding implementation.

For plant owners, the message is practical. The future of inspection and repair is not about replacing people with technology. It is about giving skilled workers better tools, data, guidance, and ways to perform complex work safely and consistently while reducing avoidable disruption.

Rather than relying on a single breakthrough, the next generation of nuclear inspection and repair will be built on the integration of advanced inspection methods, AI-assisted analysis, digital work execution, and qualified repair technologies. Together, these capabilities are helping utilities improve decision-making; reduce disruption; and support the safe, reliable operation of nuclear assets over the long term.

Michael Ruszkowski is the director of plant support at EPRI.