North Carolina State University, the New York Power Authority (NYPA), and EPRI have demonstrated a megawatt-class solid-state transformer (SST) on a live 13.2-kV utility distribution feeder, marking what the partners describe as the first independent validation of a megawatt-class SST under real-world distribution conditions.
The NC State-built prototype was connected at EPRI’s power delivery laboratory in Lenox, Massachusetts, where it supplied an electric vehicle (EV), injected reactive power for grid support, and was repeatedly energized and de-energized during a three-day field campaign in mid-June.
“This is the first independently verified, megawatt-class solid-state transformer validated on a live utility distribution feeder,” said Srdjan Lukic, principal investigator on the SST project and Lampe Distinguished Professor of Electrical and Computer Engineering at NC State. “This is a significant step in advancing power transformer technology.”
From Laboratory Prototypes to a 1-MVA Field Demonstration
The effort marks a major leap for SSTs, which researchers have pursued for decades for their potential to combine voltage transformation with functions that typically require separate power-conversion and grid-support equipment. Unlike conventional 50- or 60-Hz transformers, SSTs use semiconductor-based power converters and high-frequency isolation to regulate voltage, support bidirectional power flow, provide AC or DC output, and manage power quality. In a December 2025 white paper, EPRI noted that operating the transformer at tens of kilohertz instead of grid frequency can substantially reduce the size and weight of magnetic components, while integrated sensing and control allow the device to actively manage power flow. As POWER has reported, those capabilities have drawn interest for data centers and other large DC loads, EV charging, storage, renewable integration, and distribution-grid applications.
The concept dates much earlier. General Electric proposed an electronic transformer in the late 1960s, and EPRI introduced its Intelligent Universal Transformer in 1995 and researched the concept for almost two decades. Development has accelerated as semiconductor and materials technologies have improved, but higher-power medium-voltage systems have remained largely in prototypes and targeted demonstrations. EPRI’s December 2025 survey cataloged systems ranging from a 25-kVA distribution SST and a 400-kW EV charger to a 500-kVA hybrid design, while noting that NC State’s own FREEDM prototypes had advanced from 10 kVA to more than 20 kVA.
EPRI nevertheless characterized SST technology as “largely in pilot or prototype stages” and identified scaling power levels and proving long-term performance as work still underway.
Taking the SST to Medium Voltage
NC State’s work, meanwhile, traces to a National Science Foundation-funded program launched in 2008, when “many of the components we’re using now in this system were simply not available, even kind of in the research realm, let alone commercially,” Lukic told POWER. Over successive iterations, the FREEDM Center moved from a 10-kVA laboratory prototype to systems above 20 kVA using 15-kV silicon-carbide metal-oxide-semiconductor field-effect transistors (MOSFETs). EPRI’s 2025 review said those earlier systems demonstrated the viability of high-frequency medium-voltage conversion and informed later designs.
NYPA and NC State began collaborating on the current project in 2018, when they jointly pursued U.S. Department of Energy (DOE) support to develop a system capable of operating on a distribution feeder. The DOE Transportation Technologies Office initially funded the work under award DE-EE0008450. NYPA funded and contributed to construction and field testing, while NC State designed and built the SST’s power-electronics hardware. Ramadan Elmoudi, senior research, technology and development engineer at NYPA and the authority’s project lead, said NYPA’s role included helping move the university’s work into a utility environment.
“We won this project funded by the Department of Energy to build or to go from a lab prototype, which we had built before, to something that can actually connect to a real distribution feed and operate in real-world environment,” Lukic said. “So we kind of took the knowledge we’ve gathered over the years in these experimental prototypes and built this system that was finally deployed with the help and support of NYPA.”
NC State also used the project to train the engineers who may carry the technology forward. “Our role was really building the solid-state transformer, the actual power-electronics hardware,” Lukic said. “That was done not with engineers, but with graduate students who will now bring this technology into the real world as they move from academia back into the industry.”
Series-Stacked Modules Solve the Voltage Challenge
The technical leap announced in August, however, centered on the grid connection. While power-electronic converters are already common at 480 V and similar voltage levels, as Lukic explained, connecting one directly to a medium-voltage distribution feeder presents a different problem: an individual semiconductor switching device cannot block the full voltage imposed by the grid. NC State addressed that limitation through a modular architecture that connects multiple smaller power-electronic modules in series on the medium-voltage side, allowing the feeder voltage to be shared across them.
But series stacking created the next engineering challenge. “Now the challenge becomes: how do we coordinate these modules that are connected in series, and importantly, how do we ensure the isolation between that medium-voltage side and the low-voltage side,” Lukic said.
The team designed an isolation transformer specifically for the system. Lukic called it “one of the key innovations in the system” and “really that last hurdle that we solved with the help of NYPA in the last couple of years.” That architecture also gives the SST capabilities beyond voltage conversion. Its power electronics can control real and reactive power at the grid interface, functions the Lenox campaign would test directly on an operating distribution feeder.
Three Days on a Live 13.2-kV Distribution Feeder
Crucially, the Lenox campaign was not intended as a full-power test. Because site limitations constrained the power available to the load, the emphasis shifted to whether the SST could operate successfully at the distribution interface. “The key goal of this demonstration was simply to see: Can we connect this system to a distribution feed and deliver power to a load?” Lukic told POWER.
The SST was tied into a 15-kV-class distribution line operating at 13.2 kV line-to-line. During three days in mid-June, researchers energized and de-energized the system at least 10 times while testing different features and operating modes. The SST supplied real power to an EV at the site and injected reactive power into the feeder under operator command.
“We demonstrated that not only can we deliver real power to the vehicle, but we can also deliver reactive power back to the grid, as commanded by the system operator,” Lukic said. “We also show that the power quality delivered by the SST is quite high in terms of the harmonic distortion and the current that is being injected back in the grid.”
NYPA, meanwhile, was watching how the unfamiliar equipment behaved as those operating steps progressed. “Connecting the device, and you don’t need see sparks. That’s a great achievement,” Elmoudi said. “You can be able to connect and disconnect and see the impact on the grid. That’s very crucial. So it’s like you deploy this device on the grid and nothing happen and does not get rejected by the grid. That’s a good sign. Then you start connecting loads and see how that power flow affects the grid and affects the SST by itself.”
The exercise also gave NYPA direct experience with a newer generation of medium-voltage power electronics. Elmoudi pointed to the authority’s earlier use of stacked thyristor-based equipment for reactive-power control and power-flow management, and said advanced silicon-carbide devices can improve efficiency and reduce equipment size, potentially widening where such systems can be deployed. “As Srdjan said, by stacking those power electronics and do the innovation and the insulation between the medium voltage and these devices, that’s a big leap,” he said.
The SST’s efficiency was evaluated separately in the laboratory, where NC State could measure the power-electronic modules more precisely. “On the actual demonstration site, we didn’t focus on the efficiency of the unit from the grid to the load,” Lukic said. “The reason is simply this is something we already characterized in the lab and the measurement process and the devices that we’re using don’t have that resolution that would allow us to get better estimate or better measurement of efficiency in the field compared to what we can do in the lab.”
NC State characterized an individual module from its medium-voltage connection to the SST’s DC bus. “That efficiency peaks out at about 98%,” Lukic said. The figure applies to the single-module conversion stage, not measured grid-to-load efficiency for the complete 1-MVA system at Lenox, he noted.
Where SSTs Could Find a Market
For now, while the Lenox results point to good progress for the technology—in the field, beyond laboratory operation—Lukic distinguished between demonstrating the architecture in its target environment and producing equipment that utilities or large power users can routinely procure. “I think in terms of commercialization, the proof of concept is there,” he noted.
NC State and NYPA do not plan to commercialize the SST themselves. Lukic said the project was intended in part to answer questions that sit between SST developers and prospective customers. NYPA brought EPRI into the effort to broaden that exchange. “EPRI, they have access to many different utilities and vendors from different perspectives and different fields,” Elmoudi said. “So hopefully, this can accelerate the commercialization of this device and bring it to the level where it can be designed and manufactured in mass.”
Potential markets could extend beyond EV charging. Lukic agreed that medium-voltage distribution, data centers, renewable and storage integration, and microgrids are among the applications under consideration. He also pointed to interest in Europe in residential installations where reducing switchgear, wiring, and equipment footprint could matter. “The big one, of course, is data centers that everyone is talking about, and maybe that is the killer application that will help us move further along more quickly than we would have otherwise with this technology,” Lukic said.
For facilities that ultimately require DC power, one potential advantage could be to combine voltage transformation and power conversion in a more compact system. The university has also pointed to smaller equipment footprint as an advantage for large-scale projects where electrical infrastructure must be located close to the load.
NYPA sees a broader distribution-grid role. “We learned through the project’s positive outcomes that solid-state transformers can serve as another tool in our clean energy toolbox to enable New York State to operate and maintain a compact, efficient, and resilient grid,” Elmoudi said.
Standards and Grid Survivability Remain Open Questions
Still, moving beyond the successful three-day field campaign to routine deployment will require qualification for the conditions utilities expect medium-voltage equipment to withstand over decades of service. “We have to go from where we are, which is demonstrating that the system operates in the target environment, to a device where the end user has confidence that these units will last as long as a traditional transformer, and that really all of the fault scenarios and use cases are fully accounted for and resolved, especially of course from a safety point of view and from a reliability point of view for the solid state transformer,” Lukic said.
The isolation stage developed for the NC State unit is one R&D priority. Lukic called it “a very unique and interesting solution,” but said it remains “the new component in this system that really needs to be fully understood and fully explored in terms of performance requirements and qualification.”
Utilities must also become comfortable placing semiconductor electronics directly at the distribution-system interface after generations of experience with conventional transformers and switchgear. “Another key component is making the utility folks comfortable with this device, that it really can survive and handle all the issues that you would see on the grid,” Lukic said.
EPRI’s December white paper identified a broad test agenda that includes efficiency across load conditions, control validation, overvoltage and undervoltage tolerance, short-circuit and overload protection, lightning and switching surges, harmonic distortion, thermal cycling, accelerated aging, environmental testing, safety and compliance, communications, and cybersecurity. It also identified cost, reliability, standards, and interoperability as barriers to wider medium-voltage deployment.
Elmoudi also added standardization as one of the most immediate gaps. “I believe the next big step is going to be standardization because we don’t have any standards for SST,” he said. He suggested IEEE or another standards organization develop requirements covering “how you can design it, how you can build it, and how you can test it, and also how to commission it and maintain it.”
Protection and survivability under abnormal grid conditions are equally fundamental, he noted. “If we have a lightning, what’s going to happen to the SST? If we have through fault, what’s going to happen to the SST? All those questions need to be answered,” Elmoudi said. “We need to expose this to industry and invite them to invent and collaborate in this project.”
—Sonal C. Patel is a POWER senior editor (@sonalcpatel, @POWERmagazine).
