NC State, NYPA and EPRI Run a 1 MW Solid-State Transformer at 98% Efficiency on a Live Feeder

North Carolina State University, the New York Power Authority and EPRI reported a working 1 MW solid-state transformer on August 18, 2026. The three partners ran it on a live utility distribution feeder at EPRI’s power delivery laboratory in Lenox, Massachusetts. The unit converts high-voltage alternating current directly to low-voltage direct current. It held about 98% efficiency, against roughly 96% for the conventional transformer-and-rectifier chain it replaces. Srdjan Lukic ran the project as principal investigator. He is the Lampe Distinguished Professor of Electrical and Computer Engineering at NC State. He called the unit the first independently verified megawatt-class solid-state transformer validated on a live utility distribution feeder. The US Department of Energy funded the work through Transportation Technologies Office award DE-EE0008450. The New York Power Authority paid for construction and field testing. NC State students Oscar Montes, Nazmul Hassan and David Dadzie designed, built and tested the prototype.

What the two-point efficiency gain is worth

The partners reported the efficiency figures as percentages. Converted to kilowatts at the unit’s 1 MW rating, the conventional transformer-and-rectifier chain loses 40 kW to heat and the solid-state unit loses 20 kW. Held at full load for a year that difference comes to roughly 175 MWh, and the 20 kW that never appears also never has to be rejected by a cooling plant. Neither NC State nor EPRI put a dollar value on either saving.

The size claim carries similar weight for siting. The team describes the unit as smaller than the transformer-and-rectifier pair it replaces, without publishing dimensions or weight. Ramadan Elmoudi is a senior research, technology and development engineer at the New York Power Authority. He said the project showed solid-state transformers serve as another tool in the utility’s clean energy toolbox. Drew McGuire, senior director of transmission and distribution research at EPRI, framed the collaboration as a way to accelerate development of the technology. NC State, NYPA and EPRI named data centers, electric vehicle charging and space-constrained urban sites as the target applications.

Why It Matters

Solid-state transformers have carried a twenty-year gap between laboratory promise and utility purchase orders, and most of the demonstration record sits well below a megawatt. The field deployments that do exist cluster in rail traction, where ABB’s PETT locomotive converter proved the physics on Swiss track without ever moving the distribution market. Third-party verification of a megawatt-class unit on a real feeder addresses one narrow part of that record, the part vendor datasheets have never been able to settle.

The timing lands on a specific problem. AC-to-DC conversion inside a data center happens several times between the substation and the chip, and every step throws away power. A device that does the whole conversion once, at medium voltage, in a smaller cabinet, attacks the exact loss stack hyperscale operators are now trying to shrink. MGRID has covered the same architecture arriving from the product side. Recent examples include an 800 VDC solid-state transformer bound for the San Diego Supercomputer Center, and 3.3 kV silicon carbide modules built to cut device count in these converters.

Path to commercialization

Three numbers stay unpublished, and every one of them decides whether a utility writes a purchase order. The partners did not disclose the input and output voltage levels, which leaves the class of feeder this unit actually serves unclear. They did not say how long the prototype ran on the live feeder, and duration is what separates a demonstration from a reliability claim. They gave no installed cost. Cost is the reason iron-core distribution transformers still win nearly every procurement, including the ones where a solid-state unit is the better engineering answer.

MGRID could not independently verify the voltage class, the test duration or the cost, because the partners did not publish them. What the announcement does establish is that a university-built megawatt device survived a real feeder under independent observation. That is the step every prior solid-state transformer program had to clear before a utility would write a specification around it.

Sources

• NC State University News
• NC State Office of Research and Innovation
• Transformer Magazine
• Newswise
• EPRI, Distribution Solid-State Transformers

Related Coverage

• Hitachi Energy Commits Additional $1.5 Billion for Transformer and Solid-State Transformer Capacity in Finland, Germany and North America: Investing in solid-state transformer capacity
• SolarEdge and Infineon Announce Modular 2-5 MW Solid-State Transformer Targeting >99% Efficiency for AI Data Centers: Developing high-efficiency solid-state transformers
• ABB E-mobility Launches MCS1200 1.2-Megawatt Charger With Integrated Solid-State Transformer for Heavy-Duty Trucks: Launching solid-state transformer products
• How Eaton’s 2021 DOE $4.9 Million Solid-State Transformer Award Set Up the 2025 Resilient Power Acquisition: Funding solid-state transformer research and development