Site preparation is underway in southern Ontario for what is expected to be the first small modular reactor to come online in North America, a 300-MW unit projected to cost $7.7 billion CAD.

Much hope has been attached to SMR technology as a solution to large load power demand. But early movers are expected to pay more, as they will not benefit from the speed and cost savings that are the value prospect of serial production and construction.

Ontario Power Generation (OPG) led its first-quarter earnings report May 13 with an update on its Darlington New Nuclear Project, which five days earlier received provincial approval for start of construction.

Three subsequent SMR units are planned on the site, bringing the combined capacity to 1,200 MW. The total cost including interest, potential cost escalation and contingencies is projected at $20.9 billion CAD, or $15 billion USD at the present exchange rate. It will be borne by ratepayers.

The project has drawn opposition in part because of that price tag. By contrast, the most prominent recent example of expensive nuclear power — construction of Plant Vogtle Units 3 and 4 in Georgia — added a bit more than 2,200 MW at a cost of more than $30 billion USD.

On May 8, the same day the Ontario government approved construction of the first Darlington SMR, the Ontario Clean Air Alliance released a report stating the levelized cost of Darlington’s nuclear electricity would be up to eight times higher than onshore wind and nearly six times higher than solar.

On May 9, the Alliance blasted the construction approval, saying Ontario is rolling the dice on untested first-of-a-kind technology for a project that may cost as much as $27 billion CAD and would rely on uranium imported from the country that elected Donald Trump president.

OPG did not return a request for comment for this story.

In announcements from OPG and the Ontario Ministry of Energy and Mines, the Darlington SMR project is hailed as a groundbreaking initiative that is the first of its kind among the G7 nations, an economic boon to the region’s workforce and a project that is expected to contribute $38.5 billion CAD to the nation’s economy over 65 years.

The construction site is on the Lake Ontario shoreline 35 miles northeast of Toronto. It is adjacent to OPG’s Darlington Nuclear Generating Station, whose four units provide over 20% of Ontario’s electricity needs and are undergoing a $9.2 billion USD refurbishment expected to extend their operational lives 30 years.

OPG said construction of the first Darlington SMR would incorporate more than 7,000 lessons learned so far from the Darlington refurbishment.

This points to the promise and peril of the SMR revolution envisioned by some U.S. policymakers and energy industry leaders: SMRs could standardize and modularize the process of permitting and building nuclear stations so much that the timeline and costs are significantly reduced.

But until that standardization comes, costs will be high.

Vogtle 3 and 4 had a number of setbacks, not least of which was the bankruptcy of its contractor. But despite relying on widely used and proven technology, Vogtle had a disadvantage common to first-of-a-kind ventures: Nobody had successfully built a full-scale commercial reactor in the United States in a generation.

Some analysts maintain that if a Vogtle 5 or similar project began construction soon after Vogtle 4 was completed — while the lessons learned at such high cost were front-of-mind and still relevant — it would not see anywhere near the degree of budget and timeline overruns that Vogtle 3 and 4 suffered.

Many are skeptical that SMRs will make the leap anytime soon from first of a kind to “nth of a kind,” that subjective point when a new concept stops being new and is an accepted technology operating with economies of scale.

NextEra Energy CEO John Ketchum, for example, said during an October 2024 earnings call that he does not foresee any meaningful amount of new nuclear capacity coming online in the U.S. in the next decade.

The nearly one dozen companies pursuing SMR development are insufficiently capitalized for the most part, he said, and those SMR designs that do reach the deployment stage will be very expensive and risky at first.

SMR developers themselves seem a much more optimistic lot, recently announcing multiple rollout plans and agreements with tech giants to provide emissions-free baseload power to the data centers some expect to be built in large numbers.

Industry analyst Dean Murphy, a principal at Brattle Group, told RTO Insider that while the 10 design teams developing SMR will not all be successful — and should not be, because that would limit standardization — there can be shared learnings across the different design concepts, such as how to build a pump to withstand extreme temperatures.

So there is value in these competing efforts underway now, he said, and value in winnowing them down.

But building an SMR never will be like erecting a wind turbine or solar array, he added. There is too much complexity, even with the lightened regulatory regime the Trump administration is reported to be considering.

“So, we’re going to have to build this first one in Darlington,” Murphy said. “And there are a couple of other projects that are sort of their first of a kind, but they’re going to have to get designed and built and constructed and operate for at least a little while before people say, ‘OK, that one looks like it’s going to work.’ And then we’re going to go through the next round with light revisions to the design, which likely means you’ve got to redo the licensing.”

And then there is the siting.

Americans and their elected officials are said to be more supportive of nuclear energy than they were a generation ago, when the Three Mile Island accident was fresher in the collective mind.

But they have not been faced with the prospect of hundreds of small reactors dotting the landscape with lightened safety protocols.

This is why the regulatory process should not become superficial even if it can be expedited, Murphy said: It risks a bad reaction from the American public and a bad result from a botched project.

An SMR buildout can be pursued successfully but not quickly, he said.

“Nuclear is a really promising technology, including SMRs, really for the second half of this century. I think that’s how long it’s going to take to get a couple of times through this technology cycle before you can start building them in volumes that are enough to make a difference.”