Nuclear industry: As construction of first small modular reactor looms, prospective buyers wait for the final tally
The race to construct Canada’s first new nuclear power reactor in 40 years seems to have passed a point of no return. This summer, Ontario Power Generation completed regrading the site for its Darlington New Nuclear Project in Clarington, Ont., and started drilling for the reactor’s retaining wall, which will be buried partly underground. At a regulatory hearing, OPG’s chief executive officer Ken Hartwick, who will retire at the end of this year, promised that this reactor will be “the first of many to come.”
But that will depend on a crucial yet-to-be-revealed detail: its price tag.
It’s no exaggeration to say that the world is waiting for it. The new Darlington reactor would be the first BWRX-300, a small modular reactor (SMR) being designed by an American vendor, GE-Hitachi Nuclear Energy, and the first SMR built in any Western country. Other prospective buyers include the Tennessee Valley Authority (TVA), SaskPower and Great British Nuclear. More BWRX-300s are in early planning stages in Poland and the Czech Republic.
Crucially, however, OPG is the first and only utility worldwide to bind itself contractually to build a BWRX-300. A report published by the U.S. Department of Energy in September said American utilities are waiting to see pricing and construction schedules for early units, and would “prefer to be fifth.” SaskPower also wants to avoid the risks associated with building a “first of a kind” reactor; it won’t decide until 2029 and it hopes SMRs will be less expensive than traditional nuclear plants.
Scheduled for release this winter, the Darlington SMR’s estimated cost will speak volumes about whether SMRs can deliver on their many promises. Yet there are early indications of serious sticker shock: Recently published estimates from the TVA suggest the first BWRX-300 could cost more than five times GE-Hitachi’s original target price. How will OPG and GE-Hitachi drive pricing far below the TVA’s estimate? And if they cannot, what then will be the prospects for SMRs?
Ditching the scaling law
SMRs were conceived as an antidote to the hefty price tags that brought reactor construction to a standstill in Western countries for decades.
Previously, the nuclear industry relied heavily on something called economies of scale or the “scaling law”: As a power plant’s size increases, capital costs also rise, but in a less than linear fashion. So vendors designed ever-larger reactors. Reactors under construction today average about one gigawatt, roughly three times the BWRX-300’s output. They can cost more than US$10-billion, leaving only the largest government-backed utilities as potential purchasers.
SMRs represent a promising but untested new approach to manufacturing reactors – one that emphasizes simplification and mass production techniques. The key term is modular: Rather than building monolithic, one-of-a-kind plants, the industry hoped instead to churn out substantially identical factory-built units; repetition would help drive down costs, as it had for competing technologies such as wind turbines and solar panels.
But modularity requires multiple orders, which in turn demands competitive pricing. Through early discussions with potential customers, GE-Hitachi executives understood the BWRX-300 had to be priced low, not only in absolute terms, but also relative to other power-generation technologies. They told audiences it would cost less than US$1-billion, or US$2,250 per kilowatt hour of power generation capacity – low enough to compete with natural gas-fired power plants.
“The total capital cost of one plant has to be less than $1-billion in order for our customer base to go up,” Christer Dahlgren, a GE-Hitachi executive, said during a talk in Helskini in March, 2019.
Shrinking a giant
GE-Hitachi’s designers began by shrinking a behemoth: the 1,500-megawatt Economic Simplified Boiling Water Reactor (ESBWR). Their objective was to reduce the volume of the building housing the reactor by 90 per cent, to greatly reduce the amount of concrete and steel required during construction.
This was accomplished primarily through eliminating safety systems. Pressure relief valves, common in traditional reactors, were removed. In place of two completely separate emergency shutdown systems, as is customary, the BWRX-300 would have two systems that would propel the same set of control rods into the reactor’s core. GE-Hitachi emphasized that the BWRX-300 featured “passive” safety systems that would keep the reactor safe during an accident, and its simplicity reduced the need for redundant engineered systems.
Sean Sexstone, head of GE-Hitachi’s advanced nuclear team, said the entire facility – which includes the reactor building, the control room and the turbine hall – will measure just 145 metres by 85 metres.
“You can walk that site in a minute-and-a-half,” he said.
GE-Hitachi also sought substitutes for concrete. The reactor building is to be constructed using factory-made steel panels that will be shipped to the site, assembled into modules and lifted by crane into position. These modules essentially serve as forms into which concrete is poured. These steel plates are as strong as concrete, OPG says, yet eliminate the need to use rebar extensively. This approach “lends itself to more modularity, more work in a factory, versus more work in the field,” Mr. Sexstone explained.
The Darlington SMR will be erected using a technique called “open-top construction.” The reactor building’s roof won’t be installed until the very last. The building will be constructed upward, floor by floor, with large components lowered in by crane rather than being moved through doors and hatches.
Many of the BWRX-300’s components would be identical to those used in previous GE power plants, such as its control rods, fuel assemblies and steam separators. Its steam turbine would be the same one used in natural-gas-fired plants. And the plant could be run by as few as 75 staff, far below the nearly 1,000 employed at large single-reactor Canadian nuclear plants.
Historically, utilities tended to build bespoke nuclear plants meeting highly individualized requirements. The result: In the United States alone there are more than 50 commercial reactor designs. Few designs were built twice, limiting opportunities to learn through repetition.
GE-Hitachi intended the BWRX-300 to be highly standardized, constructible in multiple countries with as few tweaks as possible. It assembled an international coterie of utility partners, including OPG, the TVA and a Polish company named Synthos Green Energy, which last year agreed to jointly contribute to the estimated US$400-million cost of the SMR’s standardized design.
Subo Sinnathamby, OPG’s chief projects officer, acknowledged in an interview that the first SMR will be expensive. But lessons learned from building it, including newly identified opportunities for additional modularization, will be applied to three subsequent units at Darlington, bringing down overall costs.
“For us, success is going to be sticking to how we have executed megaprojects at OPG, using the same processes and principles,” she said, citing the continuing refurbishment of Darlington’s existing reactors.
“The last thing we want to do is get into construction and then stop the work force.”
GE-Hitachi’s emphasis on lowering plant costs has been validated by many independent observers, who regard it as essential to SMRs’ future prospects.
In a report published in May, Clean Prosperity, a climate policy think tank, concluded that the BWRX-300 “is the strongest candidate” among SMRs to experience continued cost reductions as more were built – but only at the right price, which it pegged at about $3.3-billion. “Cost curves will only become possible for the BWRX-300 in Ontario and beyond,” it warned, “with a final price tag that is low enough to compel additional expansion.”
In September, the U.S. Department of Energy published a report examining the prospects for widespread deployment of reactors across the U.S., an expansion it strongly supported. But to drive down costs, SMR vendors needed to move more than half of the overall spending on a project into standardized factory-like production – a tall order.
Similarly, a report published last year by the U.S. National Academy of Sciences argued that if nuclear plants are to contribute meaningfully to future electricity systems, they must be cost-competitive with other low-emission technologies. It looked at so-called overnight capital costs – what costs would be if construction were completed overnight, with no charges for financing and no consideration of how long it will last. The academy said capital costs should be US$2,000 or less per kilowatt of generating capacity. At between US$4,000 and US$6,000 a kilowatt, reactors might still be competitive if costs unexpectedly rose for renewable technologies.
Enter the TVA.
In an integrated resource plan published in September, the TVA estimated that a first light water SMR would have an overnight capital cost of nearly US$18,000 a kilowatt.
At that pricing, the first Darlington SMR would cost more than $8-billion. That’s about 10 times the cost of a similarly sized natural-gas-fired plant: SaskPower’s recently completed Great Plains Power Station, a 377 MW natural-gas-fired plant in Moose Jaw, cost just $825-million.
Oregon-based NuScale Power Corp. has already discovered what happens when pricing falls in this range. Founded in 2007, its 77-MW NuScale Power Module was the first SMR to be licensed by regulators in a Western country. But last year its flagship project, undertaken with the Utah Association of Municipal Power Systems (UAMPS), was cancelled after cost soared to about US$20,000 a kilowatt.
There are several important caveats about the TVA’s estimate.
Greg Boerschig, a TVA vice-president, described it as a “Class 5″ estimate. According to standard global practices, cost estimation is based on a five-level system. Class 5 is the least detailed and reliable and is intended for planning purposes; actual costs could be half that much, or double.
The estimate is far higher than the TVA would have liked, Mr. Boerschig said. But since OPG is further along in deploying the BWRX-300, he added, it has a better sense of the reactor’s cost.
“We’re a couple of years behind them,” Mr. Boerschig acknowledged.
Indeed, according to a presentation by Aecon Group Inc., a partner on the Darlington SMR, a Class 4 estimate had already been completed as of February this year. Ms. Sinnathamby said OPG is working on a Class 3 estimate.
“Our number is going to be very specific: What is it going to cost us to build, on this location, these four SMRs?” she said.
Another caveat is that the BWRX-300 was only one of several reactors represented in the estimate, which was based on the TVA’s experience exploring potential SMRs at its Clinch River site near Oak Ridge, Tenn., and by examining recently completed nuclear construction projects.
OPG might enjoy certain cost advantages over the TVA. The Darlington Nuclear Generating Station is a complex that was built during the 1980s and early 1990s on the shore of Lake Ontario, the proximity of which could make cooling reactors there cheaper. Clinch River is a greenfield site, whereas Darlington already has four operating reactors.
“That will automatically reduce the cost to OPG relative to TVA,” said Koroush Shirvan, a professor of energy studies at the Massachusetts Institute of Technology, who has studied the BWRX-300’s economics.
Nonetheless, opponents and skeptics of SMRs in general, and the Darlington SMR in particular, have embraced TVA’s estimate.
Chris Keefer, an emergency medicine physician, has advocated passionately for refurbishment of Ontario’s existing nuclear power plants, which are all based on Canada’s homegrown reactor design, the Candu. He has also argued for modernizing the Candu design and building more. He said the TVA’s estimates reflect a more honest assessment of SMR pricing than Canadians received in the past.
“It points to this emerging consensus that SMRs are not economic, and that shouldn’t be a surprise,” he said.
“TVA, I think they’ve got several hundreds of millions of dollars in the development process on this reactor. I wouldn’t say that those numbers are naive.”
Prof. Shirvan said his own cost estimate for the BWRX-300 reactor is “in line” with the TVA’s.
Chris Gadomski, head of nuclear research at BloombergNEF, said TVA’s estimates are discouragingly high, and imply that reactor sales might be less than anticipated. Contributing factors might include high labour costs in North America, and recent high inflation and high financing costs, factors he expects will persist.
“The nuclear people don’t operate in a vacuum, they operate in competition to other technologies,” he said.
“The cost for solar is going down. The cost of batteries, we anticipate, is going down. And so, when you’re looking at spending billions of dollars and all of a sudden the price tag gets so large, people will say: ‘Hey, listen, you’ve got to look at other options, or buy less of this.’ ”
If there is a silver lining, the TVA estimated follow-on SMRs would cost substantially less than the first, at roughly US$12,500 a kilowatt. But that’s still more than double the upper limit the U.S. National Academy of Sciences deemed necessary to support widespread SMR adoption.
We might learn in a few months whether GE-Hitachi and OPG have succeeded in bringing the BWRX-300’s cost down. But a review of regulatory applications and other documents hint at why the original US$1-billion target price might be difficult to realize.
Prof. Shirvan said GE-Hitachi’s original plan – to slim the reactor down by removing safety systems – encountered resistance from regulators in Canada and the U.S. “When you strip out most of the safety system, you have to come up with very good reasoning how that’s justified,” he said. GE-Hitachi started adding some of those systems back in, he said, which caused the BWRX-300’s reactor building’s diameter to swell.
This dramatic increase, Mr. Keefer said, has greatly reduced the BWRX-300’s economic attractiveness.
“Proportionately, you’re actually doing a lot more civil works than you would for a large reactor,” he said. “And that actually means that the whole SMR paradigm, which is to get all the work into a factory, goes away.”
(GE-Hitachi denied that the plant had grown. “While the design has matured, the overall footprint of the BWRX-300 plant has not changed significantly,” Mr. Sexstone said.)
OPG’s regulatory documents also make clear that some modular construction techniques it seeks to employ at Darlington are in their infancy. As recently as last year, most of the walls and floors of the SMR building were to have been built using a technique developed in Britain known as Steel Bricks. GE-Hitachi recently dropped Steel Bricks in favour of a similar approach known as Diaphragm Plate Steel Composite.
Moreover, OPG’s published construction plans show that the reactor building will be built largely below-grade, requiring significant excavation including into bedrock. Tunnel boring machines will be used to excavate more tunnels, tens of metres wide, to convey cooling water to and from Lake Ontario. Make no mistake, the Darlington SMR remains a complex capital project.
To date there have been no indications that pricing might derail the Darlington SMR. Ontario’s government appears willing to pay a significant premium: It hopes that as a first mover, OPG will be well-poised to sell equipment and expertise in other countries.
During a stump speech in Scarborough in December, Energy Minister Stephen Lecce said Ontario was keen to sell its technology and expertise for building SMRs abroad.
“I was just in Poland and Estonia, literally selling Canadian small modular reactors that will be built here, exported there,” he said.
Yet Mr. Lecce has also vowed to keep Ontarians’ electricity bills low, an objective high SMR price tags might compromise.
GE-Hitachi maintains its creation’s pricing will stack up favourably.
“I think we’re in a really good spot to feel very comfortable about this unit being probably the most cost competitive SMR in the market,” Mr. Sexstone said. “I think your readers will be pleasantly surprised.”
Ms. Sinnathamby, for OPG’s part, said actual costs to construct BWRX-300s should be considerably lower than TVA’s estimate.
“The TVA numbers can only come down,” she said. “That’s how conservative, in our mind, those numbers are.”
This article was first reported by The Globe and Mail
