More than a decade ago a contract was signed to build the world’s first third-generation European pressurized reactor (EPR) in Finland. The cutting-edge, 1,600-megawatt nuclear power plant, Olkiluoto 3, which its French maker Areva boasted as the most advanced safety design of the time, is still under construction today. There have been various setbacks as well as endless finger-pointing between Areva and the Finnish utility TVO, which are locked in court battle over expanding costs. Now the reactor might not be completed until at least 2017, if at all, with a price tag of $11 billion, more than double its original estimate.

The Olkiluoto 3 situation is not unique. Another Areva EPR in Flamanville, France, is also behind schedule and over budget. A recent government deal for two new EPRs in the U.K. has also come under fire.

The prospects for a nuclear power revival are no better in the U.S. Although the technology has never been cheap, cost overruns and delays are plaguing the handful of next-generation pressurized water reactors currently being built, the first since Japan’s Fukushima nuclear disaster in 2011. Even before that event, a study from Massachusetts Institute of Technology found that the cost of new nuclear plants, globally, doubled from 2002 to 2009. The third-generation reactors have safety features that should prevent a meltdown similar to Fukushima’s but political controversy, along with the high price tag means that new nuclear complexes in the U.S. and Europe could be in the single digits instead of dozens originally planned less than a decade ago.

Ironically, the experience has been markedly different in Asia. Two of Areva’s EPRs are expected to come online in China next year. China and South Korea are building the third-generation reactors with fewer construction delays and cost overruns than their Western counterparts. “They’ve been single minded about it,” says Tony Roulstone, course director for nuclear energy at the University of Cambridge. “And that single-mindedness has its advantages.” China and other Asian countries have been building nonstop for the last 30 years whereas the multiyear gap in the U.S. has resulted in a loss of construction knowledge. China also seems to have the advantage of endless manpower, and the state owns the country’s largest nuclear firm.

The one possible turn in the otherwise rough road in the West could be a shift in technology toward small, modular reactors (SMRs), which U.S. Department of Energy (DoE) Secretary Ernest Moniz said could be deployed as soon as 2022 with his agency’s support. But even SMRs face uncertainty and a long march to commercial deployment.

Bigger, but not cheaper
In the face of delays the U.S. government recently put money forward to try to jump-start the domestic nuclear power industry. The DoE issued $6.5 billion in loan guarantees for two new Westinghouse AP1000 nuclear reactors already under construction at the Vogtle Electric Generating Plant in Georgia in February. These units, pressurized water reactors, were the first new facilities to break ground in the U.S. in about 30 years.

The other two new reactors being built in the U.S., in South Carolina, are also AP1000s. The design is Westinghouse’s take on a third-generation pressurized water reactor, and each one produces about 1,000 megawatts of electricity. The two Vogtle units are slated to come online in 2017 and 2018. The South Carolina units will arrive later, being constructed at South Carolina Electric and Gas’s V. C. Summer Nuclear Station. The world’s first AP1000 is scheduled to come online in China in 2015. “It’s fair to say utilities are watching their experience,” notes Pete Lyons, assistant secretary for nuclear energy at DoE.

“A big reactor is now 1.6 gigawatts [1,600 megawatts], and they’ve done that to get the unit cost down,” Roulstone says, but the increased size had led to construction issues. In the case of the Olkiluoto 3 EPR, he notes that safety engineers designed a reactor with more pumps and valves to prevent the loss of any coolant, “but there wasn’t as much effort into designing for construction,” which is driving the overruns. The EPR has an 80,000-cubic-meter, double-containment structure that sits on a three-meter-thick concrete base. The system also has four separate cooling systems, each with its own water pumps, valves and control systems—all which add to the complexity in on-site construction.

Costs and delays are also rising for the AP1000 being built by Georgia Power, in part because the units must meet more stringent safety requirements that regulators have introduced in the wake of September 11, 2001, attacks and the Fukushima meltdown. The safety system itself is actually less complex, which should help reduce costs. Instead of more pumps, Westinghouse’s design has 50 percent fewer safety-related valves and 80 percent less safety-related pumping. “It’s an intrinsically simpler design,” says Roulstone, who compared the natural circulation system to a car radiator. The reactor has a slightly smaller footprint compared with an older reactor that only produced about half the amount of electricity, says Jeff Benjamin, senior vice president, nuclear power plants at Westinghouse.

The simplified AP1000 safety system is known as a “passive” design, in that if there were a failure similar to what happened at Fukushima, human intervention would not be required to shut down a reactor. The engineering for a passively safe reactor has been in the works for decades, Benjamin says. One new feature, for example, is a 2.8-million-liter (750,000-gallon) water tank on top of the containment vessel surrounding the core that would use gravity in the case of emergency to send water flowing over the core, cooling it to prevent a meltdown.

Although the AP1000 is an intrinsically simpler design than the EPR, they both have also faced cost overruns because they are first-of-their-kind deployments. A report by Citi Research found that the Vogtle project is about six months behind schedule and will cost at least $1 billion above original projections. “We’re learning how to do this as efficiently as we can,” says Benjamin, who admits there has been a learning curve. Finalizing the design for construction, which has been approved by American and Chinese regulators, has taken longer than anticipated. So has securing delivery on certain components, such as reactor coolant pumps, because they are being produced for the first time. “We take these lessons as we learn them,” says Benjamin, “and make sure these issues get [resolved] so we don’t face them again.”

Westinghouse hopes that working out the kinks in the first designs will drive down the time line and costs on other plants moving forward—but history shows that may not be possible. A study from Arnulf Grubler at Yale University in 2012 found that despite France’s standardization of building nuclear from the 1970s through 2000s, the French nuclear program still saw substantial cost escalation.

A scaled-down nuclear future?
If large reactors can never achieve significant cost and construction reductions via standardization, some nuclear experts think the way forward is downsized, prefabricated reactors. A number of countries—such as Korea, Russia and the U.S.—are developing SMRs. The smaller reactors are generally less than 300 megawatts in electricity output.

Unlike large reactors, SMRs would be essentially prefabricated in factories instead of on-site. Some of the designs, such as NuScale Power’s, which received DoE funding, don’t require cooling pumps at all. Instead, they are cooled with convection—a natural flow of air. DoE is investing $452 million in SMR support and thinks the units can be built cheaper, faster and with superior safety measures. “The department is very, very focused on the potential for SMRs to make a difference in our energy future,” Lyons says. Because the reactors produce less energy, they are well suited to isolated applications such as powering a small city or for some large industrial usage. Yet some industry advocates think they can supply traditional utility power needs as well.

Researchers at the University of Chicago found that SMRs have the potential of lower upfront costs, along with construction risk, which would lessen the average cost of capital and therefore make the units more cost-competitive with natural gas power plants. A study from Carnegie Mellon University, however, found the jury is still out on whether SMRs will be more cost-effective than gigawatt-sized reactors.

China is also out front in terms of deploying SMRs. One of the most advanced units is joint-venture Chinergy, Co.’s 210 megawatt HTR-PM, a high-temperature, gas-cooled reactor that is currently under construction and is expected to begin operation in 2017.

There is still uncertainty in the marketplace about how regulatory decisions will play out for SMRs, which has led to a bit of a chicken-and-egg problem in bringing them to market. Earlier this year Westinghouse scaled back work on its 225-megawatt SMR as it awaits design certification clarity from regulators. One sticking point is safety and security requirements that would be added if the units are built underground, as the DoE envisions. “It’s fair to think of these small reactors as grid-appropriate for many parts of the world,” Benjamin says. But there are many details to work out, and experts think it will be at least a decade before this technology is ready for prime time.