Out of options
A surprising culprit in the nuclear crisis
As the nuclear crisis in Japan unfolded last week, experts scrambled to understand why things were going so horribly wrong. While no one was surprised that a 9.0 earthquake and a massive tsunami had caused severe and complicated problems, critics charged that various aspects of the Fukushima Daiichi plant’s design had made the catastrophe more perilous than it had to be. Some considered the particulars: Why had the cooling system’s backup generators been installed in a way that left them vulnerable to the tsunami? Why did the reactors use a cost-saving containment vessel whose disaster-worthiness had been repeatedly questioned by scientists? Why had the pool of spent fuel rods overheated?
For those taking a longer view, however, there is a larger question looming over the disaster: Why was Japan, a nation at high risk for earthquakes and natural disasters, using a type of reactor that needed such active cooling to stay safe? And the answer to that doesn’t lie with Japan, or the way the plant was built. The problem lies deeper, and concerns the entire nuclear industry.
Japan’s reactors are “light water” reactors, whose safety depends on an uninterrupted power supply to circulate water quickly around the hot core. A light water system is not the only way to design a nuclear reactor. But because of the way the commercial nuclear power industry developed in its early years, it’s virtually the only type of reactor used in nuclear power plants today. Even though there might be better technologies out there, light water is the one that utility companies know how to build, and that governments have historically been willing to fund.
Economists call this problem “technological lock-in”: The term refers to the process by which one new technology can prevail over another for no good reason other than circumstance and inertia. The best-known example of technological lock-in comes from the 1970s, when VHS and Betamax, two different kinds of videotape, competed in the market until VHS gained a slight lead and then leveraged it to total domination. Whether the VHS format was actually superior to Betamax didn’t matter. After the lock-in, consumers no longer had a choice.
Much more is at stake in nuclear power. Some reactor designs are safer than others in an accident; some are more efficient than others in their use of fuel and produce less nuclear waste. The fact that the industry settled on light water over any number of alternatives was determined in the years after World War II, when the US Atomic Energy Commission and Navy Admiral Hyman Rickover made a series of hasty decisions that irreversibly set the course for how nuclear power plants around the world are built today.
“There were lots and lots of ideas floating around, and they essentially lost when light water came to dominate,” said Robin Cowan, a professor at the University of Strasbourg and the University of Maastricht who wrote a 1990 paper in The Journal of Economic History about the nuclear industry’s technological stagnation. “The market tends to choose a dominant design before it’s optimal, and it tends to under-explore.”
The fact that light water was used in the first American nuclear power plant in 1957 made it that much more likely that subsequent nuclear plants in the United States and around the world would use it, too, as utility companies decided, one after another, that it was in their best interest to use a well-established reactor technology instead of trying something more experimental. The result was that many potentially viable proposals — including plants that, in an emergency, wouldn’t have depended on the diesel generators that failed at Fukushima — were stifled before anyone could properly evaluate them.
Lately, concerns about fossil fuels have brought new enthusiasm for nuclear power, and led scientists back to the task of inventing new and better types of reactors. But even as innovators succeed at securing funding and attention for their research, the power of technological lock-in still hovers over the practical questions of who will design and build them. As the catastrophe in Japan inspires a new reckoning with the benefits and perils of nuclear energy, the industry’s lock-in will need to be part of the discussion.
When nuclear fission was initially harnessed for energy, its first successful use was not in power plants, but in submarines. After the conclusion of World War II, the Navy wanted a submarine fleet that could stay underwater for long periods of time without having to come up for fuel. The key would be nuclear energy. In 1946, Rickover was put in charge of figuring out what kind of reactor should be used in these submarines.
Rickover had several kinds of reactor designs to choose from besides light water, which at the time was the exclusive domain of the American manufacturing giants Westinghouse and General Electric. The most significant differences among them had to do with the materials used to cool their cores and to moderate the fission process. The so-called heavy water reactor, developed in Canada, used a coolant based on a different isotope of hydrogen from that in normal water. (Light water is just normal H2O.) Gas-cooled reactors, meanwhile, which had their origins in Great Britain, did not use liquid coolant at all. The breeder reactor, developed by
Each one had its advantages. The sodium-cooled breeder reactors were more economically efficient; the gas-cooled reactors took longer to get hot, and would probably not melt down as quickly if their power failed. But Rickover’s choice ultimately came down to size. “He needed a reactor that would fit inside his submarine,” said Charles Forsberg, executive director of the MIT Nuclear Fuel Cycle Project. Because reactors built with light water technology could be much smaller than any other kind, Rickover decided they were the Navy’s best bet.
At that point, the nuclear power industry was still anyone’s game. Though Rickover’s decision certainly gave Westinghouse reason to be optimistic about the future of its light water technology, it did not discourage the Canadians from pursuing heavy water nor the British from pursuing their gas-cooled systems. In fact, there was a brief window when it looked like the British had a big win on their hands with the completion of the world’s first commercial nuclear power plant in August 1956. But as it happened, the Americans were right behind them: A little over a year later, the first US nuclear plant went live in the town of Shippingport, Pa. It was a light water plant.
It had not been a foregone conclusion that Shippingport would be powered by light water technology. In fact, according to Cowan, some of the nuclear physicists on the US government’s payroll at the time insisted that they had not done enough research on light water to conclusively declare it was the best option available. But a few factors tipped the scale. First of all, Rickover was in charge of overseeing its construction. And the US government considered it a matter of national security to get the plant built as quickly as possible, in order to send a swift signal to the Soviet Union and the rest of the world about America’s technological supremacy. Light water was familiar, domestic, and the most likely to work immediately.
“We were competing with the rest of the world, trying to become leaders in nuclear technology, and we wanted to quickly, without straining the budget, build some demonstrations,” said Ashley Finan, a PhD candidate at MIT’s department of nuclear science and engineering who has been studying the history of innovation in nuclear technology since World War II. “The reactors that were most available were the light water reactors, so that’s what we used.”
It was after Shippingport that light water truly took off, as both Westinghouse and General Electric — which had focused all of its resources on developing water reactors after some early experiments with breeders — made an aggressive push on behalf of the United States for dominance of the global market. “They knew there were other technologies in the wings that might be better,” said Cowan, whose paper on the rise of light water technology was recently featured on the economics blog Marginal Revolution.
By 1970, according to Cowan’s research, light water had been adopted by every major consumer of nuclear power in the world except for Canada and Great Britain, who were at that point still trying to make a go of heavy water and gas-cooled, respectively. According to Finan, the federal regulations in the United States essentially assumed that plants would use light water technology, making it extremely difficult for any other type of plant to get clearance at all.
By 1986, more than 80 percent of all nuclear reactors in the works around the world — excluding the Soviet Union — were of the light water variety. Japan’s reactors were built in the 1970s based on light water designs by General Electric,
Light water may have been — and may still be — the best option available. Certainly the technology has been refined since its early days, and newer reactor designs have advanced safety features that partly address the cooling problems suffered in Japan. But the trouble with technological lock-in is that you never really know: With only one choice, it’s impossible to tell whether you might have been better off with one of the early alternatives.
“Once the bandwagon gets rolling and starts to accrue advantage, it tends to get more advantage,” said W. Brian Arthur, the economist at the Santa Fe Institute who introduced and coined the concept of technological lock-in in the early 1980s. “Even if it’s not the best to start with, if it just gets ahead by chance, then it tends to get further ahead because of all the advantages.”
As soon as one technology gets a significant leg up in the competition, it becomes extremely difficult for rivals to derail it. Apart from the VHS-Betamax competition, the effects of lock-in can be seen in the dominance of the QWERTY configuration of keyboard letters, and in the victory of internal-combustion automobiles over Stanley Steamers in the early 20th century.
In the nuclear industry, more experimental approaches were decisively frozen out before their merits could be properly tested. A good example of this happened in the mid-’70s, when an American company called General Atomics tried to break into the market with a type of “high temperature” gas-cooled reactor that could cool itself in an emergency using the natural circulation of air instead of relying on motor-powered pumps and valves. One might have thought this cooling system would be seen as a major achievement — and certainly, a useful option for sites like Japan, where there’s a high risk of natural disasters knocking out a power supply. But of the seven high temperature graphite-gas-based power plants that General Atomics had been contracted to build by 1976, all but one ended up being canceled.
“One of the, let’s say, challenges in nuclear technology is that the plants are expensive and owners are usually risk-averse,” said Mujid Kazimi, director of MIT’s Center for Advanced Nuclear Energy Systems. “Part of it, also, is that it takes a long time to get a license for a new reactor concept. If you have a new idea and you want to eventually get it into the market, you need a license, and that can take years.”
Ultimately, experts say, overcoming technological lock-in requires the deliberate participation of the federal government, at least by partnering with private companies and university researchers trying to introduce new ideas. Trying to disrupt a locked-in technology is an extremely risky proposition, and in most cases, it’s too tall an order for a single company to take on alone. As Arthur put it, “If everybody’s clapping to one beat and you try to start clapping to a different one, it’s a little hard to take over. It can be done, but you need a whole bunch of people and it needs to be loud.”
There are signs that that’s starting to happen. Over the past decade or so, concerns about climate change and fossil fuel dependence have led to something of a renaissance for the nuclear industry, which has started to look into new ideas that don’t involve light water, and some that do. In 2002, the US government unveiled an initiative aimed at helping private companies develop and bring to market new nuclear technologies. And according to Finan, the federal regulations that are used to license new nuclear technology in the United States are in the process of being rewritten so that they don’t favor light water reactors as heavily as they used to.
Meanwhile, a number of promising projects are underway around the world. Toshiba is working on a sodium-based breeder reactor. The Department of Energy is working with General Atomics and Westinghouse on a high-temperature gas-cooled one. In India, there are plans for a reactor that uses thorium and molten salt. Meanwhile, several manufacturers are exploring designs for so-called modular reactors that could radically change the industry — and make it friendlier to innovation — by allowing utilities to build small plants and expand them as necessary instead of going all in with a multibillion dollar investment right away.
Whether any of these ideas catches on will depend in part on the industry’s willingness to take risks. It will also depend on circumstances, though, just as it did back in the 1940s when light water first emerged. As the nuclear industry comes to grips with the tragedy in Japan, experts hope that one effect will be to encourage a rare, concerted push that forces a locked-in market to open up again.
“There are certainly innovative reactors available to be developed that have safety features we would value,” said Finan. “I hope we’ll see a more open-minded approach.”
Leon Neyfakh is the staff writer for Ideas. E-mail firstname.lastname@example.org.
Correction: Because of a reporting error, this article gave an incomplete account of why light-water reactors were chosen for the US nuclear submarine fleet after World War II. In addition to being compact enough, light-water reactors were also deemed more practical than the alternatives.