If They Come, Who Will Build It?
The challenges to advanced reactors aren't just technical and regulatory
By Matthew L. Wald
For good and valid reasons, most of the United States has moved away from having electricity generated, transmitted and delivered by monopolies. The reasons for the change did not primarily have to do with nuclear energy. But a nuclear renaissance could turn out to be a casualty. The old model made construction of a new power plant a shared risk; the new one, in most of the country, turns building a generator into a speculative investment.
Nearly all of the power reactors now operating in the United States got started by monopolies. Utility executives made their best estimates of future demand, and planned to add generation and transmission to meet it. Regulators in each state, usually called public service commissions, reviewed and approved the plans. The utilities, with a mostly-guaranteed revenue stream, easily sold stocks or bonds or borrowed money to build new generation.
When the new assets went into service, the money spent to build them was added to something called the “rate base.” The regulators specified what rate of return (or profit) the shareholders should receive, as a percentage of the rate base. They set the price of electricity accordingly.
Customers, who were captive, also reimbursed utilities for depreciation, as the asset was “used up” over time. These were big jobs done by big companies, who were able to do so because of the existence of this system; it’s what made them so credit-worthy.
Critics pointed out that this tended to create a bias within the utility to solve the supply/demand equation by capital investment, building more generation. As technology evolved, it became clear that total cost, and total environmental effect, could be reduced by investing in efficiency, but the classic regulatory system didn’t give utilities a financial incentive to do that.
This led, in some cases, to over-building generation. Utilities were shocked into starting construction in years of steep growth, but when growth slackened, the plants didn’t seem worth finishing. A pile-up of unfinished plants set the political climate for changing the regulatory system. And for better or worse, as a wave of “de-regulation” swept the United States in the last half of the last century, from trucking to airlines to long-distance phone service, electricity companies joined the list.
The contrast to today is huge. In the beginning of the electric era, a single company would build generating stations, transmission lines and distribution lines, and sell the electricity, a soup-to-nuts operation.
In the renovated system, distribution is still a monopoly. At a given street address, only one company will deliver electricity, and bill for it. That company will be responsible for maintaining the lines and the electric meter. But it might not even be delivering its own electricity; it may be billing on behalf of a third party.
After decades of de-regulation, consumer advocates and others are still arguing over the macro question, which was, did it make electricity less costly? But there is a subsidiary question that deserves more attention than it gets: did it bias the technology choices in a way that reduces our ability to clean up the generation system?
Build at Your Own Risk
Construction of new generating stations is mostly in the “merchant generator” category. That means that a developer (which is not quite what people think of as a utility) secures a site, cooling water, access to the grid, etc., and builds. The developer then has two streams of revenue: whenever the grid operator tells the plant to run, it runs and it sells its electricity at the price that the grid operator has set. That price varies over the course of the day (and the week and the month and the year) and if the market price falls too low, because of reduced demand or the presence of a lot of generation with zero operating cost, like sun and wind, then the merchant generator will shut down, or, if that’s not technically feasible, reduce generation to minimum levels.
The second source of revenue is that in most “deregulated” jurisdictions (but not Texas), companies that supply electricity to consumers must also make capacity payments. In effect, they “rent” capacity sufficient to meet their peak load. For generators that run only a few hundred hours a year, when demand is high, these capacity payments can be a very high portion of their revenue. For “baseload” plants that have lower operating costs and are called on to run around the clock, it is a smaller fraction.
But capacity payments are also highly variable. They are priced by auction, and changes in demand or supply can have big effects on price.
Profit is thus at the whim of the market, and will depend on fuel cost, how many other players in the area are offering electricity and capacity, and similar factors.
(Notable exceptions are the Tennessee Valley Authority, a government agency, and some utilities that are still traditionally regulated, like Georgia Power, which built the new Vogtle reactors.)
All told, the “de-regulated” financial structure makes nuclear a far less attractive option. For one thing, de-regulated financial structures exist on electric systems with lots of wind and solar, where oversupply can make prices go negative, with the grid operating giving incentives to generators to stop generating.
But most reactors are hard to shut down and re-start promptly, because of the physics of fission. They will keep running. And if they do reduce their output, which is difficult for many of them, they do not reduce their fuel costs, because they replace the fuel on a scheduled basis, like replacing the batteries in a smoke detector on a fixed date every year, whether the batteries have a few more weeks of life in them or not.
In a system where the number of hours you will operate is uncertain, the builders prefer a technology in which a huge share of the plant’s expenses are in fuel that can be saved, and not construction expense that cannot. Hence a bias to pick cheap-to-build, expensive-to-run generators. That’s the opposite of nuclear.
In this context, it’s noteworthy that one of the two leading advanced reactor projects, X-energy’s pebble bed high temperature model, was initially planned for a utility, a public power agency in the Pacific Northwest, but after hesitation by that utility, X-energy moved on to a company that isn’t a utility at all. It’s Dow, the chemical company, which has some of the attributes of an old-line utility: deep pockets, construction experience, and an obligation to the community. In this case, the obligation isn’t to provide electricity, and it’s voluntary--it’s to cut its carbon emissions.
Short Time Horizons are Favored
The structure is biased in another way: No one wants to introduce a new generator into a saturated market, and it’s hard to predict what the market will be at the end of the ten years or so it takes to build a reactor. The project sponsor is essentially building on spec, into a market too distant to predict. So, plants that can be built with less lead time are favored because that may require predicting the market only two or three years hence. Those plants tend to be fossil-powered, not because they are better for the engineering of the system, but because gas turbines pop off an assembly line and can be plunked down and installed where needed. Small modular reactors may get to that point someday, but they’re not there now. So, we pick technologies not according to what we need, but what we can get fast.
This is not a rational way to choose long-term assets. It is especially wrong in an era where the two traditional goals of regulation–providing electricity that is reliable and reasonably priced–have been joined by another goal, electricity that is clean. We have a system set up to meet last century’s goals, not this century’s.
The deregulated system not only biases the construction choices toward carbon-emitting plants. By eliminating top-to-bottom monopolies, the deregulated structure also diffuses the responsibility for assuring adequate generation. Hence the situation in Texas, where strong growth in the economy and the population, pared with deregulation, leave the state perennially skating on the edge of shortage. But electricity is vital to the prosperity, comfort and even the lives of people who live there. Arguably, Texas has a regulatory system for electricity that helps winter storms kill people. Storm Uri, in February 2021, killed hundreds in Texas.
Texas has gas and wind available, but there, and in other deregulated places, they are advancing in tandem because of artifacts of the regulatory system. More nuclear reactors would have given the state a more reliable electricity system.
Building nuclear in a deregulated market is theoretically possible. There was a big push to develop a fleet of merchant reactors, using the French EPR design, around 2007, when the price of natural gas hit a peak. At one point, gas was selling for $16 per million BTU. But then came fracking and the price of gas collapsed, dragging the price of electricity down with it. If the EPRs had been built, they would have been forced to sell their electricity way below the market price, causing severe financial pain to the builders. But they would also have taken a chunk out of carbon emissions. Developing a carbon policy that isolates builders from short-term commodity swings in fossil fuel prices is not a challenge that we have mastered, except with extremely generous subsidies to solar and wind.
Solar and wind also continue to be built because in many states they are mandated; the states have adopted a “renewable portfolio standard” that demands that the companies that deliver electricity to customers obtain a certain fraction of that from certain categories of renewables. (A clean energy standard would make a lot more sense, but that is in force in fewer places.)
A New Category of Builder
The first reactors running today were supplied by companies that were already in the power plant business. Two still exist, in modified form: Westinghouse (no longer a conglomerate, but still a nuclear company) and General Electric (now GE-Hitachi). Two are mostly gone, Combustion Engineering and Babcock & Wilcox, although their nameplates can still be found on equipment in operating plants. (And a descendant of Babcock & Wilcox, called BWXT, does extensive nuclear work for the Navy and is branching out into other areas.)
Those legacy companies pursued nuclear reactors as an extension of their existing business lines. They had relationships with foundries, component suppliers and utilities because of their decades of production of plants that ran on coal, gas, or falling water.
GE-Hitachi may emerge as the first successful SMR manufacturer, and Westinghouse may join it. Tellingly, they have smart, advanced designs, but they are technologically close to existing reactors, retaining current fuel designs and other components. The companies with the designs that are greater departures from the familiar models now in service are distinguished for never having built power plants of any kind. They have hired veterans who have long experience, but they are start-ups. This runs counter to the American pattern, where big jobs are done by big companies.
But, when the utility structure requires short time horizons and lower upfront capital costs, many companies have their hands tied. New technologies, like those pursued by small modular reactor developers, can mitigate the challenges brought about by de-regulated energy markets, but require organizations that are brave and have deep pockets to support them. That description applies to Dow, which, as previously noted, isn’t a utility, but not so much to the merchant generator companies that dominate construction in the de-regulated industry.
Start-ups can succeed in capital-intensive industries; Tesla did. But Polestar and Rivian are struggling. The market is more tolerant of start-ups when the field is a new technology. Some start-ups have made a go of it in solar panels. They have some success but the field is dominated by Chinese giants.
More non-utilities like Dow could step forward. Or a utility or group of utilities could step forward to order a series of reactors, spreading the risk and expense of the first-of-a-kind unit among them, and gearing up a supply chain for multiple units. Or the federal government could offer more help to a “first mover.”
But the current utility structure in most of the country is stacked against nuclear, advanced or otherwise.