Overcoming Obstacles to Advanced Reactor Technologies

Perspective

C O R P O R A T I O N Expert insights on a timely policy issue

Edward Geist

Whither the Nuclear Renaissance? high natural gas prices experienced prior to the 2008 financial crisis A decade ago, it appeared that the United States might be on the would continue indefinitely. Instead, the economic downturn and verge of a so-called nuclear renaissance. Bolstered by high natural the increasing supply of shale gas transformed methane from an gas prices, new standardized reactor designs, a streamlined regu- increasingly unaffordable commodity into the most cost-effective latory framework, and the prospect of stronger greenhouse gas fuel for new generating capacity, including baseload applications. regulations, U.S. utilities began ordering new nuclear power plants Second, the predicted capital costs for new nuclear plants increased for the first time since the 1970s. Proponents asserted that the new markedly, from as little as $1,100 per kilowatt a decade ago to more reactors would not merely provide a massive improvement in safety than three times that for the units now under construction, even over existing designs but also produce electricity at a cost competi- as the economic crisis tamped down demand for new generating tive with conventional alternatives. Nuclear fission would finally capacity.1 The U.S. government’s failure to put a price on carbon become a viable business proposition, offering a profitable avenue to wean the American economy off of fossil fuels. 1 The U.S. Energy Information Administration’s Annual Energy Outlook 2004 An unanticipated confluence of economic and political factors with Projections to 2025 predicted that Westinghouse’s AP1000 reactors would cost $1,580 per kilowatt for first-of-a-kind units, dropping to $1,081 in 2019 (U.S. smothered the nuclear renaissance. The economic case for new Energy Information Administration, 2004, p. 58). As of early 2015, the projected nuclear power plants had been premised on the notion that the completion costs of the two AP1000s under construction by Georgia Power are emissions deprived nuclear fission of a critical economic advantage fuels, hold out the possibility of being both cheaper and safer than over fossil fuels. The accident at Japan’s Fukushima Daiichi nuclear current nuclear plants. Unfortunately, the NRC is ill-equipped to plant and waning enthusiasm for nuclear power in Congress and assess the safety of these novel nuclear plants, in large part because state governments coincided with U.S. Nuclear Regulatory Com- the institutions charged with encouraging the development of these mission (NRC) decisions to increase the licensing requirements for technologies and demonstrating their operational characteristics new nuclear plants, imposing costly delays on the plants under con- have failed to do so effectively. Therefore, institutional reforms are struction. As a consequence, all but two of the new reactor projects needed if nuclear fission is to play a greater role in ensuring U.S. in the United States were either canceled or put on hold. energy security and forestalling climate change. This paper aims to These economic and political factors were merely proximate outline the policy challenges that must be met in order to establish causes for the failure of the nuclear renaissance. The fundamental the commercial viability of advanced nuclear technologies while reason for the economic uncompetitiveness of the current genera- developing a regulatory framework that can ensure the safety of the tion of nuclear plants lies in the imperfect safety characteristics of public without imposing prohibitive costs on new reactors.2 the large light-water reactors (LWRs) that have historically formed the basis of U.S. civilian nuclear technology. Although these The Physical Underpinnings of Nuclear Safety LWRs are more affordable and less dangerous than reactor designs The unfavorability of the economics of current nuclear power commercialized in some other nations, the physics of these large plants ultimately stems from the potential consequences of extreme LWRs and the zirconium-clad uranium oxide fuel they employ accidents. The core of a typical nuclear power reactor contains an requires the inclusion of costly engineered safety systems to protect immense radiological inventory, equivalent to that released by a the public from the consequences of severe accidents. The nuclear large thermonuclear explosion. The release of merely a small frac- renaissance hinged on the assumption that standardized design and tion of these radioactive isotopes to the environment can result in modular construction would reduce the expense of these systems to radiation exposures demanding the evacuation of local popula- manageable levels, but experience has not borne out this hope. tions and billions of dollars in economic losses. Therefore, the goal If nuclear power is to play a greater role in the U.S. energy sup- of nuclear safety is to forestall this outcome both by reducing the ply in coming years, it must move beyond the current paradigm of likelihood of accidents as much as possible and by attempting to large LWRs. Numerous candidates for alternative nuclear technolo- prevent the release of radioactive material even in the event of a gies, ranging from miniature, mass-produced versions of exist- severe accident. ing LWRs to exotic possibilities, such as reactors utilizing liquid

2 We use the term advanced nuclear reactor to refer to a reactor using technolo- $7.4 billion, or about $3,300 per kilowatt, the increase being caused by mounting gies other than those used in the large LWRs that are currently operating in the construction delays (Overton, 2015). United States.

2 Current LWRs achieve these goals through the use of a fuel cladding fails at a much lower temperature and experiences defense-in-depth strategy that aims to provide multiple barriers a catalytic reaction with heated steam producing large amounts between the radiological inventory in the fuel and the environ- of potentially explosive hydrogen, as the Fukushima Daiichi ment. The first, and most important, of these barriers consists of accident dramatically illustrated. To avoid these pitfalls, the fuel the fuel assemblies themselves. All nuclear power plants currently must be adequately cooled at all times, including when the reac- operating in the United States utilize fuel that consists of uranium tor is shut down. Although LWRs are essentially immune to the dioxide pellets jacketed in a zirconium-alloy cladding (the best- kind of runaway power excursion that caused the explosion of known brand being Zircaloy).3 These are assembled into carefully Chernobyl Nuclear Power Plant unit 4 in April 1986, the decay of engineered fuel bundles with the geometric, chemical, and nuclear short-lived radioisotopes in the core produces so much heat that, characteristics essential to survive within the punishing environ- in the absence of cooling water, it can cause the fuel assemblies ment of a nuclear reactor core. So long as the fuel assemblies to lose integrity within minutes, resulting in a nuclear meltdown. remain intact, little or no radioactive material will escape. The sec- Unfortunately, the use of highly pressurized water as coolant makes ond layer of defense is found in the massive steel reactor vessel that maintaining cooling under accident conditions particularly chal- surrounds the fuel assemblies. Designed to operate at high pressure lenging. A break in the reactor loop can allow the water to flash and temperature, the reactor vessel places several inches of steel into steam, potentially exposing the fuel assemblies, and introduc- between the fuel and the third layer, the containment building. ing additional cooling water requires either substantial energy, the This ferroconcrete structure, which varies considerably in design depressurization of the reactor vessel, or both. between power plants, aims to prevent the release of radioactive LWRs therefore need engineered safety systems to avoid a material even in case of an accident compromising the integrity of nuclear meltdown, but these do not come cheap. Having multiple one of the reactor’s primary coolant loops. redundant emergency cooling arrangements designed to address Typical LWR fuel assemblies have problematic physical and different accident conditions increases the construction and opera- chemical properties that pose serious challenges in the case of an tional complexity of LWRs, swelling their costs. Furthermore, in accident. Although the uranium oxide fuel pellets have a high heat older plant designs, these systems require considerable amounts of resistance, their poor thermal conductivity allows their interior electrical power to function, but the Fukushima Daiichi accident to approach the material’s melting point of 2,865 degrees Celsius proved the naïveté of assuming the availability of emergency power. even under ordinary operating conditions. The zirconium-alloy Plant designers have sought to overcome these challenges by pursuing economies of scale, but increasing the size of LWRs compounds 3 Although mixed-oxide fuel containing both enriched uranium and plutonium some of their safety problems. The larger the reactor core, the is used in LWRs abroad and was tested at the Catawba Nuclear Station in South lower the ratio of the surface area of the reactor vessel to the core Carolina between 2005 and 2008, it is not currently used in the U.S. nuclear industry. volume, increasing the difficulty of passive cooling. Furthermore,

3 large nuclear plants demand precision, Containment and Safety in a Light-Water Reactor high-quality construction on a scale that severely taxes the contractors tasked with building them, often increasing costs because of the need to redo work first performed incorrectly. Finally, larger plants have longer construction Steam line Walls made of concrete and Containment times, and those help inflate financing steel 3–5 ft. (1–1.5 m) thick cooling system costs to prohibitive levels. Despite these challenges, the gen- Steam eral technical consensus is that current generator

LWR designs, such as Westinghouse’s Reactor Control vessel rods AP1000 and Areva’s EPR, are safe Turbine generator enough to offer an extremely high assurance that even an extreme accident would pose negligible danger Heater Condenser to the public.4 Sixty years of experi- Condensate ence designing and operating LWRs, pumps Coolant thousands of reactor-years of cumulative loop operation, and the best efforts of thou- Feed pumps sands of nuclear engineers, however, have not produced a design for a large LWR that is both safe and affordable Demineralizer Reactor coolant Pressurizer enough to compete with conventional pumps Containment Emergency water sources of electricity. If nuclear power is structure supply systems to play a major role in America’s energy SOURCE: NRC, 2015. RAND PE156-1

4 The cancellation of Constellation Energy’s plans to build an EPR in Maryland leaves the AP1000 as the only new nuclear plant design under construction in the United States. For an assessment of the safety characteristics of the AP1000 produced by the UK government, see UK Health and Safety Executive, undated.

4 future, new reactor designs with intrinsically better safety charac- Regulating Beyond Light Water teristics must be found. Unfortunately, because of the historical predominance of LWRs Thankfully, large LWRs are far from the only way to exploit in the United States, the NRC is poorly equipped to evaluate the the potential of fission. Historically, pressurized-water reactors and safety of alternative technologies. Practical experience with a type boiling-water reactors were merely two of a wide range of power of reactor is essential to develop appropriate regulations, and all reactor designs investigated by the Atomic Energy Commission but a handful of early U.S. nuclear power plants utilized LWRs. (AEC), and, arguably, they came to predominate for circumstantial The NRC’s institutional procedures evolved over time on the basis reasons.5 Today, a variety of alternative nuclear power paradigms of growing experience constructing and operating nuclear power are under investigation both in the United States and abroad. plants, on several occasions as a reaction to near misses that called All but the least exotic of these proposals take a very different its regulatory assumptions into question. These events, includ- technical approach to nuclear safety from that of current LWRs. ing the 1975 Browns Ferry fire and the 1979 meltdown of Three Aiming to eliminate the reliance on costly engineered safety sys- Mile Island unit 2, stemmed from the dismissive attitude toward tems that crippled the economic viability of present-day designs, safety risks that the NRC’s institutional predecessor, the AEC, these reactors posit combinations of fuel and coolant that promise demonstrated. Determined to spur the widespread commercial high safety margins even in the absence of emergency power or adoption of nuclear fission, the AEC discounted its own employees’ operator intervention. In short, they seek safety in physics rather warnings that the technical assumptions undergirding its safety than in engineering. But without experience operating pilot plants, regulations remained untested and might be egregiously in error the NRC is ill-equipped to evaluate power reactors that deviate (Walker, 2004, pp. 54–62). Most critically, AEC regulators focused substantially from light-water designs, and the current regulatory on a narrow range of possible accidents—in particular, a so-called framework makes assumptions that might not be appropriate for design-basis accident involving the sudden breakage of the largest novel nuclear concepts. For reasons explained below, both practical pipe in the primary coolant loop—and convinced themselves that experience with alternative nuclear technologies and institutional safety measures designed for this extreme scenario would prevent reforms are necessary before the United States can contemplate severe core damage in any likely accident.6 Because the design-basis deploying these new reactors. accident did not incorporate external initiating events, such as earthquakes, and the possibility of operator error, nuclear engineers sought to combat it largely by means of redundant engineered safety systems. 5 An influential 1990 article by economist Robin Cowan argues that, although “light water is considered inferior to other technologies,” the heavy investment of the U.S. Navy in LWRs for naval propulsion allowed it to become entrenched 6 For a discussion of the AEC’s philosophy regarding the design-basis accident, before alternatives could enter the market (Cowan, 1990). see Thompson and McCullough, 1973.

5 Although the AEC’s overconfidence in this area later caused array of possible accidents, but it had two associated pitfalls: It considerable grief for the U.S. nuclear industry, the organization addressed only those risks considered by the analysts, and its esti- was actually overly conservative in other areas of nuclear safety. mates could never be any more accurate than the probabilities the Uncertain how melting fuel would interact with reactor vessels analysts employed, even though, in many cases, these values were and containment buildings and unwilling to undertake expensive no more than semi-educated guesses. and politically risky physical experiments in this field, the AEC As a consequence, the PRA initially received a mixed response assumed that substantial fuel damage would likely cause unac- from both the nuclear industry and skeptics of nuclear power. ceptable radioactivity releases to the environment (Geist, 2014). Producers of nuclear energy found it much less straightforward This belief compounded the emphasis on engineered safety systems to design and operate plants to account for the many accident because LWR fuel would require some sort of cooling during an sequences envisioned by the PRA than to address a single design- accident. basis accident (Guibert, 1983). Opponents of nuclear power, mean- A public scandal resulting from the AEC’s attempt to squelch while, dismissed the PRA as a cynical attempt to silence criticism concerns about the potential inadequacy of the emergency core- by using dodgy statistics to assert that nuclear reactors were “safe cooling systems in plants then under construction helped impel enough.”8 As a consequence, the NRC only gradually integrated the establishment of the independent NRC in 1975 (Walker, 2004, the PRA into its regulatory activities over the subsequent decades; p. 62). Shortly after the NRC’s formation, the new agency began even today, it continues to employ a combination of empirical and exploring alternatives to the deterministic approach to nuclear risk-based techniques.9 safety embodied in the perspective of the design-basis accident. The NRC did not apply its shifting regulatory requirements The influential Reactor Safety Study adapted the probabilistic risk equally to all plants. In accordance with the so-called backfit rule, assessment (PRA), a technique first pioneered in the aerospace modified regulations apply differently to facilities depending on industry, to estimate the probability that a large LWR would the date on which their construction or operating licenses were experience an accident resulting in severe core damage.7 The PRA issued.10 In practice, the logic of the backfit rule has led to absurd assigns probabilities to a large number of events that might occur during the course of a nuclear accident, then utilizes these figures 8 For an important early example of this, see Union of Concerned Scientists, 1977. to calculate probabilities of sequences threatening core integrity. 9 For the NRC’s account of the increasing use of risk-based techniques in its The PRA represented an immense improvement over the older, regulatory activities between the 1970s and the 1990s, see NRC, 2013. In recent deterministic technique because it can account for a much broader years, the NRC has increasingly explored the possibility of making a full transition to what it calls a Risk Management Regulatory Framework that would complete the transition to risk-informed approaches. See Risk Management Task Force, 2012. 7 On the history of the PRA for civilian nuclear facilities, see Keller and Modarres, 2005. 10 For the NRC’s regulatory definition of the backfit rule, see NRC, 2014a.

6 results, which are illustrated by the handful of nuclear power plants currently under construction in the United States. Although the [I]t is entirely possible for the nuclear four AP1000s being built in Georgia and South Carolina are the industry to be simultaneously overregulated only nuclear power plants that have begun construction in the in some respects and underregulated in United States since the 1970s, another plant is currently nearing others. completion—the Tennessee Valley Authority’s (TVA’s) Watts Bar unit 2 in Tennessee. Although the AP1000s experienced costly 13 construction delays when the NRC imposed new regulations for victim of “regulatory capture.” These positions are not mutu- resistance to aircraft impact, Watts Bar was exempt from these ally exclusive—it is entirely possible for the nuclear industry to be measures—because it has an active construction license that the simultaneously overregulated in some respects and underregulated AEC approved during the Nixon administration.11 in others. Finally, the NRC is underfunded and understaffed to The NRC’s current regulatory regime is a source of intense carry out all parts of its mandate, despite the small number of dissatisfaction for all stakeholders, including the nuclear industry, facilities under construction and the shrinking number of older 14 its critics, and the NRC itself. The builders and operators of nuclear plants that remain in service. It struggles to compete for qualified power plants consider themselves to be subject to arbitrary and personnel with private industry and other sectors of the govern- inconsistent regulations, capriciously applied by an overbearing ment, and it is torn between a fee-for-service funding model that bureaucracy.12 Antinuclear activists, meanwhile, regularly assert makes it dependent on the industry it regulates and politicians who that the NRC too often bows to industry demands and is the meddle in its decisions (U.S. Government Accountability Office [GAO], 2012, p. 30). In short, the history of civilian nuclear regulation in the United States offers relatively few positive lessons for how to craft a regulatory foundation for a qualitatively different type of civil- 11 For these regulations regarding aircraft impact, see NRC, 2014b. Watts Bar units 1 and 2 began construction in 1973, but work ceased in 1988. In 1996, TVA completed unit 1, which is the most recent nuclear plant to enter service in the United States. Because of a peculiarity of its charter, which forbade it from 13 Shortly after the accident at Fukushima Daiichi, Princeton professor Frank N. writing off its partially completed nuclear plants like all private utilities ultimately von Hippel penned a New York Times op-ed making this accusation (von Hippel, did, TVA is the only utility with outstanding construction licenses predating the 2011). NRC. In addition to building Watts Bar, TVA also has construction licenses for 14 Although NRC staffing increased considerably in the 2000s in response to Bellefonte units 1 and 2 in Alabama, and it has seriously considered completing increasing retirements and anticipation of a nuclear renaissance, these personnel these units as well. lacked the experience of those departing and did not necessarily have the particu- 12 Congressional nuclear energy proponents grilled Allison Macfarlane on NRC lar esoteric skills that the regulator needed. At present, the NRC is proposing to “overregulation” at her 2013 confirmation hearings (Nuclear Energy Institute, reduce its staff from 3,677 today to about 3,400 in 2020 (“US NRC Could Shrink 2013). by 10% in Five Years,” 2011).

7 ian nuclear technology. It does, however, provide several important constructed an experimental reactor on the territory of its own lessons: facilities (Rosenthal, 2010, pp. 14–23). In others, such as that of • Because of its hybrid of deterministic and risk-informed the short-lived Hallam Nuclear Generating Station in Nebraska, approaches, the NRC’s current regulatory regime might be too the AEC partnered with private industry to construct and oper- closely tailored to the large LWRs currently in use to evaluate ate pilot commercial plants (U.S. Department of Energy Office of other reactor technologies fairly. Legacy Management, 2008, pp. 2–4). These experiments provided • The risks of any new technology need to be assessed prior to the operational experience essential both to develop commercially the design of the regulatory framework, rather than following viable reactors and to establish necessary regulatory norms. The commercialization, as occurred with LWRs. AEC’s role as both promoter and regulator of civilian nuclear • This assessment should include an analysis of off-site conse- power, however, created a serious conflict of interest. An increasing quences of the maximum credible accident whose assumptions perception that the AEC had pressed for the aggressive adoption have been verified by empirical data.15 of large nuclear reactors for power generation without establishing their safety characteristics first helped impel the breakup of The Neglected Department of Energy Role in New Reactor the AEC into the NRC and the Energy Research and Develop- Development ment Administration (ERDA) in 1975. As implied by its moniker, In theory, the NRC is supposed to work in concert with the ERDA inherited the AEC’s mission to develop and promote new Department of Energy and industry interests to develop, license, civilian nuclear technologies (Walker, 2004, pp. 31–34). and commercialize new nuclear power reactor technologies, but Although ERDA—and, from 1977, its successor, the Depart- this arrangement has never worked as intended. Prior to its disso- ment of Energy—aimed to continue the AEC’s earlier precedent lution in the mid-1970s, the AEC aggressively sought the com- of maintaining effective partnerships between government and mercialization of nuclear reactors for civilian power generation. In industry to bring new reactor technologies into commercial use, addition to evaluating the LWRs that ultimately saw widespread both political and market pressures conspired to prevent this. Pri- adoption, which were originally developed for military applications, vate industry invested heavily in nuclear energy during the 1960s the AEC investigated a wide array of reactor technologies envi- and 1970s for two reasons. Firstly, strong growth in electric power sioned solely for civilian roles. In some cases, such as the aqueous consumption throughout the postwar period established a percep- homogenous reactor at Oak Ridge National Laboratory, the AEC tion that electricity demand would continue to increase rapidly for the foreseeable future. With so much future demand, it seemed that 15 The NRC employs three levels of PRA, of which only the level 3 PRA attempts new forms of electricity generation would be essential to meet it a full analysis of off-site consequences. Although the NRC has been contemplat- all—and utilities invested in nuclear fission accordingly. Secondly, ing expanded use of level 3 PRA in its own regulation, the last such assessment it sponsored was NUREG-1150 in 1990 (NRC, 1990; Sheron, 2012). until the mid-1970s, nuclear power appeared to be the cheapest

8 form of new generation.16 Thanks to these assumptions, during the inspire intense scrutiny of the department’s research programs, par- AEC era, both utilities and plant vendors had been willing to fund ticularly in controversial areas, such as nuclear power. The depart- the construction and operation of commercially unviable prototype ment could legally construct and operate exotic research reactors on nuclear plants in the belief that the experience they gained would its own territory without being subject to NRC regulations because pay handsome dividends down the road. By the end of the 1970s, it inherited from the AEC the right to self-regulate its own nuclear changing economic conditions exploded both of these rationales for facilities, but there has been little budgetary or political support to commercial interests to risk investment in untried nuclear tech- exercise this privilege.17 nologies. The 1973 energy crisis and its 1979 reprise, along with a Given these unfavorable conditions, it is of little surprise that stagnant economy, slowed growth in electricity demand to a frac- the Department of Energy’s activities developing new types of tion of what had been expected a few years before. Furthermore, power reactors have largely been continuations of research efforts massive cost overruns and indifferent operational performance that were particularly advanced at the time of the AEC’s dissolu- made the large LWRs under construction much less economically tion. These programs have emphasized two types of reactors: liquid attractive than projected. Where the AEC had coaxed utilities to sodium–cooled fast reactors and high-temperature gas-cooled invest in nuclear power with visions of ample future profits, they reactors (HTGRs). During the AEC era, many in the technical found themselves struggling to recoup their multidecade invest- community believed that future energy demand would ultimately ment in the technology (Walker, 2004, pp. 7–9). Understandably feeling burned by this experience, utilities showed little subsequent Given these unfavorable conditions, it interest in taking chances on exotic nuclear reactors. Nor did the Department of Energy prove a particularly is of little surprise that the Department enthusiastic advocate for the development of new types of nuclear of Energy’s activities developing new power plants. Charged with activities ranging from maintaining types of power reactors have largely been the nation’s nuclear arsenal to conducting basic scientific research, continuations of research efforts that were the Department of Energy found itself overburdened with more- pressing responsibilities. Increasing awareness of the lamentable particularly advanced at the time of the environmental legacy produced by careless practices during the AEC’s dissolution. early years of U.S. nuclear weapon development both consumed an ever-increasing fraction of the department’s budget and helped 17 Given the Department of Energy’s imperfect safety and environmental record, there have been several proposals that an agency, such as the NRC, should have 16 For a contemporary assertion of these views, see Atlantic Council of the United oversight of the department’s nuclear facilities. For instance, see Haughney et al., States, 1976. 1999.

9 demand the widespread deployment of nuclear breeder reactors, that made more-efficient use of resources inherited from the AEC. which produce more usable nuclear fuel than they consume, Instead of attempting to construct an astronomically expensive and fast reactors utilizing molten sodium as coolant appeared to pilot plant from scratch, which the French and Soviets were doing offer the most promising gateway to a future so-called plutonium on a grander scale anyway, the integral-fast-reactor program aimed economy. As a consequence, the AEC lavished a disproportionate to demonstrate critical new technologies in facilities originally built share of its attention and resources on these reactors, including in the 1960s. Modifications to the Experimental Breeder Reactor II test reactors on its own territory (such as the Experimental Breeder aimed to demonstrate both the reactor’s passive safety characteris- Reactor), and pilot commercial plants (such as the Enrico Fermi tics and on-site reprocessing of metallic fuel using electrorefining Nuclear Generating Station on Lake Erie) (Waltar and Reynolds, (Chang, 1989). Opposition to the project in Congress, however, 1981, pp. 24–25). resulted in its cancellation in 1994 before some of these goals had At the time the AEC broke up, the centerpiece of this program been met (Westfall, 2004, pp. 25–27). was the Clinch River Breeder Reactor project, which aimed to In addition to the stillborn dream of fast reactors, the Depart- build a 350-megawatt prototype fast reactor plant in Oak Ridge, ment of Energy pursued a second advanced reactor technology— Tennessee. Although safety and proliferation concerns stoked HTGRs. Like research on the breeder, research into this concept increasing political opposition to this effort in the late 1970s, was already well advanced when the department was established. runaway cost inflation sounded the project’s ultimate death knell. The AEC had worked with the Philadelphia Electric Company Where the AEC had expected costs to be split evenly between gov- to build a pilot HTGR that went online in 1966, and a powerful ernment and industry investors, as the budget swelled, commercial industrial player, Gulf General Atomic, was aggressively promot- interests refused to increase their pledged contributions. As a result, ing a standardized HTGR of its own design (International Atomic this single project cannibalized most of the Department of Energy’s Energy Agency [IAEA], 2001, p. 6). Most orders for these plants budget for advanced nuclear development in its early years, and were canceled, but a single unit, Fort St. Vrain, operated in Colo- more than $1 billion in government money was expended by 1981 rado from 1979 until 1989 (Copinger and Moses, 2004). Whereas even though construction never progressed beyond the preliminary growing concerns about safety and proliferation, as well as low phase (Boudreau, 1981, p. 119). By the time Congress canceled the uranium prices, made fast metal-cooled reactors an increasingly project in 1983, the General Accounting Office predicted that the dubious proposition in the 1980s, they increased the appeal of reactor would require a total of $3.6 billion to complete (Bowsher, HTGRs. Bolstered by the coalition of commercial interests dubbed 1982, p. i). Gas Cooled Reactor Associates, the Department of Energy spon- Following the cancellation of the Clinch River Breeder Reac- sored a research program to develop what it called a Modular High- tor project, the Department of Energy pursued the integral fast Temperature Gas-Cooled Reactor (MHTGR) that built on earlier reactor, an alternative sodium-cooled reactor development program experience. Thanks to this industry interest, a preliminary safety

10 information document about a 350-megawatt-thermal MHTGR program, with the ultimate aim “to demonstrate the generation of was submitted to the NRC in 1986. The NRC issued a generally electricity and/or hydrogen with a high-temperature nuclear energy favorable draft safety evaluation report on the proposed design in source”—in this case, by completing a prototype HTGR at Idaho 1989, but regulatory progress then stalled because of what briefly National Laboratory by 2021. Between fiscal years 2006 and 2010, seemed like a fortuitous development for the MHTGR program the Department of Energy expended $528.4 million on the NGNP (IAEA, 2001, p. 9). program, but, in 2011, plans to move forward with what it called Although it initially envisioned the MHTGR as a useful source phase 2 and build the pilot plant were put on hold for budgetary of process heat for industrial applications, at the end of the 1980s, reasons (U.S. Department of Energy Office of Nuclear Energy, the Department of Energy became interested in the possibility of 2010, pp. i, iii; GAO, 2014). Despite this impasse, in 2012, the building a highly modified MHTGR on the territory of one of its department selected an Areva HTGR design utilizing prismatic own research facilities to produce the tritium needed to maintain block fuel and a steam cycle for the prototype facility, although it the U.S. nuclear arsenal. Dubbed the New Production MHTGR, did not provide Congress with initial design parameters as required the design changes necessary for tritium production inspired doubt under the Energy Policy Act (“Areva Modular Reactor Selected for within the NRC as to the applicability of the earlier safety review; NGNP Development,” 2012). in any case, the predicted final cost of $3.6 billion made the project In 2014, GAO reported that progress toward constructing the a political nonstarter. The New Production MHTGR program NGNP had effectively stalled. The inability to formulate a cost- was canceled in 1992, followed by the interruption of the Depart- sharing agreement acceptable to both government and industry ment of Energy’s other HTGR development efforts in 1994 (IAEA, poses an imposing obstacle to the NGNP, but the Department 2001; Lartonoix, 2014, p. 158). of Energy’s unwillingness to specify exactly what it aims to build Near the end of the Clinton administration, the Department makes both budgetary and regulatory challenges impossible to of Energy revived its HTGR research efforts, which led to the resolve. Wary of the department’s well-established reputation for still-ongoing Next Generation Nuclear Plant (NGNP) program. cost inflation and missed deadlines, industry players are loath to In 1999, the department launched the Nuclear Energy Research invest in the project without a clear guarantee that their contribu- Initiative, which aimed to develop nuclear reactors with better tions will both remain predictable and result in an operational safety, proliferation resistance, and economic characteristics than facility. Meanwhile, the NRC “has reassigned staff from its NGNP LWRs (World Nuclear Association, 2015). Because HTGRs operat- work and engages with [the Department of Energy’s Office of ing on a once-through fuel cycle appeared to meet these goals and Nuclear Energy] on a minimal basis” because NRC officials believe looked comparatively technologically mature, the department soon “that they cannot proceed substantively further in developing a chose to emphasize this particular technology. The Energy Policy licensing framework until [the Office of Nuclear Energy] has devel- Act of 2005 (Pub. L. 109-58) formally established the NGNP oped a specific design for an advanced reactor technology” (GAO,

11 2014, pp. 32–33). In light of these circumstances, GAO reached their resources on technologies that are already comparatively the sensible conclusion that “it is not clear if or when” the Depart- mature, particularly HTGRs. Yet, even after decades of intense ment of Energy will actually build the NGNP (GAO, 2014, p. 32). interest from both the department and industry, the NRC has yet The institutional procedures of both the Department of to specify what the safety requirements for HTGRs should be; Energy and the NRC contribute to a vicious cycle that impedes the without substantial guidance on this issue, it is impossible to design development of new types of nuclear reactors in the United States. the pilot plant and predict how much it will cost. Without institu- During its four decades of existence, the Department of Energy has tional reforms of some kind, this pattern is likely to repeat itself ad spent billions of dollars on research into non-LWR technologies, infinitum, preventing the construction and licensing of non-LWR without many tangible results. Technologically, the selected design nuclear plants in the United States. for the NGNP is not all that different from the stillborn 1980s Such reforms could take a variety of forms, but only a few of MHTGR. Without experience operating new types of reactors, them are politically or economically realistic. One possibility would it is difficult to learn enough about their operational and safety be to recreate the conditions under which nuclear power was originally developed by significantly relaxing regulatory requirements characteristics to ascertain their commercial viability. Govern- for prototype plants, but this seems highly implausible, as does the ment investment in experimental reactors is acceptable only with prospect of vastly increased federal research and development fund- substantial investment from private sources, but these interests fear ing for the development of new types of nuclear reactors. For the a repeat of the Clinch River Breeder Reactor scenario, in which the foreseeable future, the likeliest means of licensing non-LWR plants project consumed far more than the original budget despite never in the United States might be to draw more effectively on the coming anywhere near completion. Furthermore, private industry experience of other countries that are aggressively pursuing such is obviously interested in contributing only to projects with clear technologies. By forging effective partnerships with these nations, commercial potential, yet costly pilot facilities must be built before the United States could influence their research priorities toward the marketability of new reactors can be confirmed. As a conse- issues of particular interest to U.S. regulators and industry. Foreign quence, the Department of Energy and industry have concentrated partners would benefit from this arrangement because the NRC has a reputation of being particularly thorough compared with its The institutional procedures of both the counterparts elsewhere, so a plant approved by the NRC would Department of Energy and the NRC probably benefit from greatly improved international marketabil- contribute to a vicious cycle that impedes the ity. The downside of emphasizing such partnerships is that they would betray the longstanding goal of the Department of Energy’s development of new types of nuclear reactors research efforts to maintain U.S. industry leadership in civilian in the United States. nuclear energy. Without operational experience, however, there is

12 little chance that the United States will be able to attain leadership ily than other isotopes. To represent a meaningful improvement in the field of advanced reactor technologies. over current LWR technology, new nuclear reactor designs, such as those listed in the table, need to provide a significantly more Envisioning Alternative Nuclear Futures cost-effective strategy for preventing the escape of such radioactive The Chernobyl and Fukushima Daiichi disasters established that materials during an accident. Although proposed new reactor concepts pursue this end with means ranging from minor variations on the primary hazard that nuclear power plant accidents pose to the well-established technologies to highly exotic deviations from past public results from the escape of cesium and iodine radioisotopes practice, all of them sit uncomfortably in the present regulatory (United Nations General Assembly, 2013, p. 8). Although these paradigm. two elements make up only a limited fraction of the radiological The most mundane of the current contenders for the future of inventory of a power reactor, their volatility allows them to escape nuclear power is a small, modular variant on current LWRs. The and disperse in the surrounding environment much more read- comparatively small size of these reactors—ranging from tens to a Some Possible Future Reactor Types few hundred megawatts of generating capacity, in contrast to more

Fuel Format than a gigawatt for a typical LWR—offers a variety of safety advan-

Coolant Solid Liquid tages. It is much more straightforward to design passive emergency cooling systems requiring no electrical power to operate for a small Gas HTGR or pebble-bed modular reactor LWR, and the reactors can be made compact enough to site under-

Fast gas reactor ground, obviating the NRC’s concerns about resistance to aircraft impact. The reactors can also be assembled in a factory rather than Liquid metal Liquid-metal fast breeder reactor on site, leading to higher quality assurances and shorter construc-

Svintsovo-Vismutovyi tion times. Small LWRs also benefit from decades of naval experi- Bystryi Reaktor, or lead- bismuth fast reactor ence and utilize the same fuel and materials as existing plants. Yet this attribute is a mixed blessing because it also means that small, Lead-cooled fast reactor (BREST [Bystry modular LWRs can achieve only a limited improvement in their Reaktor so Svintsovym Teplonositelem, or cost and safety performance over their larger brethren, particularly fast reactor with lead coolant]) because they cannot take advantage of the same economies of scale.

Molten salt Advanced high- Liquid-fluoride MSR Furthermore, despite their relatively conventional designs, even temperature reactor these reactors are hamstrung by the fact that the NRC currently NOTE: Blue = in operation. Tan = undergoing commercialization. licenses nuclear plants on a per-reactor basis, with no discount for Rust = in development. MSR = molten-salt reactor. smaller capacities.

13 The most–technologically mature alternatives to LWRs at considerable associated cost savings. A miniature fast reactor could present are fast reactors utilizing molten metal as coolant. Histori- easily be sited underground, alleviating concerns about terrorist cally, these reactors were the subject of much enthusiasm in both attack and some kinds of natural disasters. The thermal proper- the United States and abroad because of the possibility that they ties of lead bismuth also mean that such reactors could be highly could be constructed to breed more usable nuclear fuel than they resilient against operator error and the loss of external power. consumed, but lower-than-envisioned uranium prices and prolifera- Unfortunately, lead bismuth suffers from economic and operational tion concerns have caused most nations to abandon this pursuit. disadvantages as well. Enormous quantities of bismuth are required Fast reactors present a set of hazards that differs from that of LWRs relative to current uses of that element, and the lead-bismuth eutec- because, without careful core design, they could be subject to the tic is much harder on reactor materials than either sodium or water, kind of positive void and thermal coefficients of reactivity that despite being free of elemental sodium’s violent chemical reactivity caused the Chernobyl disaster. The use of molten metal as coolant with water. Russia has declared its intention to build a prototype is arguably preferable to using water because of the former’s high of a small, modular lead-bismuth fast reactor in the next few years, thermal capacity and high boiling point. This allows the reactor to and an American start-up, Gen4 Energy, is also attempting to com- operate at high temperature and low pressure. Unlike an LWR, in mercialize the technology. which a pipe break can allow the cooling water to flash to steam Although metal-cooled reactors aim to make nuclear power as it depressurizes, potentially exposing the fuel, the coolant in a safer by using a coolant that is difficult to boil away, HTGRs pur- metal-cooled fast reactor can be at atmospheric pressure, and it can sue this end by using fuel assemblies that are extremely difficult to absorb substantial amounts of decay heat during an accident before melt. In theory, ceramic fuel assemblies could retain their integrity emergency cooling is needed to prevent damage to the fuel. without emergency cooling even in case of a severe accident. Like The particular safety characteristics of such a reactor, however, liquid metal–cooled fast reactors, this concept has a long history depend on the exact choice of fuel and coolant. The combination both inside and outside of the United States. As noted above, the used in Russia’s BN-600 and BN-800 reactors—uranium oxide only non-LWR power reactor that ever operated in the United fuel with molten-sodium coolant—is both familiar and relatively States under NRC auspices, the Fort St. Vrain Generating Station, inexpensive, but, although the sodium interacts well with reactor was an HTGR, and Department of Energy research efforts have materials, its violent reaction to water has caused serious opera- disproportionately focused on these reactors. tional problems. Another option, employed by the Soviet Union in In more-recent decades, much attention has been paid to a its Alpha attack submarine, is the use of lead-bismuth eutectic as variety of HTGR known as the pebble-bed modular reactor, in coolant. This offers a low melting point and extraordinary thermal which the ceramic fuel and graphite moderator are combined into capacity, which allows lead-bismuth fast reactors to be extremely spherical pebbles. These were pioneered in Germany between the compact relative to their power output—with the potential for 1960s and 1980s; today, the Chinese are constructing a two-unit

14 prototype at Shidaowan in Shandong province of their own varia- ing could cause them to fail and release dangerous radioisotopes. tion of the pebble-bed concept, dubbed the High Temperature As a result, to withstand a lengthy period without active cooling, Reactor—Pebble Module (HTR-PM). This plant’s designers assert the pebbles need to be able to eject heat to the environment—but that it will be so safe as to “practically exclude the need for [an] off- the sort of steel and concrete containment structures normally used site emergency plan.”18 This safety depends heavily on the expected in LWRs would prevent this. The HTR-PM therefore features a high performance of the reactor’s tristructural-isotropic fuel. “low-pressure vented containment” (which probably does not meet Although the fuel pebbles are intended to withstand temperatures the NRC’s definition of containment as an “essentially leak-tight in excess of 1,600 degrees Celsius during an accident, further heat- barrier against the uncontrolled release of radioactivity to the

Coated Particle Fuel in a Pebble-Fuel Element environment” [NRC, 2014c, criterion 16]). It would therefore be difficult to license such reactors under the NRC’s current defense- Fuel sphere in-depth paradigm, even if a risk-based assessment found that the Diameter: 60 mm HTGR posed smaller hazards to the public than present-day LWR designs.19

5-mm graphite layer A still more exotic type of reactor aims to overcome the challenges of meltdowns by starting with molten fuel. Although, at first Coated particles embedded in graphite matrix glance, such an approach appears highly counterintuitive, reactors employing molten salt as fuel could provide substantial safety Pyrolytic carbon 40 microns (µ) Silicon-carbide barrier coating 35 µ advantages over solid-fuel designs. Like liquid metal–cooled fast Section Inner pyrolytic carbon 40 µ Porous carbon buffer 95 µ reactors, MSRs can operate at high temperature and low pressure and can be extremely compact relative to water- or gas-cooled TRISO-coated reactors with the same power output. Employing the fuel itself as particle a heat-transfer medium, as well as the ability to use simple passive Diameter: 0.92 mm Fuel kernel Diameter: 0.55 mm measures, such as freeze plugs, to automatically drain the fuel into Uranium dioxide a passively cooled noncritical configuration in case of an emer- NOTE: TRISO = tristructural isotropic. SOURCE: Moses, 2010. RAND PE156-2 19 In a 2011 report, Idaho National Laboratory noted the gap between the LWR definition of containment and the sort of functional containment featured in an HTGR (Idaho National Laboratory, 2011, p. 7). In this case, the regulatory gap 18 This is characterized as reducing the “[p]ossibility with accident consequence results in part from the consideration that LWR containments are a legacy of the at site boundary exceeding 50mSv [millisieverts] less than 10–6 per reactor year” AEC’s deterministic approach to regulation and establishing that the confluence (Sun, 2013, p. 14). of features used in an HTGR to achieve the same goal is far less straightforward.

15 Components of the Reactor and Steam Generator in a cessing capabilities that might pose serious safety and proliferation Single Vessel concerns.20 Furthermore, their safety and operational characteris-

Control-rod tics are likely to be linked strongly to their exact fuel chemistries drive lines Pressurizer region and core designs, which could take many forms. MSRs are so alien to the current NRC regulatory paradigms Steam outlet Reactor coolant that it is difficult to see a path to licensing them in the United pump (1 of 8) States, particularly given the Department of Energy’s historical

Helical-coil steam struggles to construct a prototype of the vastly more-conventional generator (1 of 8) HTGR. The chief executive officer of a start-up developing an MSR complained in congressional testimony in December 2014 Barrel Feedwater inlet that, “right now, there is no viable pathway for bringing advanced nuclear reactor designs beyond laboratory-scale development” and that, as a result, Core region Shield plates the current system incentivizes reactor designers to develop their first products outside of the US. In fact, this has already happened: some existing nuclear reac- SOURCE: “Nuclear Energy,” 2002. tor design companies are planning on building their first RAND PE156-3 power plants overseas . . . because they do not think it will be possible to build an advanced reactor in the US gency, could substantially or totally eliminate the need for the kind under the current regulatory system. (Dewan, 2014, of complicated safety systems that make LWRs so expensive. In p. 2) contrast to the reactors described above, however, practical experience with MSRs is too limited at present to ascertain how easily NRC chair Stephen G. Burns essentially admitted the inability they could realize these promises. Oak Ridge National Labora- of his agency to formulate a regulatory framework for advanced tory operated two experimental MSRs during the 1960s and early reactors at the NRC’s 27th annual Regulatory Information Confer- i 1970s, and, although the results showed considerable potential, ence in March 2015. Noting the interest of f rms, such as Trans- they could not clarify many of the practical issues that must be 20 Although MSR technology has never matured to the point of being a serious understood before MSRs can be commercialized (Haubenreich and proliferation risk, these reactors could produce weapon-usable fissile materials, such as U-233. Theoretical investigations of proliferation-resistant MSR variants Engel, 1970). A mind-boggling diversity of MSR designs is theo- date back to the 1970s and continue today (Ahmad, McClamrock, and Glaser, retically possible, with some of them incorporating online repro- 2015).

16 atomic and Gen4, in receiving NRC feedback on exotic reactor designs, Burns insisted that the commission was “willing and able” [N]ext-generation reactors face a potentially to work on this problem but that, “without a specific applicant, insurmountable chicken-and-egg problem. and with intense pressure on resources and budget, it is challeng- ing for the NRC to be too forward-leaning and expend significant try to develop new nuclear reactor technologies and demonstrate resources on the development of a new regulatory framework” their commercial viability, but this arrangement has never worked (“No Mid-Life Crisis for NRC,” 2015). Therefore, next-generation as intended. Planned prototype plants, such as the Clinch River reactors face a potentially insurmountable chicken-and-egg prob- Breeder Reactor and the MHTGR, consumed vast sums without lem: Investors will fund the development of the technology only ever coming to fruition. Scarred by these experiences, industry is with a viable path to commercialization, but the NRC will develop skeptical of cost-sharing arrangements with the Department of the requisite regulatory framework only in response to a serious Energy to build pilot facilities, but, without these, building the applicant. prototype reactors needed to understand exotic nuclear technologies well enough to draft commercial designs is politically infea- Nuclear Safety from the Outside In sible. Meanwhile, the NRC is unwilling to issue a clear judgment If fission is to play a larger role in America’s energy future, funda- on the safety requirements for advanced nuclear reactors without a mental reforms are needed in the U.S. nuclear regulatory regime. fully specified design requiring years of effort and a massive invest- The current regulatory paradigm is questionable even for the large ment to produce. This expectation is tantamount to a demand that LWRs in use today, as the NRC’s decades-long attempt to transi- companies invest tens of millions of dollars developing advanced tion to a consistent risk-informed approach attests, and it is not at all adapted to significantly different reactor designs. Although such plant designs based on no more than an educated guess about what reforms pose a massive challenge, they also present an opportunity the ultimate regulatory requirements will be. Given this mutually to escape some of the dubious assumptions and preoccupations that reinforcing tangle of institutional dysfunction, it is little wonder have bedeviled the U.S. nuclear industry since the AEC era. that advanced nuclear reactor technologies are little closer to com- Although the NRC has failed to articulate an appropriate mercialization in the United States today than they were a quarter- regulatory framework for advanced reactors, the Department of century ago. Energy and industry share the blame for the United States’ difficul- Since the 1950s, the United States has conceptualized the ties developing these technologies. Sensibly, the NRC feels that, problem of nuclear power plant safety “from the inside out.” Safety without substantial hands-on experience with new types of reac- regulations emphasized preventing or limiting damage to reactors tors, it cannot determine how to regulate them appropriately. In themselves on the assumption that this approach would protect the theory, the Department of Energy is supposed to work with indus- public most effectively. This monomania for preventing, rather than

17 managing, accidents led to embarrassing failures, such as the Three to create an accident with characteristics more favorable to effec- Mile Island accident, and the same attitude catalyzed the catastro- tive emergency management. Instead of perpetuating the hubristic, phes at Chernobyl and Fukushima Daiichi. In each of these cases, and infeasible, ambition to somehow prevent all accidents, the aim the belief that serious accidents would never occur encouraged a should be to prevent accidents from becoming catastrophes. neglect of emergency planning that crippled authorities’ ability to To encourage the development of new civilian nuclear tech- take effective action to protect the public (Geist, 2014). nologies, the United States should forge relationships with other The objective of nuclear safety regulation should be to mini- nations to develop the operational experience and technical data mize the health and economic impacts of large-scale accidents, necessary to commercialize non-LWR nuclear plants. Much of rather than to minimize the theoretical incidence of damage these data can be acquired only through the construction and oper- to reactors.21 Rather than focus on saving the reactor, the goal ation of pilot facilities, but, even with substantially increased fund- should be to protect the public from the outside in. New nuclear ing and political support, it seems unlikely that the Department technologies, such as liquid-fuel reactors, could enable formerly of Energy will construct more than one type of prototype reactor. inconceivable means of achieving this end.22 Reactors designed to To face the possible energy challenges of the 21st century, the sacrifice themselves in an accident in order to minimize the chance department has a responsibility to explore all potentially promising of radiological releases might prove not only feasible but vastly energy technologies, and this is feasible only by partnering with more cost-effective than present-day LWR designs. Plants might be nuclear research programs in states that are aggressively developing designed so that, in the worst-case scenario, they fail elegantly so as advanced reactors. Although a nuclear energy policy emphasizing international cooperation is contrary to the past goal of maintain- 21 For instance, NRC review of the NGNP still set goals in terms of core damage ing U.S. industry leadership, if present conditions continue, the frequency, rather than assessments of offsite consequences (NRC, undated). United States risks having little influence over the future of nuclear 22 A liquid-fuel format offers the intriguing possibility of radically altering the fuel chemistry during an accident to reduce the volatility of radioisotopes, such power at all. Working with international partners, instead of com- as I-131 and Cs-137. For instance, an MSR could incorporate a passive safety peting with them, will help ensure that U.S. values regarding safety arrangement utilizing blocks of salt designed to melt if the reactor exceeds a certain temperature to effect such a change in the fuel. Even though the resulting and proliferation resistance will be reflected in future nuclear plants fuel mixture might cause irreparable damage to the reactor, the simplicity of the wherever they are built and potentially produce the knowledge base arrangement might make the reactor so cheap as to be essentially disposable— needed to commercialize advanced nuclear reactors in the United achieving a far better combination of economics and safety than that of current LWRs. States.

18 References GAO—See U.S. Government Accountability Office. Ahmad, Ali, Edward B. McClamrock, and Alexander Glaser, “Neutronics Geist, Edward Moore, “What Three Mile Island, Chernobyl, and Fukushima Can Calculations for Denatured Molten Salt Reactors: Assessing Resource Teach About the Next One,” Bulletin of the Atomic Scientists, April 28, 2014. As of Requirements and Proliferation-Risk Attributes,” Annals of Nuclear Energy, June 22, 2015: Vol. 75, January 2015, pp. 261–267. http://thebulletin.org/ what-three-mile-island-chernobyl-and-fukushima-can-teach-about-next-one7104 “Areva Modular Reactor Selected for NGNP Development,” World Nuclear News, February 15, 2012. As of June 23, 2015: Guibert, J. C., Potential Impact of NRC Safety Goals on the Nuclear Industry, Palo http://www.world-nuclear-news.org/ Alto, Calif.: Nuclear Safety Analysis Center, 1983. NN-Areva_modular_reactor_selected_for_NGNP_development-1502124.html Haubenreich, Paul N., and J. R. Engel, “Experience with the Molten-Salt Reactor Atlantic Council of the United States, Nuclear Fuels Policy Working Group, Experiment,” Nuclear Applications and Technology, Vol. 8, No. 2, February 1970, Nuclear Fuels Policy: Report of the Atlantic Council’s Nuclear Fuels Policy Working pp. 118–136. Group, Boulder, Colo.: Westview Press, 1976. Haughney, C. J., P. A. Rathbun, L. C. Suttora, H. E. Thompson, K. L. Winsberg, Boudreau, Jay, “The American Breeder Reactor Program Gets a Second Chance,” F. M. Costello, and R. G. Summers, External Regulation of Department of Energy Los Alamos Science, Vol. 2, No. 2, Summer–Fall 1981, pp. 118–119. Nuclear Facilities: A Pilot Program, Washington, D.C.: U.S. Nuclear Regulatory Commission, NUREG-1708, August 1999. As of June 23, 2015: Bowsher, Charles A., Analysis of the Department of Energy’s Clinch River Breeder http://pbadupws.nrc.gov/docs/ML0731/ML073190349.pdf Reactor Cost Estimate: Report by the Comptroller General of the United States, Washington, D.C.: U.S. General Accounting Office, GAO/RCED-83-74, IAEA—See International Atomic Energy Agency. December 10, 1982. As of June 23, 2015: Idaho National Laboratory, NGNP Project Regulatory Gap Analysis for Modular http://gao.gov/products/RCED-83-74 HTGRs, Idaho Falls, Idaho: U.S. Department of Energy, Office of Nuclear Chang, Yoon I., “The Integral Fast Reactor,” Nuclear Technology, Vol. 88, No. 2, Energy, INL/EXT-11-23216, September 2011. As of June 23, 2015: November 1989, pp. 129–138. https://inis.iaea.org/search/search.aspx?orig_q=RN:43009640 Copinger, D. A., and D. L. Moses, Fort Saint Vrain Gas Cooled Reactor International Atomic Energy Agency, Current Status and Future Development of Operational Experience, Washington, D.C.: Division of Systems Analysis and Modular High Temperature Gas Cooled Reactor Technology, IAEA-TECDOC-1198, Regulatory Effectiveness, Office of Nuclear Regulatory Research, U.S. Nuclear February 2001. As of June 23, 2015: Regulatory Commission, NUREG/CR-6839, ORNL/TM-2003/223, January http://www-pub.iaea.org/books/IAEABooks/6124/Current-Status-and-Future- 2004. As of June 23, 2015: Development-of-Modular-High-Temperature-Gas-Cooled-Reactor-Technology http://www.nrc.gov/reading-rm/doc-collections/nuregs/contract/cr6839/ Keller, William, and Mohammad Modarres, “A Historical Overview of Cowan, Robin, “Nuclear Power Reactors: A Study in Technological Lock-In,” Probabilistic Risk Assessment Development and Its Use in the Nuclear Power Journal of Economic History, Vol. 50, No. 3, September 1990, pp. 541–567. Industry: A Tribute to the Late Professor Normal Carl Rasmussen,” Reliability Engineering and System Safety, Vol. 89, No. 3, September 2005, pp. 271–285. Dewan, Leslie, cofounder and chief executive officer, Transatomic Power Corporation, “The Future of Nuclear Energy,” testimony before the U.S. House Lartonoix, David K., “Potential LEU Encumbrances and Alternatives for the U.S. of Representatives Committee on Science, Space, and Technology, Energy Production of Tritium for National Security Purposes,” in Sarah Weiner, ed., Subcommittee, December 11, 2014. As of June 23, 2015: Nuclear Scholars Initiative: A Collection of Papers from the 2013 Nuclear Scholars http://science.house.gov/sites/republicans.science.house.gov/files/documents/ Initiative, Washington, D.C.: Center for Strategic and International Studies, hearings/2014%2012%2011%20-%20Dewan%20Testimony%20-%20Nuclear. January 2014, pp. 143–161. As of June 23, 2015: pdf http://csis.org/files/publication/ 140109_Weiner_NuclearScholarsInitiative2013_WEB.pdf

19 Moses, David L., Very High-Temperature Reactor (VHTR) Proliferation Resistance Sun, Yuliang, “HTR Development Status in China,” International Atomic Energy and Physical Protection (PR&PP), Oak Ridge, Tenn.: Oak Ridge National Agency Technical Working Group on Gas Cooled Reactors meeting, Vienna, Laboratory, ORNL/TM-2010/163, August 2010. As of July 8, 2015: March 5–7, 2013. As of June 23, 2015: http://info.ornl.gov/sites/publications/files/Pub25470.pdf https://www.iaea.org/NuclearPower/Downloadable/Meetings/2013/ 2013-03-05-03-07-TWG-NPTD/Day_1/3.Sun.pdf “No Mid-Life Crisis for NRC,” World Nuclear News, March 11, 2015. As of June 23, 2015: Thompson, Theos Jardin, and C. R. McCullough, “The Concepts of Reactor http://www.world-nuclear-news.org/RS-No-mid-life-crisis-for-NRC-1103157.html Containment,” in Theos Jardin Thompson and James G. Beckerley, eds., The Technology of Nuclear Reactor Safety, Vol. 2: Reactor Materials and Engineering, NRC—See U.S. Nuclear Regulatory Commission. Cambridge, Mass.: MIT Press, 1973, pp. 785–789. “Nuclear Energy: Assuring Future Energy Supplies,” Oak Ridge National UK Health and Safety Executive, Her Majesty’s Nuclear Installations Laboratory Review, Vol. 35, No. 2, 2002. As of July 14, 2015: Inspectorate, “New Reactor Build: Westinghouse AP 1000 Step 2 PSA http://web.ornl.gov/info/ornlreview/v35_2_02/assuring.shtml Assessment,” Merseyside, undated. As of June 22, 2015: Nuclear Energy Institute, “Senators Question Macfarlane on ‘Overregulation,’” http://www.onr.org.uk/new-reactors/reports/ap1000psa.pdf May 23, 2013. As of June 23, 2015: Union of Concerned Scientists, The Risks of Nuclear Power Reactors: A Review of http://www.nei.org/News-Media/News/News-Archives/ the NRC Reactor Safety Study, WASH-1400 (NUREG-75/014), Cambridge, Mass., Senators-Question-Macfarlane-on-%E2%80%98Overregulation%E2%80%99 1977. Overton, Thomas, “Even More Delays and Cost Overruns for Vogtle Expansion,” United Nations General Assembly, Report of the United Nations Scientific Power, February 2, 2015. As of June 22, 2015: Committee on the Effects of Atomic Radiation: Sixtieth Session (27–31 May 2013), http://www.powermag.com/ New York, A/68/46, 2013. As of June 23, 2015: even-more-delays-and-cost-overruns-for-vogtle-expansion/ http://www.unscear.org/docs/GAreports/A-68-46_e_V1385727.pdf Public Law 109-58, Energy Policy Act of 2005, August 8, 2005. As of June 23, U.S. Department of Energy Office of Legacy Management, Long-Term 2015: Surveillance Plan for the Decommissioned Hallam Nuclear Power Facility, Hallam, http://www.gpo.gov/fdsys/pkg/STATUTE-119/pdf/STATUTE-119-Pg594.pdf Nebraska, LMS/HAL/S03478-1.0, revised June 25, 2008. Risk Management Task Force, A Proposed Risk Management Regulatory U.S. Department of Energy Office of Nuclear Energy, Next Generation Nuclear Framework: A Report to NRC Chairman Gregory B. Jaczko from the Risk Plant: A Report to Congress, April 2010. As of June 23, 2015: Management Task Force, NUREG-2150, April 2012. As of June 22, 2015: http://www.energy.gov/sites/prod/files/4.4_NGNP_ReporttoCongress_2010.pdf http://pbadupws.nrc.gov/docs/ML1210/ML12109A277.pdf U.S. Energy Information Administration, Annual Energy Outlook 2004 with Rosenthal, Murray W., An Account of Oak Ridge National Laboratory’s Thirteen Projections to 2025, Washington, D.C., DOE/EIA-0383(2004), January 2004. As Nuclear Reactors, Oak Ridge, Tenn.: Oak Ridge National Laboratory, ORNL/ of June 22, 2015: TM-2009/181, revised March 2010. As of June 23, 2015: http://www.eia.gov/forecasts/archive/aeo04/ http://info.ornl.gov/sites/publications/files/Pub20808.pdf U.S. Government Accountability Office, Nuclear Regulatory Commission: Natural Sheron, Brian W., director, Office of Nuclear Regulatory Research, “Update on Hazard Assessments Could Be More Risk-Informed, Washington, D.C., GAO-12- Staff Plans to Apply the Full-Scope Site Level 3 PRA Project Results to the NRC’s 465, April 2012. As of June 23, 2015: Regulatory Framework,” memorandum for the commissioners, SECY-12-0123, http://www.gao.gov/products/GAO-12-465 September 13, 2012. As of June 23, 2015: http://www.nrc.gov/reading-rm/doc-collections/commission/secys/2012/ 2012-0123scy.pdf

20 ———, Advanced Reactor Research: DOE Supports Multiple Technologies, ———, “§ 50.150 Aircraft Impact Assessment,” last updated July 10, 2014b. As but Actions Needed to Ensure a Prototype Is Built—Report to the Chairman, of June 23, 2015: Subcommittee on Energy and Water Development, Committee on Appropriations, http://www.nrc.gov/reading-rm/doc-collections/cfr/part050/part050-0150.html U.S. Senate, Washington, D.C., GAO-14-545, June 2014. As of June 23, 2015: ———, “Appendix A to Part 50: General Design Criteria for Nuclear Power http://www.gao.gov/products/GAO-14-545 Plants,” last updated December 1, 2014c. As of June 24, 2015: “US NRC Could Shrink by 10% in Five Years, Says Report,” World Nuclear News, http://www.nrc.gov/reading-rm/doc-collections/cfr/part050/part050-appa.html February 19, 2015. As of June 23, 2015: ———, “Pressurized Water Reactors,” last updated January 15, 2015. As of July 8, http://www.world-nuclear-news.org/ 2015: C-US-NRC-could-shrink-by-10-percent-in-five-years-19021501.html http://www.nrc.gov/reactors/pwrs.html U.S. Nuclear Regulatory Commission, Office of New Reactors, Summary von Hippel, Frank N., “It Could Happen Here,” New York Times, March 23, Feedback on Four Key Licensing Issues: Next Generation Nuclear Plant Project 0748, 2011. As of June 23, 2015: undated. As of June 23, 2015: http://www.nytimes.com/2011/03/24/opinion/24Von-Hippel.html?_r=0 http://pbadupws.nrc.gov/docs/ML1417/ML14174A774.pdf Walker, J. Samuel, Three Mile Island: A Nuclear Crisis in Historical Perspective, ———, Severe Accident Risks: An Assessment for Five U.S. Nuclear Power Plants, Berkeley, Calif.: University of California Press, 2004. Washington, D.C., NUREG-1150, December 1990. As of June 23, 2015: http://www.nrc.gov/reading-rm/doc-collections/nuregs/staff/sr1150/ Waltar, Alan E., and Albert B. Reynolds, Fast Breeder Reactors, New York: Pergamon Press, 1981. ———, “History of the NRC’s Risk-Informed Regulatory Programs,” last updated July 18, 2013. As of June 22, 2015: Westfall, Catherine, “Vision and Reality: The EBR-II Story,” Nuclear News, http://www.nrc.gov/about-nrc/regulatory/risk-informed/history.html Vol. 47, No. 2, February 2004, pp. 25–32. As of June 23, 2015: http://www.ne.anl.gov/About/reactors/EBR2-NN-2004-2-2.pdf ———, “Backfitting,” last updated July 10, 2014a. As of June 23, 2015: http://www.nrc.gov/reading-rm/doc-collections/cfr/part050/part050-0109.html World Nuclear Association, “US Nuclear Power Policy,” updated March 2015. As of June 23, 2015: http://www.world-nuclear.org/info/Country-Profiles/Countries-T-Z/ USA--Nuclear-Power-Policy/

21 Abbreviations AEC Atomic Energy Commission

ERDA Energy Research and Development Administration

GAO U.S. Government Accountability Office

HTGR high-temperature gas-cooled reactor

HTR-PM High Temperature Reactor—Pebble Module

IAEA International Atomic Energy Agency

LWR light-water reactor

μ micron

MHTGR Modular High-Temperature Gas-Cooled Reactor

MSR molten-salt reactor

NGNP Next Generation Nuclear Plant

NRC U.S. Nuclear Regulatory Commission

PRA probabilistic risk assessment

TVA Tennessee Valley Authority

22 About the Author

Edward Geist has a Ph.D. and M.A. in history from the University of North Carolina at Chapel Hill and a B.A. in history from the College of William and Mary. He is a MacArthur Foundation Nuclear Security Fel- low with the Center for International Security and Cooperation at Stanford University.

23 About This Perspective Eisenhower appointed Stanton to a committee convened to develop the first comprehensive plan for the survival of the United States following a nuclear This perspective examines the institutional and technical obstacles to the attack. Stanton led the effort to develop plans for national and international commercialization of advanced nuclear reactors for electrical power gen- communication in the aftermath of a nuclear incident. Stanton also served eration in the United States. The so-called nuclear renaissance that seemed as chair (1961–1967) and member (1957–1978) of the RAND Corpora- imminent ten years ago has failed to materialize, in considerable part tion Board of Trustees. The Stanton Foundation aims, through its support of because of the failure of large light-water reactors (LWRs) to achieve the the Stanton Nuclear Security Fellows Program, to perpetuate his efforts to envisioned improvement in capital costs. If nuclear fission is to play a sub- meet these challenges. For more information about the Stanton Fellowship stantial role in the future of the U.S. energy supply, a more cost-effective at RAND, visit www.rand.org/about/edu_op/fellowships/stanton-nuclear. type of nuclear power plant must be commercialized. This piece examines the underlying technical reasons LWRs require expensive engineered safety systems to protect the public. It then explores the institutional barriers that make it difficult for the U.S. Nuclear Regulatory Commission (NRC) to evaluate non-LWR nuclear plants and discourage the U.S. Department of Energy and industry from investing the time and resources needed to establish the operational and safety characteristics of these technologies. Finally, it provides an overview of several candidate reactor designs that might offer alternatives to the current technological paradigm and outlines steps policymakers can take to overcome the barriers to the commercialization of next-generation nuclear reactors.

Thoughtful feedback from former NRC commissioner Victor Gilinsky Limited Print and Electronic Distribution Rights and Tom LaTourrette of RAND was instrumental in the development of this This document and trademark(s) contained herein are protected by law. This representa- perspective. I would also like to thank Brian A. Jackson, Andrew R. Hoehn, tion of RAND intellectual property is provided for noncommercial use only. Unauthor- and Sarah Harting of RAND, and the Stanton Foundation for helping make ized posting of this publication online is prohibited. Permission is given to duplicate this document for personal use only, as long as it is unaltered and complete. Permission is this piece possible. required from RAND to reproduce, or reuse in another form, any of our research docu- The research reported here was conducted as part of the Stanton ments for commercial use. For information on reprint and linking permissions, please visit Nuclear Security Fellows Program at the RAND Corporation. The Stanton www.rand.org/pubs/permissions.html. Nuclear Security Fellows Program was created to stimulate the develop- The RAND Corporation is a research organization that develops solutions to public policy ment of the next generation of leaders on nuclear security by supporting challenges to help make communities throughout the world safer and more secure, health- ier and more prosperous. RAND is nonprofit, nonpartisan, and committed to the public interdisciplinary research that will advance policy-relevant understanding interest. of the issues. Each fellow carries out a yearlong period of independent RAND’s publications do not necessarily reflect the opinions of its research clients and spon- research, collectively producing studies that contribute to the general body sors. R® is a registered trademark. of knowledge on nuclear security. For more information on this publication, visit www.rand.org/t/PE156. The Stanton Foundation, a creation of Frank Stanton, former president of CBS and a pioneering executive who led the television network for 25 years, supports Stanton fellows. In 1954, President Dwight D.

PE-156-TSF (2015)