Nuclear Reactors: Generation to Generation

Next Steps

Back to table of contents
Stephen M. Goldberg and Robert Rosner
Global Nuclear Future

After enduring the usual reliability growing pains, Gen I and Gen II nuclear reactors have proven to be economically successful. According to the Nuclear Energy Institute, U.S. nuclear power plants in 2006 supplied the second-highest amount of electricity in the industry’s history while achieving a record-low average production cost of 1.66 cents/kWh. Because the capital costs of many Gen I and Gen II reactors have been paid off, average production costs have been below 2 cents/kWh for the past seven years. Capacity factors have remained higher than 90 percent. Self-financing (essentially paid off the balance sheet) is a key factor, leading to not having to pay any capital charges and resulting in very low costs to operate these plants. Power upgrades and improvements in operational efficiency over the past decade have yielded the equivalent of multiple new nuclear plants. Whether this performance platform can be extrapolated to the Gen III and III+ designs is uncertain19 because of the significant overnight cost20 investment for the GEN III/III+ plants.

The Move to Small Modular Reactors

As President Barack Obama pushes to revive the domestic nuclear power industry amid mounting concerns about fossil-fired electricity generation, a new type of small reactor is about to enter the market. Several firms are working on Gen III and Gen III+ designs that are smaller in scale than the current designs and in several cases also make use of modular construction techniques. This small modular reactor (SMR) architecture is based on significant learning-by-doing efficiencies. The vendors are planning to apply for NRC design certification pursuant to either 10 CFR Part 50 or 10 CFR Part 52. It is our understanding that because of new policy issues (e.g., revised emergency planning zone and accident scenarios) updated or new regulatory criteria and guides may be necessary. One example of SMR architecture is the mPower 125 MW module (Figure 2), an integral, advanced LWR of modular design, in current development through an alliance of Babcock and Wilcox Nuclear Energy Inc. (B&W NE) and Bechtel Power Corporation. The reactor is significantly smaller than most operating PWRs but is scalable and incorporates already existing LWR technology, a fully passive safety design, industry-standard PWR fuel, 60-year used fuel storage, and a four-to-five-year refueling cycle. The small B&W NE reactor is 75 feet tall and 15 feet wide. Unlike steam generators in traditional nuclear facilities, its steam generator (the cylindrical structure seen along the center axis of the reactor vessel in Figure 2) is integrated within the reactor vessel.

Figure 2. The Babcock and Wilcox mPower Reactor

Figure 2

Copyright and reprinted with permission from Babcock & Wilcox Nuclear Energy, Inc.

Other notable examples include a 45 MW integral, scalable, modular LWR currently in development by NuScale Power; and the International Reactor Innovative and Secure (IRIS), a 300 MW scalable, modular LWR reactor developed by an international consortium led by Westinghouse.21 The design for NuScale Power’s LWR calls for advanced passive safety features and for the entire nuclear steam supply system to be prefabricated and sited below ground.22

These new reactors—smaller than a rail car and one-tenth the cost of a big plant—could be built quickly and installed at the dozens of existing nuclear sites, or they could replace existing coal-fired plants that do not meet current federal air quality emission standards. “We see significant benefits from the new, modular technology,” said Donald Moul, vice president of nuclear support for First Energy, an Ohio-based utility corporation.23

Smaller reactors (defined by the International Atomic Energy Agency as those capable of generating less than 300 MW) are the logical choice for smaller countries or countries with a limited electrical grid. Small reactors are now in different stages of development throughout the world. They are attractive because of their simplicity and enhanced safety and because of the relatively limited financial resources required to build them. But acclaim is not universal. Detractors argue that SMR designs are not economical because of economies of scale. Capital construction costs (price per one thousand watts of electric capacity, or $/kWe) of a nuclear reactor decrease with size, but the economy of scale applies only if reactors are of a very similar design, as has historically been the case. The design characteristics of SMRs, however, are significantly different from those of large reactors. SMRs approach the economies of scale problem by achieving significant cost savings elsewhere. For example, SMR designs seek to streamline safety and safeguard requirements by replacing (at least some) security guards with concrete security barriers and/ or by building underground, streamlining the requirements for operators, and streamlining emergency planning zone requirements. Awareness and realization of the economic potential of small, modular reactors have grown significantly in the last few years. Argonne National Laboratory (ANL) is conducting a significant, detailed economic analysis that will address the economic competitiveness of these reactors.

What these designs might not do, however, because of deep-seated American Society of Mechanical Engineers In-service Inspection (ISI) and Nondestructive Engineering Division (NDE) requirements, is to stretch out the refueling schedule over decades. The new designs do stretch out refueling schedules, from 18 months to possibly 3–5 years and potentially to as long as 10 years (subject to ISI testing and monitoring). But longer-term refueling cycles (such as are commonly associated with so-called battery reactors) are currently left to the future. The following is a list of SMRs being researched and/or developed in the United States:

  • Water-cooled reactors with small coated particle fuel without on-site refueling: AFPR (PNNL)
  • Sodium-cooled small reactor with extended fuel cycles: 4S (Westinghouse/ Toshiba); PRISM (GE); ARC (ARC)
  • Lead- or lead-bismuth-cooled small reactors with extended fuel cycles: HPM (Hyperion); LFR/SSTAR and its variations such as STAR- LM, STAR-H2, and SSTAR (ANL, LLNL and LANL); ENHS (UC Berkeley)
  • Gas-cooled thermal neutron spectrum reactor: MHR (GA); PBMR (Westinghouse); ANTARES (AREVA-U.S.)
  • Gas-cooled fast neutron spectrum reactor with extended fuel cycle: EM2 (GA)
  • Salt-cooled small reactor with pebble-bed fuel: PB-AHTR (UC Berkeley); SmAHTR (ORNL).

Looking Past Gen III and Gen III+

Nuclear scientists have left implementation of the Gen III+ and SMR designs to the engineers, believing them to be within the current state-of-the-art, and have instead focused on nuclear alternatives—commonly called Gen IV—that still require considerable fundamental research.

Conceptually, Gen IV reactors have all of the features of Gen III+ units, as well as the ability, when operating at high temperature, to support economical hydrogen production, thermal energy off-taking, and perhaps even water desalination. In addition, these designs include advanced actinide management.24

Gen IV reactors include:

  • High temperature water-, gas-, and liquid salt–based pebble bed thermal and epithermal reactors.
  • Liquid metal–cooled reactors and other reactors with more-advanced cooling. One such design is the Power Reactor Innovative Small Module (PRISM), a compact modular pool-type reactor developed by GE-Hitachi with passive cooling for decay heat removal.
  • Traveling wave reactors that convert fertile material into fissile fuel as they operate, using the process of nuclear transmutation being developed by TerraPower. This type of reactor is also based on a liquid metal primary cooling system. It is also being designed with passive safety features for decay heat removal, and has as a major design goal minimization of life cycle fuel costs by both substantially increasing the burnup percentage and internally breeding depleted uranium.
  • Hyperion Power Module (25 MW module). According to Hyperion, uranium nitride fuel would be beneficial to the physical characteristics and neutronics of the standard ceramic uranium oxide fuel in LWRs.25

Gen IV reactors are two-to-four decades away, although some designs could be available within a decade. As in the case of Gen III and Gen III+ designs in the United States, Gen IV designs must be certified by the NRC pursuant to 10 CFR Part 52, based on updated regulations and regulatory guides.

The U.S. Department of Energy (DOE) Office of Nuclear Energy has taken responsibility for developing the science required for five Gen IV technologies (Table 3 summarizes their characteristics and operating parameters and also provides information on two versions of the molten salt reactor, which the United States is not currently researching). Funding levels for each of the technology concepts reflects the DOE’s assessment of the concept’s technological development stage and its potential to meet national energy goals. The Next Generation Nuclear Plant (NGNP) project is developing one example of a Gen IV reactor system, the Very High Temperature Reactor, which is configured to provide high-temperature heat (up to 950°C) for a variety of co-products, including hydrogen production. The NRC is working with DOE on a licensing approach. The earliest potential date for a COL application is the middle of this decade.

Table 3. Characteristics and Operating Parameters of the Eight Generation IV Reactor Systems under Development

Neutron Spectrum
Coolant Temperature
Pressure* Fuel Fuel Cycle Size(s) (MWe) Uses
Gas-cooled fast reactors Fast Helium 850 High U-238† Closed, on site 1,200 Electricity & Hydrogen‡
Lead-cooled fast reactors Fast Lead or lead- bismuth 480–800 Low U-238† Closed, regional 20–180** 300–1,200 600–1,000 Electricity &Hydrogen‡
Molten salt fast reactors Fast Fluoride salts 700–800 Low UF in salt Closed 1,000 Electricity & Hydrogen‡
Molten salt reactor— Advanced high temperature reactors Thermal Fluoride salts 750–1,000 UO2 particlesin prism Open 1,000– 1,500 30-150 Hydrogen‡
Sodium- cooled fast reactors Fast Sodium 550 Low U-238 & MOX Closed 300–1,500 1,000– 2,000 300–700 Electricity
Traveling wave reactors Fast Sodium ~510 Low U-238 metal with U-235 igniter seed Open 400–1,500 Electricity
Supercritical water-cooled reactors Thermal or fast Water 510–625 Very high UO2 Open (thermal)closed (fast) 1,000– 1,500 Electricity
Very high temperature gas reactors   Thermal Helium 900–1,000 High UO2 prism or pebbles Open 250–300 Electricity & Hydrogen‡

* High = 7–15 megapascals
† = With some U-235 or Pu-239
** ‘Battery’ model with long cassette core life (15–20 years) or replaceable reactor module
‡ Such plants can efficiently produce hydrogen because of their high operating temperature characteristic, a characteristic that is also useful for providing process heat to, for example, refineries that would also utilize the hydrogen as a feedstock to upgrade the energy characteristics of the processed fossil fuels.
Used with permission from the World Nuclear Association, “Generation IV Reactors” (World Nuclear Association, June 2010),

In general, Gen IV systems include full actinide recycling and on-site fuel-cycle facilities based on advanced aqueous, pyrometallurgical, or other dry-processing options.26 Fast reactor research has been active in the United States and more active in China, France, India, and the countries of the former Soviet Union.27

One rationale for closing the fuel cycle with fast reactors is the potentially limited supply of uranium. However, given the current significant economic supplies of uranium, from both primary and secondary stores, the objective of breeding civil plutonium cannot be currently based on commercial needs. This leaves the depletion of the fertile and fissile content of used fuel, including plutonium, the remaining fissile uranium, and the minor actinides, as the remaining key rationale for closing the fuel cycle. Of course, in the final analysis, the ultimate metric for success will be the economics of what is being proposed, all within the constraints imposed by meeting all of the safety, security, and nonproliferation concerns—concerns that are likely to be reinforced by the recent events at Fukushima, Japan.


19. The ratings agencies have continually downgraded new nuclear investments, until there is an operating history for these new plants.

20. The overnight cost of a large capital project is the cost of the project (i.e., all capital costs, including owner’s costs) without financing costs, as if the project could be completed overnight.

21. Westinghouse plans to announce details of a new design evolved from the IRIS design.

22. At the time of release of this paper, NuScale Power abruptly halted its operations after the U.S. Securities and Exchange Commission began a civil action against the Michael Kenwood Group, the main investor in the company.

23. “Small Reactors Generate Big Hopes,”

24. An actinide is an element with an atomic number from 89 (actinium) to 103 (lawrencium). The term is usually applied to elements heavier than uranium. Actinides typically have relatively long half-lives. Plutonium and the minor actinides (plutonium and uranium are the “major” actinides) will be largely responsible for the bulk of the radiotoxicity and the heat load within a repository for the used fuel wastes from LWRs.

25. Uranium nitride has beneficial traits such as higher thermal conductivity, which results in less retained heat energy. These characteristics make it preferable to oxide fuels when used at temperature regimes greater than the 250–300°C temperatures that characterize LWRs. By operating at higher temperatures, steam plants can operate at a higher thermal efficiency.

26. The exceptions are Gen IV designs that focus on intrinsically higher burnup and possible breeding of fertile fuel (such as depleted uranium), and thus potentially no refueling of the reactor core—with the consequence that no recycling or reprocessing would be contemplated. The TerraPower Traveling Wave Reactor (TWR) concept is an example of such a Gen IV design.

27. Small fast reactor facilities (BR-1, BR-2, and BR-5) were constructed in the 1950s and 1960s in the former Soviet Union, and operational results were used to refine the design and construction of larger plants. The large power plant facilities currently developed in this program are the BOR-60, BN-350, BN-600, and BN-800. Japan has built one demonstration FBR, Monju, adding to the research base developed by its older research FBR, the Joyo reactor. Monju is a sodium-cooled, MOX-fueled loop-type reactor with three primary coolant loops, producing 280 MW. Monju began construction in 1985 and was completed in 1991, achieving criticality on April 5, 1994; it was closed in December 1995 after a sodium leak and fire in a secondary cooling circuit. The reactor was restarted in June 2010 (about two years after its expected restart date of 2008). France’s first fast reactor, Rapsodie, first achieved criticality in 1967. Rapsodie was a loop-type reactor with a thermal output of 40 MW and no electrical generation facilities; it closed in 1983. France’s second fast reactor was the 233 MW Phénix, grid connected since 1973 and still operating as both a power reactor and, more important, as the center of work on reprocessing nuclear waste by transmutation. Superphénix, 1,200 MWe, entered service in 1984 and as of 2006 remains the largest FBR built; it was shut down in 1997 because of a political commitment the left-wing government of the time had made to competitive market forces. At the time of the shutdown, the power plant had not produced electricity for most of the preceding decade. The fast reactor EBR-I (Experimental Breeder Reactor-1) in Idaho became operational on December 20, 1951, when it produced enough electricity to power four light bulbs. The next day it produced enough power to run the entire EBR-I building. This was a milestone in the development of nuclear power reactors in the United States. The next generation of EBR was the Experimental Breeder Reactor-2, which went into service at the Idaho National Engineering and Environmental Laboratory in 1964 and operated until 1994. The EBR-2 was designed to be an “integral” nuclear plant, equipped to handle fuel recycling onsite. It typically operated at 20 MW out of its 62.5 MW maximum design power and provided the bulk of heat and electricity to the surrounding facilities. The world’s first commercial Liquid Metal Fast Breeder Reactor, and the only one built in the United States, was the 94 MW Unit 1 at Enrico Fermi Nuclear Generating Station. The plant went into operation in 1963 but shut down on October 5, 1966, because of high temperatures caused by a loose piece of zirconium that was blocking the molten sodium coolant nozzles. After restarting, it ran until August 1972 when its operating license renewal was denied. India’s first 40 MW Fast Breeder Test Reactor (FBTR) attained criticality on October 18, 1985. India was the sixth nation to demonstrate the technology to build and operate an FBTR (after the United States, United Kingdom, France, Japan, and the former USSR). India has developed the technology to produce plutonium-rich U-Pu mixed carbide fuel. The Chinese Experimental Fast Reactor (CEFR) was designed in 2003 and built near Beijing by Russia’s OKBM Afrikantov in collaboration with OKB Gidropress, NIKIET, and the Kurchatov Institute. The reactor achieved criticality in July 2010, can generate 20 MWe, and will be grid connected in 2011.