In This Issue
Fall Issue of The Bridge on Nuclear Energy Revisited
September 15, 2020 Volume 50 Issue 3
The desire to reduce the carbon intensity of human activities and strengthen the resilience of infrastructure key to economic prosperity and geopolitical stability shines a new spotlight on the value and challenges of nuclear energy.

Regulatory Innovation to Support Advanced Reactors

Tuesday, September 15, 2020

Author: Richard A. Meserve

All commercial nuclear power plants operating in the United States are light water reactors (LWRs), in which the coolant and the moderator are ordinary water. Many vendors seek to commercialize different types of reactors—so-called “advanced reactors”—that are radically different from existing LWRs. Some advanced reactors (ARs) will use gas, liquid metal, or molten salt as a coolant and simplified, inherent, passive, or other innovative means to accomplish their safety functions. Some will have a fast neutron spectrum (LWRs have a thermal neutron spectrum), some will operate at or near atmospheric pressure, and some will be much smaller than current-generation LWRs.

Advanced reactors promise lower cost per kWh, higher operating temperatures (providing greater thermodynamic efficiency and enabling expanded process heat applications), longer or more flexible operating cycles, and reduced waste production. But the unique aspects associated with these designs present a challenge because the existing regulatory system focuses on ensuring the safety of LWRs.[1]

Licensing Procedures

Existing Licensing Approaches under the Code of Federal Regulations

All of the existing US reactors, with the exception of two now under construction in Georgia, were licensed by the Nuclear Regulatory Commission (NRC) under a regulatory scheme defined in the Code of Federal Regulations, 10 CFR Part 50 (NRC 2018a). Under this licensing paradigm, an applicant first obtains a construction permit and then, while construction is underway, seeks an operating license.

The regulatory procedures associated with a construction permit involve a review of the suitability of the site and of the general appropriateness of the reactor technology. A thorough review of this technology is part of the evaluation of an application for an operating license. At both stages of the process, affected indi­viduals and organizations can challenge the NRC staff’s proposed decisions, which can result in extensive ­hearings before the commission’s Atomic Safety and Licensing Board (ASLB), followed by review by the NRC staff and the courts.

Under the Part 50 process the NRC can deny an operating license or require substantial and expensive retrofits of an already built reactor. This occurred for reactors that were under construction at the time of the Three Mile Island accident (Walker and Wellock 2010). As a result, in the late 1980s the NRC established a second licensing process (Part 52) to reduce the financial and delay risks associated with Part 50.

Under Part 52 (the regulatory approach used for the two reactors in Georgia), the licensing process can involve three components (NRC 2018a). A vendor of a reactor technology can pursue a design certification (DC) for the full design of the plant’s nuclear island or a standard design approval (SDA) for a significant portion of a design. After review of the adequacy of the design to achieve safety requirements, the NRC can promulgate a rule certifying it, which may occur long before there is a commitment to actually construct the design.

A prospective licensee can also obtain an early site permit (ESP), which defines the “environmental envelope” to be satisfied by a reactor at a particular site. An ESP can be sought before selection of the reactor technology or even a firm commitment to pursue construction.

Finally, a prospective licensee can obtain a combined license (COL) that authorizes both construction and operation (10 CFR 52.97(b)). However, the licensee is not permitted to load fuel and commence operations until the NRC determines that the inspections, tests, analyses, and acceptance criteria (ITAAC) specified in the license have been satisfied (10 CFR 52.103(g)). A COL applicant need not have an ESP and a DC (or an SDA), but if it does, these authorizations can be incorporated in the application.

Each of the Part 52 processes enables an applicant for construction to avoid regulatory risk. Matters resolved in connection with an ESP or DC cannot be reexamined (absent new and significant information that could call into question the previous resolution of an issue), which limits the scope of the licensing proceeding for a COL.

A COL by itself serves to avoid much of the regulatory risk associated with a Part 50 license because it is issued before safety-related construction starts (absent construction authorization by the NRC), reducing the danger of regulatory changes. The risk is not eliminated because of the need to satisfy the ITAAC, and in any event the NRC can always require “backfits” to conform to new regulatory requirements at any reactor if necessary to provide adequate protection of public health and safety or if the weighing of comparative costs and benefits justifies a change (10 CFR 50.109).

A drawback of Part 52 is that a DC or COL approval “freezes” the design under circumstances in which the first construction of a new design may expose the need for changes that were not anticipated in the approval process. Changes in the approval can result in expense and delay.

Neither Part 50 nor Part 52, as originally contemplated, meets the licensing needs of those pursuing advanced technologies. Moreover, other aspects of current regulatory processes present challenges.

Staged Licensing

Part 50 can present unacceptable financial risks because the determination of whether a given design can be licensed by the NRC is resolved only when an operating license is pursued. An applicant pursuing approval of an AR design confronts significant regulatory risk after substantial cost has been incurred because of the absence of precedents as to how the NRC will view novel design features.

Furthermore, although a DC under Part 52 provides some earlier certainty, it requires a complete design (or a significant portion of the design in the case of an SDA) to be defined in the application. The vendor cost for a DC is very large (many times the NRC fees) because of the necessity for submission of a complete design for NRC review, along with all the necessary test data and analyses (10 CFR 50.43(e)). Design certification thus involves a formidable front-loaded investment (SEAB 2016).

Investments in advanced technologies are typically made in stages or graduated steps as risks are retired. Some of the risks associated with the pursuit of an AR technology are technical, some reflect market risk, and some are regulatory. Regulatory risk arises because the NRC might reject a new approach or impose requirements that make the design unattractive in the ­market, or because the cost and delay of NRC review may be more than the applicant can bear. Regulatory risk is inimical to investment because it may be difficult for an applicant to assess (SEAB 2016).

For these reasons, some have urged a regulatory process in which issues are resolved in a stepwise fashion, to be compatible with a staged series of investments (Finan 2016; Merrifield 2016). In 2019 Congress passed the Nuclear Energy Innovation and Moderni­zation Act, which directs the NRC to proceed with staged licensing and, by 2027, to promulgate an optional new licensing pathway (Part 53) that is technology-inclusive, risk-informed, and performance-based (NEIMA 2019).

While the NRC prepares for the rulemaking, it is adapting current licensing pathways to achieve staged licensing using existing regulatory vehicles—­technical reports, topical reports, exemption approvals, white papers, and possibly generic environmental impact statements—to provide early guidance to vendors. It has encouraged vendors to consult with the staff to surface important issues at an early stage and to establish a licensing project plan that reflects a common understanding of the responsibility of each party and sets a licensing schedule (NRC 2019a, 2020a).

Although there still may be regulatory ­uncertainty—the NRC staff, the commission, the Advisory Committee on Reactor Safeguards, the ASLB, and the courts are not legally bound by some of these early staff ­determinations—vendor concerns about regulatory risk have been reduced (INRC 2020a).


Under existing law the NRC must recover 90 percent of its budget from fees charged to current licensees (e.g., annual fees for various classes of licensees) and through hourly charges for work to benefit a specific licensee or applicant. Many advanced reactors are much smaller than existing reactors and, in recognition of this fact, the NRC has completed a rulemaking to adjust its annual fees once such plants are in operation (NRC 2016a).

But recent experience shows that the hourly fees can present a serious challenge, as reflected in the costs for review of LWRs. The NRC review of DC applications has resulted in fees from $14 million to almost $68 million, a COL can involve fees from $22 million to $55 million, and an ESP may result in fees from $5 million to $14 million (NRC 2020b). It may be hoped that the simplicity of some of the AR designs and the promise of increased safety will reduce the cost of review, but these designs present unique issues that may make this unlikely, at least in their first regulatory encounter. Moreover, these costs reflect only NRC fees and do not include the much greater costs that applicants confront in collecting the data, completing necessary analyses, and assembling the case for licensing.

 The fees present a particular challenge for applicants with advanced approaches because many vendors are small companies whose resources must be carefully husbanded. Some cost sharing was provided by the Department of Energy (DOE) for the DC fees of a small modular reactor (SMR) and broader cost sharing of fees may be essential on an ongoing basis (INRC 2020b). Indeed, Congress recognizes that cost sharing of the even greater overall cost of the early stages of developing and building an advanced reactor may be necessary to set the stage for commercial exploitation. Congress appropriated $230 million in the FY2020 ­budget for the DOE to start a demonstration program for advanced reactors, including $160 million for the first year of funding to build two AR demonstrations (DOE/NE 2019, 2020).

Prototype Plants

Data to establish the safety of an advanced reactor design, including in particular the examination of the interaction of subsystems (so-called integral effects), may be insufficient to allow licensing of a design in its contemplated commercial configuration. If an applicant determines that sufficient data are not available from component, integral, and separate effects testing to demonstrate safety features, an applicant may propose that the planned first-of-a-kind reactor be licensed and tested as a prototype plant (NRC 2017a).

A prototype plant may be identical to a proposed standard plant design in all features and size but include additional safety features to protect the public and the plant staff from the possible consequences of accidents during the testing period. The plant can be used to test new or innovative design or safety features and com­puter models. The resulting reduction in uncertainty can then be used to justify less restrictive reactor protection systems, higher operating powers, higher operating temperatures, or longer operating cycles for subsequent plants of the same design.

A prototype plant can thus be a transitional step between the development of a particular reactor technology and full commercial deployment. The construction of a prototype to support NRC licensing has not been undertaken, but may be an attractive approach for some vendors (Buongiorno et al. 2018).

Regulatory Approach to Safety

The NRC is tasked to provide reasonable assurance of adequate protection of public health and safety and of the environment. This objective is achieved by ensuring the means to control reactivity, remove heat from the reactor and waste stores, and limit the release of radio­active material. A reactor design must provide high confidence that there are means to prevent and mitigate any failure to achieve these fundamental safety functions.

A central element in design and operations is a ­philosophy of defense in depth (DID)—layers of diverse, independent, and redundant protections and barriers to prevent or minimize a radioactive release. In addition to careful design, safety depends on close attention to safety culture, radiation protection, quality assurance, operating experience, training, maintenance and surveillance, operational excellence, and emergency preparedness (INSAG 1999).

The detailed means to achieve the safety objective for most operating LWRs were based on ­deterministic analyses and judgments, resulting in prescriptive requirements promulgated in the 1960s and ’70s. The capacity to undertake sophisticated probabilistic analyses of accident sequences was subsequently developed, along with quantitative health objectives (NRC 1986). The probabilistic analyses provide a means to determine whether requirements for meeting safety objectives should be enhanced or can be relaxed.

At the same time, the regulatory philosophy evolved to emphasize outcomes rather than prescriptive requirements (Walker and Wellock 2010). Today the early regulatory requirements for LWRs are supplemented with risk-informed and performance-based requirements (Kadambi et al. 2019; NRC 1995).[2]

In fulfillment of an NRC (2007) feasibility study, a profound new approach to the determination of the regulatory requirements for advanced reactors is being formulated with support from industry and DOE. The Licensing Modernization Project seeks to use probabilistic insights for the selection of licensing basis events (considered in the design and licensing of a plant); for the classification of structures, systems, and components (SSCs) to ensure that safety-significant components can each fulfill their function; and for the determination of DID adequacy (NEI 2019; NRC 2019b,c, 2020c). Judgment still plays an important role, but principally to provide margin to deal with uncertainty.

Figure 1 

The process is guided by a frequency-consequence curve (figure 1) to ensure that more frequent event sequences have low consequences; somewhat greater consequences can be permitted for infrequent (or rare) event sequences. The aim is a risk-informed, performance-based, and technology-inclusive means to guide licensing by way of a logical, systematic, and reproducible process. The NRC has endorsed the process and is allowing vendors to use it, rather than the detailed licensing guides established for LWRs, to support the licensing of advanced reactors (NRC 2020d).

Specific Technical Challenges

Because advanced reactors can present very different risks from those presented by LWRs, the designer and the NRC must confront specific technical issues in licensing, such as the following.

Safety Systems

Existing reactors have DID systems to ensure safety. For example, all LWRs have independent systems to inject water into the reactor and cool the core in the event of a major pipe break (10 CFR 50.46). These systems typically depend on “active” components (e.g., pumps, automatic valves, and safety-related AC power) to fulfill their function.

One common characteristic of both advanced reactors and advanced LWRs (such as those nearing completion in ­Georgia) is reliance on passive systems that use gravity, natural convection, or pressure gradients to meet the safety objective. Such systems simplify the reactor design in ways that may reduce cost. They also can have important spin-off impacts, such as the determination that the passive safety capabilities of the NuScale design (a light water SMR) justified relaxation of the safety requirements for the electrical systems that provide emergency power (NRC 2019d).

Detailed analyses and data are required to show the effectiveness of passive systems. Moreover, while AR designs may eliminate the need to consider some LWR-based accident concerns, some designs may present new safety issues, such as sodium-water reactions in sodium-cooled fast reactor designs.


The behavior and potential releases and ­consequences of events and accidents at advanced reactors may differ significantly from those of large LWRs. Many advanced designs have relatively small cores as well as other features (such as passive decay heat removal) that are anticipated to result in smaller and slower accident releases. This means that advanced reactors might allow reduced distances to exclusion area boundaries and low-population zones, and ­potentially increased proximity to population centers (NRC 2016b, 2017b, 2019e). This opens the prospect that advanced reactors might replace fossil plants, which are often located in the vicinity of dense populations, and thereby make use of existing energy transmission infrastructure.


The current physical security framework for large LWRs is designed to protect the plant features that provide fundamental safety functions. Advanced reactor designs are expected to include attributes that result in smaller and slower releases of fission products in the event of any loss of safety functions (NRC 2018b).

There is an opportunity with advanced reactors to consider security requirements in the design to a ­greater extent than was the case with LWRs (NEI 2016). For example, protection from an aircraft attack can be enhanced through below-grade installation of safety-significant SSCs. Similarly, enhanced safety systems could limit or delay the radiological risk arising from an attack. These changes may improve security and reduce reliance on security personnel—a meaningful part of the operating cost at existing LWRs—to prevent or mitigate attempted radiological sabotage. The NRC is pursuing a limited-scope rulemaking to address this issue (NRC 2019f).


Much of the construction cost for LWRs is associated with the massive reinforced concrete structures that are intended to provide the final barrier to the release of radioactive material in the event of an accident (­Buongiorno et al. 2018). The operating conditions, coolants, and fuel forms of non-LWR technologies differ from those of LWRs and may allow or possibly require different types of containments. If a design can retain radioactive materials by using other barriers, the building enclosing the reactor may not be necessary to fulfill the containment function for some or all event categories (NRC 2018c, 2019g).

NRC staff are applying functional containment performance criteria, opening the prospect of avoidance of the significant cost of existing containments for AR designs that can provide alternative means for preventing or mitigating large radioactive releases (NRC 2020c).


Existing LWRs use uranium-oxide pellets enriched in U-235 to about 5 percent, with a zirconium alloy fuel cladding. Several fuel types are proposed for advanced reactors, including tristructural isotropic (TRISO) particle and metallic fuels, enriched in U-235 in some cases to nearly 20 percent (GNI 2019). Some of the contemplated molten salt reactors even have the nuclear fuel dissolved in the molten salt coolant.

The NRC requires that all fuels display accident tolerance while meeting other performance standards, such as retention of fission products and cladding-coolant compatibility. One of the challenges that must be overcome is the limited experimental data on some non-LWR fuel types. This is likely to be a particular challenge for some designs because of the need for extensive irradiation to provide the data necessary to support the safety case.


There is great interest in the commercialization of advanced reactors, but their licensing presents serious regulatory challenges. The business case for many of the new designs assumes that many existing regulatory requirements can be relaxed or modified in light of their inherent safety features.

Beyond procedural and technical challenges, there are needs for NRC staff training on unfamiliar technologies, the development of analytical tools, ­advances in computer codes and standards, and coordina­tion among industry, DOE, national laboratories, and international organizations (NRC 2020a). Fortu­nately, the commercial sector, DOE, and NRC are working to address these challenges and their complicated dimensions.


Buongiorno J, Corradini M, Parsons J, Petti D. 2018. The Future of Nuclear Energy in a Carbon-Constrained World. Cambridge: Massachusetts Institute of Technology.

DOE/NE [Department of Energy, Office of Nuclear Energy]. 2019. Analysis: FY2020 Spending Bill Points to Nuclear Resurgence. Washington.

DOE/NE. 2020. Request for Information/Notice of Intent (DE-FOA RFI-0002271): Information Request on the Advanced Reactor Demonstration Program. Washington.

Finan AE. 2016. Enabling Nuclear Innovation: Strategies for Advanced Reactor Licensing. Cambridge MA: Nuclear Innovation Alliance.

GNI [Global Nexus Initiative]. 2019. Advancing Nuclear Innovation: Responding to Climate Change and Strengthening Global Security.

INRC [Inside NRC]. 2020a. Advanced reactor developers’ focus shifts to deployment: US NIC. Inside NRC 42(8):3.

INRC. 2020b. NRC review fees “much too high” for micro­reactors, says USNIC. Inside NRC 42(11):1.

INSAG [International Nuclear Safety Advisory Group]. 1999. Basic Safety Principles for Nuclear Power Plants (75-INSAG-3 Rev 1; INSAG-12). Vienna: International Atomic Energy Agency.

Kadambi NB, Wallace E, O’Brien J, Youngblood R. 2019. Risk-informed, performance-based safety: Past, present, and future. Nuclear News 45.

Merrifield JS. 2016. Issue brief on the framework for advanced reactor licensing modernization. US Nuclear Infrastructure Council. Online at 804492aec73c4284b0577281d5b3 a5a7.pdf.

NEI [Nuclear Energy Institute]. 2016. Proposed physical security requirements for advanced reactor technologies. ­Washington. 

NEI. 2019. Risk-informed performance-based technology guidance for non-light water reactor licensing basis development (NEI 18-04 Rev 1). Washington.

NEIMA [Nuclear Energy Innovation and Modernization Act]. 2019. Public Law 115-439, 132 Stat. 5565. Online at 512/text.

NRC [US Nuclear Regulatory Commission]. 1986. Safety goals for the operation of nuclear power plants: Policy statement (republication). 51 FR 28044. Washington.

NRC. 1995. Use of probabilistic risk assessment methods in nuclear regulatory activities: Final policy statement. 60 FR 42622.

NRC. 2007. Feasibility study for risk-informed and performance-based regulatory structure for future plant licensing (NUREG-1860).

NRC. 2016a. Variable annual fee structure for small modular reactors (final rule). 81 FR 32617.

NRC. 2016b. Accident source terms and siting for small modular reactors and non-light water reactors (SECY 16-0012).

NRC. 2017a. A regulatory review roadmap for non-light water reactors (ML17312B567).

NRC. 2017b. Siting considerations related to population for small modular and non-light water reactors (preliminary draft).

NRC. 2018a. Backgrounder on nuclear power plant licensing process.

NRC. 2018b. Options and recommendations for physical security for advanced reactors (SECY 18-0076).

NRC. 2018c. Functional containment performance criteria for non-light-water reactors (SECY 18-0096).

NRC. 2019a. Non-light water review strategy (draft staff white paper).

NRC. 2019b. Technology-inclusive, risk-informed, and performance-based methodology to inform the licensing basis and content to applications for licenses, certifications, and approvals for non-light-water reactors (SECY-19-0117).

NRC. 2019c. Guidance for a technology-inclusive, risk-informed, and performance-based methodology to inform the licensing basis and content of applications for licenses, certifications, and approvals for non-light-water reactors (DG-1353).

NRC. 2019d. NuScale safety evaluation report, ch 8.

NRC. 2019e. In the matter of Tennessee Valley Authority (Clinch River Nuclear Site Early Site Permit Application) (CLI-19-10).

NRC. 2019f. Physical security for advanced reactors (proposed rule). 84 FR 33861.

NRC. 2019g. Functional containment performance criteria technology-inclusive, risk-informed, performance-based approach (SECY 18-0096, Enc 2).

NRC. 2020a. Report to Congress on approaches for expediting and establishing stages in the licensing process for commercial advanced nuclear reactors.

NRC. 2020b. New reactors business line fee estimates.

NRC. 2020c. Non-light-water reactor implementation action plan: Progress summary and future plans.

NRC. 2020d. Report to Congress on increasing the use of risk-informed and performance-based evaluation techniques and regulatory guidance in licensing commercial advanced nuclear reactors.

SEAB [Secretary of Energy Advisory Board]. 2016. Final Report of the Task Force on the Future of Nuclear Power. Washington: Department of Energy.

Walker JS, Wellock TR. 2010. A short history of ­nuclear regulation, 1946–2009, Rev 2 (NUREG/BR-0175). ­Washington: US Nuclear Regulatory Commission.


[1]  A summary by the Nuclear Regulatory Commission of its various activities to deal with the regulatory challenges is available at A survey of issues is at

[2]  The history of risk-informed regulation is available at history.html.

About the Author:Richard Meserve (NAE) is senior of counsel at Covington & Burling LLP and former chair of the US Nuclear Regulatory Commission.