Don’t scrap ALARA—modernize it

For decades, U.S. radiation protection has rested on two pillars: the linear no-threshold (LNT) model and the principle of "as low as reasonably achievable" (ALARA). Together, these have shaped occupational dose management, regulatory limits, and emergency planning across the nuclear industry.

Today, both are under renewed scrutiny-not because the underlying science has changed in any fundamental sense, but because the legal, institutional, and policy environment has shifted. Several recent developments illustrate this shift.

Executive direction, including Executive Order 14300, calls on the Nuclear Regulatory Commission to reconsider its use of both LNT and ALARA. The EO reflects a broader policy view that current radiation protection standards may impose significant costs and could constrain nuclear energy expansion that President Trump seeks to promote.

In January 2026, the Department of Energy eliminated ALARA from its radiation protection framework without adopting a direct replacement standard. While the health impacts of low-dose exposures remain persistently uncertain and difficult to detect, the DOE cited concerns about reducing "economic and operational burdens" as part of its rationale for moving away from long-held practice. Although the DOE operates under a different statutory framework from the NRC, the move illustrates that pressure to reconsider ALARA and LNT is moving from theory to institutional change.

Simultaneously, the legal framework governing NRC decision-making has shifted in ways that will shape the commission's upcoming decisions on ALARA/LNT. The 2024 ADVANCE Act imposes two key mandates: (1) the NRC must regulate in a manner that is "efficient" and that "does not unnecessarily limit" the societal benefits of nuclear technology, and (2) it must remain "consistent with the policies of the Atomic Energy Act . . . including to provide reasonable assurance of adequate protection to public health and safety."

Meanwhile, the Supreme Court's decision in Loper Bright Enterprises v. Raimondo eliminated so-called Chevron deference, under which courts generally deferred to reasonable agency interpretations of ambiguous statutes. Courts no longer defer to agency interpretations of ambiguous terms-such as what "efficient" regulation or "adequate protection" mean under the ADVANCE Act-but instead exercise their own independent judgment regarding the "best" interpretation. The NRC must now demonstrate that its decisions are not only reasonable and scientifically grounded, but also legally defensible under this standard. To reduce the risk of reversal by courts, the NRC must interpret the ADVANCE Act in a manner that is consistent with likely judicial interpretations, while retaining the ability to justify its decisions as required by the Administrative Procedure Act.

What has changed, therefore, is not the existence of uncertainty in radiation science but the institutional requirement that such uncertainty be addressed explicitly within a legal framework that demands justification rather than reliance on judicial deference.

This raises two central questions:

1. Is the ALARA decision-making framework consistent with the mandates of the ADVANCE Act for both efficiency and adequate protection?

2. How should uncertainty in radiation protection be addressed to satisfy legal and technical requirements for both efficiency and adequate protection?

This article addresses both questions and outlines a path forward.

On the ALARA decision-making framework

The ADVANCE Act requires the NRC to regulate efficiently and in a manner that does not unnecessarily limit the societal benefits of nuclear technology. Efficiency can be understood in three dimensions:

Technological efficiency-performance of nuclear systems.

Organizational efficiency-NRC internal processes and timeliness.

Judicial/regulatory efficiency-whether regulatory decisions are justified through structured cost-benefit reasoning.

Although the NRC has historically relied on technological and organizational dimensions of efficiency, the judicial/regulatory dimension has become increasingly important under the legal framework emerging from the ADVANCE Act and the Supreme Court's Loper Bright decision.¹ Loper Bright makes it clear that courts must exercise independent judgment in interpreting statutory terms rather than deferring to agency interpretations. Courts will not simply defer to the NRC's interpretation of what "efficiency" means under the act. Instead, NRC decisions invoking "efficiency" will be evaluated based on traditional tools of statutory construction-tools that have historically led courts to read efficiency as requiring cost-benefit analysis.

Engaging with judicial/regulatory efficiency increases the predictability of outcomes for both the NRC and regulated entities. Ignoring the judicial/regulatory dimension creates uncertainty because NRC licensing and regulatory actions become more vulnerable to judicial reversal. Courts are also likely to scrutinize the reasoning underlying NRC decisions rather than accept interpretations based on agency expertise alone.

ALARA is a decision-making framework for managing radiation exposures under conditions of uncertainty. In U.S. nuclear regulation, ALARA is embedded in 10 CFR §20.1101(b), which requires licensees to maintain exposures "as low as is reasonably achievable," taking into account economic, technological, and societal factors. This aligns with the International Commission on Radiological Protection's system of radiological protection, which defines optimization as a continuing process of balancing radiation risk against broader considerations such as cost, feasibility, and societal benefit. At its core, it is best understood as an optimization principle rather than a mandate to minimize dose at any cost. ALARA programs rely on engineering controls (shielding, automation, remote handling), administrative controls (time minimization, scheduling, access limits), radiation monitoring, and structured planning and review. These are implemented through formal procedures in which trade-offs among dose, cost, and feasibility are managed using expert judgment and, in some cases, semiquantitative cost-benefit analysis.

In practice, however, ALARA implementation has often relied on qualitative or implicitly structured assessments within procedural radiation protection programs. Over time, the balancing embedded in "reasonably achievable" has weakened. Observers have argued that ALARA application has drifted in practice toward "as low as possible," with NRC implementation increasingly emphasizing dose minimization.² The consequence is that NRC decision records do not always clearly demonstrate how dose reduction, cost, and feasibility are systematically balanced. In a legal environment shaped by the ADVANCE Act and Loper Bright, such gaps in justification may become increasingly difficult to defend.

Importantly, this does not imply that ALARA should be abandoned. Rather, it suggests that the optimization framework underlying ALARA should be more explicit, systematic, and transparent-in closer alignment with the efficiency mandate and evolving expectations for reasoned decision-making.

If ALARA is implemented through a clearly defined cost-benefit structure, then safety and efficiency can be addressed within a unified decision framework. In this sense, and as we have argued elsewhere,¹ efficiency is not in tension with safety. ALARA also has the potential to support the "adequate protection" mandate of the ADVANCE Act. Within the range of cost-benefit-justified options, "reasonably achievable" implies a structured optimization directive. The phrase "as low as" introduces a protective orientation by favoring lower-exposure outcomes among otherwise acceptable alternatives.

Consider two benefit-justified technologies for reducing exposure. One reduces dose by 1 mrem, while the other reduces dose by 2 mrem. Both are benefit-justified and both are acceptable under the efficiency mandate, so efficiency alone does not determine which should be selected. Without ALARA, selection may depend on arbitrary or implicit assumptions. ALARA, however, directs decision-makers toward the more protective option, providing a structured basis for favoring lower exposures among otherwise benefit-justified alternatives and thereby helping to satisfy the adequate protection mandate.

ALARA helps reconcile the NRC's dual mandate: to regulate efficiently while still ensuring reasonable assurance of adequate protection. When implemented as a structured optimization procedure, ALARA provides a mechanism for satisfying both requirements simultaneously. Eliminating ALARA would remove an established decision-making framework without resolving the underlying need for structured management of the ADVANCE Act's efficiency and adequate protection mandates.

On addressing uncertainty in radiation protection

To support the efficiency and adequate protection mandates, it is not sufficient to rely solely on the ALARA decision-making framework. It is also necessary to explicitly address the uncertainties associated with the cost and benefit inputs to that framework. Two broad categories of uncertainty affect the cost and benefit inputs used in ALARA decision-making: (1) model-form uncertainty in dose-response relationships and (2) control effectiveness and human performance uncertainty.

At low doses, generally below roughly 100 millisieverts, epidemiological studies lack sufficient statistical power to distinguish among competing dose-response models. Linear, threshold, and linear-quadratic relationships have all been proposed, but available data do not permit reliable discrimination among them in this range. Accordingly, there remains persistent uncertainty regarding health effects at low doses, and no model can be reliably distinguished using available epidemiological data at these dose levels. The absence of observed effects may therefore reflect limitations of detection rather than confirmation or rejection of any specific dose-response relationship.

Because of this indeterminacy, nuclear regulatory systems have historically adopted conservative default assumptions to support decision-making under uncertainty. It is in this context that LNT and ALARA are often treated as inseparable, although they are conceptually distinct.

The LNT model is a dose-response hypothesis that assumes cancer risk increases linearly with dose, with no threshold. It is used as a default assumption in several regulatory contexts, including by the Environmental Protection Agency for radiation risk assessment and more broadly for carcinogenic risk evaluation. In contrast, ALARA is a decision-making framework for managing exposures under uncertainty. It requires a dose-response model as an input, but is itself model-agnostic; it does not prescribe a specific functional relationship between dose and risk.

In practice, however, ALARA implementation frequently relies on LNT-based assumptions when quantifying potential health impacts from small changes in exposure. When quantitative cost-benefit analysis is applied, reductions in collective dose (person-rem) are commonly converted into monetary terms using factors developed in NRC regulatory analysis guidance, including NUREG/BR-0058 and Regulatory Guide 1.110. Health impacts are typically valued using the Value of a Statistical Life, consistent with federal regulatory practice and Office of Management and Budget guidance, as well as EPA methodologies.

Conceptually, this process separates into two components: a dose-response model used to estimate cancer risk per unit dose, and a monetary valuation framework used to assign economic value to that cancer risk. This yields a dollar-per-person-rem conversion factor that enables comparison of radiation protection options on a common economic basis. While the dose-response assumptions influence estimated risk per unit dose, the monetary valuation framework is defined independently. Consequently, different dose-response models might change the numerical conversion factor but not the underlying ALARA decision-making framework for balancing protection, cost, and feasibility.

Model-form uncertainty in dose-response relationships and control effectiveness uncertainty jointly affect ALARA outcomes. Dose-response assumptions determine estimated health risk per unit exposure, while control effectiveness determines the actual distribution of exposure reductions achieved in practice. Current regulatory practice, however, adopts a single model assumption-that is, LNT-rather than systematically addressing uncertainty across alternative plausible dose-response models.

Control effectiveness and human performance uncertainty arise because radiation protection measures-including engineering controls, administrative procedures, and human performance-introduce probabilistic variability in realized dose reductions. In ALARA applications, these measures are often treated deterministically or credited at nominal effectiveness levels, despite well-established variability in real-world implementation.

Historically, these uncertainties were addressed implicitly within regulatory practice. Justifying such regulatory decisions was substantially easier before Loper Bright; now, however, regulatory decisions are more likely to be evaluated based on consistency with what a court determines to be the best reading of the statutory mandates, including how key assumptions underlying statutory terms such as "efficient" are treated. This increases the importance of explicitly articulating how uncertainty is incorporated into ALARA-based cost-benefit analysis and of aligning regulatory practice with statutory language.

A central question, therefore, is how these uncertainties can be explicitly incorporated into ALARA-based cost-benefit analysis in a structured and transparent manner.

As discussed previously,¹ probabilistic methods provide a structured approach to quantifying uncertainty in both costs and benefits, enabling expected outcomes to be estimated in a transparent and reproducible way. When uncertainty is not explicitly quantified, deterministic or worst-case assumptions may instead be used. While such deterministic approaches may be appropriate in limited contexts, they must be technically and legally justified as bounding cases and may systematically bias decisions toward conservatism if applied broadly.

Importantly, eliminating ALARA would not eliminate uncertainty. It would instead remove a structured decision-making framework through which uncertainty can be analyzed and incorporated into regulatory justification. By contrast, strengthening ALARA to explicitly address uncertainties using probabilistic methods would improve its ability to meet both the adequate protection objective and emerging legal expectations for reasoned and efficient decision-making.

A path forward

The best solution is not to abandon ALARA but to restore and strengthen its original function as an optimization framework capable of supporting transparent, defensible decision-making under uncertainty. This will require situating it explicitly within the ADVANCE Act and the NRC's other statutory mandates.

1. Fix the ALARA decision-making framework to support both efficiency and adequate protection. When properly implemented, ALARA offers a mechanism for aligning the new statutory mandate of efficiency with the NRC's ongoing mandate for adequate protection. To ensure alignment with the NRC's statutory mandate of efficiency, ALARA should explicitly incorporate structured cost-benefit analysis as a decision-support tool for determining reasonableness, rather than relying primarily on implicit or purely qualitative judgment.

Efficiency in this context does not mean minimizing radiation exposure at any cost. Rather, it requires identifying levels of protection that appropriately balance reductions in radiological risk against economic, technological, and operational considerations under conditions of uncertainty.

In addition to satisfying the efficiency mandate, the NRC remains legally obligated to ensure adequate protection in radiation regulation. Consistent with this obligation, ALARA should retain an explicit protective orientation among alternatives deemed justified through structured cost-benefit analysis.

Accordingly, ALARA should operate as a structured optimization framework that: (i) identifies benefit-justified alternatives through cost-benefit analysis in order to satisfy the efficiency mandate; and (ii) favors the most protective option (i.e., the lowest exposure) among those benefit-justified alternatives in order to support the adequate protection mandate.

Under this structure, ALARA can serve as a practical mechanism for reconciling the agency's dual mandate-advancing regulatory efficiency while maintaining reasonable assurance of adequate protection.

2. Incorporate risk-informed methods to address uncertainty within dose-­response models and ALARA control processes in order to satisfy legal and technical requirements for efficiency and adequate protection. To support these dual mandates, uncertainties in both cost and benefit inputs to the ALARA decision-making framework should be explicitly addressed. These include uncertainties in dose-response relationships as well as control effectiveness and human performance.

At low doses, multiple dose-response models remain scientifically plausible, and no single functional form-including the LNT model-can be empirically distinguished with high confidence. Accordingly, ALARA decision-making should not rely on a single assumed model but instead should evaluate outcomes across a set of plausible alternatives, explicitly representing model-form uncertainty. If model-form uncertainty analysis is not conducted, LNT may be used as a conservative default assumption; however, this can systematically bias decisions toward conservatism.

Existing NRC guidance already recognizes the importance of treating epistemic uncertainty in regulatory analysis. For example, NUREG-1855 Revision 1 provides a structured framework for identifying and characterizing model-form uncertainty in probabilistic risk assessments; however, this treatment has not yet been systematically integrated into ALARA-based radiation protection decision-making.

In addition, control effectiveness and human performance introduce substantial variability in realized dose reductions. Engineering controls, administrative procedures, and behavioral factors all influence actual exposure outcomes, often in ways that differ from nominal or design assumptions. These uncertainties can be explicitly represented using established methods, such as probabilistic risk assessment (PRA) and human reliability analysis (HRA). If uncertainties in control effectiveness and human performance are not explicitly quantified using probabilistic methods, worst-case assumptions that conservatively bound outcomes with respect to safety may instead be applied. While such deterministic approaches may be appropriate in limited contexts, they require technical and legal justification as bounding cases and may introduce systematic bias if used broadly.

While PRA and HRA are well developed in the context of reactor risk analysis, their application to occupational and public radiation protection decisions under ALARA remains limited. Extending these risk-informed methods into radiation protection would align ALARA more closely with the NRC's broader and long-standing risk-informed, performance-based (RIPB) regulatory framework. The ADVANCE Act also directs the use of RIPB approaches in multiple NRC licensing and rulemaking activities.

Integrating these tools would also provide a more explicit, quantitative, and defensible basis for evaluating cost, benefit, and uncertainty in ALARA decision-making-thereby strengthening its compatibility with the efficiency expectations emerging under the ADVANCE Act and post-Loper Bright judicial review.

Conclusion

The debate over LNT and ALARA is often framed as scientific. In practice, it is becoming increasingly legal and institutional.

The issue is not whether uncertainty exists-it always has-but how that uncertainty is addressed within a legal framework that now requires satisfaction of two statutory mandates-efficiency and adequate protection-as well as reasoned decision-making under the Administrative Procedure Act. The problem is not ALARA itself, but the way it has often been implemented.

Abandoning ALARA would remove a structured approach without resolving the underlying challenge. A more defensible path is restoration and modernization. ALARA should be explicitly re-established as a structured decision-making framework that identifies benefit-justified alternatives through cost-benefit analysis while preserving a protective orientation by favoring the most protective option (i.e., the lowest exposure) among those alternatives. Its implementation should incorporate risk-informed and probabilistic methods that explicitly address uncertainty in dose-response relationships and radiation protection controls.

In this form, radiation protection can better reflect the scientific reality of low-dose uncertainty and the legal need for reasoned, efficient regulation. At the same time, it can continue to provide adequate protection.

References

1. G. Joslin et al., "Nuclear Power's New Rule Book: Managing Uncertainty in Efficiency, Safety, and Independence," Nuclear Newswire, Nov. 21, 2025.
2. J. C. Wagner et al., Reevaluation of Radiation Protection Standards for Workers and the Public Based on Current Scientific Evidence (INL/RPT-25-85463, Idaho National Laboratory, July 2025).

George Joslin is a graduate research assistant and Ph.D. student in the Department of Nuclear, Plasma, and Radiological Engineering (NPRE) and the Socio-Technical Risk Analysis (SoTeRiA) Research Laboratory at the University of Illinois-Urbana-Champaign (UIUC). Arden Rowell is a professor in the College of Law at UIUC. Seyed Reihani is a senior research scientist in the NPRE Department and the SoTeRiA Research Laboratory at UIUC. Zahra Mohaghegh is a professor in the NPRE Department at UIUC, where she leads the SoTeRiA Research Laboratory.