The CANDU Design and Its Inherent Licensing Advantages

CANDU reactors are pressurized heavy-water reactors that use deuterium oxide as both moderator and coolant. Because heavy water absorbs very few neutrons, the reactor can sustain a chain reaction using natural uranium without enrichment. This fundamental design choice creates a cascade of safety and operational features with direct implications for licensing. The use of hundreds of horizontal pressure tubes rather than a single large pressure vessel means that a failure in one channel does not propagate structurally to the rest of the core. This distributed geometry has been a cornerstone of the safety case submitted to regulators in Canada, South Korea, Romania, China, India, Argentina, and other countries that operate CANDU or CANDU-derivative units. Online refueling—where fuel bundles are pushed through the channels by remote machines at full power—eliminates periodic shutdowns and reduces reactivity transients. Regulators have historically viewed this as a positive factor for grid stability and operational flexibility, especially as more countries explore coupling nuclear with variable renewables.

CANDU reactors also offer a high degree of fuel-cycle flexibility. They can burn natural uranium, slightly enriched uranium, mixed oxide fuel, thorium-based fuels, and even spent fuel from light-water reactors through the DUPIC cycle. This versatility means that a CANDU unit can be licensed for different fuel strategies over its lifetime, provided each new fuel type undergoes a thorough safety review. Regulators in Canada, for example, have already assessed advanced fuel bundles through the Canadian Nuclear Safety Commission’s risk-informed framework, setting precedents for how changes in fuel composition can be handled without requiring a completely new site license.

From a structural integrity standpoint, the use of double containment buildings, a dousing system for rapid pressure suppression, and two independent fast-acting shutdown systems (shut-off rods and liquid poison injection) give CANDU stations a strong probabilistic safety profile. The CNSC’s public hearings on life extension and refurbishment projects at Bruce Power and Ontario Power Generation’s Darlington and Pickering stations have repeatedly validated that the defense-in-depth philosophy meets modern safety objectives, even as the reactors age. This continuous regulatory oversight, rather than a one-time stamp of approval, is a hallmark of Canadian nuclear regulation that many other nations now seek to emulate.

Evolving Regulatory Frameworks in a Decarbonizing World

From Design-Basis to Risk-Informed Licensing

Nuclear licensing traditionally rested on deterministic design-basis accidents: a set of prescribed fault scenarios that a plant must withstand without releasing radioactive material beyond limits. While DBAs still form the backbone of CANDU safety analysis, regulators globally are shifting toward risk-informed, performance-based approaches. The CNSC’s REGDOC‑2.5.2 explicitly embraces probabilistic safety assessment as a complement to deterministic methods. For CANDU reactors, extensive operational data—some units have clocked more than 600,000 hours—can be fed into living PSA models. These models provide a dynamic, quantitative picture of risk that helps prioritize maintenance, aging management, and design modifications. As nations demand that power plants demonstrate resilience to beyond-design-basis events, including those triggered by climate change (extreme flooding, loss of ultimate heat sink), the risk-informed approach allows CANDU operators to justify targeted upgrades without an open-ended, cost-prohibitive resubmission of the entire safety case.

Post-Fukushima Enhancements and Climate Resilience

The 2011 Fukushima Daiichi accident prompted a fundamental re-evaluation of reactor licensing worldwide. For CANDU units, regulators now require demonstration that the plant can withstand severe accident scenarios, including station blackout, prolonged loss of cooling, and multi-unit events. The CNSC’s Integrated Action Plan required all power reactors to implement diverse emergency power supplies, hardened containment venting, and enhanced instrumentation for accident monitoring. These modifications were licensed through a structured process that re-analyzed containment response to hydrogen combustion and added passive autocatalytic recombiners. Climate resilience adds another layer: new licensing applications must account for site-specific projections of extreme weather—higher flood levels, more intense wildfires, and increased ambient temperatures that could degrade ultimate heat sink performance. CANDU operators in Canada have raised flood protection barriers and diversified water sources; these measures have been incorporated into updated safety reports reviewed by the CNSC.

Harmonization with International Standards

The International Atomic Energy Agency has been instrumental in developing safety standards such as SSR‑2/1 and GSR Part 4. CANDU licensing now routinely draws on these documents to demonstrate equivalence, facilitating regulatory collaboration between countries. The IAEA’s safety standards serve as a common reference. For CANDU new builds in countries like China and South Korea, a hybrid licensing approach blending IAEA principles with local regulation was adopted. Today, as several nations consider CANDU-based small modular reactors, there is renewed push to harmonize licensing processes through forums like the IAEA’s Nuclear Harmonization and Standardization Initiative and the CANDU Owners Group. COG members exchange safety analysis methodologies, aging management programs, and operational experience, effectively building a pre-reviewed library that can streamline a new unit’s licensing. The World Nuclear Association highlights that such cooperation reduces time and cost by avoiding duplication of generic assessments.

Integrating Security, Safeguards, and Environmental Impact

In the age of energy transition, a reactor license is no longer just a nuclear safety document. It must satisfy increasingly stringent requirements for security, physical protection, and environmental sustainability. The physical protection regime for CANDU facilities follows the IAEA’s INFCIRC/225/Rev.5 guidance and, in Canada, the Nuclear Security Regulations. Because CANDU stations store multiple years’ worth of spent fuel in water-filled bays surrounded by reinforced concrete, they have historically been judged robust against external threats. However, licensing new units or life extensions now includes cyber-security assessment of safety-critical digital assets. The CNSC and its foreign counterparts have developed technical guidance to evaluate the security of digital instrumentation and control systems, which are increasingly being retrofitted even in older CANDU 6 units.

Environmental assessments under legislation like the Canadian Impact Assessment Act demand that licensing account for the entire lifecycle—from uranium mining through waste disposal. CANDU’s natural-uranium fuel cycle reduces the front-end environmental footprint by avoiding enrichment, but the back-end challenge of used fuel management remains. The deep geological repository concept, as pursued by Canada’s Nuclear Waste Management Organization, must demonstrate that CANDU spent fuel bundles can be safely isolated for hundreds of thousands of years. Site-specific environmental impact statements, public hearings, and Indigenous engagement are now integral to any new CANDU licensing application, reflecting society’s demand for transparent and inclusive decision-making.

Challenges Specific to the Current Energy Transition

Public Acceptance and the Social License

No reactor design earns a regulatory license in a vacuum; the social license conferred by host communities is equally critical. The energy transition has paradoxically both boosted nuclear’s appeal as a firm, low-carbon source and intensified scrutiny of its risks. In Canada, public trust in the CNSC’s independence has been a stabilizing factor, but CANDU projects in other jurisdictions have faced delays when local populations raised concerns about seismic safety, emergency planning, or waste storage adequacy. The licensing process increasingly incorporates open hearings, community liaison committees, and participatory environmental monitoring. This shift toward deliberative governance, while time-consuming, can strengthen the safety case by exposing weaknesses early and encouraging operators to go beyond compliance. The experience of the Cernavodă units in Romania, where an independent expert review panel complemented the national regulator, illustrates how public engagement can be folded into the licensing framework without undermining technical rigor.

Aging Infrastructure and Life Extension

Many of the world’s CANDU reactors are entering or already in their second license renewal periods. Age-related degradation mechanisms—flow-accelerated corrosion in feeder pipes, pressure-tube deformation, steam-generator fouling—must be justified to continue operation for a 60-year or even 70-year lifetime. The licensing process for life extension (often called refurbishment in the CANDU context) is a massive undertaking involving replacement of core components like all 480 fuel channels and the calandria tubes. Canadian regulators require a robust safety case showing that the refurbished core meets the same safety objectives as a new plant, even if the design basis was written in the 1970s. The CNSC’s risk-informed approach allows operators to demonstrate that certain original conservatisms can be relaxed if sufficient operational data exist. Nonetheless, the scale of documentation—thousands of pages—and multi-year hearing process represent a significant regulatory cost. For countries considering new build CANDU, life-extension precedent provides a proven regulatory pathway but also highlights the importance of designing for long-term maintainability from the start.

Waste Management and Repository Licensing

The back-end of the fuel cycle remains one of the most contentious licensing hurdles for any nuclear technology. CANDU spent fuel differs from light-water reactor fuel in bundle geometry, burnup, and heat load, affecting both interim storage and final disposal. In Canada, the NWMO’s site selection process for a deep geological repository involves a separate licensing step under the CNSC, and the repository design must specifically accommodate CANDU bundles. Licensing new CANDU units now typically includes a condition that the operator must have a credible waste management plan, including financing for eventual decommissioning. Some regulators, such as the U.S. Nuclear Regulatory Commission, require a waste confidence determination even for reactor license renewals. For CANDU operators, demonstrating that waste can be safely emplaced in a repository—regardless of whether that repository is yet built—is essential to obtaining a construction or operating license. The success of the NWMO’s licensing process in Canada will set a powerful precedent for other CANDU-host countries.

Supply Chain Resilience and Fuel Cycle Security

The global energy transition has exposed vulnerabilities in nuclear fuel supply chains, particularly enrichment services. CANDU’s ability to operate on natural uranium insulates its operators from disruptions in enrichment capacity, a strategic advantage gaining recognition. However, the heavy water supply itself is concentrated: Canada and some other nations produce reactor-grade heavy water, and any new CANDU licensing application must demonstrate a reliable source for both initial inventory and ongoing makeup. Regulators may also require contingency plans for heavy water loss, including the possibility of converting to light-water cooling in an emergency. As interest in thorium-fueled CANDU designs grows, licensing will need to address the supply of thorium, which is more abundant than uranium but lacks established mining and processing infrastructure. The IAEA’s nuclear fuel cycle guidance can help standardize safety and safeguards assessments for alternative fuel cycles, but national regulators will need to invest in expertise to review unique characteristics of each fuel type.

Innovations That Could Reshape CANDU Licensing

CANDU-Based Small Modular Reactors

Small modular reactors incorporating heavy-water moderation, natural-uranium fueling, and online refueling are moving from concept to design stage. The Canadian government’s SMR Action Plan explicitly includes heavy-water-cooled designs. Licensing an SMR presents an opportunity to build a modern safety case from scratch, using probabilistic methods, digital twins, and advanced accident simulation tools. Regulators may apply graded approaches, relaxing some siting requirements if the smaller source term and lower decay heat reduce the off-site emergency planning zone. However, new questions arise: how does a reactor designed for factory fabrication and shipment to site interface with national safety codes for transport and installation? What measurement frameworks prove modular fabrication achieves the same quality as on-site construction? Early engagement between SMR developers and regulators, including the CNSC’s pre-licensing vendor design review of the CANDU‑SMR concept, is laying the groundwork for a licensing framework that could be replicated internationally.

In the United States, a CANDU‑type SMR may eventually interact with the NRC’s Part 53 rulemaking for advanced reactors, which aims to create a technology-inclusive, risk-informed framework. The World Nuclear Association notes that collaboration through the IAEA’s SMR Regulators’ Forum can help align technical expectations, so that a CANDU SMR licensed in Canada could be rapidly accepted by other nations.

Advanced Fuel Cycles and On-Site Waste Management

Licensing a CANDU reactor to burn advanced fuels—such as thorium-plutonium mixtures or reprocessed uranium from light-water reactors—requires a thorough reassessment of reactivity coefficients, accident source terms, and fuel-handling procedures. The DUPIC cycle demonstrated that CANDU physics can accommodate a fuel chemically and radiologically very different from natural uranium. Future licensing applications for such cycles will need to show that the modified fuel does not compromise the two-shutdown-system philosophy, that thermal performance during a loss-of-coolant accident remains within accepted limits, and that the waste form is compatible with existing disposal concepts. Regulators are increasingly open to performance-based assessments using modeling and experimental data to justify deviations from standard fuel specifications, provided monitoring during operation is sufficiently rigorous. This could unlock a scenario where existing CANDU fleets become effective waste burners, reducing the volume and radiotoxicity of high-level waste destined for deep geological repositories.

Digitalization and the Safety Case of the Future

Digital twins—high-fidelity simulation models mirroring the physical plant in real time—are beginning to influence the regulatory dialogue. A CANDU operator might propose using a digital twin to monitor pressure-tube sag, feeder thinning, or steam-generator tube plugging, feeding data into a probabilistic model that auto-updates the risk profile. If regulators accept this as a valid living safety case, it could reduce the frequency of manual inspections and allow condition-based maintenance. Pilot projects in Canada are exploring how machine-learning algorithms can predict flow-accelerated corrosion, a chronic issue in CANDU outlet feeders, potentially preempting failures and improving justification for extended operating intervals. Licensing such systems requires a new kind of engineering judgment: the algorithm itself becomes a safety-significant component, and its training data, bias, and failure modes must be scrutinized. The CNSC’s discussion paper on regulation of artificial intelligence in nuclear safety signals that this is a live issue that will only grow in importance.

Modular Construction and Standardized Design Review

The shift toward factory fabrication of reactor modules presents both opportunities and regulatory challenges. For CANDU-derived SMRs, the ability to build and test components off-site can reduce construction risk and improve quality assurance. Licensing must account for the fact that a module manufactured in one country may be assembled in another, raising questions about delegation of regulatory oversight. Initiatives like the IAEA’s regulatory coherence project aim to establish mutual recognition of quality assurance programs and inspection results. In the CANDU context, COG has developed standardized procurement specifications aligning with Canadian nuclear codes, and these could be extended to cover modular components. Regulators may issue design certification or standard design approval applying to a generic CANDU‑SMR, with site-specific licensing limited to external hazards and emergency planning. This approach, used for light-water SMRs, could be adapted to heavy-water designs, potentially cutting years from the licensing timeline for a first-of-a-kind unit.

Case Studies: CANDU Licensing Across Borders

China: Qinshan Phase III

The two CANDU 6 units at Qinshan, commissioned in 2002‑2003, were the first CANDU reactors built outside Canada since Wolsong in the 1990s. The licensing process involved China’s National Nuclear Safety Administration applying a combination of IAEA safety standards and its own developing regulatory framework. The experience highlighted the challenge of transferring a Canadian safety design to a country with different seismic hazard profiles, grid codes, and emergency response cultures. A key lesson was the value of a comprehensive, up-front environmental site assessment, as local regulations required a more detailed aquatic ecology study than earlier CANDU projects had undertaken. Despite the learning curve, the two units have operated with strong capacity factors and provided a baseline for further CANDU cooperation in China.

Romania: Cernavodă Units 1 & 2 and Expansion Plans

Romania’s CANDU units at Cernavodă represent a unique example of a nation embracing the technology as a pillar of energy independence. Initial licensing was overseen by the National Commission for Nuclear Activities Control with extensive Canadian expert input. Over time, CNCAN developed indigenous capacity for CANDU safety assessment, and in 2023‑2024 it approved life-extension of Unit 1 to 60 years using a risk-informed review incorporating lessons from Bruce and Darlington refurbishments. Plans for Units 3 and 4, partially constructed since the 1980s, are being revived with potential international financing, and a novel licensing approach is discussed: using the original site license as a basis but updating the safety case to reflect post-Fukushima requirements, including severe accident management guidelines and hardened containment venting.

Argentina: Atucha and Embalse

Argentina’s Atucha I and II, and the Embalse CANDU 6, form part of a diversified nuclear program. Licensing under the Nuclear Regulatory Authority has evolved alongside the country’s changing economic and energy context. Embalse’s life-extension project completed in 2019 replaced all 380 pressure tubes, calandria tubes, and steam generators—a first for Latin America—and yielded a new regulatory precedent: ARN permitted a phased approval that kept the plant online in one channel while replacement proceeded in another, thanks to online refueling design. This success encouraged Argentina to examine new CANDU-derived designs, including an indigenous heavy-water reactor concept for desalination and hydrogen production, which will face licensing scrutiny merging nuclear safety with chemical safety.

India: CANDU-Derived PHWRs and Indigenous Licensing

India’s pressurized heavy-water reactors are based on CANDU technology but evolved independently over decades. Licensing under the Atomic Energy Regulatory Board has included novel features such as thorium‑uranium‑233 fuel in experimental units. India’s experience demonstrates how a CANDU-derived design can adapt to local conditions, including a hot, humid climate and growing need for load-following. The AERB has adopted a risk-informed approach emphasizing in-service inspection and probabilistic safety assessment, similar to the Canadian model. Construction of 700 MWe PHWR units at Kakrapar and Rawatbhata has been licensed with high domestic regulatory oversight, and lessons learned are shared through COG. India’s success in licensing large numbers of PHWRs with high capacity factors offers a valuable example for countries seeking to build heavy-water reactors without relying on foreign vendors for licensing support.

Balancing Deep Decarbonization and Nuclear Safety Goals

The global energy transition demands a dramatic increase in clean firm capacity, and CANDU reactors—existing, refurbished, and new—will likely be part of that portfolio. However, a licensing regime that is too rigid could stall deployment to the point where the climate benefit is lost. Conversely, relaxation of standards could erode public trust and cause setbacks across the entire industry. The path forward lies in a risk-informed, performance-based regulatory model that uses the wealth of operational data from the global CANDU fleet to continuously refine safety analyses. International organizations, particularly the IAEA and the CANDU Owners Group, can accelerate this by creating generic design assessment frameworks accepted by multiple regulators. The Canadian Nuclear Safety Commission’s modernized approach—increasingly transparent, open to digital innovation, and committed to Indigenous engagement—serves as a potential template. But each host country will need to adapt the licensing package to local seismic, environmental, and social conditions.

CANDU’s fuel flexibility is a strategic asset in a world where supply chains are being weaponized and enrichment capacity remains concentrated. A reactor running on indigenous natural uranium, or even spent fuel from other reactors, aligns with energy sovereignty goals that many nations now pursue alongside decarbonization. Licensing authorities must prepare for applications operating with fuel cycles that, while technically feasible, require a new level of regulatory sophistication—chemically complex fuel matrices, higher burnups, and altered waste signatures. The investment in building that competence now will pay dividends when the first CANDU-based SMR or advanced fuel-cycle unit seeks a license.

Regulatory Efficiency and Timely Deployment

One pressing challenge is aligning regulatory timelines with climate targets. A typical new-build licensing process can take five to eight years, and refurbishment projects require three to five years of regulatory review before construction begins. Initiatives like the CNSC’s Vendor Design Review allow preliminary safety assessments before site application, saving time. For CANDU‑SMRs, a pre-licensing design review by multiple regulators—similar to the UK’s Generic Design Assessment—could compress the schedule. The IAEA’s Milestones Approach provides a roadmap for countries new to nuclear power, and CANDU vendors can leverage this to streamline licensing. However, regulators must also be adequately resourced: the complexity of reviewing a full CANDU safety case, with its unique pressure-tube design and online refueling, demands specialized staff. Building regulatory capacity in developing countries will be essential if CANDU technology is to contribute to global decarbonization at scale.

Outlook for CANDU Licensing in a Net‑Zero World

Looking ahead, by 2035 several CANDU units could be in advanced stages of construction or commissioning in nations currently without nuclear power, particularly in Eastern Europe, the Middle East, and parts of Southeast Asia where heavy-water reactor technology offers a workable path to nuclear entry without domestic enrichment capability. Licensing these units will be a litmus test for whether the global institutional architecture can deliver safe approvals on schedules matching climate targets. The IAEA’s milestones approach for new entrants, combined with COG’s pre-competitive information sharing, can lower barriers. However, ultimate responsibility rests with national regulators that must withstand political pressures while projecting long-term confidence in plant safety.

The existing fleet’s refurbishment programs provide another dimension of licensing activity. Bruce Power’s Major Component Replacement and OPG’s Darlington Refurbishment are among the largest energy infrastructure projects in North America. Their licensing iterations—covering environmental assessments, worker safety, novel inspection techniques, and waste minimization—are generating a body of knowledge that could be codified into international guidance. If regulators can move toward a consensus that a refurbished CANDU can be licensed as safely as a new build, the economic case for continued operation will be strengthened, preserving clean generation already paid for.

In the background, emerging reactor designs inheriting CANDU’s heavy-water DNA but scaling down to SMR sizes may transform the licensing conversation from “how do we adapt an old design?” to “how do we build a new, inherently licensable one?” Those designs could offer standardized, sealed core units with long-life fuel, reducing refueling-related accidents and shrinking emergency planning zones. If regulators accept a bounding safety case for a generic heavy-water SMR, site-specific licensing could become a matter of verifying siting parameters and emergency plans—a much faster process. The Canadian SMR Action Plan explicitly aims to achieve timely regulatory reviews, and the results will be watched closely by the global nuclear community.

A Stable Center in a Transforming Grid

As variable renewables expand, the grid increasingly values generation that can maintain frequency, provide inertia, and adjust output without becoming uneconomic. CANDU reactors, once criticized for high fixed costs, are now being re-evaluated as a decarbonization asset delivering exactly those services. Their licensing must not only ensure safety but also enable flexible operation—load-following, coupling with hydrogen production, and stabilizing transmission systems. The technical capacity exists; the regulatory challenge is to write operating rules that permit this flexibility without compromising the safety case originally built for steady-state base-load operation.

CANDU reactor licensing is neither a static relic of the 20th century nor an insurmountable barrier to climate action. It is a dynamic, evidence-based process that, when properly resourced and internationally coordinated, can evolve in tandem with the technology itself. The age of global energy transition demands that licensing authorities embrace risk-informed methods, digital tools, and transparent public engagement. If they do, the CANDU design—in its full-size and small-module variants—will continue to provide clean, reliable power for decades, helping nations meet climate commitments while maintaining the highest standards of nuclear safety.