The Enduring Design Principles of CANDU Reactors

Developed by Atomic Energy of Canada Limited in the 1950s and 1960s, the CANDU reactor represents a distinct philosophy in nuclear fission engineering. The defining choice is the use of heavy water (deuterium oxide, D₂O) as both the moderator and the coolant. This provides an exceptional neutron economy that allows the core to sustain a controlled chain reaction using natural uranium fuel—containing just 0.71% fissile uranium-235. By contrast, the light-water reactors that dominate global nuclear capacity require enrichment to 3–5% U-235, a process that demands costly centrifuge infrastructure and carries proliferation sensitivity. A country choosing CANDU technology bypasses the enrichment step entirely, gaining fuel-cycle independence that many nations find strategically attractive.

The core layout reinforces this flexibility. The CANDU design employs a horizontal pressure-tube configuration. A large, low-pressure cylindrical tank called the calandria holds the heavy-water moderator. Through it run hundreds of horizontal pressure tubes, each containing 12 short fuel bundles. The coolant (heavy water at high pressure) flows through these tubes, transferring heat from the fuel. The moderator in the calandria remains at near-atmospheric pressure and relatively low temperature, acting as a large heat sink. The physical separation of coolant and moderator is a fundamental safety feature: even if a pressure tube fails and coolant is lost, the moderator can absorb significant decay heat from the fuel, preventing core melting. Two independent, fast-acting shutdown systems (one injecting neutron-absorbing rods, the other a liquid poison) provide diverse and redundant reactor protection.

Operationally, the pressure-tube design enables online refuelling. Two robotic refuelling machines, one at each end of a fuel channel, simultaneously insert fresh bundles at one face while pushing spent bundles out the opposite end—all at full reactor power. This eliminates the lengthy refuelling outages that light-water reactors require every 12–24 months. As a result, CANDU units consistently achieve lifetime capacity factors above 80%, often exceeding 90% in well-run stations. Grid operators value this steady, dispatchable power; Ontario's Bruce and Darlington stations routinely run as baseload, supplying roughly 60% of the province's electricity.

Natural uranium fuel also yields a unique fuel bundle design. Each bundle is a compact assembly of 37 zirconium-alloy-clad fuel elements arranged in concentric rings. The bundle geometry and the use of natural uranium mean that the fuel achieves a relatively low burnup—approximately 7–8 MW·d/kg U, compared to 40–50 MW·d/kg for light-water reactors. This leads to more spent fuel bundles per unit of electricity generated, a point we will revisit in the context of waste management.

The Global CANDU Fleet: Refurbishments and a Pause in New Builds

Today about 30 CANDU-derived reactors operate across seven nations. Canada and Romania host the largest fleets, with additional units in China, South Korea, Argentina, Pakistan, and India. India's 16 pressurised heavy-water reactors are not direct Canadian exports but share the same design heritage. The Canadian fleet comprises the multi-unit stations at Bruce (8 units), Darlington (4 units), plus single CANDU-6 units at Point Lepreau (New Brunswick) and Gentilly-2 (Quebec, now permanently shut down).

Canada is executing its most ambitious nuclear refurbishment program in history. Ontario Power Generation (OPG) is completely retubing and overhauling all four Darlington units—a project that will extend their licensed operating lives to approximately 2055–2065. The work replaces reactor core components: all 480 fuel channels, associated feeder pipes, calandria tubes, and key safety systems. It is effectively a reactor rebuild from the inside. The project cost is estimated at C$12.8 billion, a sum OPG justifies through long-term clean electricity contracts that recognise nuclear's zero-carbon baseload value. Bruce Power is pursuing a similar life-extension program for its eight-unit facility, with work staggered across multiple units through the 2020s and 2030s.

Internationally, Romania's Cernavodă plant operates two CANDU-6 units. Unit 1 underwent a refurbishment completed in 2019; Unit 2 is scheduled for similar work in the 2020s. Romania is actively evaluating construction of Units 3 and 4 with international partners, a project that has been discussed for years but remains contingent on financing and political will. No entirely greenfield CANDU construction has started since Cernavodă Unit 2 began commercial operation in 2007. The global market has shifted predominantly toward large light-water reactors (e.g., AP1000, EPR, VVER‑1200) and, increasingly, small modular reactors. For a country without existing heavy-water production capability, building that infrastructure from scratch adds significant capital cost and timeline risk, limiting new CANDU builds to nations with a pre-existing industrial base and trained workforce.

Nuclear Fusion: A Transformative Promise with a Long Timeline

Nuclear fusion generates energy by fusing light atomic nuclei—typically deuterium and tritium—into helium, releasing energy through mass conversion. Recreating the conditions found in the sun’s core requires confining hydrogen plasma at temperatures exceeding 100 million degrees Celsius. The leading approach is magnetic confinement in a tokamak, a doughnut-shaped vessel using powerful magnetic fields. The international ITER project, under construction in Cadarache, France, is the flagship experiment. ITER is designed to produce 500 MW of thermal fusion power from 50 MW of input heating, achieving a Q value of 10—a net energy gain. First plasma is anticipated in the mid-2020s, with full deuterium-tritium operation around 2035. ITER will not produce electricity, but will test key technologies including tritium breeding blanket modules.

Parallel to ITER, a growing number of private companies are pursuing more compact designs. Commonwealth Fusion Systems (CFS) is building SPARC, a tokamak using high-temperature superconducting (HTS) magnets to achieve a much higher magnetic field, allowing a smaller, lower-cost device. SPARC targets Q>2, and if successful, CFS plans a follow-on commercial pilot plant called ARC in the 2030s. TAE Technologies uses a field-reversed configuration, while Tokamak Energy and General Fusion explore spherical tokamaks and magnetised target fusion. These private roadmaps are aggressive, but major engineering hurdles remain: materials must survive intense 14 MeV neutron bombardment; a tritium breeding blanket system that achieves self-sufficiency (tritium breeding ratio >1.1) must be demonstrated; and the power exhaust (handling the plasma's heat flux, which is orders of magnitude higher than in a fission core) must be solved. Commercial fusion electricity is unlikely before the early 2040s under any credible scenario, with wide-scale deployment probably not occurring until the 2050s or later. This long timeline is critical for framing the relationship between fusion and existing fission assets.

The Critical Overlap: CANDU Operations and the Arrival of Fusion

The central strategic question is straightforward: will fusion arrive soon enough to displace CANDU reactors? The answer depends on geography and grid composition. Canadian CANDU units undergoing life extension are licensed to operate until approximately 2055–2065. If the first commercial fusion plants reach the grid in the early 2040s—an optimistic projection—they will initially be high-cost, first-of-a-kind units producing small amounts of power, likely supported by carbon pricing or clean electricity mandates rather than competing on raw market economics. They will not instantaneously replace the 14–15 GW of CANDU capacity in Ontario alone.

Even under an accelerated fusion timeline, the next 25 years will see existing fission plants remain the backbone of low-carbon baseload power in the jurisdictions that host them. Fusion will first complement, not substitute, fission generation. Both technologies provide steady, non-intermittent power that modern grids critically need as they integrate higher shares of variable renewables. In regions like Ontario, where nuclear already supplies about 60% of electricity, CANDU units and early fusion plants can coexist in the same market. The pressing requirement is to maintain CANDU fleet reliability and extend operational life while fusion technology matures, not to accelerate premature phase-out. A premature loss of fission capacity would force reliance on natural gas or costly battery-backed renewables, driving up both emissions and consumer costs.

Could CANDU Infrastructure Support a Fusion Economy?

Rather than becoming stranded assets, elements of CANDU infrastructure might be repurposed to serve a growing fusion industry. The most compelling immediate link is tritium supply. A fusion power plant using deuterium-tritium fuel requires a starting inventory of tritium. While future reactors are designed to breed tritium in a lithium blanket, the first generation of fusion plants—and any operational shortfalls in breeding ratio—will need an external source. CANDU reactors naturally produce tritium as a by-product of neutron capture in the heavy-water moderator and coolant. Tritium production rates at a 700 MW(e) CANDU-6 unit are approximately 120–140 grams per year. Canada already operates extraction facilities at its CANDU stations, and the fleet's ongoing tritium output could provide a strategic bridge supply for early fusion deployment. This would monetise a waste stream that otherwise must be managed and stored, creating a valuable symbiotic link between the two technologies.

A more speculative concept is the fusion-fission hybrid. This uses fusion neutrons to drive a subcritical fission blanket, enabling transmutation of long-lived actinides from fission waste or breeding fissile fuel. While most hybrid ideas focus on fast-spectrum or light-water blankets, academic literature has explored coupling a fusion neutron source to a heavy-water-moderated fission zone. The CANDU lattice's ability to operate with natural or slightly enriched fuel, combined with its flexible fuel-cycle options such as thorium, makes it an interesting candidate. In theory, a fusion source could transmute actinides or breed U-233 from thorium, extending the value of heavy-water infrastructure. However, such concepts remain at a low technology readiness level and face enormous engineering and regulatory hurdles; they are unlikely to influence near-term CANDU decisions but may shape long-range research programs within Canadian Nuclear Laboratories.

Safety, Waste, and Shifting Public Narrative

Public acceptance has always influenced nuclear policy. Fusion energy, with its inherent safety characteristics (no runaway chain reaction, no high-level waste requiring geological disposal) and minimal long-lived waste (mainly activated structural materials with half-lives of decades, not millennia), is often framed as a clean break from fission's perceived risks. As fusion demonstrations draw closer, this narrative could increase political pressure to move away from fission, even if the technical basis for doing so prematurely is unsound. Elected officials may champion fusion while imposing tighter regulatory burdens on existing fission plants, including CANDU units, under the banner of a just transition.

On waste management, CANDU reactors produce more spent fuel per unit of electricity than light-water reactors because natural uranium fuel achieves lower burnup. A typical 700 MW(e) CANDU-6 unit produces about 2,400 spent fuel bundles (approximately 19 tonnes of heavy metal) per year, compared to about 30 tonnes for an equivalent light-water reactor. Canada's long-term waste plan, led by the Nuclear Waste Management Organization, involves a deep geological repository; site selection is expected in the mid-2030s. Fusion's waste profile is simpler, consisting primarily of activated structural materials that can be recycled or disposed of in near-surface facilities after decay. The presence of an operating CANDU fleet for decades means that spent-fuel management will remain a central policy issue. If fusion enters public consciousness as the clean alternative, it could reframe the debate in a way that accelerates calls for fission phase-out, even if fusion cannot yet fill the generation gap.

The counterargument is economic and practical. A grid losing 15 GW of steady nuclear output before fusion is ready would face massive expansion of gas or renewables, driving up both emissions and consumer costs. A responsible strategy involves clearly communicating the timeline mismatch, investing in CANDU life extension, and allowing fusion to prove itself commercially before disconnecting a single fission plant. The IAEA has noted that fission and fusion will likely share the grid for many decades.

Policy, Economics, and the Competitive Landscape

Canada's commitment to net-zero emissions by 2050 shapes the environment in which CANDU reactors operate. Federal and provincial governments view existing nuclear assets as indispensable for meeting interim climate targets. The Darlington refurbishment is backed by favourable financing and long-term clean electricity contracts that recognise the value of a known, low-carbon asset. At the same time, Canada invests in fusion research through Canadian Nuclear Laboratories (CNL) and participates in the ITER project via a collaborative agreement. CNL has also launched a fusion energy program, studying tritium handling, materials science, and waste management for fusion.

Internationally, other CANDU users face different pressures. Romania sees its CANDU units as a pillar of energy security and is actively pursuing new builds, understanding that fusion will not be accessible to a typical middle-income country for many decades. China operates two CANDU-6 reactors at Qinshan, using the experience to advance its broader heavy-water research, while its main nuclear focus is on rapid deployment of PWRs, SMRs, and a domestic fusion program. Argentina is refurbishing its CANDU-based Atucha and Embalse plants, with fusion seen as a distant horizon. In all cases, policy choices are shaped by national security, indigenous fuel cycles, and the immediacy of clean power needs, relegating fusion-induced displacement to a long-term strategic consideration.

Technical Pathways to Extend CANDU Relevance

Even in a future where fusion begins to contribute significant power, advanced CANDU configurations could fill specific roles. Canadian researchers have studied the Advanced CANDU Reactor 1000 (ACR-1000), which uses slightly enriched uranium (1.5–2% U-235) and light-water coolant while retaining a heavy-water moderator. This design reduces heavy-water inventory and uses a compact core with 520 fuel channels, lowering capital costs. Although no orders were placed, the design demonstrated pathways to improved fuel utilisation. An advanced CANDU able to burn thorium or recycled uranium from reprocessed light-water fuel could provide waste management services, acting as a bridge between the current fission fleet and a pure fusion future.

The flexible fuel cycle of CANDU also makes it suitable for consuming surplus plutonium or actinide burning in a scenario where nations wish to reduce fissile inventories. A fusion-fission hybrid, if it matures, would likely rely on a subcritical fission blanket that could be based on heavy-water thermal-spectrum assemblies. While speculative, this underscores the fundamental value of the heavy-water platform: its neutron economy provides options that light-water systems cannot easily replicate. Additionally, CANDU plants can cogenerate high-temperature steam for industrial processes: district heating, desalination, or hydrogen production via thermochemical cycles. In a future grid with substantial fusion capacity, a thermal CANDU plant could be dedicated to process heat, complementing fusion electricity production. Such repurposing requires significant retrofitting and regulatory approval, but the concept has been studied in Canada and could gain relevance as the energy landscape evolves.

Preserving Human Capital for a Nuclear Future

A critical, often overlooked dimension is the nuclear workforce. The skilled engineers, operators, and technicians who run and maintain CANDU reactors today represent a reservoir of operational experience, regulatory knowledge, and safety culture that will be invaluable as new nuclear technologies—including fusion—emerge. According to the Canadian Nuclear Association, the industry employs over 30,000 people directly and supports tens of thousands more. A significant portion of this workforce is approaching retirement age; the loss of institutional knowledge is a real risk. Prematurely dismantling CANDU capabilities would scatter this talent pool and erode the industrial ecosystem that supports nuclear research and regulation.

By contrast, a deliberate strategy that sustains CANDU operations through 2050 and beyond ensures that when fusion plants arrive, there is a cohort of professionals ready to design, build, and operate them. Fusion will not be built in a vacuum; it requires the same rigorous safety culture, materials science expertise, and project management discipline that have been honed in the fission industry. The CANDU fleet can serve as a training ground and technology incubator, preserving institutional knowledge essential for both existing reactor safety and future fusion development. Programs like the Nuclear Engineering and Reactor Physics course at Ontario Tech University and the CANDU-specific training at the McMaster Nuclear Reactor provide the pipeline. Sustaining this human capital is arguably the most important long-term contribution the CANDU fleet can make to the fusion horizon.

Conclusion: A Managed Transition

The future of CANDU reactors in the context of fusion is not a binary choice between fission and fusion, but a managed transition spanning several decades. CANDU life-extension projects are sound investments that secure low-carbon baseload power through the critical window when fusion technology is still maturing. Meanwhile, the heavy-water infrastructure offers unique assets—tritium production, flexible fuel cycles, and a highly skilled workforce—that can support the early fusion economy. Energy planners should resist calls to prematurely decommission existing CANDU capacity in favour of a fusion promise that remains years away from commercial demonstration. Instead, a prudent strategy combines sustained CANDU operation with steady fusion R&D funding, allowing the two technologies to coexist and eventually complement each other on the path to a fully decarbonized electricity system.