The Distinctive Engineering Challenges of CANDU Reactors

To grasp the complexities of decommissioning a CANDU reactor, one must first understand its unique design. Unlike light‑water reactors, CANDU units use a horizontal calandria—a large cylindrical tank—that contains hundreds of zirconium‑alloy pressure tubes. These tubes hold natural uranium fuel bundles and are surrounded by a separate heavy‑water moderator system. The ability to refuel while operating means the calandria and its internals are exposed to prolonged neutron flux, leading to significant activation. Heavy water in both the primary heat transport and moderator systems generates high concentrations of tritium, a radioactive isotope that complicates contamination management. Furthermore, the scale of a typical CANDU calandria—over seven meters in diameter and length—combined with intricate heavy‑water piping, demands specialized remote handling, precise segmentation strategies, and bespoke tooling rather than standard cut‑and‑ship approaches.

Key Decommissioning Hurdles

Spent Fuel and High‑Activity Waste Management

While spent fuel is usually removed before full decommissioning begins, managing residual fuel inventory and any on‑site storage remains critical. CANDU reactors produce a higher volume of spent fuel bundles per unit of energy than enriched‑uranium reactors because natural uranium achieves lower burnup. Each bundle is small, simplifying handling but increasing the count of elements requiring long‑term oversight. At many CANDU stations, spent fuel has been transferred to dry storage canisters or concrete modules. Decommissioning plans must ensure these systems remain secure and allow eventual transfer to a deep geological repository. The Canadian Nuclear Safety Commission (CNSC) mandates that all radioactive waste, including spent fuel, be managed to protect health and safety for generations. The Nuclear Waste Management Organization (NWMO) continues site selection for a repository, with a decision expected in the late 2020s, adding timing uncertainty to every decommissioning schedule.

Radioactive Contamination of Core Components

Over decades of operation, internal surfaces of the primary heat transport system—pumps, steam generators, piping—accumulate activated corrosion products and fission product deposits. The calandria, pressure tubes, and end fittings become heavily contaminated and activated. Decontamination before disassembly is essential to lower worker dose rates and enable proper waste classification. Traditional chemical flushing with acids or oxidizing agents is effective but generates large secondary liquid waste volumes. Mechanical methods like grit blasting create airborne radioactive dust that requires robust containment. Each facility’s operational history dictates the optimal balance. For example, Bruce A’s primary system has unique corrosion product compositions due to extended high‑temperature operation, necessitating site‑specific protocols.

Tritiated Heavy Water Handling

Tritiated heavy water is perhaps the most distinctive complication in CANDU decommissioning. Neutron capture in deuterium produces tritium throughout the moderator and heat transport systems. After shutdown, this heavy water must be drained and residual moisture removed to prevent airborne tritium release during cutting. Facilities like Ontario Power Generation’s Tritium Removal Facility can detritiate heavy water, recovering clean deuterium and immobilizing tritium. However, not all sites have on‑site capabilities, and transporting tritiated water presents logistical safety challenges. Residual tritium absorbed in metal surfaces can off‑gas, so crews must use continuous air monitoring and local exhaust ventilation. At Pickering A, engineers developed a heated nitrogen purging system that strips tritium from piping before cutting, reducing airborne levels by over 95%.

Volume and Diversity of Low‑ and Intermediate‑Level Waste

CANDU decommissioning generates a vast array of low‑ and intermediate‑level wastes: concrete biological shielding, piping, insulation, cables, and structural steel. Although much material has only surface contamination, the sheer volume creates logistical and financial burdens. Non‑radiological hazards like asbestos insulation or lead‑based paint in older stations add complexity. Operators implement aggressive waste minimization programs—decontamination for free release, metal melting and recycling, supercompaction—to divert material from repositories. At the Douglas Point prototype, over 80% of concrete was reused as clean fill or crushed for aggregate after rigorous radiological surveys.

Activated Structural Materials: Calandria and Pressure Tubes

The calandria and pressure tubes experience the highest neutron flux and become among the most activated components. After 30–40 years of operation, Zircaloy‑2 pressure tubes and the stainless steel calandria shell contain long‑lived radionuclides like carbon‑14, cobalt‑60, and nickel‑63. Removing these requires remote cutting in heavily shielded environments. Strategies include plasma arc cutting, abrasive water jet, or mechanical saws operated by robotic manipulators. Abrasive water jets produce minimal dross and no heat‑affected zone, ideal for thin‑walled tubes; plasma arc is faster for thick calandria shells but generates fumes requiring extraction.

Regulatory and Cost Uncertainties

Decommissioning in Canada operates under a comprehensive CNSC framework requiring detailed plans, environmental impact statements, and financial guarantees. Public hearings and Indigenous engagement add transparency but also potential delays. Cost estimates for multi‑unit CANDU stations can reach billions of dollars over decades. Inaccurate forecasting of waste volumes, contamination levels, or regulatory timelines quickly inflates costs. The national policy for radioactive waste management, including repository siting, directly affects economic viability. As of 2024, NWMO continues site selection; until that facility is operational, some waste streams remain in interim storage, prolonging the care‑and‑maintenance period. The CNSC now requires probabilistic modeling for cost estimates, an industry‑wide best practice.

Innovative Solutions and Proven Practices

Advanced Decontamination Technologies

Modern systems move beyond single‑stage chemical flushes toward multi‑cycle, closed‑loop processes tailored to specific radionuclide compositions. The CAN‑DECON process uses a dilute regenerative chemical solution that selectively dissolves contaminated oxide layers without aggressive base‑metal attack, minimizing corrosion and secondary waste while achieving dose rate reductions of 10 to 30 times. Foam‑based or gel‑based agents cling to vertical walls and complex geometries, reducing liquid waste volume. Electrochemical techniques with controlled current strip contamination precisely. At Bruce B, a foam decontamination trial on steam generator channel heads achieved a 97% dose rate reduction while generating less than 200 liters of liquid waste.

Robotics and Remote Handling

Robotics are transforming CANDU decommissioning. Custom‑designed robotic arms and crawlers enter the reactor vault to cut and package components under remote surveillance. For instance, the Gentilly‑1 calandria segmentation used a remotely operated manipulator with plasma and abrasive water jet cutting heads. Robots navigate cluttered spaces, cut large components, and place them into shielded containers. Force feedback, advanced vision algorithms, and automated guidance reduce cutting times and improve precision. Data collected feeds 3D digital twins, enabling planners to simulate and refine workflows. Snake‑arm robots reaching tight spaces between feeder pipes allow cutting without removing large adjacent sections.

Waste Minimization and Recycling

Operators increasingly apply circular economy principles. Surface‑contaminated metals are sent to licensed recycling facilities where radioactive slag is separated, leaving clean metal for reuse. Concrete from biological shields is crushed, surveyed, and used as backfill or non‑structural aggregate. Steel and copper cabling are similarly processed. The CNSC sets clearance levels for materials to be released from regulatory control; modern monitoring portals enable rapid screening of large volumes. Many projects achieve diversion rates exceeding 90% for bulk materials. OPG’s Western Waste Management Facility processes decommissioning metals, with melted steel ingots sold for shielding blocks, demonstrating closed‑loop operations.

Dry Storage for Spent Fuel

Although spent fuel is not a decommissioning waste per se, its safe interim storage is a prerequisite for dismantling. CANDU stations have adopted robust dry storage systems with passive cooling. Modular above‑ground concrete containers—OPG’s Dry Storage Containers or NB Power’s CANSTOR modules—seal spent fuel bundles in welded steel baskets. These systems are licensed for decades, bridging the gap until a permanent repository is available. Dual‑purpose casks rated for storage and eventual transportation simplify final disposal logistics. Transition to dry storage at nearly all defueled CANDUs reduces site costs and frees spent fuel bays for other tasks. At Pickering, draining the wet storage bay after fuel transfer allowed removal of the liners, which were significant sources of occupational dose.

Integrated Decommissioning Strategies

Successful decommissioning relies on selecting optimal timing between immediate dismantling and deferred dismantling (SAFSTOR). Many CANDU operators choose a deferred approach—placing the facility in safe storage for 30–50 years to allow shorter‑lived radionuclides to decay, reducing source terms and worker doses. This strategy was used for Pickering A units A2 and A3 and is planned for others. During safe storage, the reactor is defueled, fluids drained, and systems stabilized; active monitoring continues at a fraction of the operating cost. When dismantled later, reduced radiation fields allow simpler techniques and lower waste costs. Life‑cycle modeling with probabilistic risk assessments informs unit‑by‑unit decisions. Gentilly‑1’s deferred strategy is projected to have saved over $200 million compared to immediate dismantling.

International Collaboration and Knowledge Sharing

Canada shares CANDU decommissioning lessons with countries like South Korea (Wolsong), Romania (Cernavodă), and Argentina (Embalse). IAEA decommissioning projects and bilateral agreements accelerate tool development. Experience at Douglas Point—one of the first CANDUs dismantled—provided data on tritium behavior, waste characterization, and remote cutting shared internationally. Joint research on laser cutting, chemical decontamination of heavy water piping, and robotic assembly prevents costly trial‑and‑error. Recent bilateral projects between Canada and South Korea focus on modeling long‑term corrosion of pressure tubes in storage, a key input for safe storage duration decisions.

Case Study: Gentilly‑1 Decommissioning Progress

Gentilly‑1 in Bécancour, Quebec, a 250‑MWe prototype, operated briefly from 1971 to 1979, but its decommissioning journey spans decades. Fuel and heavy water were removed in the 1980s, and the station entered monitored safe storage. In recent years, Hydro‑Québec and its contractor undertook detailed dismantling of the calandria using remotely operated cutting tools. Workers entered via an engineered access platform after extensive environmental remediation and air‑quality control. Hydro‑Québec reports indicate physical dismantling of the reactor core is on track for completion this decade, yielding valuable data on waste volumes, actual dose uptake, and cost performance. The project validated deferred dismantling after long safe storage and informed plans for larger stations like Pickering and Bruce. The worker collective dose was less than 10% of the original regulatory projection, thanks to extended cooling and advanced robotics.

Emerging Technologies and Future Directions

Next‑generation CANDU decommissioning will likely incorporate laser‑based cutting and in‑situ spectroscopy for material characterization. Laser cutting offers narrow kerfs, reduced secondary waste, and contact‑free operation. Researchers at the University of Ontario Institute of Technology demonstrated a fiber‑laser system cutting activated calandria tube sheets at speeds comparable to plasma arc, with no airborne contamination due to localized heating. Autonomous drones with gamma spectrometers are being tested for rapid radiological mapping, reducing pre‑entry surveys. Digital twin technology fed by real‑time sensor data enables dynamic optimization of cutting sequences to balance dose minimization with packaging efficiency. Machine learning for waste classification promises further cost and timeline reductions. The CNSC has issued guidance on autonomous systems in decommissioning, paving the way for broader adoption.

The Path Forward: Sustainable Decommissioning

As Canada manages the end‑of‑life phase of its CANDU fleet, a coordinated national strategy is emerging. New regulatory documents like CNSC REGDOC‑2.11.1 clarify requirements for cost estimates, financial guarantees, and environmental assessments. Advances in waste processing—including a potential near‑surface disposal facility for low‑level waste—reduce reliance on on‑site storage. Practical experience from early projects is being codified into best‑practice guides by the Canadian Nuclear Association and the University Network of Excellence in Nuclear Engineering. Indigenous and local communities are engaged earlier, ensuring cultural and environmental concerns are addressed. The ultimate vision is a systematic, transparent process that returns sites to greenfield status or repurposes them for new energy infrastructure, demonstrating that the nuclear industry can responsibly close the fuel cycle.

Decommissioning CANDU reactors is a nuanced undertaking requiring specialized techniques for heavy water, tritium, multiple fuel channels, and activated calandria. Yet through remote robotics, advanced decontamination chemistry, intelligent waste segregation, and patient deferred dismantling, operators are steadily reducing hazards and uncertainties. Lessons from Gentilly‑1, Pickering, and Douglas Point shape a new generation of standards applied across Canada and the global CANDU family. As technology evolves, future projects will benefit from greater efficiency and lower doses. Successful decommissioning is not only an engineering challenge—it is a commitment to environmental stewardship and intergenerational equity, closing the nuclear energy lifecycle with the same rigor that marked its operation.