The Canadian supercritical water-cooled nuclear reactor (SCWR) is a 2540 MWth channel-type SCWR concept that employs 336 fuel channels in the reactor core. Each fuel channel includes a pressure tube that is submerged in a heavy water moderator and contains a removable fuel assembly. The fuel assembly is designed so that all in-core components exposed to high radiation fields (other than the pressure tube) are part of the fuel assembly, which is removed from the reactor core as part of the assembly after three operating cycles. This design feature significantly reduces the likelihood of component failures due to radiation damage. To achieve high (>45%) power conversion efficiency, the Canadian SCWR operates at a supercritical water pressure (25 MPa) and high temperatures (350 °C at the inlet, 625 °C at the outlet). These conditions lead to fuel cladding temperatures close to 800 °C. Because of the reduced material strength at this temperature and higher fission gas production of the fuel, collapsible fuel cladding is selected over internally pressurized cladding. To increase heat transfer and to reduce cladding temperatures, turbulence-inducing wire-wraps are employed on fuel elements. Numerical models have been developed to analyze the thermal-structural behavior of Canadian SCWR fuel at normal and accident conditions. It was found that axial ridging, a possible failure mechanism with collapsed fuel cladding, can be avoided if the cladding thickness is larger than 0.4 mm. Detailed numerical analysis showed that the maximum fuel cladding temperature for the worst-case accident scenario is below the melting point by a small margin. This result was obtained with conservative assumptions, suggesting that the actual margin is greater. Hence, one of the design goals, the exclusion of the possibility of melting of the fuel, which is called the “no-core-melt” concept, seems attainable. However, this needs to be demonstrated more rigorously by removing the conservative assumptions in the analysis and performing supporting experimental work. This paper presents a description of the Canadian SCWR fuel assembly concept, its unique features, the rationale used in the concept development and the results of various numerical analyses demonstrating the performance and characteristics of the Canadian SCWR fuel channel.
The supercritical water-cooled reactor (SCWR) is one of the six Generation IV reactor technologies pursued by the Generation IV International Forum countries . Currently, four Generation IV International Forum members are signatories for the SCWR reactor: Japan, the European Union, Russia, and Canada. In addition, China is actively working on developing its own SCWR reactor. Preconceptual designs of pressure-vessel-type SCWRs have been developed by Japan  and the European Union , while Canada has developed a pressure-tube-type SCWR [4–6].
The basic premise of the SCWR is the increased efficiency in power conversion as a result of the increased thermodynamic efficiency attainable at supercritical water conditions, elimination of large nuclear components (such as steam generators or steam separators), smaller reactor, and turbine buildings and process simplifications due to the single-phase behavior of the supercritical water coolant.
Canadian Supercritical Water-Cooled Reactor
The Canadian SCWR core concept, safety systems, and the plant layout are documented in the previous publications [4–6]. The reactor core design concept is illustrated in Fig. 1. It consists of a pressurized inlet plenum, a low-pressure calandria vessel that contains a heavy water moderator and pressure-tube fuel channels that are attached to a common outlet header. Fuel channels are connected to the inlet plenum at the tubesheet and extend into the calandria vessel where they are immersed in the low-pressure heavy water moderator. The reactor core is oriented vertically for ease of batch refueling. The coolant flows into the inlet plenum, around the outside of the outlet header and enters into the fuel assembly. Within the fuel assembly, the coolant is directed to the lower end of the core through a central flow tube, is reversed to return vertically through the fuel elements and discharged through the outlet header.
The reactor core includes 336 fuel channels, generating 2540 MW of thermal power. The number of fuel channels is selected to be a multiple of 12 so that quarter symmetry is maintained in a three-position fuel reshuffling arrangement. Each fuel channel includes a pressure tube that is submerged in heavy water and contains a removable fuel assembly. The fuel assembly includes a 64-element two-ring fuel bundle, a central flow tube, a ceramic insulator encapsulated by an inner liner and an outer liner, and a locking mechanism that allows quick installation and removal of the fuel assembly. Schematics of the fuel channel and the fuel assembly are shown in Fig. 2. In this figure, the length of the channel is shortened to illustrate the details.
As shown in Fig. 2(a), the fuel channel consists of a pressure tube connected to the tubesheet in a leak-tight fashion, a guide tube (pressure tube extension) that extends the fuel channel into the inlet plenum, and an expansion bellows that connects the guide tube to the outlet header. The penetration holes in the tubesheet and at the outlet header were chosen such that the removal of all fuel channel components is possible. This feature allows a single fuel channel replacement, as well as a complete core replacement.
The fuel assembly, shown in Fig. 2(b), includes all in-core components exposed to high radiation fields. It remains in the reactor core for a total of three operating cycles. Figure 2(c) shows the installed fuel assembly in the fuel channel. There is an annular gap between the fuel assembly and the pressure tube to accommodate manufacturing tolerances and to facilitate installation and removal of the fuel assembly. This gap is a critical feature of the Canadian SCWR fuel channel concept for meeting the no-core-melt requirement and is optimized to minimize heat conductance during normal operating conditions. Under accident conditions, as the fuel assembly expands due to rising temperatures and the gap size becomes smaller, core decay heat can be transferred to the moderator more effectively.
Pressure Tube and Pressure Tube Extension.
The pressure tube forms the fuel channel pressure boundary and contains the supercritical water coolant pressure at 25 MPa. Pressure tubes are submerged in the heavy water moderator that is contained by the calandria vessel. At normal conditions, the heavy water moderator is maintained at about 100 °C and 0.35 MPa(a), resulting in pressure tube temperatures of approximately 120 °C and 150 °C at the outside surface and inside surface, respectively.
The selected pressure tube material is Excel, a zirconium alloy developed by Atomic Energy of Canada Limited in the 1970s. Excel has high strength and high creep resistance at this operating temperature, in addition to having a small neutron absorption cross section . The pressure tube is connected to the tubesheet via a welded joint to ensure leak-tightness. The tubesheet is made from SA508 (a stainless steel (SS) alloy). It is known that zirconium alloys cannot be reliably welded to stainless steel components; hence, the top section of the pressure tube is transitioned to a stainless steel alloy that can be welded to the SA508 tubesheet. The Excel and SS sections of the pressure tube are joined in a continuous manner either by co-extrusion (that results in molecular diffusion of two materials at the joint line) or by an additive manufacturing method such as laser consolidation. Past work on co-extrusion of Zirconium alloy and SS provided insights that there is a good chance of success with this process. Laser consolidation is new—some of the preliminary tests indicated that a transition material, such as Titanium, may be needed to make the process work. With both processes, extensive qualification testing (tensile, toughness, fatigue, creep, and irradiation) would be required before the pressure tube can be used as proposed. The transition zone is more than 0.5 m above the top of the fuel line, and hence, is not exposed to high neutron flux. Above the tubesheet, a guide tube (also called the pressure tube extension) and an expansion bellows extend from the tubesheet to the outlet header to guide the fuel assembly into the pressure tube. The guide tube has inlet ports that admit flow into the fuel assembly. The inlet ports on the guide tube are sized to match the channel flow rate to the channel power so that the channel outlet temperatures across the reactor core are reasonably uniform. Expansion bellows provide axial and radial flexibility that is needed to accommodate the differential thermal expansion of the inlet plenum and the outlet header. The pressure tube extension components are not part of the pressure boundary and are outside the high neutron fields. Hence, their design requirements are less demanding.
The fuel assembly, shown schematically in Fig. 2(b), includes a fuel bundle, a central flow tube, and a ceramic insulator encapsulated by inner and outer liners. Above the core, the assembly includes a cross-over piece that admits inlet flow (blue arrows) into the central flow tube while allowing outlet flow (red arrows) to the outlet header via a double-walled vertical chimney, and a seal that isolates the outlet flow from the inlet flow. Figure 3 shows the solid model of the fuel assembly with cut-outs at various locations as well as the paths of inlet and outlet flow streams.
The fuel assembly includes a ceramic insulator that is mounted in the fuel assembly, separated from the fuel by the inner liner and encapsulated by the outer liner, as shown in Fig. 4. Together, the fuel-bundle, liner tubes, and the insulator are the in-core components of the fuel assembly and are replaced or shuffled as a whole during refueling activities.
The fuel bundle consists of 64 fuel elements arranged in a two-ring configuration, with 32 fuel elements in each ring, laid out circumferentially around the central flow channel. The size of the flow channel, ring, and fuel element diameters were determined through an iterative solution process of physics and thermalhydraulics analyses. Figure 4 shows a section of the fuel channel bottom showing the fuel bundle details and associated fuel assembly components. A cross-sectional schematic of the fuel bundle and other fuel channel components are shown in Fig. 5.
The fuel for the Canadian SCWR concept is similar to existing typical water-cooled reactor fuels in that a ceramic pellet produces heat, which is transferred through the metallic cladding to the primary coolant. The fuel element, shown in Fig. 6, consists of a 6.5 m long cladding tube housing the fuel pellets, a gas plenum at the upper end of the fuel stack to accommodate fission gas release, an inner filler tube in the plenum area to prevent collapse under external pressure, and a spring inside the gas annulus inner tube to hold the pellets in place while allowing pellet expansion. Each end of the fuel element is enclosed and sealed with an end plug, which is welded to the cladding tube. At normal conditions, the supercritical water (SCW) coolant pressure on the gas plenum is supported by an inner tube filled with fully stabilized zirconia (ceramic) “frit” pellets with 50% void space. In keeping with the collapsible cladding concept, the Canadian SCWR fuel concept adopts the “standard” CANDU®2 type pellet configuration . Pellets are manufactured from powder, sintered to high density, double-dished and chamfered. High density pellets, with 2–5% void, negate problems associated with in-reactor sintering (shrinkage), double-dishes negate the problems associated with axial expansion stresses due to radial variations in pellet thermal expansion, and the chamfers avoid problems with pellet-end chipping, ease pellet loading and ensure pellet axial expansion is transferred via the (cooler) periphery of the fuel. A critical factor is the diametral clearance tolerance to ensure fuel cladding collapse without longitudinal ridging. The diametral clearance of 0.03–0.1 mm has been proven achievable in the fabrication of CANDU fuel and, because the tubes used for the CANDU fuel cladding are fabricated in long (6 m) lengths and then cut to the required 50 cm length, it is reasonable to assume the same tolerance can be achieved in this fuel-assembly concept.
Calculations have been performed to assess the buildup of internal gas pressure using the conservative assumptions of 40 MWd/kgHE3 burnup and 75% release of total gas (mostly xenon and krypton) production at a gas temperature of 1000 °C. The analysis results suggest that a 0.5 mm thick cladding material with yield strength of 50 MPa at a worse-case accident temperature will not experience cladding rupture in a sudden depressurization event.
To increase heat transfer and to reduce cladding temperature, turbulence-inducing wire-wraps are considered. The wire-wraps are also used as spacers between fuel rods. With heat transfer enhancing wire-wraps, the cladding temperature is expected to be closer to 750 °C. A feature of the current Canadian SCWR fuel element concept is the adoption of “CANLUB” (colloidal graphite) coating of the internal surface of the cladding [9,10]. The CANLUB coating of standard CANDU PT-HWR fuel cladding has been proven to provide additional margin against internal stress-corrosion cracking. While the mechanism of this protection is not clearly understood, the most popular theories involve either CANLUB acting as a “getter” for volatile corrosive fission products or providing a physical barrier between the fuel pellet and the cladding, protecting the cladding from fission fragment damage.
The high outlet temperatures of the Canadian SCWR, combined with the reduced heat-transfer coefficient of SCW (compared to water), result in high fuel cladding temperatures compared with current typical reactor fuels. Fuel cladding temperatures close to 800 °C are predicted for the proposed assembly without considering any heat-transfer enhancements, such as wire-wrapped spacers . At this temperature, material strengths of most candidate materials are significantly less than at the typical cladding temperatures of 400 °C found in existing water-cooled reactors. Hence, internally pressurized fuel cladding is not a feasible option without a very thick cladding material that would result in significant neutron absorption and loss of fuel utilization. However, collapsible fuel cladding, as used in CANDU reactors, is a viable option. Numerical models developed to model the thermal-structural behavior of Canadian SCWR fuel indicate that cladding thickness as small as 0.4 mm is feasible at the Canadian SCWR operating conditions, if a Ni-based cladding material (Alloy 625 or Alloy 800 H) is used . Work is ongoing in this area to identify the most appropriate cladding material and cladding thickness .
Central Flow Tube.
The central flow tube shown in Figs. 4 and 5 admits the inlet flow stream to the bottom of the fuel channel before the flow stream reverses direction and enters the fuel bundle. This flow path is selected over the alternative flow path through a peripheral annulus enveloping the fuel bundle due to its enhanced safety and its considerable neutronics advantage. From a safety perspective, the no-core-melt safety goal of the Canadian SCWR (see “No-core-melt” section) would not likely be met with the peripheral downward inlet flow. This is because during a depressurization transient following a large primary pipe break, the annulus could be filled with steam, hindering the fuel cooling mechanism that relies on transferring residual decay fuel heat to the moderator. From a neutronics perspective, the central inlet flow provides significant neutron moderation, resulting in increased power produced at the central pins as well as a large negative coefficient of void reactivity (CVR), which is about −30 mk. As a result of this, the maximum achievable exit burnup is increased, and a highly desirable negative CVR value, which causes the reactor to reduce power with sudden voiding, is achieved. The CVR value will be optimized at the detailed design stage by changing (reducing) the size of the central flow channel.
A ceramic insulator is used as a part of the fuel assembly to reduce heat losses to the moderator and to shield the pressure tube from the high-temperature coolant. The insulator material is selected to be yttria-stabilized zirconia that has low neutron absorption, good thermal resistance, moderate mechanical stability under neutron irradiation and a low corrosion rate at SCW temperatures. To reduce thermal stresses caused by the temperature difference between the inner surface and the outer surface (and the resultant differential growth), the insulator is circumferentially segmented into three pieces. This arrangement allows the insulator pieces to freely expand with the thermally expanding inner liner and, thus, results in significantly reduced mechanical stresses. A similar approach is adopted in the axial direction, where the insulator is segmented in the axial direction to reduce thermal stresses and accommodate the differential axial thermal expansion of the insulator and other metallic fuel assembly components. In addition, the insulator segments are encapsulated by an inner liner and an outer liner so that, in the unlikely scenario of insulator cracking, the insulator pieces will be retained by the liners.
Figure 7 depicts the exploded view of the bottom of the fuel assembly showing the various insulator segments. Canmet-Materials of Hamilton, ON, Canada, has a patented technology  for the fabrication of porous ceramic materials that can be formed into complex shapes. Canmet-Materials can also manufacture denser yttria-stabilized zirconia components, which are used in the Canadian SCWR fuel assembly.
Inlet Cross-Over Piece.
The inlet cross-over piece is an integral part of the fuel assembly. The inlet flow stream enters the fuel assembly first through inlet ports on the pressure tube extension (see Fig. 2(a), and then, through the cross-over piece (see Fig. 2(b)). Four connecting tubes attach the fuel assembly liner (outer cylindrical structure) to the central flow tube. While the cross-over piece directs the inlet flow stream into the central flow tube, it also allows the hot outlet flow stream to flow in the opposite direction, around the connecting tubes. These connecting tubes see the inlet flow stream at 350 °C on the inside surface and the outlet flow stream at 625 °C on the outside surface, and hence, are subject to a considerable thermal stress. To evaluate these thermal stresses, a finite element analysis of this component was performed using the ANSYS®4 software. Figure 8 shows the results of this analysis. A peak (von Mises) stress value of 214 MPa was calculated. This value is less than 2/3rd of the material yield stress of Alloy 800 H at 600 °C, which is 415 MPa. It was found that stresses developed due to the large thermal gradients across component walls and, also, due to the differential thermal expansion of the inner liner (which forms the larger diameter peripheral structure of the cross-over piece) and the central flow tube. Hence, these stresses can be reduced to a certain extent by using more flexible connecting tubes and physically separating the inlet and outlet streams using a double-wall construction in regions of high thermal gradients. As an alternative proposal, a novel flow exchange piece is proposed. This concept is shown in Fig. 9. In this concept, inlet and outlet streams are completely decoupled from each other with gaps between flow passages. These gaps are filled with an insulating material, preventing the formation of large thermal gradients across the walls of flow passages. This component can be manufactured by either hydroforming or through an additive manufacturing process.
Fuel Assembly Seal.
The fuel assembly is supported and retained in place by a locking mechanism located at the upper end of the pressure tube extension. As illustrated in Fig. 2(c), the support also incorporates a seal, which prevents the mixing of inlet and outlet flow streams. The seal is a metallic “E” ring incorporated in the fuel assembly support. This configuration requires low forces in order to activate the seal, which is provided by the weight of the fuel assembly as well as a spring incorporated in the locking mechanism. The pressure difference across this seal is less than 100 kPa, which is determined by the pressure drop of the coolant in the fuel channel.
The “no-core-melt” feature is one of the important design goals of the Canadian SCWR. This is an attainable goal because of the submergence of the fuel channels in the heavy water moderator and the use of the moderator as a cooling system. In fact, two cooling systems are implemented in the Canadian SCWR. At normal operating conditions, the moderator is circulated by an active moderator cooling system (AMCS) that is capable of removing all the heat (gamma heating and heat gained through conduction/convection from fuel channels) transferred to the moderator and maintaining the moderator at a substantially subcooled state. In addition to the AMCS, there is also a passive moderator cooling system (PMCS) connected to the moderator on a standby state. The PMCS is a passive system with no valves or pumps, requiring an increase in moderator temperature to activate. It is composed of inlet and outlet lines connected to the calandria vessel, a set of condensers immersed in the reserve water pool located above the calandria and a head tank as described in references  and . At normal conditions there is very little driving force, which is determined by the density difference of the fluid at the cold and hot legs and elevation difference between the heat source (reactor core) and the sink (condensers), to circulate flow on the PMCS. Hence, the PMCS does very little to remove moderator heat. In a total loss of station power (station blackout) and the loss of AMCS functionality, moderator temperature rises steadily to the boiling temperature. The resultant density change caused by boiling promotes a vigorous flow of the moderator through PMCS lines and automatically activates the PMCS cooling system. The activation of the PMCS system, as well as the transfer of decay heat from fuel elements to the moderator, provides a means to cool the fuel elements and makes the no-core-melt goal a possibility. A numerical model of the fuel assembly, shown in Fig. 10, was developed using ANSYS to evaluate the consequence of a worst-case accident scenario, where complete loss of all coolant-side safety systems, sudden loss of all light-water coolant (LOCA), and loss of all station power were modelled. A two-dimensional approximation for the thermal-structural ANSYS model was considered since the heat transfer and deformation in the circumferential and radial direction are assumed to be more dominant than those in the axial direction (fuel assembly is free to expand in the axial direction). The 64-element fuel bundle with two rings of fuel elements, a center flow tube, inner-ring fuel elements, outer-ring fuel elements, an inner liner tube, segmented insulator, outer liner tube, and pressure tube are included in the analysis. In LOCA scenarios, the assembly is assumed to be void of coolant, and the fuel channel is heated by the fuel elements primarily through radiation. In the fuel channel, radial gaps were modeled between the inner liner tube and the insulator, between the insulator and the outer liner tube, and between the outer liner tube and the pressure tube. The contact modeling and detection capability in ANSYS was used to evaluate the interactions between various component when they expand and deflect with temperature.
In the simulation, heat was transferred by conduction from the fuel pellet to the fuel cladding (perfect contact between the fuel pellet and the cladding was assumed), and heat was transfer by radiation from the fuel elements to the inner liner tube (instant voiding of coolant was assumed). Heat was transferred by conduction within the inner liner tube, the insulator, the outer liner tube, the pressure tube, and through gaps/contacts between these components. Between the fuel channel components, heat was transferred by both radiation and the conduction in the radial gaps. A contact heat transfer model was used at locations where components are detected to be in contact. An atmospheric internal channel pressure was applied on both inside and outside surfaces of the inner liner tube, the insulator, the outer liner tube, and the inside surface of the pressure tube. Finally, convective heat transfer from the outside surface of the pressure tube to the heavy water moderator is modeled. In order to avoid rigid body motion, displacement constraints on the fuel channel components were specified on selected nodes.
It was assumed that the shutdown systems successfully shut down the reactor, and the evaporation of the coolant and heat-up of the moderator from 100 °C to full PMCS activation took about 1 min, during which the reactor decay heat was reduced to 3% of full power. Calculations were performed for a constant heat generation of 3% (in reality, the decay heat reduces exponentially as a function of time). As a baseline case and for comparison purposes, calculations were conducted for an unsegmented insulator. This baseline case is axisymmetric and lends itself to an analytical solution using a one-dimensional fluid network software cathena (Canadian Algorithm for Thermal Hydraulic Network Analysis). Therefore, it was also used for the verification of the numerical model. The predicted temperature distribution from the fluid network (CATHENA) and numerical (ANSYS) models of the test case are shown in Fig. 11. The difference in fuel cladding temperature is due to the simplifying assumption in the analytical model that the cladding temperature does not vary in the circumferential direction. Hence, the numerical model should be considered more accurate.
Under LOCA conditions, the analysis results show that the radial temperature profile from the inner liner to the outside of the pressure tube ranged from 453 °C to 133 °C for the average power case and from 598 °C to 143 °C for the highest channel power case. The fuel channel deformed due to thermal expansion, and the size of the gaps varied circumferentially. For the modeled gap sizes in this simulation, the inner liner tube expanded more than other components and pushed the insulator segments toward the outer liner. As a result, Gap 1 is mostly closed, while Gap 2 is partially closed. With a segmented insulator, a significant reduction in stresses of the inner liner tube and the insulator were obtained as compared to the nonsegmented insulator: the maximum stress of the inner liner tube was reduced from 262 MPa to 121 MPa, while the maximum stress of the insulator was reduced from 197 MPa to 60 MPa. These numbers are significantly smaller than the yield stresses for the liner tube (alloy 800 H) and the insulator (Zirconia), which are ∼415 MPa and greater than 500 MPa at 600 °C, respectively. The fuel sheath temperature was between 973 °C and 1194 °C for the average power channel, which remains below its melting temperature (between 1357 °C and 1385 °C) with some margin. Meanwhile, the fuel sheath temperatures were between 1062 °C and 1340 °C for the highest power channel, which still remains below its melting temperature, but with a small margin. Hence, a more refined analysis with reduced conservatism is needed to better understand the safety margin. Conservative assumptions can be eliminated by taking credit for the cooling effect of water evaporation during LOCA depressurization and by modeling the exponentially decaying behavior of the residual (decay) heat.
This paper presented a description of the Canadian SCWR fuel assembly concept and the results of various numerical analyses demonstrating the performance and characteristics of the Canadian SCWR fuel channel. Some of the important features of the Canadian SCWR fuel assembly can be summarized as below:
Negative CVR: Incorporation of the central flow tube, as a conduit for the higher density inlet flow, increases neutron moderation and results in a large negative CVR. This is a desirable safety feature.
Time-Limited Radiation Damage of In-Core Components: All in-core components except the pressure tube are incorporated in the fuel assembly design, which is shuffled twice in the reactor core after the initial installation. Hence, none of the core components other than the pressure tube are exposed to high radiation fields for more than three operating cycles. This feature limits the radiation exposure of the in-core components and, as a result, significantly reduces the likelihood of radiation-induced component failures.
Single Fuel Channel Replacement: One of the design goals of the Canadian SCWR is the ability to replace a single fuel channel if a defect or a material degradation issue requires the replacement of a fuel channel. For this purpose, the sizes of the fuel holes at the tubesheet and at the bellows/outlet header connections are selected such that a degraded or failed pressure tube and its pressure tube extension components can be removed without removing the outlet header. For a quick fuel channel replacement, a scheme is developed that does not require the removal of fuel assemblies in other fuel channels. As an extension of this capability, a complete core replacement is also possible.
No-Core-Melt: It has been shown by detailed numerical analysis that the maximum fuel cladding temperature for the worst-case accident scenario is below the melting point of the cladding material with a small margin. This result was obtained with conservative assumptions, suggesting that the actual margin is greater. Hence, the no-core-melt goal seems attainable, but the safety margin needs to be evaluated more rigorously by removing conservative assumptions in the analysis and performing additional experimental work.
Financial support of the Canadian Gen-IV National Program has been provided by the Office of Energy Research and Development (OERD) at Natural Resources Canada, the Natural Sciences and Engineering Research Council (NSERC), and Canadian Nuclear Laboratories.
CANDU is a registered trademark of Atomic Energy of Canada Limited.
Historically, MWd/kgHE has been used in the pressurized heavy water reactor (CANDU) industry as a unit for fuel utilization (burnup), while MWd/kgHM has been used by the light water reactor community. HE stands for heavy element, while HM stands for heavy metal but both means actinides like thorium, uranium, and plutonium.
ANSYS is a registered trademark of ANSYS, Inc., or its subsidiaries in the United States or other countries.