We can use thorium

As early as the 1950s, the use of thorium was identified as a promising fuel cycle in Atomic Energy of Canada Limited’s (AECL’s) CANDU development programme. This was due to its anticipated improved fuel performance [1-4] (e.g., reduced fission gas release, due to thorium’s enhanced thermal and chemical properties) and the relative abundance of thorium. The “Valubreeder” proposed by W.B. Lewis [5] in 1968 was an early theoretical concept of a CANDU reactor as a near-breeder using thorium-uranium fuel.

Thorium is a fundamental part AECL’s fuel cycle vision for CANDU, and represents a low-uranium-consumption fuel cycle option. A thorium-fuelled CANDU reactor is particularly attractive to countries with thorium reserves but no uranium – addressing their need for energy self-reliance.

The current expansion in nuclear plant new build and planning – the nuclear renaissance – puts increasing pressure on nuclear fuel reserves. This has heightened active interest in advanced fuel cycles. In Canada, and around the world, both “open” (i.e., not involving reprocessing) and “closed” (i.e., involving reprocessing) fuel cycles have and are being evaluated as ways of extending nuclear fuel resources, reducing waste and enhancing proliferation-resistance. AECL is proposing thorium fuel cycles as options for its CANDU 6/Enhanced CANDU 6 (EC6) reactors [6], and the Generation III+ ACR-1000 [7], which would start with a first- pass, open cycle, that would generate potential fuel material as a fissile material resource for initiating later closed fuel cycles.

AECL is also developing other fuel options [8, 9] for CANDU – including low-enriched uranium development (LEU), recovered uranium from conventional reprocessing, variations of MOX, and actinide waste. The CANFLEX [10] fuel bundle is the optimal carrier for these different fuels. CANFLEX has a nominal 43 elements, with varying element sizes dependent on fuel type.

Fuel cycle vision

The most attractive CANDU thorium fuel cycles would be closed, with reprocessing of the used fuel and recycling of the separated uranium-233 and unconverted thorium. Small quantities of fissile material (driver fuel) must be added to the thorium to initiate and sustain the chain reaction and to breed U-233 through neutron capture and subsequent beta decay. With successive recycle campaigns, the initiating driver fuel would be progressively displaced by recycled U-233. AECL estimates that a thorium-capable version of the present CANDU 6 could, upon reaching equilibrium fuel cycle conditions, generate up to 80% of its energy output from thorium. Later designs, optimized from the ground up for enhanced neutron economy, could be expected to improve on this figure. Fissile driver fuel material can be provided by low enriched uranium (LEU) or by plutonium (Pu) or recovered uranium extracted from spent light water reactor fuel. In the long term, plutonium or U-233 produced in fast breeder reactors (FBRs) could provide the driver fuel.

Short-term strategy

Thorium-rich countries could elect to first deploy a first-pass fuel in the near term in CANDU reactors, to acquire experience in thorium fuel-cycle technology and to build a strategic resource of U-233 safeguarded in the used fuel, without committing to its future recovery. In a CANDU reactor, the single-pass, Th-LEU fuel cycle increases uranium utilization by up to 30% relative to natural uranium fuel, and would provide a low-cost insurance policy against future shortages of uranium as well as provide a bridge to the closed thorium cycles of the future.

AECL is considering several thorium fuel configurations:

• Heterogeneous Th-LEU bundle

• Homogeneous Pu/Th mixture

• Mixed-bundle thorium (some fuel pins containing ThO₂, others UO₂ fuel)

Long-term strategy

For the long term, AECL is proposing the use of CANDU reactors with the near-Self-Sufficient Equilibrium Thorium (SSET) fuel cycle, which breeds enough U-233 that – through recycling – can keep the fuel cycle running indefinitely, with minimal need for additional fissile material. This closed cycle approach, which includes reprocessing, is the approach that will provide real future economic benefits. In the future, the CANDU-FBR synergism will allow a few fuel-generating, though expensive, FBRs to supply the fissile requirements of less-expensive, high-conversion-ratio CANDU reactors operating on the thorium cycle.

Past experience

Supporting its fuel cycle vision, AECL has more than 50 years of experience with thorium based fuel irradiations, with burnups to 47 MWd/kgHE, and powers ranging up to 77 kW/m. Twenty five thorium-based irradiation tests have been performed in the Nuclear Power Demonstration reactor (NPD), and AECL’s experimental reactors: National Research Experimental (NRX), National Research Universal (NRU) and Whiteshell Reactor 1 (WR1).

The composition of the fuels ranged from natural ThO₂ to ThO₂ + 30 wt% UO₂, although the majority of experiments were in the range ThO₂ + (1-3 wt%) enriched UO₂, where the UO₂ is enriched to 70 – 93 wt% U-235 in total U. Some experiments used Pu as the fissile material. Pellet geometries were generally standard (l/d ˜ 1.3), but several experiments examined short pellets (l/d ˜ 0.5), and a variety of pellet dish designs, including inter-pellet graphite discs. Figure 1 shows a post-irradiation section featuring the novel graphite disc design: plenum, 2 mm disc and pellet. Burnup was 43 MWd/kgHE, with a maximum linear power of 75 kW/m.

Early experiments showed great promise for ThO₂ based fuels, with fuel performance parameters generally superior to UO₂ under similar operating conditions. These results created incentive for numerous experiments in the late 1970s and early 1980s.

While most of the thorium test irradiations were performed in experimental reactors, of particular note is the NPD 51 irradiation, which was performed in a pre-commercial CANDU power reactor (NPD) from 1977 to 1987, and demonstrated the technical feasibility of using ThO2 fuels, and provided technical experience with these fuels. The experiment irradiated four standard NPD 19 element bundles (two containing 2.6 wt% UO₂, and two with 1.45 wt% UO₂; the UO₂ was enriched to 93 wt% U-235 in total U) at low powers (< 30 kW/m) to a maximum outer element burnup of 47 MWd/kgHE. The bundles operated successfully, and demonstrated features typical of high-burnup, low-power fuel. In this wide variety of test irradiations, AECL successfully operated thorium-based fuels at high powers to extended burnups, with good fuel performance. This demonstration supports the requirements for economic recycling of U-233. Higher thermal conductivity resulted in low fission gas releases through reduced operating temperatures. Grain growth was also reduced (compared with UO₂), with rare evidence of columnar grain growth. Reduced fission gas release and lower grain growth, have allowed thorium to perform without increasing sheath strain. Sheath strains generally were bounded by -0.6 to +0.6%, despite high burnups and powers, varying pellet geometries and different fuel compositions. Figure 2 shows a post-irradiation cross section from a fuel element from an experiment with (Th,U)O₂ (1.5% U-235). Burnup was 41 MWd/kgHE at a maximum linear power of 48 kW/m.

Sheath hydriding/deuteriding show no significant difference between thorium and uranium oxide fuels. CANLUB (AECL’s patented fuel sheath internal lubricant) retention in thorium based fuel is consistent with that observed in UO₂ fuels. Power ramp experience with thorium based fuel is limited, but suggests improved defect thresholds over UO₂, provided CANLUB coatings are used. Internal and external sheath oxidation is comparable to that observed in UO₂ fuels. Defected thorium fuel elements released significantly less fuel and fission products into the coolant system, as expected from the more stable, cooler ThO₂ matrix.

Additionally, AECL has initiated R&D programs on the management of thorium wastes [23].

Overall, AECL experience with thorium fuel is extensive, covering a wide range of fuel compositions, pellet geometries and operating conditions. Thorium fuel has always shown either comparable or superior performance relative to UO₂. One outcome of AECL’s comprehensive test programme has been the establishment of an optimum thorium pellet fabrication route, ensuring reproducible irradiation performance superior to that of UO₂ fuels.

References
1. Kingery W.D., Francl J., Coble R.L. and Vasilos T., (1954), “Data for Several Pure Oxide materials Corrected to Zero Porosity”, J. A. Cer. S., Vol 2, p.37.

2. Jeffs A.T., (1969) “Thermal Conductivity of ThO2-PuO2 Under Irradiation”, AECL Report, AECL-3294.

3. J.M. Fink, M.G. Chasanov and L. Leibowitz, “Thermal Conductivity and Thermal Diffusivity of Solid UO2”, Argonne National Laboratory Report, ANL-CEN-RSD-81-3 (1981).

4. J. Belle and R.M. Berman, “The High-temperature Ex-reactor Thermal Conductivity of Thoria and Thoria-Urania Solid Solutions (LWBR/AWBA Development Program)”, Report WAPD-TM-1530 (1982).

5. Lewis W.B., (1968), “Superconverter or Valubreeder: A Near-Breeder Uranium-Thorium Nuclear Fuel Cycle”, AECL report AECL-3081.

6. Azeez S. and Girouard, P., (2006) “Enhanced CANDU 6 Reactor: Status”, 15PBNC, Sydney, Australia, October 15-20.

7. Petrunik K., (2008), “ACR-1000: Technology for the Future”, 16PBNC, Aomori City, Japan, October 13-18.

8. Boczar P.G., (1998), “CANDU Fuel Cycle vision”, IAEA Technical Committee Meeting on Fuel Cycle Options for LWRs and HWRs, Victoria, British Columbia, Canada, April 28 – May 01.

9. Chan P., Dyck G., Kuran S. Ivanco M. Hopwood J. and Hastings I.J., (2008) “Fuel Cycles—Secure Supply, Reduced waste, Enhanced Non-Proliferation” 16PBNC, Aomori City, Japan, October 13-18.

10. Hastings I.J., Boczar P.G., and Lane A.D., (1989)“CANFLEX—An Advanced Bundle Design for Introducing Slightly Enriched Uranium into CANDU”, Proc. Intl Symposium on Uranium and Electricity, Saskatoon, SK, September 18-21, also reprinted as AECL-9766.

11. Rao S.V.K., (1963) “Investigation of ThO2 – UO2 as a Nuclear Fuel”, AECL Report, AECL 1785.

12. Milgram M.S., (1982) “Once Through Thorium Cycles in CANDU Reactors”, AECL Report, AECL-7516.

13. Veeder J. and Didsbury R., (1985) “A Catalogue of Advanced Fuel Cycles in CANDU-PHW Reactors”, AECL Report AECL-8641 (1985).

14. Dastur A.R., Meneley D.A. and Buss D.B, (1995) “Thorium Cycle Options in CANDU Reactors”, Proc. Global 95, Versailles, France.

15. Boczar P.G., Chan P.S.W., Ellis R.J., Dyck G.R., Sullivan J.D., Taylor P. and Jones R.T., (1988) “A Fresh Look at Thorium Fuel Cycles in CANDU Reactors”, Proc. 11th PBNC, Banff, Canada, May 3-7.

16. Torgerson D.F., Fehrenbach P.J., Hopwood J.M., Duffey R., Boczar P.G., Love I., Kuran S., Ivanco M.J., Dyck G.R., Chan P.S.W. and Tyagi A.K., (2006) “CANDU Reactors with Thorium Fuel Cycles”, 15PBNC, Sydney, Australia, October 15-20.

17. Jeffs A.T., (1967) “Irradiation Behaviour of Experimental Fuel Assemblies of UO2 PuO2 and ThO2 UO2”, AECL Report AECL 2674.

18. Bain A.S., (1968) “Crack Healing and Void Movement During Irradiation of ThO2 2 wt% UO2”, AECL Report AECL 3008.

19. Bain A.S., Christie J. and Hastings I.J., (1977) “Performance of ThO2-UO2 Fuel Irradiated in the NRU Reactor at CRNL”, Trans. ANS 27(1977)307.

20. Hastings I.J., Celli A., Onofrei M. and Swanson M.L., (1982) “Irradiation Performance of (Th,U)O2 Fuel designed for Advanced Cycle Applications”, Proc. 3rd CNS Annual Conf., also reprinted as AECL Report AECL-7697.

21. Smith A.D., Walsworth J.A., Donders R.E and Fehrenbach P.J., (1985) “In Reactor Measurement of Operating Temperature in (Th,U)O2 Fuel: Effect of Pellet Microstructure”, Proc. 6th Annual Conference of the Canadian Nuclear Society, Ottawa.

22. Mao J., Chan P.S.W, and Kuran S., (2009) “Fuel Management Simulations of Thorium Fuel Cycle in CANDU 6 Reactors” , Advances in Nuclear fuel management IV (ANFM-IV), Hilton Head, SC, USA, April 12-15.

23. Taylor P., Hocking W.H. et al, (1994) “Waste Management Aspects of (Th,Pu)O2 Fuels”, Presented at the IAEA Technical Meeting on “Unconventional Options for Pu Disposition”, Obninsk, Russia, November.

Ian Hastings, director of operations, reactor development, AECL. Email: hastingsi@aecl.ca.