Reprocessing and recycling

A client from overseas has asked Meridian for a briefing on whether recycling used fuel could make sense for a future Canadian fleet. Amara hands Hana a folder of public reports from the World Nuclear Association, the OECD Nuclear Energy Agency and the NWMO. "Canada has never reprocessed commercial fuel," she says. "Natural uranium is cheap, and CANDU fuel has little fissile content left when it comes out. But the question keeps coming back, so we need to understand the arguments on both sides."

Open and closed fuel cycles

In an open, or once-through, fuel cycle, used fuel is stored and then disposed of directly in a repository. Canada, the United States, Finland and Sweden follow this policy. In a closed fuel cycle, used fuel is reprocessed to recover the uranium and plutonium it still contains, and these are recycled into new fuel, while the remaining fission products and minor actinides are immobilised as high-level waste. France has followed this policy for decades; the United Kingdom reprocessed for many years and has now stopped; Japan has built a reprocessing plant at Rokkasho; Russia and China operate or are building reprocessing facilities.

Recycling in today's reactors is partial. Recovered plutonium is made into mixed oxide (MOX) fuel and used once in light water reactors; the used MOX fuel is then stored, because recycling it again in thermal reactors gives diminishing returns. A fully closed cycle, in which plutonium and some of the minor actinides are recycled repeatedly, would need fast reactors, which are discussed in the advanced reactors course.

What reprocessing does

Used light water reactor fuel is about 95 % uranium, about 1 % plutonium, and about 4 % fission products and minor actinides, as lesson 5 described. Commercial reprocessing uses a solvent-extraction process called PUREX, short for plutonium uranium reduction extraction, to separate reusable uranium and plutonium from the rest. The process has been used on an industrial scale for more than half a century. At the level of a policy briefing, three points matter: the uranium and plutonium recovered can be made into new fuel; the remaining fission products and minor actinides are turned into glass for disposal; and the plant handles intensely radioactive material, so it operates remotely behind heavy shielding and under strict international safeguards.

The largest commercial plant is Orano's La Hague facility in Normandy, France, which has reprocessed used fuel from French reactors and, in the past, from several other countries. The recovered plutonium is made into MOX fuel at Orano's Melox plant and used in some French reactors. The United Kingdom's Sellafield site reprocessed oxide fuel at its THORP plant until 2018 and metal Magnox fuel until 2022, and is now focused on cleanup. Canada has no reprocessing industry, though AECL carried out research on fuel cycle options and the NWMO's studies have looked at whether reprocessing would change repository needs.

The waste argument

Supporters of recycling argue that it reduces the volume and long-term radiotoxicity of the waste that must go to a repository. Vitrified high-level waste from reprocessing takes up much less volume than the used fuel it came from, and, because most of the plutonium has been removed, its long-term radiotoxicity falls faster. Critics answer that the reduction in the volume of high-level waste is partly offset by intermediate-level wastes from the plant itself, including the metal hulls and end fittings of the fuel, process wastes and decommissioning wastes; that the used MOX fuel still has to be stored or disposed of, and it produces more heat than ordinary used fuel; and that the size of a repository is set mainly by heat rather than volume.

# Waste volumes per terawatt-hour for an open and a recycling fuel cycle (illustrative public figures).
# The point is the relative size of the streams, not exact values.
twh = 1.0
open_cycle = {"spent fuel (packaged for disposal)": 2.5}
recycle = {"vitrified high-level waste": 0.35, "compacted hulls and end fittings": 0.5,
           "spent MOX (stored for later use or disposal)": 0.3}
print("cubic metres per TWh")
for name, v in open_cycle.items():
    print(f"  open cycle   {name:44s} {v*twh:4.2f}")
for name, v in recycle.items():
    print(f"  recycling    {name:44s} {v*twh:4.2f}")
saving = 1 - sum(recycle.values()) / sum(open_cycle.values())
print(f"high-level waste volume reduction with recycling: about {saving*100:.0f} %")
print("uranium savings from MOX and recycled uranium are commonly quoted at 10 to 25 %")

cubic metres per TWh
  open cycle   spent fuel (packaged for disposal)           2.50
  recycling    vitrified high-level waste                   0.35
  recycling    compacted hulls and end fittings             0.50
  recycling    spent MOX (stored for later use or disposal) 0.30
high-level waste volume reduction with recycling: about 54 %
uranium savings from MOX and recycled uranium are commonly quoted at 10 to 25 %

The output shows a large reduction in the volume of the most active waste, about half when the hulls and used MOX are included, but not an elimination of the need for a repository. Studies by the OECD Nuclear Energy Agency and others have found that recycling in light water reactors alone gives modest benefits for repository size, because the fission products that dominate heat in the first centuries are still there. Larger reductions in heat and radiotoxicity need repeated recycling of plutonium and minor actinides in fast reactors, which is the subject of continued research rather than current commercial practice.

The resource argument

Recycling also recovers energy. Recovered plutonium in MOX fuel and recycled uranium re-enriched for further use can replace a share of the natural uranium a reactor fleet needs. Public estimates commonly quote savings of 10 to 25 % of the uranium requirement. Whether this is worth the cost depends on uranium prices. When uranium is cheap, as it has been for most of the past four decades, the cost of reprocessing and MOX fabrication is higher than the cost of the uranium saved, so the once-through cycle is cheaper. Studies by MIT, the OECD and others have concluded that, at current and projected uranium prices, the open cycle is likely to remain the lower-cost option for some time.

The proliferation argument

The most important objection to reprocessing is that it separates plutonium from the highly radioactive fission products that otherwise make used fuel hard to handle. Although plutonium from commercial used fuel is not well suited to weapons, separated plutonium of any kind is treated by the IAEA as direct-use material and must be protected and safeguarded to the highest standard. Stocks of separated civil plutonium in several countries are large, and their security and safeguarding is a continuing international concern. Lesson 12 describes the safeguards system.

Canada's non-proliferation policy, which requires that Canadian uranium is used only for peaceful purposes and that any reprocessing of it needs Canada's consent under bilateral agreements, reflects these concerns. The once-through CANDU cycle, with no separated plutonium, fits well within that policy.

Recent developments

Interest in recycling has returned with advanced reactor designs. Some proposed reactors would use fuels made from recycled used fuel, and some developers have proposed processes that do not separate pure plutonium. In Canada, Moltex Energy has proposed a process to recover material from CANDU used fuel for use in a molten salt reactor, and the federal government has said that any reprocessing proposal would have to meet strict safety, security and non-proliferation requirements. Any such facility would need a CNSC licence and would be subject to IAEA safeguards.

Vitrified waste

The high-level waste from reprocessing contains most of the fission products and minor actinides from the used fuel. It is turned into glass, a process called vitrification, by mixing it with glass-forming materials and melting the mixture at high temperature. The molten glass is poured into stainless steel canisters, where it solidifies, and the canisters are sealed by welding. Borosilicate glass is used because it can hold a wide range of elements in its structure and dissolves very slowly in water. Each standard canister from La Hague holds about 400 kg of glass and is about 1.3 m tall. The canisters are stored in ventilated vaults to cool for several decades before disposal.

Vitrification is also used for defence waste. In the United States, the Defense Waste Processing Facility at the Savannah River Site has vitrified high-level waste since 1996, and the Waste Treatment and Immobilization Plant at Hanford has started up a facility to vitrify low-activity tank waste. These programs show that the technology is mature, though large and costly, and that the waste form is suitable for geological disposal.

Comparing studies of fuel cycle options

Many national and international studies have compared fuel cycle options. Their conclusions are broadly consistent. Direct disposal of used fuel in a geological repository is technically feasible and safe, and is currently the least expensive option. Recycling in thermal reactors reduces the use of uranium and the volume of high-level waste, but costs more and adds separated plutonium to the cycle. Multi-recycling in fast reactors could greatly increase the energy obtained from uranium and reduce long-term radiotoxicity, but would need a large fleet of fast reactors and reprocessing plants, decades of development and strong safeguards. None of the options removes the need for a repository.

For Canada, the economics of recycling are particularly unfavourable with current CANDU fuel, because natural uranium fuel contains very little fissile material when discharged: a CANDU bundle at discharge contains less fissile uranium than natural uranium itself, and only a small amount of plutonium. Any future recycling would most likely be tied to new reactor types. The NWMO's repository design is based on direct disposal, but the APM approach keeps the option of retrieving used fuel if future generations decide to recycle it.

Thorium and other alternative cycles

Some proposals for closed cycles use thorium rather than uranium. Thorium-232 is not fissile, but it absorbs neutrons to form uranium-233, which is. Thorium is about three times as abundant as uranium in the Earth's crust, and India, which has large thorium reserves and limited uranium, has made a three-stage program based on thorium a long-term national goal. Thorium cycles produce much less plutonium and fewer minor actinides, but they need fissile material to start, they produce uranium-233, which must itself be safeguarded, and they need reprocessing to recover the uranium-233 for reuse. CANDU reactors, with their high neutron economy, have been studied as thorium burners, and AECL and international partners carried out research on thorium fuels over several decades.

For now, these alternative cycles remain topics of research and development rather than commercial practice. They illustrate an important point for policy: the choice of fuel cycle is not only a technical one but also an economic, environmental and security decision, made by governments over decades with input from industry, regulators and the public.