Thorium fuel cycle

A sample of thorium

The thorium fuel cycle is a nuclear fuel cycle that uses an isotope of thorium, ²³²
Th
, as the fertile material. In the reactor, ²³²
Th
is transmuted into the fissile artificial uranium isotope ²³³
U
which is the nuclear fuel. Unlike natural uranium, natural thorium contains only trace amounts of fissile material (such as ²³¹
Th
), which are insufficient to initiate a nuclear chain reaction. Additional fissile material or another neutron source is necessary to initiate the fuel cycle. In a thorium-fuelled reactor, ²³²
Th
absorbs neutrons to produce ²³³
U
. This parallels the process in uranium breeder reactors whereby fertile ²³⁸
U
absorbs neutrons to form fissile ²³⁹
Pu
. Depending on the design of the reactor and fuel cycle, the generated ²³³
U
either fissions in situ or is chemically separated from the used nuclear fuel and formed into new nuclear fuel.

The thorium fuel cycle has several potential advantages over a uranium fuel cycle, including thorium's greater abundance, superior physical and nuclear properties, reduced plutonium and actinide production, and better resistance to nuclear weapons proliferation when used in a traditional light water reactor though not in a molten salt reactor.

History

Concerns about the limits of worldwide uranium resources motivated initial interest in the thorium fuel cycle. It was envisioned that as uranium reserves were depleted, thorium would supplement uranium as a fertile material. However, for most countries uranium was relatively abundant and research in thorium fuel cycles waned. A notable exception was India's three-stage nuclear power programme. In the twenty-first century thorium's claimed potential for improving proliferation resistance and waste characteristics led to renewed interest in the thorium fuel cycle. While thorium is more abundant in the continental crust than uranium and easily extracted from monazite as a side product of rare earth element mining, it is much less abundant in seawater than uranium.

At Oak Ridge National Laboratory in the 1960s, the Molten-Salt Reactor Experiment used ²³³
U
as the fissile fuel in an experiment to demonstrate a part of the Molten Salt Breeder Reactor that was designed to operate on the thorium fuel cycle. Molten salt reactor (MSR) experiments assessed thorium's feasibility, using thorium(IV) fluoride dissolved in a molten salt fluid that eliminated the need to fabricate fuel elements. The MSR program was defunded in 1976 after its patron Alvin Weinberg was fired.

In 1993, Carlo Rubbia proposed the concept of an energy amplifier or "accelerator driven system" (ADS), which he saw as a novel and safe way to produce nuclear energy that exploited existing accelerator technologies. Rubbia's proposal offered the potential to incinerate high-activity nuclear waste and produce energy from natural thorium and depleted uranium.

Kirk Sorensen, former NASA scientist and Chief Technologist at Flibe Energy, has been a long-time promoter of thorium fuel cycle and particularly liquid fluoride thorium reactors (LFTRs). He first researched thorium reactors while working at NASA, while evaluating power plant designs suitable for lunar colonies. In 2006 Sorensen started "energyfromthorium.com" to promote and make information available about this technology.

A 2011 MIT study concluded that although there is little in the way of barriers to a thorium fuel cycle, with current or near term light-water reactor designs there is also little incentive for any significant market penetration to occur. As such they conclude there is little chance of thorium cycles replacing conventional uranium cycles in the current nuclear power market, despite the potential benefits.

Nuclear reactions with thorium

In the thorium cycle, fuel is formed when ²³²
Th
captures a neutron (whether in a fast reactor or thermal reactor) to become ²³³
Th
. This normally emits an electron and an anti-neutrino (
ν
) by
β⁻
decay to become ²³³
Pa
. This then emits another electron and anti-neutrino by a second
β⁻
decay to become ²³³
U
, the fuel:

${\ce {{\overset {neutron}{n}}+{^{232}_{90}Th}->{^{233}_{90}Th}->[\beta ^{-}]{^{233}_{91}Pa}->[\beta ^{-}]{\overset {fuel}{^{233}_{92}U}}}}$

Fission product waste

Nuclear fission produces radioactive fission products which can have half-lives from days to greater than 200,000 years. According to some toxicity studies, the thorium cycle can fully recycle actinide wastes and only emit fission product wastes, and after a few hundred years, the waste from a thorium reactor can be less toxic than the uranium ore that would have been used to produce low enriched uranium fuel for a light water reactor of the same power. Other studies assume some actinide losses and find that actinide wastes dominate thorium cycle waste radioactivity at some future periods. Some fission products have been proposed for nuclear transmutation, which would further reduce the amount of nuclear waste and the duration during which it would have to be stored (whether in a deep geological repository or elsewhere). However, while the principal feasibility of some of those reactions has been demonstrated at laboratory scale, there is, as of 2024, no large scale deliberate transmutation of fission products anywhere in the world, and the upcoming MYRRHA research project into transmutation is mostly focused on transuranic waste. Furthermore, the cross section of some fission products is relatively low and others - such as caesium - are present as a mixture of stable, short lived and long lived isotopes in nuclear waste, making transmutation dependent on expensive isotope separation.

Actinide waste

In a reactor, when a neutron hits a fissile atom (such as certain isotopes of uranium), it either splits the nucleus or is captured and transmutes the atom. In the case of ²³³
U
, the transmutations tend to produce useful nuclear fuels rather than transuranic waste. When ²³³
U
absorbs a neutron, it either fissions or becomes ²³⁴
U
. The chance of fissioning on absorption of a thermal neutron is about 92%; the capture-to-fission ratio of ²³³
U
, therefore, is about 1:12 – which is better than the corresponding capture vs. fission ratios of ²³⁵
U
(about 1:6), or ²³⁹
Pu
or ²⁴¹
Pu
(both about 1:3). The result is less transuranic waste than in a reactor using the uranium-plutonium fuel cycle.

Transmutations in the thorium fuel cycle
														237Np
														↑
		231U	←	232U	↔	233U	↔	234U	↔	235U	↔	236U	→	237U
		↓		↑		↑		↑
		231Pa	→	232Pa	←	233Pa	→	234Pa
		↑				↑
230Th	→	231Th	←	232Th	→	233Th
Nuclides with a yellow background in italic have half-lives under 30 days Nuclides in bold have half-lives over 1,000,000 years Nuclides in red frames are fissile

²³⁴
U
, like most actinides with an even number of neutrons, is not fissile, but neutron capture produces fissile ²³⁵
U
. If the fissile isotope fails to fission on neutron capture, it produces ²³⁶
U
, ²³⁷
Np
, ²³⁸
Pu
, and eventually fissile ²³⁹
Pu
and heavier isotopes of plutonium. The ²³⁷
Np
can be removed and stored as waste or retained and transmuted to plutonium, where more of it fissions, while the remainder becomes ²⁴²
Pu
, then americium and curium, which in turn can be removed as waste or returned to reactors for further transmutation and fission.

However, the ²³¹
Pa
(with a half-life of 3.27×10⁴ years) formed via (n,2n) reactions with ²³²
Th
(yielding ²³¹
Th
that decays to ²³¹
Pa
), while not a transuranic waste, is a major contributor to the long-term radiotoxicity of spent nuclear fuel. While ²³¹
Pa can in principle be converted back to ²³²
Th by neutron absorption, its neutron absorption cross section is relatively low, making this rather difficult and possibly uneconomic.

Uranium-232 contamination

²³²
U
is also formed in this process, via (n,2n) reactions between fast neutrons and ²³³
U
, ²³³
Pa
, and ²³²
Th
:

${\begin{aligned}{}\\{\ce {{^{232}_{90}Th}->[+n]{^{233}_{90}Th}->[\beta ^{-}]{^{233}_{91}Pa}{\text{ }}->[\beta ^{-}]{^{233}_{92}U}->[+n-2n]{^{232}_{92}U}}}\\{\ce {{^{232}_{90}Th}->[+n]{^{233}_{90}Th}->[\beta ^{-}]{^{233}_{91}Pa}{\text{ }}->[+n-2n]{^{232}_{91}Pa}->[\beta ^{-}]{^{232}_{92}U}}}\\{\ce {{^{232}_{90}Th}->[+n-2n]{^{231}_{90}Th}->[\beta ^{-}]{^{231}_{91}Pa}{\text{ }}->[+n]{^{232}_{91}Pa}->[\beta ^{-}]{^{232}_{92}U}}}\\{}\end{aligned}}$

Unlike most even numbered heavy isotopes, ²³²
U
is also a fissile fuel fissioning just over half the time when it absorbs a thermal neutron. ²³²
U
has a relatively short half-life (68.9 years), and some decay products emit high energy gamma radiation, such as ²²⁰
Rn
, ²¹²
Bi
and particularly ²⁰⁸
Tl
. The full decay chain, along with half-lives and relevant gamma energies, is:

The 4n decay chain of ²³²Th, commonly called the "thorium series"

²³²
U
decays to ²²⁸
Th
where it joins the decay chain of ²³²
Th

${\begin{aligned}{}\\{\ce {^{232}_{92}U ->[\alpha] ^{228}_{90}Th}}\ &\mathrm {(68.9\ years)} \\{\ce {^{228}_{90}Th ->[\alpha] ^{224}_{88}Ra}}\ &\mathrm {(1.9\ year)} \\{\ce {^{224}_{88}Ra ->[\alpha] ^{220}_{86}Rn}}\ &\mathrm {(3.6\ day,\ 0.24\ MeV)} \\{\ce {^{220}_{86}Rn ->[\alpha] ^{216}_{84}Po}}\ &\mathrm {(55\ s,\ 0.54\ MeV)} \\{\ce {^{216}_{84}Po ->[\alpha] ^{212}_{82}Pb}}\ &\mathrm {(0.15\ s)} \\{\ce {^{212}_{82}Pb ->[\beta^-] ^{212}_{83}Bi}}\ &\mathrm {(10.64\ h)} \\{\ce {^{212}_{83}Bi ->[\alpha] ^{208}_{81}Tl}}\ &\mathrm {(61\ m,\ 0.78\ MeV)} \\{\ce {^{208}_{81}Tl ->[\beta^-] ^{208}_{82}Pb}}\ &\mathrm {(3\ m,\ 2.6\ MeV)} \\{}\end{aligned}}$

Thorium-cycle fuels produce hard gamma emissions, which damage electronics, limiting their use in bombs. ²³²
U
cannot be chemically separated from ²³³
U
from used nuclear fuel; however, chemical separation of thorium from uranium removes the decay product ²²⁸
Th
and the radiation from the rest of the decay chain, which gradually build up as ²²⁸
Th
reaccumulates. The contamination could also be avoided by using a molten-salt breeder reactor and separating the ²³³
Pa
before it decays into ²³³
U
. The hard gamma emissions also create a radiological hazard which requires remote handling during reprocessing.

Nuclear fuel

As a fertile material thorium is similar to ²³⁸
U
, the major part of natural and depleted uranium. The thermal neutron absorption cross section (σ_a) and resonance integral (average of neutron cross sections over intermediate neutron energies) for ²³²
Th
are about three and one third times those of the respective values for ²³⁸
U
.

Advantages

The primary physical advantage of thorium fuel is that it uniquely makes possible a breeder reactor that runs with slow neutrons, otherwise known as a thermal breeder reactor. These reactors are often considered simpler than the more traditional fast-neutron breeders. Although the thermal neutron fission cross section (σ_f) of the resulting ²³³
U
is comparable to ²³⁵
U
and ²³⁹
Pu
, it has a much lower capture cross section (σ_γ) than the latter two fissile isotopes, providing fewer non-fissile neutron absorptions and improved neutron economy. The ratio of neutrons released per neutron absorbed (η) in ²³³
U
is greater than two over a wide range of energies, including the thermal spectrum. A breeding reactor in the uranium–plutonium cycle needs to use fast neutrons, because in the thermal spectrum one neutron absorbed by ²³⁹
Pu
on average leads to less than two neutrons.

Thorium is estimated to be about three to four times more abundant than uranium in Earth's crust, although present knowledge of reserves is limited. Current demand for thorium has been satisfied as a by-product of rare-earth extraction from monazite sands. Notably, there is very little thorium dissolved in seawater, so seawater extraction is not viable, as it is with uranium. Using breeder reactors, known thorium and uranium resources can both generate world-scale energy for thousands of years.

Thorium-based fuels also display favorable physical and chemical properties that improve reactor and repository performance. Compared to the predominant reactor fuel, uranium dioxide (UO
₂), thorium dioxide (ThO
₂) has a higher melting point, higher thermal conductivity, and lower coefficient of thermal expansion. Thorium dioxide also exhibits greater chemical stability and, unlike uranium dioxide, does not further oxidize.

Because the ²³³
U
produced in thorium fuels is significantly contaminated with ²³²
U
in proposed power reactor designs, thorium-based used nuclear fuel possesses inherent proliferation resistance. ²³²
U
cannot be chemically separated from ²³³
U
and has several decay products that emit high-energy gamma radiation. These high-energy photons are a radiological hazard that necessitate the use of remote handling of separated uranium and aid in the passive detection of such materials.

The long-term (on the order of roughly 10³ to 10⁶ years) radiological hazard of conventional uranium-based used nuclear fuel is dominated by plutonium and other minor actinides, after which long-lived fission products become significant contributors again. A single neutron capture in ²³⁸
U
is sufficient to produce transuranic elements, whereas five captures are generally necessary to do so from ²³²
Th
. 98–99% of thorium-cycle fuel nuclei would fission at either ²³³
U
or ²³⁵
U
, so fewer long-lived transuranics are produced. Because of this, thorium is a potentially attractive alternative to uranium in mixed oxide (MOX) fuels to minimize the generation of transuranics and maximize the destruction of plutonium.

Disadvantages

There are several challenges to the application of thorium as a nuclear fuel, particularly for solid fuel reactors:

In contrast to uranium, naturally occurring thorium is effectively mononuclidic and contains no fissile isotopes; fissile material, generally ²³³
U
, ²³⁵
U
or plutonium, must be added to achieve criticality. This, along with the high sintering temperature necessary to make thorium-dioxide fuel, complicates fuel fabrication. Oak Ridge National Laboratory experimented with thorium tetrafluoride as fuel in a molten salt reactor from 1964 to 1969, which was expected to be easier to process and separate from contaminants that slow or stop the chain reaction.

In an open fuel cycle (i.e. utilizing ²³³
U
in situ), higher burnup is necessary to achieve a favorable neutron economy. Although thorium dioxide performed well at burnups of 170,000 MWd/t and 150,000 MWd/t at Fort St. Vrain Generating Station and AVR respectively, challenges complicate achieving this in light water reactors (LWR), which compose the vast majority of existing power reactors.

In a once-through thorium fuel cycle, thorium-based fuels produce far less long-lived transuranics than uranium-based fuels, some long-lived actinide products constitute a long-term radiological impact, especially ²³¹
Pa
and ²³³
U
. On a closed cycle,²³³
U
and ²³¹
Pa
can be reprocessed. ²³¹
Pa
is also considered an excellent burnable poison absorber in light water reactors.

Another challenge associated with the thorium fuel cycle is the comparatively long interval over which ²³²
Th
breeds to ²³³
U
. The half-life of ²³³
Pa
is about 27 days, which is an order of magnitude longer than the half-life of ²³⁹
Np
. As a result, substantial ²³³
Pa
develops in thorium-based fuels. ²³³
Pa
is a significant neutron absorber and, although it eventually breeds into fissile ²³⁵
U
, this requires two more neutron absorptions, which degrades neutron economy and increases the likelihood of transuranic production.

Alternatively, if solid thorium is used in a closed fuel cycle in which ²³³
U
is recycled, remote handling is necessary for fuel fabrication because of the high radiation levels resulting from the decay products of ²³²
U
. This is also true of recycled thorium because of the presence of ²²⁸
Th
, which is part of the ²³²
U
decay sequence. Further, unlike proven uranium fuel recycling technology (e.g. PUREX), recycling technology for thorium (e.g. THOREX) is only under development.

Although the presence of ²³²
U
complicates matters, there are public documents showing that ²³³
U
has been used once in a nuclear weapon test. The United States tested a composite ²³³
U
-plutonium bomb core in the MET (Military Effects Test) blast during Operation Teapot in 1955, though with much lower yield than expected.

Advocates for liquid core and molten salt reactors such as LFTRs claim that these technologies negate thorium's disadvantages present in solid fuelled reactors. As only two liquid-core fluoride salt reactors have been built (the ORNL ARE and MSRE) and neither have used thorium, it is hard to validate the exact benefits.

Thorium-fueled reactors

Thorium fuels have fueled several different reactor types, including light water reactors, heavy water reactors, high temperature gas reactors, sodium-cooled fast reactors, and molten salt reactors.

List of thorium-fueled reactors

From IAEA TECDOC-1450 "Thorium Fuel Cycle – Potential Benefits and Challenges", Table 1: Thorium utilization in different experimental and power reactors. Additionally from Energy Information Administration, "Spent Nuclear Fuel Discharges from U. S. Reactors", Table B4: Dresden 1 Assembly Class.

Name	Operation period	Country	Reactor type	Power	Fuel
NRX & NRU	1947 (NRX) + 1957 (NRU); Irradiation–testing of few fuel elements	Canada	MTR (pin assemblies)	020000 20 MW; 200 MW (see)	Th+235 U , Test Fuel
Dresden Unit 1	1960–1978	United States	BWR	300000 197 MW(e)	ThO2 corner rods, UO2 clad in Zircaloy-2 tube
CIRUS; DHRUVA; & KAMINI	1960–2010 (CIRUS); others in operation	India	MTR thermal	040000 40 MWt; 100 MWt; 30 kWt (low power, research)	Al+233 U Driver fuel, ‘J’ rod of Th & ThO2, ‘J’ rod of ThO2
Indian Point 1	1962–1965	United States	LWBR, PWR, (pin assemblies)	285000 285 MW(e)	Th+233 U Driver fuel, oxide pellets
BORAX-IV & Elk River Station	1963–1968	United States	BWR (pin assemblies)	002400 2.4 MW(e); 24 MW(e)	Th+235 U Driver fuel oxide pellets
MSRE ORNL	1964–1969	United States	MSR	007500 7.5 MWt	233 U molten fluorides
Peach Bottom	1966–1972	United States	HTGR, Experimental (prismatic block)	040000 40 MW(e)	Th+235 U Driver fuel, coated fuel particles, oxide & dicarbides
Dragon (OECD-Euratom)	1966–1973	UK (also Sweden, Norway and Switzerland)	HTGR, Experimental (pin-in-block design)	020000 20 MWt	Th+235 U Driver fuel, coated fuel particles, oxide & dicarbides
AVR	1967–1988	Germany (West)	HTGR, experimental (pebble bed reactor)	015000 15 MW(e)	Th+235 U Driver fuel, coated fuel particles, oxide & dicarbides
Lingen	1968–1973	Germany (West)	BWR irradiation-testing	060000 60 MW(e)	Test fuel (Th,Pu)O2 pellets
SUSPOP/KSTR KEMA	1974–1977	Netherlands	Aqueous homogeneous suspension (pin assemblies)	001000 1 MWt	Th+HEU, oxide pellets
Fort St Vrain	1976–1989	United States	HTGR, Power (prismatic block)	330000 330 MW(e)	Th+235 U Driver fuel, coated fuel particles, Dicarbide
Shippingport	1977–1982	United States	LWBR, PWR, (pin assemblies)	100000 100 MW(e)	Th+233 U Driver fuel, oxide pellets
KAPS 1 &2; KGS 1 & 2; RAPS 2, 3 & 4	1980 (RAPS 2) +; continuing in all new PHWRs	India	PHWR, (pin assemblies)	220000 220 MW(e)	ThO2 pellets (for neutron flux flattening of initial core after start-up)
FBTR	1985; in operation	India	LMFBR, (pin assemblies)	040000 40 MWt	ThO2 blanket
THTR-300	1985–1989	Germany (West)	HTGR, power (pebble type)	300000 300 MW(e)	Th+235 U Driver fuel, coated fuel particles, oxide & dicarbides
TMSR-LF1	2023; operating license issued	China	Liquid fuel thorium-based molten salt experimental reactor	002000 2 MWt	Thorium-based molten salt
Petten	2024; planned	Netherlands	High Flux Reactor thorium molten salt experiment	060000 45 MW(e)	?

See also

Nuclear technology portal Energy portal

• Thorium
• Thorium-232
• Occurrence of thorium
• Thorium-based nuclear power
• List of countries by thorium resources
• List of countries by uranium reserves
• Advanced heavy-water reactor
• Alvin Radkowsky
• CANDU reactor
• Fuji MSR
• Peak uranium
• Radioactive waste
• Thorium Energy Alliance
• Weinberg Foundation
• World energy resources and consumption

Further reading

• Kasten, P. R. (1998). "Review of the Radkowsky Thorium reactor concept" Science & Global Security, 7(3), 237–269.
• Duncan Clark (9 September 2011), "Thorium advocates launch pressure group. Huge optimism for thorium nuclear energy at the launch of the Weinberg Foundation", The Guardian
• Nelson, A. T. (2012). "Thorium: Not a near-term commercial nuclear fuel". Bulletin of the Atomic Scientists. 68 (5): 33–44. Bibcode:2012BuAtS..68e..33N. doi:10.1177/0096340212459125. S2CID 144725888.
• B.D. Kuz'minov, V.N. Manokhin, (1998) "Status of nuclear data for the thorium fuel cycle", IAEA translation from the Russian journal Yadernye Konstanty (Nuclear Constants) Issue No. 3–4, 1997
• Thorium and uranium fuel cycles comparison by the UK National Nuclear Laboratory
• Fact sheet on thorium Archived 2013-02-16 at the Wayback Machine at the World Nuclear Association.
• Annotated bibliography for the thorium fuel cycle Archived 2010-10-07 at the Wayback Machine from the Alsos Digital Library for Nuclear Issues

External links

• International Thorium Energy Committee