Recently Oliver Tickell, a
British journalist, author and environmental campaigner, wrote a public opinion
piece attempting to sway public opinion and dissuade growing public interest in
thorium. The piece is entitled “Thorium:
Not ‘green’, not ‘viable’, and not likely” and represents the most direct
attack on thorium used in the liquid-fluoride reactor thus far. Mr. Tickell’s
alternative is not overtly made clear until the final paragraph of the article,
which merits recitation at this point in the discussion:
“Far better to invest in the renewable technologies that are already shaping our national and global future, and whose cost is rapidly falling – in the process developing valuable UK-based expertise and technologies, and accelerating the renewables revolution.”
Despite telling us that solar and wind power are better choices than thorium, Mr. Tickell yields a fair number of our points throughout his piece:
“[Thorium] is three to four times more abundant in the Earth’s crust than uranium, and is especially plentiful in Australia, Norway, India, the USA and China.”
“[Molten-salt reactors] offer several potential advantages over conventional reactor designs: the reactor and its cooling circuits operate at near atmospheric pressure, reducing the chance of any explosion, [and] in the event of a reactor overheating, the fuel can simply drain out into a secondary container and the fission chain reaction will halt, reducing the risk of reactor meltdowns such as those experienced at Chernobyl and Fukushima.”
“thorium reactors do not produce plutonium”
“the design of molten salt reactors does indeed mitigate against reactor meltdown and explosion”
“important elements of the LFTR technology were proven during the 1970s”
“LFTRs are theoretically capable of a high fuel burn-up rate”
“the thorium fuel cycle does indeed produce very low volumes of plutonium and other long-lived actinides so long as only thorium and 233U are used as fuel.”
On these points, we agree with the author that the liquid-fluoride thorium reactor, which is currently under development by Flibe Energy, offers many potential advantages over current energy generation technologies, both conventional and nuclear.
“Far better to invest in the renewable technologies that are already shaping our national and global future, and whose cost is rapidly falling – in the process developing valuable UK-based expertise and technologies, and accelerating the renewables revolution.”
Despite telling us that solar and wind power are better choices than thorium, Mr. Tickell yields a fair number of our points throughout his piece:
“[Thorium] is three to four times more abundant in the Earth’s crust than uranium, and is especially plentiful in Australia, Norway, India, the USA and China.”
“[Molten-salt reactors] offer several potential advantages over conventional reactor designs: the reactor and its cooling circuits operate at near atmospheric pressure, reducing the chance of any explosion, [and] in the event of a reactor overheating, the fuel can simply drain out into a secondary container and the fission chain reaction will halt, reducing the risk of reactor meltdowns such as those experienced at Chernobyl and Fukushima.”
“thorium reactors do not produce plutonium”
“the design of molten salt reactors does indeed mitigate against reactor meltdown and explosion”
“important elements of the LFTR technology were proven during the 1970s”
“LFTRs are theoretically capable of a high fuel burn-up rate”
“the thorium fuel cycle does indeed produce very low volumes of plutonium and other long-lived actinides so long as only thorium and 233U are used as fuel.”
On these points, we agree with the author that the liquid-fluoride thorium reactor, which is currently under development by Flibe Energy, offers many potential advantages over current energy generation technologies, both conventional and nuclear.
With regard to the thorium fuel cycle, reprocessing and
proliferation, however, they are still all too willing to spread half-truths
and untruths that have the semblance of plausibility because of their
relationships to the limitations of legacy uranium fuel cycle technologies.
Thorium advocates have previously met personally with two of the contributors
on this article and extended an open invitation to discuss any concerns they may
have about LFTR technology. Many of their environmental advocate associates
have elected to keep an open mind about the potential of thorium and LFTR
technology as a safe, sustainable, cost effective form of nuclear energy. This
openness was not reciprocated by the author of this article.
To be clear, we do not advocate that thorium and LFTR immidiately displace non-carbon energy generation, but rather displace coal-fired power generation and other forms of fossil fuel dependence. Let both industries develop their best possible solutions to our shared problems and then let the market realities and fundamentals of physics determine the extents of the respective roles in the world’s energy future.
Nonetheless, we thank the authors for elevating the general public awareness and public discussion of thorium and LFTR technology (ug, no we don’t, this makes them sound important).
Why, then, does he tell us that LFTR is not “viable”, not “green”, and not “likely”?
Here is a brief summary of that statements used against LFTR (excerpts from Tickell's article will be in italics):
1. Introduction
“With uranium-based nuclear power continuing its decades-long economic collapse, it’s awfully late to be thinking of developing a whole new fuel cycle whose problems differ only in detail from current versions.”
Amory Lovins, Rocky Mountain Institute, March 2009.
To be clear, we do not advocate that thorium and LFTR immidiately displace non-carbon energy generation, but rather displace coal-fired power generation and other forms of fossil fuel dependence. Let both industries develop their best possible solutions to our shared problems and then let the market realities and fundamentals of physics determine the extents of the respective roles in the world’s energy future.
Nonetheless, we thank the authors for elevating the general public awareness and public discussion of thorium and LFTR technology (ug, no we don’t, this makes them sound important).
Why, then, does he tell us that LFTR is not “viable”, not “green”, and not “likely”?
Here is a brief summary of that statements used against LFTR (excerpts from Tickell's article will be in italics):
1. Introduction
“With uranium-based nuclear power continuing its decades-long economic collapse, it’s awfully late to be thinking of developing a whole new fuel cycle whose problems differ only in detail from current versions.”
Amory Lovins, Rocky Mountain Institute, March 2009.
Yes, the better time to more fully develop the thorium fuel cycle would have been 40 years ago when Dr. Alvin Weinberg and his team successfully demonstrated a technology that is vastly superior to current legacy reactor technologies. Fortunately, it is never too late to develop a sustainable long-term solution to world energy needs. The need to provide inexpensive, reliable baseload power for billions, while also minimizing environmental impact, makes now an ideal time to develop the technology to access the world’s greatest reserve of stored energy, thorium.
A number of commentators have argued that most of the problems associated with nuclear power could be avoided by both:
- using thorium fuel in place of uranium or plutonium fuels
- using ‘molten salt reactors’ (MSRs) in place of conventional solid fuel reactor designs.
The combination of these two technologies is known as the Liquid Fluoride Thorium Reactor or LFTR, because the fuel is in form of a molten fluoride salt of thorium and other elements.
The authors have correctly appreciated that LFTR fits within the family of molten salt reactors and that it is the combination of LFTR technology with thorium that uniquely leads to the many benefits claimed.
In this Briefing, we examine the validity of the optimistic claims made for thorium fuel, MSRs and the LFTR in particular. We find that the claims do not stand up to critical scrutiny, and that these technologies have significant drawbacks including: the very high costs of technology development, construction and operation.
Very high costs compared to what? Consider the treatment by Mr. Tickell of the costs associated with his large-scale renewable plan:
“That does not mean that the transition to a renewable energy world will be easy or straightforward. We will need to reconfigure power grids so they operate as networks accepting high volumes of ‘embedded generation’, not just as distribution systems; to build new long-distance electricity links to smooth out fluctuations in supply and demand; to develop the technologies to convert electrical power into liquid fuels for road vehicles and aviation; to create ‘smart grids’ in which the demand for power responds to the available supply; to find ways to store surplus power for those days or weeks when the wind isn’t blowing and the sun isn’t shining; and to waste less of the energy that we produce. All of this will require considerable investment in research, development, manufacture and installation – and will incidentally create many millions of jobs.
All the more reason then not to throw our finite national capital into the bottomless pit of nuclear subsidies.”
(http://www.theecologist.org/News/news_analysis/1482669/renewable_revolution_or_nuclear_nightmare.html)
Mr. Tickell is obviously more willing to embrace massive costs for “considerable investment in research, development, manufacture and installation,” for reworking the entire electrical grid to tolerate intermittent, diffuse power sources. Wind and solar definitely have a place in future energy generation, but is more likely as a significant supplement to reliable baseload power than as that baseload itself.
Mr. Tickell is founder of “Nuclear Pledge” and, along with his contributors and sponsors, has a clear agenda as pro-solar/wind and anti-nuclear. He sees the energy future as a zero sum game. For example Mr. Tickell states, “As for renewables and efficiency, once the bills are paid for decommissioning old power stations, discharging nuclear waste liabilities and subsidising new nuclear build, there’s just not much money left!”(http://www.nuclearpledge.com/)
The original MSRE was designed and built in four years and operated for 20,000 hours for the modest adjusted present-day cost of $80 million US. A modest modern LFTR demonstration could be achieved with several hundred million dollars. Mass production and elimination of the need for an oversized pressure containment structure can greatly reduce the initial capital costs of factory-built commercial units. With LFTR, no complex fuel fabrication is required, the fuel is fully consumed, and the Brayton-cycle power conversion system is more efficient, all leading to significant reductions in cost of operation over legacy reactors.
marginal benefits for a thorium fuel cycle over the currently utilised uranium / plutonium fuel cycles
The benefits of switching to thorium in legacy solid-fueled reactors are indeed marginal, whereas the benefits of using thorium in liquid-fueled LFTR are remarkable.
serious nuclear weapons proliferation hazards
LFTR’s end-to-end fuel cycle offers greatly reduced proliferation risks in large part due to its liquid fuel form and consumption of thorium. Once operational, LFTR’s do not require any further transport, storage or offline preparation or reprocessing of uranium. Elimination of the need for preparation, storage and transportation of resupply fissile material and of storage of unconsumed fissile material in the waste stream are significant advances in non-proliferation relative to legacy fuel cycles. Unlike legacy reactor life-cycles, neither the consumable thorium input nor the waste stream poses any proliferation risk. The thorium fuel can be safely transported and the byproducts safely stored. Other non-proliferation measures are addresses in a later section.
the danger of both routine and accidental releases of
radiation, mainly from continuous ‘live’ fuel reprocessing in MSRs
Most industrial processes involve potentially dangerous chemicals, reactions and byproducts that deserve respect and conscientious engineering and operation. Nuclear energy applications are no different. There is no reason that handling of LFTR’s salts, fuel and fission products cannot be engineered to safety levels commensurate to the actual risks posed. The fluoride salts are well-suited to safe handling of fission products, allowing for ready extraction of and safe handling of gaseous fission products which can be applied to useful industrial processes or safely retained until they have stabilized. Other fission products are safely extractable through chemical processes for medical or industrial use, with the remainder being amenable to vitrification in an inert glassy form for long-term storage.
the very long lead time for significant deployment of LFTRs of the order of half a century — rendering it irrelevant in terms of addressing current or medium term energy supply needs
Any long-term solution could be considered irrelevant for myopic planning horizons, however, LFTR could be demonstrated within the decade and mass production could begin soon thereafter. The proposed “very long lead time” unduly discounts the results of the MSRE and the significant materials innovations of the last 40 years. It is imprudent to focus single-mindedly on near-term solutions to the wholesale exclusion of potentially more significant contributors to sustainable long-term solutions.
1. Background
1.1 What is thorium?
Thorium is a heavy metal named after Thor, the Nordic God of thunder. The naturally occurring isotope, 232Th, is mildly radioactive with a very long half life of 14 billion years. Thorium presents a health hazard mainly from inhalation of dust, and from emissions of the powerfully radioactive gas radon (220Rn).
This should more accurately indicate that the aggregate decay of vast amounts of thorium and uranium within the Earth sustain life on this plant, and that this decay keeps the earth’s interior warm, driving the magnetic fields that shield our atmosphere from the solar winds and driving volcanism and originating geo-thermal energy. This decay also naturally produces radon, but there is no more or less radon generated by operation of LFTRs. Similarly, adoption of LFTR technology will not require additional mining of thorium as it is already a surplus waste product of existing mining operations. In fact, the annual thorium “waste product” from a single typical rare-earth mine could power all of Europe each year.
Much more thorium and uranium dust are released into the atmosphere from combustion of coal than from mining or any other source. Replacement of coal-fired generation with LFTR would actually reduce the release of thorium into the environment as LFTR’s thorium remains entrained in the blanket salt until it is converted and ultimately consumed.
It occurs mainly in deposits of rare earth metals. . As it has few uses requiring more than minimal volumes of material it is considered as radioactive waste — and requires careful and expensive handling to prevent contamination. It is three to four times more abundant in the Earth’s crust than uranium, and is especially plentiful in Australia, Norway, India, the USA and China.
Thorium is abundant throughout the Earth’s crust with higher grade concentrations naturally occurring with rare earth elements. India has entire beaches of thorium monazite sands and Turkey has mines where thorium is present in larger crystalline form. It is rumored that China has been stockpiling its thorium byproducts from mining rare earth ores for a decade, potentially representing a century or more worth of readily accessible energy with molten salt reactor technology.
Thorium has a half life of 14 billion years, which translates into an extremely low level of radioactive decay. Thus, thorium, like natural uranium is not a handling risk in terms of direct contact or proximity. Nonetheless, neither should be inhaled or ingested and so appropriate handling is still required.
Although thorium can be used to make nuclear fuel, it is not fissile. But it is ‘fertile’ – that is, it can be transformed into fissile material. Under neutron irradiation, typically provided by the fission of uranium or plutonium, it breeds the fissile uranium isotope 233U. Thus any thorium fuel cycle needs to be initiated by a supply of existing fissile material.
Yes, each LFTR will require an initial fissile charge to start the thorium fuel cycle, with thorium as the sole input thereafter. In contrast, existing nuclear reactors require fabrication and transport for continual resupply of new fissile uranium fuel. Effectively, LFTR allows us to treat the world’s fissile reserves as a catalyst to utilize the world’s fertile reserves without diminishing the overall fissile inventory or power generation capacity of future generations. Stated otherwise, with LFTR, fissile material represents power generation capacity while fertile material represents consumable stored energy. For example, with LFTR, a two foot sphere of fissile material represents 1GW of power generation capacity in perpetuity. By the same token, a similarly sized beach ball worth of thorium represents the annual consumable input needed to sustain that 1GW of power generation. In contrast, legacy reactors simply burn a portion of the fissile material once, relegating the rest to spent nuclear fuel stockpiles, which proportionately reduces the world’s energy generation capacity. Once-through consumption of fissile material without at least iso-breeding of replacement fissile material from fertile material is akin to wholesale logging without replanting seedlings.
For example, fission of the UK’s 114 tons of plutonium already extracted from legacy spent nuclear fuel could produce sufficient neutrons through fission to convert 114 tons of thorium into 114 tons of 233U. This could in turn provide enough fissile starter charges of 233U to put 114 GW of LFTR power on the UK grid in perpetuity, with inexpensive, abundant thorium as the only ongoing consumable. Consider the vast economic and environmental advantages of decades of resupply of 114 GW with thorium versus solid-uranium fuels.
Other once-through proposals would consume this plutonium only partially and but once to produce a mere fraction of the 114 GWs of energy generation potential, and even this would require increasingly difficult incremental reprocessing to capture ever diminishing yields.
The thorium-uranium fuel cycle has some advantages over the dominant uranium-plutonium cycle, in terms for example, of the reduced production of long-lived actinides and somewhat diminished radio-toxicity overall. However, it also creates new hazards of its own. As far as radioactive fission products are concerned, there is little to choose between the two.
It is surprising to see how little treatment is given to an advantage as significant as exclusion of long-lived actinides/transuranics from the waste stream. This dramatically reduces the time required for long-term waste storage from tens of thousands of years to a few hundred years. “Reduced production” is an unfair characterization of “nearly eliminated.”
Instead, the authors rush to refer to unspecified new hazards and the equivalence of fission product inventories. Yes, most any fission produces a similar inventory of fission products. What is different about LFTR is that the chemistry can be controlled to exclude the precursors that would otherwise absorb neutrons from fission to produce long-lived actinides/transuranics.
The fluoride salt chemistry also allows for extraction and useful industrial and medical application of fission products that would otherwise end up in the waste stream. For example, one fission product in particular, Mo-99, is desperately needed by the medical industry as the precursor for Tc-99m used in over 30 million medical procedures a year. While 5-6 percent of all fissions in any type of reactor produce Mo-99, it has a mere 66 hour half-life, rendering it essentially unrecoverable from conventional solid-fuel rods. In contrast, LFTR’s liquid fuel allows for timely extraction of Mo-99 for tens of millions of life-saving medical imaging procedures, while reducing the waste stream by five percent.
1.2 What are molten salt reactors?
Unlike conventional nuclear reactors which use solid fuel in the form of rods or pellets, molten salt reactors (MSRs) use fuel in the form of a complex mixture of fluoride salts in a molten state. The salt mixture includes the fissile material (fissile isotopes of plutonium and/or uranium), together with any fertile material (such as thorium or 238U) together with other elements.
Unlike conventional nuclear reactors which use solid fuel in the form of rods or pellets, molten salt reactors (MSRs) use fuel in the form of a complex mixture of fluoride salts in a molten state. The salt mixture includes the fissile material (fissile isotopes of plutonium and/or uranium), together with any fertile material (such as thorium or 238U) together with other elements.
Actually, the FLiBe salts are not that complex and are far less complex than any existing solid fuel forms.
The preferred initial fissile charges are 233U, then 235U, with plutonium being very much the least preferred. 238U is excluded from the salts to avoid formation of plutonium-239 from neutron capture. 238U is more useful in a fast breeder reactor than in a LFTR. 238U has only one third the neutron absorption of 232Th and would so affect the neutronics of a LFTR as to render it in operable. In fact, 238U could be used to readily denature the liquid fuel salts, effectively destroying the fuel’s usefulness for any purpose, as a further non-proliferation measure.
The molten fluoride salt serves as the primary coolant, carrying heat away from the reactor, and delivering it to a secondary cooling circuit and ultimately to the steam turbines that generate electricity.
LFTR are preferably coupled to a high-efficiency Brayton-cycle gas turbine. Conventional steam turbines cannot take advantage of the higher temperatures achievable by a LFTR.
In principle, MSR’s offer several potential advantages over conventional reactor designs:
- the reactor and its cooling circuits operate at near atmospheric pressure, reducing the chance of any explosion
- In the event of a reactor overheating, the fuel can simply drain out into a secondary container and the fission chain reaction will halt, reducing the risk of reactor meltdowns such as those experienced at Chernobyl and Fukushima
LFTR’s negative temperature coefficient would naturally prevent overheating during operation. The freeze plug is useful in the event of complete power loss or to downshift the reactor if needed for any reason.
But before ‘production’ MSRs can be built, there are significant technical problems to be overcome, among them:
- the development of corrosion-resistant materials capable of surviving for decades in a uniquely hostile environment — highly corrosive and subject to intense radiation including neutron bombardment
While there is still testing needed to round out the qualification data, the modified Hastelloy-N used in the MSRE proved up to the task of handling the high temperature salts, radiation and neutron bombardment for over 20,000 hours of MSRE operation. A modern LFTR will employ many advanced materials technologies that have been developed in the 40 years since the MSRE. Nuclear materials researchers have created newer alloys that promise superior robustness in molten salt applications.
- and, the continuous fuel reprocessing that MSRs demand, requiring the development of hazardous, complex and currently experimental pyro-processing and electro-refining technologies on a production scale.
Some proposed pyro-processing associated with Integral fast reactor (IFR) development would employ molten salt chemistry. LFTRs, however, do not employ pyro-processing or electro-refining, but rather employ well-understood fluorination, reduction, distillation and electrolysis processes that are already conducted on tonnage scales in the nuclear industry using available equipment. Admittedly, some refinement of chemical processes will be required to selectively extract fission products for industrial applications.
If these technologies are successfully developed – and
it cannot be taken for granted that they ever will — they are likely to be very
expensive. Moreover, reprocessing will always represent a weak link from a
safety and proliferation perspective.
Actually, the fluorination, reduction, distillation and electrolysis processes are already successfully developed and safely operated in other nuclear and industrial applications around the world. In a LFTR, these processes are conducted within the reactor containment in a closed-cycle in which the reagents are continuously regenerated, with none of the fissile material leaving the reactor containment. While fission products may be further processed offline, there is no offline reprocessing of any fissile materials, dramatically reducing proliferation concerns relative to legacy reactor fuel cycles.
Current State of Play
2.1 Actual thorium reactors
Thorium fuel has so far been used in about 30 operational reactors in conjunction with fissile uranium (235U / 233U) or plutonium (239Pu) to initiate the fuel cycle. Most of these were located in the USA, Germany, Netherlands and India. A single example operated in the UK, from 1965 to 1976: the Dragon Reactor at Winfrith, a helium-cooled test reactor evaluating fuel and materials for the European high-temperature reactor programme. It is currently partially decommissioned.
Most thorium reactors have been of conventional designs originally intended for uranium fuel, such as pressurised water reactors, boiling water reactors and pressurised heavy water reactors. But thorium has also been included in more exotic designs, notably the molten salt breeder experiment (MRSE) reactor (see 2.3, below), the thermal breeder reactor (USA), and the liquid metal fast reactor (India).
The only operational thorium reactors today are in India, which possesses abundant thorium reserves but little uranium. These are all solid fuel reactors. As of 2010, India had used only a small amount of thorium – approximately one tonne – in its reactors.
2.2 Planned thorium reactors
In December 2011, India announced its plans for a new generation of Advanced Heavy Water Reactors using a plutonium / uranium / thorium MOX (mixed oxide) fuel. The programme would begin with an initial test reactor whose construction could commence in 2013. Again, this would not be a molten salt reactor but would use conventional solid fuel.
Norway’s Thor Energy is also intending to develop a thorium-plutonium MOX nuclear fuel, aimed at replacing conventional fuels in light water reactors (LWRs). It is currently seeking investment to irradiate thorium-plutonium oxide fuel pins in simulated LWR conditions in the Halden fuel-testing reactor. A separate project is to optimise thorium-plutonium fuels for boiling water reactors, while maximising the breeding of 233U. Thor Energy anticipates that 25-30% of power output could arise from the thorium.
Proposals have been made to construct LFTR reactors in China, Japan and the US (see 2.3). Initially these would be research reactors and the first ‘production’ LFTR would appear to be 20-30 years away (see 2.4).
The development and deployment timelines are highly subject to funding and political will. We propose to dramatically accelerate both of the former by mustering sufficient quantities of the later.
2.3 Actual molten salt reactors
The molten salt reactor was originated in the 1950s as a potential power source for the USAF’s fleet of high altitude nuclear bomber aircraft. A working reactor was produced (under the Airborne Reactor Experiment) programme, but never commissioned.
The aircraft reactor program required a new reactor design that was simple and safe enough to operate onboard an airplane. These unique constraints ultimately led to consideration of liquid fuels and the greatly simplified fuel cycle that this fuel form enabled.
The technology was further developed at Oak Ridge National Laboratory in the 1960s under its MSRE (Molten Salt Reactor Experiment). The 7MW reactor employed fluoride salts of uranium and plutonium as fuel. In the 1970s, Oak Ridge built its Molten Salt Breeder Reactor (MSBR), which used as fuel fluoride salts of uranium, thorium and plutonium as its fuel.
2.1 Actual thorium reactors
Thorium fuel has so far been used in about 30 operational reactors in conjunction with fissile uranium (235U / 233U) or plutonium (239Pu) to initiate the fuel cycle. Most of these were located in the USA, Germany, Netherlands and India. A single example operated in the UK, from 1965 to 1976: the Dragon Reactor at Winfrith, a helium-cooled test reactor evaluating fuel and materials for the European high-temperature reactor programme. It is currently partially decommissioned.
Most thorium reactors have been of conventional designs originally intended for uranium fuel, such as pressurised water reactors, boiling water reactors and pressurised heavy water reactors. But thorium has also been included in more exotic designs, notably the molten salt breeder experiment (MRSE) reactor (see 2.3, below), the thermal breeder reactor (USA), and the liquid metal fast reactor (India).
The only operational thorium reactors today are in India, which possesses abundant thorium reserves but little uranium. These are all solid fuel reactors. As of 2010, India had used only a small amount of thorium – approximately one tonne – in its reactors.
2.2 Planned thorium reactors
In December 2011, India announced its plans for a new generation of Advanced Heavy Water Reactors using a plutonium / uranium / thorium MOX (mixed oxide) fuel. The programme would begin with an initial test reactor whose construction could commence in 2013. Again, this would not be a molten salt reactor but would use conventional solid fuel.
Norway’s Thor Energy is also intending to develop a thorium-plutonium MOX nuclear fuel, aimed at replacing conventional fuels in light water reactors (LWRs). It is currently seeking investment to irradiate thorium-plutonium oxide fuel pins in simulated LWR conditions in the Halden fuel-testing reactor. A separate project is to optimise thorium-plutonium fuels for boiling water reactors, while maximising the breeding of 233U. Thor Energy anticipates that 25-30% of power output could arise from the thorium.
Proposals have been made to construct LFTR reactors in China, Japan and the US (see 2.3). Initially these would be research reactors and the first ‘production’ LFTR would appear to be 20-30 years away (see 2.4).
The development and deployment timelines are highly subject to funding and political will. We propose to dramatically accelerate both of the former by mustering sufficient quantities of the later.
2.3 Actual molten salt reactors
The molten salt reactor was originated in the 1950s as a potential power source for the USAF’s fleet of high altitude nuclear bomber aircraft. A working reactor was produced (under the Airborne Reactor Experiment) programme, but never commissioned.
The aircraft reactor program required a new reactor design that was simple and safe enough to operate onboard an airplane. These unique constraints ultimately led to consideration of liquid fuels and the greatly simplified fuel cycle that this fuel form enabled.
The technology was further developed at Oak Ridge National Laboratory in the 1960s under its MSRE (Molten Salt Reactor Experiment). The 7MW reactor employed fluoride salts of uranium and plutonium as fuel. In the 1970s, Oak Ridge built its Molten Salt Breeder Reactor (MSBR), which used as fuel fluoride salts of uranium, thorium and plutonium as its fuel.
The MSRE, with over 20,000 hours of highly successful operation, remains the only reactor to have operated on each of the three fissile fuels and still holds the record for the longest period of continuous reactor operation. (See: www.energyfromthorium.com/pdf, ORNL semi-annual MSRE reports).
The MSBR, on the other hand, never made it off the drawing board.
Planned molten salt reactors
There are a number of proposals to build MSRs:
In January 2011, the Chinese Academy of Sciences announced plans to develop the LFTR technology into commercial reactors. But it warned that 20 to 30 years of research and development would probably be needed before an LFTR was operational.
There are a number of proposals to build MSRs:
In January 2011, the Chinese Academy of Sciences announced plans to develop the LFTR technology into commercial reactors. But it warned that 20 to 30 years of research and development would probably be needed before an LFTR was operational.
The Chinese are leading an aggressive molten salt reactor development effort with over half a billion dollars and several hundred personnel already committed to the effort, and plans for significant growth over the next five years. The original MSRE was designed, built and began operation in a mere four years with barely 200 personnel and $80 million US present day equivalent. The Chinese are building on the existing MSRE designs and will likely achieve criticality within the decade. At this rate, it is unlikely it will take 20 years for the Chinese to begin mass-production of molten salt reactors.
Flibe Energy was set up in 2011 to develop LFTR technology in the USA and worldwide. Its initial intention is to build a small test reactor. Ultimately, it aims to bring about a world with many thousands of LFTRs.
Flibe Energy’s mission is to supply the world with sustainable, affordable energy, water, medical isotopes and fuel. In the near term, Flibe Energy is pursuing development and demonstration of small LFTR as a precursor to larger modular commercial reactors. Ultimately, LFTR’s can be used not only to produce electricity, but to desalinate sea water, synthesize liquid fuels, and produce a range of lifesaving medical isotopes.
The FUJI LFTR project in Japan is attempting to raise ¢G300 million to build a 10 MW ‘MiniFUJI’ research reactor. Following the 2011 Fukushima catastrophe, the project has a low chance of attracting the necessary finance.
the UK’s Weinberg Foundation was established in September 2011 to act as a communications, debate and lobbying hub to promote thorium energy and the LFTR in particular. There are currently no plans in the UK to build an actual LFTR.
Despite the resurgence of interest in the MSR / LFTR technology, there are no concrete plans to build even a single such reactor. China currently appears most likely to provide the funding necessary to develop LFTR technology due to that country’s relatively large nuclear programme and the government’s willingness to invest in new energy generation technologies. But even there any production-scale LFTR is unlikely to materialise for 20-30 years.
2.5 New-found interests — why?
Several factors underlie the current vogue of interest in thorium reactors. Perhaps the most important is the desire for energy and nuclear independence in countries with large thorium reserves and little uranium, or which have concerns about long-term price of uranium and its availability. This would appear to apply to India, China, the USA and Norway.
Noting the large volumes of surplus thorium produced as waste in the mining of valuable rare earth metals, there is also a clear commercial interest among the mining companies concerned to give value to this waste. However, we have no evidence of any efforts by mining companies to drive forward the thorium project.
A more significant factor is perhaps a deeply-rooted techno-optimism in human psychology — the desire to believe that one or other technology provides ‘the answer’ to deep-rooted problems. Faced with the prospect of ‘peak oil’ and accelerating climate change from the burning of fossil fuels, those who are sceptical about the potential of renewable energy sources will naturally incline towards some other answer. For some, it would seem that thorium fills that particular ‘desire gap’.
“Techno-optimism” vs. “luddite-pessimism.” It is not a matter of optimism, pessimism or skepticism, the fundamentals of physics dictate that nuclear fission offers one-million times the energy density of combustion and can be accessed on demand, day and night, wind or calm, and without production of CO2.
The established nuclear industry in the UK has little interest in thorium as such, since any use of thorium would create far more cost than it ever saved. However, the mere idea that there exists a notionally ‘green’ version of nuclear power could be seen by the nuclear industry as positive in public relations terms, and useful in promoting the persistence of nuclear power in the UK’s electricity mix.
What is the basis for this “far more cost than it ever saved?” There will obviously be an upfront development cost required, but LFTR offers greatly reduced manufacturing costs by obviating the massive pressure containment vessel, nearly eliminates fuel fabrication costs, achieves nearly 50% better power conversion efficiency, and reduces the amount and length of waste storage required. These costs advantages will offset the development costs within a reasonable timeframe.
The existing nuclear industry establishment is not advocating or embracing LFTR as it is a fair deviation from their existing business models and technologies that are based on fabrication of solid-uranium fuel assemblies.
“useful in promoting the persistence of nuclear power in the UK’s electricity mix” is the author’s first clear recognition of the perceived threat to his agenda.
Thorium claims — and the reality
Numerous advantages for thorium as a nuclear fuel and for the LFTR design over conventional solid fuel reactors have been claimed. In this section we consider each of these claims in turn.
Abundance of thorium relative to uranium
Claim: Thorium is several times more abundant in the Earth’s crust than uranium.
Response: Thorium (232Th) is indeed more abundant than uranium, by a factor of three to four. But whereas 0.7% of uranium occurs as fissile 235U, none of the thorium is fissile. The world already possesses an estimated 1.2 million tonnes of depleted uranium (mainly 238U), like thorium a fertile but non-fissile material. So the greater abundance of thorium than uranium confers no advantage, other than a very marginal advantage in energy security to those countries in which it is abundant.
Claim: Thorium is several times more abundant in the Earth’s crust than uranium.
Response: Thorium (232Th) is indeed more abundant than uranium, by a factor of three to four. But whereas 0.7% of uranium occurs as fissile 235U, none of the thorium is fissile. The world already possesses an estimated 1.2 million tonnes of depleted uranium (mainly 238U), like thorium a fertile but non-fissile material. So the greater abundance of thorium than uranium confers no advantage, other than a very marginal advantage in energy security to those countries in which it is abundant.
232Th is one neutron capture away from being directly useful for energy generation. LFTRs can receive natural thorium, which is as abundant and inexpensive as lead, and convert it into fissile uranium 233 that is as useful in a reactor as 235U, which, in contrast, is as rare as platinum and similarly costly. LFTR can then turn that fissile fuel into electricity worth several multiples more, all while generating additional valuable byproducts. One ton of initial fissile charge, about a two foot sphere, could kick start the thorium fuel cycle to produce 1GW of electricity in perpetuity with fertile thorium as the sole input thereafter. It is disingenuous to attempt to marginalize the usefulness and tremendous value of thorium simply because is not immediately fissile.
Thorium is the ultimate flex fuel being uniquely consumable in both a thermal spectrum reactor and a fast spectrum reactor. 238U on the other hand has only one third the neutron absorption of thorium and thus is not useful in LFTR. In fact, substitution of 238Ufor 232Th would make a LFTR inoperable and could even be used to denature the fuel to render the fuel useless for all purposes.
The 238U, like thorium, is fertile, but is best utilized in a fast spectrum breeder reactor, which are not as safe, simple, or cost-effective as the thermal spectrum LFTR. For context, there are many operating and planned thermal spectrum reactors, whereas there have been very few fast breeders ever successfully operated. The abundance of diesel would be little argument for doing away with gasoline when most of today’s cars run on gasoline.
Relative utility of thorium and uranium as fuel
Claim: 100% of the thorium is usable as fuel, in contrast to the low (~0.7%) proportion of fissile 235U in natural uranium.
Response: Thorium must be subjected to neutron irradiation to be transformed into a fissile material suitable for nuclear fuel (uranium, 233U). The same applies to the 238U that makes up depleted uranium, which as already observed, is plentiful. In theory, 100% of either metal could be bred into nuclear fuel. However, uranium has a strong head start, as 0.7% of it is fissile (235U) in its naturally-occurring form.
238U and 232Th are only on equal ground and of equal utility in the limited context of fast spectrum reactors, the less desirable spectrum for power generation. 232Th is far superior as a nuclear fuel in that it can be fully consumed in the more desirable thermal spectrum. The appropriate evaluation of consumability of natural thorium and natural uranium is in the context of the thermal spectrum, which is where virtually all of today’s power generating reactors operate. Again, the abundance of diesel would not be an argument for discontinuing use of gasoline, when most of today’s vehicles use gasoline.
Nuclear weapons proliferation
Claim: thorium reactors do not produce plutonium, and so create little or no proliferation hazard.
Response: thorium reactors do not produce plutonium. But an LFTR could (by including 238U in the fuel) be adapted to produce plutonium of a high purity well above normal weapons-grade, presenting a major proliferation hazard.
A LFTR is by definition configured to run on thorium and would not run on 238U. LFTR would be configured to exclude 238U to avoid formation of plutonium and would be configured such that any substitution of non-fissile 238U in place of 232Th or fissile 233U would cause the reactor to shutdown. In fact, 238U is what would be used in an emergency to rapidly denature the fuel to render the fuel useless and effectively shut the reactor down until a new initial fissile charge and fresh fuel salt could be provided.
“Risks” are a composite measure of the severity of a hypothetical outcome and the likelihood of that outcome. In this case, “major proliferation risk” would seem to imply both a serious outcome and a high likelihood of that outcome. While nuclear proliferation obviously presents serious outcomes, there are reasonable engineering and security measures available to minimize the likelihood of those outcomes. While some have taken hypothetical scenarios to the extreme to postulate serious outcomes, none have proffered a reasonable basis for any practical likelihood of those outcomes. LFTR can be engineered to minimize any proliferation risks and proper regulatory oversight and security can mitigate any remaining risk.
Of the tens of thousands of warheads in the world’s arsenal, none are based on the thorium fuel cycle, and for good reason.
the need for fissile material (plutonium or uranium) to
initiate the thorium fuel cycle, which could be diverted
LFTR requires a one-time initial fissile charge to initiate the thorium fuel cycle. In contrast, the world’s many uranium-based reactors require constant fabrication, storage, and transport of resupply fissile uranium. Many of the benefit claims of LFTR are made in comparison to legacy nuclear technologies and the benefit is often a great reduction in risk, not its complete absence. Highly enriched uranium targets are routinely delivered to medical isotope reactors around the globe, always with suitable military security measures. Proportionate measures would likely be required for delivery of each one-time initial fissile charge. This is a dramatic reduction in the level and frequency of security measures currently required by constant fissile resupply demands and storage of fissile material remaining in spent fuel stockpiles for legacy reactors.
This may not be sufficient to appease hardline anti-nukes, but all energy production has some forms of risks, and LFTR reduces the inherent risks of nuclear more than any other technology. Each of the three largest hydro-electric dam failures killed more people than have all of the nuclear incidents combined or than would a perfect storm of highly improbable hypothesized worst-case nuclear-energy-based scenarios. Similarly, coal-fired production releases more uranium, thorium, mercury and toxins into the biosphere than nuclear energy ever has or likely ever will. The risks must be considered alongside the benefits and the risks of alternatives.
Nuclear energy offers a million-to-one the benefit of combustion of any form and similar magnitude gains in reliability over existing renewable energy forms. The magnitude of these benefits merits some level of tolerance of a sufficiently mitigated corresponding set of risks. A position of no tolerance of any risk of any kind would paralyze all energy generation and society. We engineer dams to avoid floods, bridges to avoid collapse and we can engineer LFTRs to suitably mitigate corresponding risks, both real and perceived.
the production of fissile uranium 233U.
Yes, LFTR produces fissile 233U fuel, which is all retained inside the reactor containment and does not require offline reprocessing or storage at any point. LFTR only produces as much fissile material as it consumes, in a closed self-sustaining cycle, without increases in the quantity of the fissile inventory. Stated otherwise, LFTRs merely replenish the fissile material that is consumed. Thus, LFTRs cannot be used to produce surplus fissile material for diversion. LFTRs would be configured such that any attempted diversion of fissile material would cause the reactor to shutdown, which would be readily detectable by regulators. Also, during operation, the fuel salt is at extreme temperatures and very radiologically hot. It would take significant additional remotely operated equipment and expertise and a considerable amount of time to accomplish any extraction of fissile material from an operational LFTR. Again, the likelihood of any serious hypothetical outcome from the diversion of replenishment 233U within a LFTR is so small as to render the practical risk extremely small.
Claim: the fissile uranium (233U) produced by thorium reactors is not ‘weaponisable’ owing to the presence of highly radiotoxic 232U as a contaminant.
Response: 233U was successfully used in a 1955 bomb test in the Nevada Desert under the USA’s Operation Teapot and so is clearly weaponisable notwithstanding any 232U present. Moreover, the continuous pyro-processing / electro-refining technologies intrinsic to MSRs / LFTRs could generate streams of 233U very low in 232U at a purity well above weapons grade as currently defined.
233U is fissile and so was tested along with other fissile materials and found impractical for weapons purposes. The lone relevant Teapot test was based on additional of 233U to a plutonium weapon and the yield was reportedly so greatly diminished that not a single weapon attempt or existing weapon has been based on the thorium fuel cycle since. It is far more likely that a weapons effort would be based on gaseous diffusion or centrifuge enrichment of natural uranium.
LFTRs high temperatures and newly-formed highly-radioactive fission products effectively protect the 233U from any nefarious efforts in the short term and the formation of 232U and its hard-gamma emitters strongly disincentivise any nefarious efforts in the long term. Again, the issue of risk is not just one of potential severity of an outcome, but of the remoteness of the likelihood of that outcome. There are a number of characteristics of LFTR and its thorium-derived fuel that render the likelihood of any diversion extremely unlikely and the risk correspondingly low.
LFTRs do not employ pyro-processing or electro-refining and the closed-cycle LFTR fuel replenishment system does not produce a divertible stream, but rather merely replaces the fissile that is being consumed. Again, there is no surplus production of fissile fuel and it is disingenuous and misleading to imply that LFTRs would produce an accessible stream of fissile material.
Safety
Claim: LFTRs are intrinsically safe, because the reactor operates at low pressure and is and incapable of melting down.
Response: the design of molten salt reactors does indeed mitigate against reactor meltdown and explosion. However, in an LFTR the main danger has been shifted from the reactor to the on-site continuous fuel reprocessing operation — a high temperature process involving highly hazardous, explosive and intensely radioactive materials. A further serious hazard lies in the potential failure of the materials used for reactor and fuel containment in a highly corrosive chemical environment, under intense neutron and other radiation.
The two largest risks of existing nuclear power plants are depressurization events and meltdowns. These two risks have been essentially eliminated by LFTR’s low pressure operation and liquid fuel form and these benefits cannot be overstated. The scale of the major accident risks eliminated by LFTR and the scale of the risks of fission product handling are vastly different. The challenges of radioactive fission products are inherent in any form of fission, however, the chemical stability of LFTRs salts, the suitability of fluoride chemistry to fission product handling and good engineering can sufficiently mitigate fission product handling risks.
The “on-site” fuel reprocessing takes place within a containment structure with sufficient engineered safety systems.
The materials tested during the MSRE proved surprisingly robust during the 20,000 hours of operation of the MSRE. More modern materials show promise of being sufficiently robust even in the high-temperature, high-neutron bombardment, and high-radiation environment of a nuclear reactor. There are stringent engineering tests to be performed to qualify these materials for nuclear applications. They will only be employed when the engineers and regulators are satisfied that no serious hazards remain in using those materials in those nuclear applications.
State of technology
Claim: the technology is already proven.
Response: important elements of the LFTR technology were proven during the 1970s Molten Salt Breeder Reactor (MSBR) at Oak Ridge National Laboratory. However, this was a small research reactor rated at just 7MW and there are huge technical and engineering challenges in scaling up this experimental design to make a ‘production’ reactor.
LFTR technology is very scalable, with successful demonstrations of 2.5 and 7.5 MWth reactors. The original ORNL MSRE team prepared detailed plans for a 250 MWe reactor and preliminary plans for a multi-gigawatt reactor. The MSRE team did demonstrate many key aspects of molten salt reactor operation. They also identified the gaps that remained for scalability of the technology and expressed confidence that each of these gaps had reasonable engineering solutions and that none of the gaps presented an insurmountable obstacle to scalability. The remaining original MSRE personnel continue to affirm their confidence in the viability and scalability of LFTR technology.
Specific challenges include:
developing materials that can both resist corrosion by liquid fluoride salts including diverse fission products, and withstand decades of intense neutron radiation;
developing materials that can both resist corrosion by liquid fluoride salts including diverse fission products, and withstand decades of intense neutron radiation;
Modern materials (1 to 2% niobium modified Hastelloy-N) show great promise for robustness even in the intense environment of a high-temperature nuclear reactor. The original MSRE team has already overcome many of the materials issued identified in the early MSRE reports and modern materials and materials engineers are up to the remaining challenges.
scaling up fuel reprocessing techniques to deal safely and reliably with large volumes of highly
radioactive material at very high temperature;
“Fuel reprocessing” in the context of legacy solid spent fuel is indeed a difficult and complicated process, leading to the many thousands of tons of unprocessed spent nuclear fuel alongside existing reactors. With LFTR, however, the thorium fuel cycle is conducted as a closed cycle within the reactor containment. Extraction of fission byproducts and offline fission product separation for commercialization or storage are scalable processes that will demand due respect and conscientious engineering and operation.
keeping radioactive releases from the reprocessing operation to an acceptably low level;
All fission product separation processes and facilities will have offgas systems with appropriate sequestration technologies and processes to meet applicable standards.
achieving a full understanding of the thorium fuel cycle.
The extensive MSRE research and other international thorium fuel cycle research provide a significant technological base and understanding for utilizing the thorium fuel cycle.
Nuclear waste
Claim: LFTRs produce far less nuclear waste than conventional solid fuel reactors.
Response: LFTRs are theoretically capable of a high fuel burn-up rate, but while this may indeed reduce the volume of waste, the waste is more radioactive due to the higher volume of radioactive fission products. The continuous fuel reprocessing that is characteristic of LFTRs will also produce hazardous chemical and radioactive waste streams, and releases to the environment will be unavoidable.
Claim: LFTRs produce far less nuclear waste than conventional solid fuel reactors.
Response: LFTRs are theoretically capable of a high fuel burn-up rate, but while this may indeed reduce the volume of waste, the waste is more radioactive due to the higher volume of radioactive fission products. The continuous fuel reprocessing that is characteristic of LFTRs will also produce hazardous chemical and radioactive waste streams, and releases to the environment will be unavoidable.
Per unit of fission energy released, all reactors produce similar quantities of fission products and radioactivity. The real difference with LFTR and its enhanced liquid fuel consumption is that they require a greatly reduced overall volume of fuel input to achieve the same level of energy release and attendant fission product inventory. Thus, LFTR’s waste per unit of energy release is not more radioactive, its waste stream is simply greatly reduced in volume and more concentrated for lack of dilutive unburned uranium. In contrast, spent fuel rods from legacy reactors still contain the majority of the original unused uranium in addition to all of the fission products accumulated from that portion of the fuel that was fissioned. Thus, it is rather misleading of the author to say that LFTR’s waste is more radioactive; rather LFTRs fission products are simply more efficiently packaged.
Moreover, a portion of LFTRs fission product stream can be separated into useful isotope streams for medical and industrial applications, eliminating these products from LFTR’s long-term waste stream. The remaining fission products are amenable to vitrification in a stable glass form suitable to retain the fission products until they have suitably decayed to background radiation levels.
Claim: Liquid fluoride thorium reactors generate no high-level waste material.
Response: This claim, although made in the report from
the House of Lords, has no basis in fact. High-level waste is an unavoidable
product of nuclear fission. Spent fuel from any LFTR will be intensely
radioactive and constitute high level waste. The reactor itself, at the end of
its lifetime, will constitute high level waste.
This particular instance of a claimed LFTR benefit unfortunately confuses the elimination of long-lived actinides/transuranics with the broader classification of high-level wastes. All fission will produce some high-level wastes.
Claim: the waste from LFTRs contains very few
long-lived isotopes, in particular transuranic actinides such as plutonium.
Response: the thorium fuel cycle does indeed produce very low volumes of plutonium and other long-lived actinides so long as only thorium and 233U are used as fuel. However, the waste contains many radioactive fission products and will remain dangerous for many hundreds of years. A particular hazard is the production of 232U, with its highly radio-toxic decay chain.
Response: the thorium fuel cycle does indeed produce very low volumes of plutonium and other long-lived actinides so long as only thorium and 233U are used as fuel. However, the waste contains many radioactive fission products and will remain dangerous for many hundreds of years. A particular hazard is the production of 232U, with its highly radio-toxic decay chain.
The significance of elimination of long-lived actinides/transuranics from the waste stream merits more discussion. Spent nuclear fuel (SNF) from legacy reactors requires geologic storage for tens of thousands of years due to the long-lived transuranics remaining in the SNF. LFTR fully consumes it fuel and the fuel chemistry is controlled to minimize production of these long-lived products. Secondly, LFTR’s liquid fuel form enables ready separation of the fission products into useful units for medical and industrial applications, further reducing the volume and radioactivity of the waste stream. The majority, by mass, of the remaining waste stream will decay to stability within 10 years and the remaining minority will decay to stability within hundreds rather than tens of thousands of years. LFTR’s residual waste stream is amenable to storage in a vitrified glassy form that is completely inert and safely storable until it drops to background radiation levels. Of these benefits, the exclusion of transuranics from LFTR waste streams is perhaps the most significant differentiator over legacy spent nuclear fuel.
It should be noted that many fission products have great societal value, particularly owing to their radioactivity, while others find use after they have stabilized. Mo-99 (66 hour half life) is the precursor for Tc-99m (6 hour half life) used in tens of millions of medical imaging procedures annually. Similarly, the thorium fuel cycle produces a family of alpha-emitters that have been shown in clinical trials to be highly effective against dispersed cancers and virtually any other disease for which a targeted delivery mechanism is available. A few short years after LFTRs begin to generate power for society, they will also generate a number of highly-effective cancer fighting radioisotopes. Many industrial processes and products rely on use of radioisotopes or their stable daughter products. Many common elements are formed through the natural radioactive decay of once highly-radioactive material. For example, Bismuth 209, the active ingredient in Pepto-Bismol is effectively beneficial nuclear waste as the natural end product of a long nuclear decay chain.
Claim: LFTRs can ‘burn up’ high level waste from conventional nuclear reactors, and stockpiles of plutonium.
Response: if LFTRs are used to ‘burn up’ waste from conventional reactors, their fuel now comprises 238U, 235U, 239Pu, 240Pu and other actinides. Operated in this way, what is now a mixed-fuel molten salt reactor will breed plutonium (from 238U) and other long lived actinides, perpetuating the plutonium cycle.
LFTRs would not be configured to directly consume waste fuel from conventional reactors, but would preferably consume clean 233U fuel bred in more specialized reactors through fission of unburned fissile material remaining in spent nuclear fuels. A waste-burning molten salt reactor would be configured differently than a clean 233U- or 235U-fueled LFTR. Such a molten salt reactor would not be a typical power generating LFTR but would serve primarily to breed clean 233U initial fissile charges for other conventional LFTRs. For example, a few centralized waste-burning molten salt reactors would fission the remaining fissile materials from spent nuclear fuel, using the released neutrons to convert thorium into clean 233U, free from any of the long-lived actinides from the original fuel. In combination with these specialized waste-burning reactors, LFTRs fuel cycle can utilize the remaining fissile materials and potentially the remaining fertile materials, reducing the volume of legacy spent nuclear fuel, reducing the overall radioactive term of legacy spent nuclear fuel, and proportionately reducing the long-term geologic storage requirements for legacy spent nuclear fuel.
Cost of electricity
Claim: the design of LFTRs tends towards low construction cost and very cheap electricity.
Response: while some elements of LFTR design may cut costs compared to conventional reactors, other elements will add cost, notably the continuous fuel reprocessing using high-temperature ‘pyro-processing’ technologies. Moreover, a costly experimental phase of ~20-40 years duration will be required before any ‘production’ LFTR reactors can be built.
Again, LFTR does not require pyro-processing. LFTR does indeed lower construction costs by eliminating the need for massive pressure containment vessels, use of more-compact Brayton-cycle turbines, and with factory production of LFTR modules. The closed-cycle fuel process is conducted within the reactor containment and is configured to regenerate its own reactants. The LFTR fuel extraction and reconstitution processes largely rely on established chemical processes using available equipment. While the high temperatures will require some refinement of these processes and modification to the equipment, these costs will be nominal compared to the cost-savings of LFTR construction and operation over those of legacy reactors.
It is very hard to predict the cost of the technology that finally emerges, but the economics of nuclear fuel reprocessing to date suggests that the nuclear fuel produced from breeder reactors is about 50 times more expensive than ‘virgin’ fuel. It therefore appears probable that any electricity produced from LFTRs will be expensive.
Indeed, reprocessing of legacy spent fuels to recoup only a further fraction of the available energy in a legacy reactor is an expensive and difficult process with increasing difficulty and diminishing returns with each successive reprocessing.
LFTR does not require reprocessing of spent fuel in the conventional sense of the word as there is no fuel left once it is spent. Instead, LFTR continuously replenishes the 233U being consumed with simple thorium being the primary consumable and all of the fuel-cycle reagents being readily regenerated in a closed system.
Again, the author appreciates a fair number of the other
benefits of LFTR, but with regard to the thorium fuel cycle, is still
unfortunately wedded to lingering biases against the limitations of legacy
uranium fuel cycle technologies.
We must also consider the prospect that relatively novel or immature energy sources, such as photovoltaic electricity and photo-evolved hydrogen, will have become well established as low-cost technologies long before LFTRs are in the market.
How is this not the very techno-optimism that was so recently maligned? Each of these technologies should be pursued far enough to determine their viability, as should LFTR technology.
Timescale
Claim: Thorium and the LFTR offer a solution to current and medium-term energy supply deficits.
Response: The thorium fuel cycle is immature. Estimates from the UK’s National Nuclear Laboratory and the Chinese Academy of Sciences (see 4.2 below) suggest that 10-15 years of research will be needed before thorium fuels are ready to be deployed in existing reactor designs. Production LFTRs will not be deployable on any significant scale for 40-70 years.
The immediacy of a solution should not be the primary metric of the viability or desirability of the solution. No one is under any delusion about immediate resolution of current energy issues with thorium and LFTR technology. Technology development and deployment timelines are compressible to a point depending on funding and public support. A slow-rolling research approach to advanced fuel cycle and reactor technologies by national laboratories (NNL) and academic institutions (CAS) could easily be stretched out for many decades as proposed. A focused private development effort, by contrast, could deliver a LFTR demonstration well within the decade and mass production within the next. For example, consider the stark contrast between the developmental track records both in terms of cost and time between Space-X and NASA.
Thorium / LFTR prospects
4.1 Timescales for thorium fuel
The thorium fuel cycle is immature and unready for production-scale deployment. Although thorium fuels have been used in approximately 30 reactors, their nuclear dynamics and operational performance remain poorly characterised.
Technology is never mature until it is matured through technology development efforts. Obviously a modern demonstration of a molten salt reactor such as LFTR would precede any production-scale deployment. This demonstration reactor would effectively answer many of the questions posed by the authors and others about suitability of materials and actual costs of the fuel cycle and other aspects of LFTR technology. We can reasonably arrive at answers to these questions at the demonstration reactor stage, before committing significant funds to any production-scale deployment or industry transitions. The MSRE provided tremendous developmental returns on a modest $80 million US equivalent investment.
India is already deploying thorium in its reactors as a component of mixed oxide (MOX) fuels comprising plutonium / uranium, and plans more of the same in its forthcoming Advanced Heavy Water Reactors. However, Norway’s Thor Energy and the UK’s National Nuclear Laboratory (NNL) both believe that considerable research, development and testing lies ahead before thorium fuels will be ready for operational use.
Yes, it would require considerable research, development and testing to transition an extremely complex solid-fuel cycle from one fuel type to another. This has no bearing on use of thorium in liquid fuel forms.
As the NNL states, “Thorium reprocessing and waste management are poorly understood. The thorium fuel cycle cannot be considered to be mature in any area.” It estimates that 10-15 years work is required before thorium fuels will be ready for use in current reactor designs, and that their use in new types of reactor is at least 40 years away. [The Thorium Fuel Cycle - An independent assessment, NNL, August 2010]
The author of the NNL paper on thorium readily admitted that he did not consider use of thorium in liquid fuel forms or in molten salt reactors. His experience is rooted primarily in Westinghouse’s solid uranium fuels and it was presumptuous for him to offer, at his own initiative, an authoritative paper with so broad and self-validating a title when he clearly only considered thorium in the narrow context of solid fuels. Moreover, the NNL paper and its nauseatingly self-validating introduction reveal that the NNL author is, at best, an expert on solid-uranium fuel cycles, and not an expert on thorium fuels, liquid or solid.
LFTR lead time: half a century
The assessment of the Chinese Academy of Sciences as it embarks on its LFTR programme is that a production LFTR is 20-30 years in the future — rather shorter than the NNL’s estimate of 40 years (see 4.1).
CAS’ estimates vary depending on the focus on molten salt cooled reactors and molten salt fueled reactors, both of which are under development by CAS. China demonstrated criticality with a salt-based pile in the 1970s and is now seeking to duplicate much of the original 1960’s MSRE development. China has an acute need for electricity for a growing power-hungry population and the political will and agility to accelerate nuclear technology development.
Given the hazards, such as the potential failure of reactor materials under intense neutron irradiation and chemical corrosion, risk-averse utilities and investors would want to observe the performance of any such full-scale LFTR for at least a decade and probably more, before embarking on any substantial LFTR programme. The lead time for nuclear construction is of the order of a decade, so this could add a further 20-30 years before production LFTRs were deployed at full scale.
Well-established materials testing techniques can accurately extrapolate and monitor suitability of materials for extended use under a given set of conditions. Sophisticated investors have technical advisers capable of independently interpreting and extrapolating materials performance data and making rational investment recommendations. Sophisticated investors will also seek unbiased advice from competent financial advisers.
Also, the improved economics of modular LFTRs will likely allow for initially shorter lifecycles than legacy reactors. With further development, the lifespan and operating temperatures can be optimized to further improve economics. LFTRs modules can be produced in a factory, which modules are then readily installable on a prepared site much like gas power plants today. Ten-year build times are another lingering notion from the authors’ conventional nuclear mindsets.
The total lead time for LFTRs would therefore be a minimum of 40 years on the shortest estimates, or 70 years based on more conservative figures.
Yes, due to the many advantages of thorium and LFTR technology, the industry will eventually, slowly move that direction. A private venture could make this a revolutionary step-change rather than an evolutionary process. Again, consider the disparity in timelines and budgets between Space-X’s Falcon 9 rocket and comparable NASA projects.
Thorium and LFTRs — investment outlook
The development of thorium / LFTR technologies represents a poor investment for national governments, utilities and private investors given:
the marginal benefits to be derived from using thorium fuels in existing reactor designs;
Agreed, there is little economic case for using thorium in existing solid-uranium fueled legacy reactor designs.
the very long-term nature of any benefit that may be realised from LFTRs, of the order of half a century;
Again, the immediacy is not reasonably the primary metric for viability or desirability of a technology, and a focused private venture could realize a LFTR demonstration within the decade. At that point, governments, utilities and further private investors can make up their own minds.
the uncertainty as to whether the very significant technical challenges of the LFTR will ever be overcome;
The primary uncertainty is bureaucratic, not technical. The fundamental physics, operation, and maintenance were sufficiently demonstrated and documented during the Molten Salt Reactor Experiment’s successful 20,000 hours of operation.
In the last semi-annual report to address the two-fluid design, the MSRE team recognized that certain aspects would require further engineering and development, however, they maintained that they had not uncovered “any aspects which indicated that major technological discoveries would be required to design a two-fluid molten-salt reactor power station” such as the modern LFTR.
the possibility that the materials used for reactor
construction may degrade more rapidly than anticipated, causing early
shut-down;
President Kennedy recognized in his initial moon mission speech that the unique requirements of space travel would require development of new alloys and materials. LFTR’s unique requirements will require similar vision and similarly achievable technological advances in materials and high-temperature power conversion systems. There are several good modern materials candidates for each relevant LFTR component and the materials industry is up to the challenge of modifying these or creating new materials as needed to achieve suitable reactor life spans. Those materials and components that are most susceptible to degradation will be configured to be replaceable to ensure continued safe operation. Also, modular construction allows for ready replacement of entire LFTR module as needed to ensure continued power generation.
the likely very high cost of LFTR electricity — especially when compared against the anticipated low future cost of electricity from renewable sources, solar in particular, over the applicable time frame.
LFTR’s true costs are more readily ascertainable than legacy nuclear plants’ as LFTR offers a fully closed fuel cycle. Again, once a readily achievable demonstration LFTR is operational, the true costs of construction and operation can be better defined and the utilities can then decide between a range of technologies. Coal-fired generation will not decline until there is a technology that can compete with coal on price. Solar energy and thorium do appear to be some of the most promising extended long-term solutions and LFTR and PVC may well be highly-complementary technologies in the future. Considering the numerous ways in which LFTR reactors would improve economics over current nuclear plants (no enrichment and fuel fabrication, higher thermodynamics efficiency, no downtime for refueling, the possibility of load following, no high pressure vessel needed – the most expensive item in a LWR, only low pressure containment needed, no need for large cooling towers), its completely unsubstantiated to say that the cost of LFTR electricity would be “likely very high”.
As NNL states: ‘thorium is competing with the uranium/plutonium fuel cycle which is already very mature. To progress to commercial deployment would demand major investments from fuel vendors and utilities ‘LWR and PHWR utilities would be unlikely to invest in thorium fuels to the levels required under current market conditions. The potential savings that thorium fuels offer and other claimed benefits are insufficiently demonstrated and too marginal to justify the technical risk that the utility would be exposed to.’
This quote belies the NNL paper’s predisposition towards solid fuels and the inapplicability of NNL’s conclusions to the present topic. Thankfully, LFTR development timelines are not dependent on LWR or PHWR utilities as their business model is centered on fuel fabrication, which is obviated by LFTR’s greatly simplified liquid fuel form.
We therefore see little prospect that LFTRs will present an economic solution if and when they are ever ready for large scale deployment. Any money invested in LFTRs, whether by governments, utilities or other investors, is likely to be wasted.
A modest modern demonstration of LFTR technology could answer many of the author’s questions and greatly deflate the stated uncertainties. Once this modest demonstration has been made, governments, utilities, and investors can make up their own minds regarding large-scale deployment. In the meantime, we look forward to open discussions with those who are rightfully curious about the role of thorium LFTR and nuclear energy in society’s long-term energy outlook.
Far better to invest in the renewable technologies that are already shaping our national and global future, and whose cost is rapidly falling – in the process developing valuable UK-based expertise and technologies, and accelerating the renewables revolution.
The clear agenda of the authors and this article.
Appendix 1 – The thorium fuel cycle
Thorium is not itself fissile, however it is ‘fertile’. That is to say that, under neutron irradiation, it can be used to breed fissile material. In any thorium reactor, the naturally occurring 232Th is irradiated with neutrons from fissile material (for example, 235U, 233U or 239Pu). Some of the thorium nuclei capture a neutron and become 233Th. This isotope then undergoes beta decay to 233Pa (protactinium 233) which in turn beta decays to 233U, a fissile isotope of uranium.
So in a thorium reactor, the fissile material is in fact uranium. The 233U behaves like the more familiar naturally occurring 235U. It has a fairly long half life of 160,000 years, and like 235U, 233U is fissionable and can create and sustain a nuclear fission chain reaction, in which the neutrons emitted by one fission event trigger further fission events in other 233U nuclei. When 233U undergoes fission, it produces similar fission products as 235U, but in different proportions.
Thorium fuel does possess some advantages over conventional uranium / plutonium fuels:
the 232Th is more likely than the 238U to capture thermal neutrons;
Accordingly, LFTR will be configured to run on 232Th, whereas 238U is best used in a vastly different fast spectrum breeder reactor. While both are fertile, they are hardly interchangeable as their vastly different neutron absorptions significantly affect reactor operation.
the resulting 233U is more likely to fission following neutron capture than is 239Pu;
fissioning 233U produces more neutrons to sustain the nuclear chain reaction.
233U produces only enough neutrons to iso-breed, meaning to replenish the consumed fissile with newly bred fissile, without a net loss or gain of fissile material.
These factors combine to create a more efficient ‘neutron economy’ for thorium than for conventional nuclear fuels, making smaller reactors more viable. They also mitigate against the formation of long-lived transuranic isotopes such as plutonium.
This last benefit is unique to LFTR with very significant implication s for long-term waste storage.
There is also one important disadvantage: the breeding of 232U, a non-fissile but strongly radioactive uranium isotope. This arises when the 233Pa absorbs a neutron before it decays to 233U. The resulting 234Pa may then expel a pair of neutrons to make 232Pa, which then undergoes beta decay to 232U.
This isotope is typically present in small quantities with a 232U:233U ratio of well under 1%. But it presents a considerable hazard due to its short half life of under 70 years and the rapid decay chain which follows, culminating in an ultra-hard 2.6 MeV gamma ray – capable of passing through a metre of lead. This powerful gamma irradiation creates a hazard to personnel and to unshielded electronic control systems. Consequently, thorium fuel requires far more shielding, and more stringent remote handling techniques than conventional nuclear fuels.
LFTR shielding and reactor containment will be engineered to be more than adequate to contain all of the high-energy emissions from fission, fission products, and decay of 232U, of which, gamma emissions from 232U are not the most significant. Nor are the 232U –related gamma emissions any more energetic than high-energy emissions within legacy reactors which are suitably shielded. Reactor vessel shielding and remote handling within the containment will be comparable to existing nuclear installations. There is no 232U in the waste stream as the 232U generated remains in the core until it absorbs another neutron and is consumed, or decays. All operations within the reactor containment will be remotely performed once the reactor is operational and 232U will never leave the reactor, so the posed threat of 232U to personnel is overstated.
The nuclear industry routinely performs remote operation and maintenance of hundreds of reactors all over the world. The complete thorium fuel cycle can be sustained entirely within the reactor containment. There is no offline reprocessing needed to refuel LFTR. Fortunately, once the initial fissile charge is placed in the reactor, the 233U and 232U need only be removed from the reactor containment very infrequently, e.g., when the reactor is retired and the fuel salt is transitioned to a new LFTR.
Fortunately, these same hard-gamma emissions from 232U are a significant deterrent to removal of the fuel from the reactor. It should be counted a benefit that these gamma emissions would harm any technicians and electronics in the unlikely event of a misguided attempt to weaponize any 233U. Moreover, these hard gamma emissions are readily detectable by regulators and security forces, as would be unplanned shutdown of a LFTR.
But the greatest problem with the thorium fuel cycle is our relative inexperience of it, compared to the conventional uranium / plutonium fuel cycle. According to the UK’s National Nuclear Laboratory, “Thorium reprocessing and waste management are poorly understood. The thorium fuel cycle cannot be considered to be mature in any area.” The NNL estimates that 10-15 years of research and development will be required before thorium fuels are ready for production deployment in conventional reactors:
“Starting from fabrication of a commercially-relevant mass of ThO2 fuel, which might take 1 or 2 years, the subsequent irradiation to full burnup would likely take 4 to 5 years. Subsequent post-irradiation examination might take another 1 to 2 years, so the overall timescale will be of the order of ~10 years. In practice, a gradual ramp-up to commercial scale loading might be necessary, leading to a more realistic timescale of about 15 years for commercial demonstration. This is comparable to the timescale that was required to commercialise MOX in LWRs.”
MOX solid fuel does entail a very complex and unattractive reprocessing scheme that is not a reasonable comparison to the greatly simplified liquid thorium fuel form. The repeated reference to thorium-oxide (ThO2) and mixed-oxide fuels (MOX) reveal the solid-fueled paradigms that underlie the NNL paper’s conclusions and predictions, which are of little value in a discussion of use of thorium in liquid fuel form in a LFTR. See earlier comments about the NNL author’s admitted omission of liquid fuels from consideration in writing an authoritative position paper titled “The Thorium Fuel Cycle.”
“Because nobody has done it yet” is never a good argument to overlook promising technologies, especially when molten salt reactors have been done, and successfully demonstrated for over 20,000 hours.
Appendix 2 — Molten salt reactors
A2.1 History of the molten salt reactor
MSRs were first developed in the early 1950s as the US Airforce sought a novel power source for its fleet of high altitude nuclear bombers. Although a working reactor was developed under the ARE (Airborne Reactor Experiment) programme, it was never deployed.
In the 1960s, the ARE technology was taken up by the Oak Ridge National Laboratory which conducted its own MSRE (molten salt reactor experiment) from 1965 to 1969. This was based on a graphite-moderated reactor using fluoride salts of uranium and plutonium as fuel. Subsequently Oak Ridge built a Molten Salt Breeder Reactor (MSBR), which operated from 1970 to 1976. In its initial phase the MSBR used as fuel fluoride salts of 235U and thorium, later followed by using the 233U it had bred in the first phase, also with thorium. It was also tested using plutonium fuel (mostly 239Pu).
The MSRE was run for 20,000 hours between 1965 and 1972. The MSBR was designed to an extent but not constructed or operated. It is likely that the author has confused separate fuel trials with the MSRE with the proposed MSBR.
The experiment demonstrated that the reactor design was viable. Particular successes included the breeding of 233U from the initial thorium; the subsequent use of the 233U to re-initiate the thorium fuel cycle; and the de-gasification of the molten salt fuel, extracting unwanted gases such as xenon (135Xe), an important neutron sink that would otherwise slow down or indeed halt the fission chain reaction.
The MSBR also highlighted some unexpected hazards. For example, the nickel-molybdenum alloy used to build the reactor became brittle under thermal neutron irradiation, and suffered extensive surface cracking due to the presence of the fission product tellurium. This highlights the importance of developing materials capable of surviving the highly corrosive environment of an LFTR environment, and to withstand the intense neutron bombardment, over a multi-decadal timescale. It also raises the prospect that any material used may degrade well before its anticipated end of life and cause premature reactor closedown.
The tellurium micro-cracking at the alloy grain boundaries was first observed when the MSRE was being shut down after 20,000 hours of operation. The original MSRE team recognized how to readily avoid this in the future with a simple periodic additive to better control the salt chemistry.
It should also be pointed out that the power output of the MSBR was limited to just 7 MW. Any production MSR built for power generation would be expected to have a thermal output closer to 500 MW, two orders of magnitude greater, with correspondingly greater fluxes of neutrons. This would create of host of challenges in engineering design, materials science and fuel reprocessing.
LFTR is very scalable and a host of engineers can arrive at reasonable solutions to any challenges presented during scale-up. The MSRE team demonstrated 2.5 and 7.5MWth designs, completed detailed 250MW designs and prepared preliminary multi-gigawatt designs. The introduction to the last report on the 250MW design reaffirmed the MSRE team’s confidence in the scalability of the technology without the need for major new technological discoveries.
A2.2 Molten salt processing
A key benefit of MSRs is that they provide the ability to clean the fuel of unwanted fission products on a continuous basis. In conventional solid fuel reactors, fission products build up in the fuel rods or pellets, and some of these are powerful neutron absorbers, like the Xenon isotope 135Xe. It is the accumulation of these neutron-absorbing fission products that ultimately limits the lifetime of solid fuels by reducing the efficiency of the reactor’s ‘neutron economy’ until the nuclear fission chain reaction slows down or halts.
Xenon decays away in about nine hours reducing its neutron absorption qualities. Legacy fuel pins are retired due to a combination of the accumulation of fission products, swelling of the solid uranium oxide pellets, and oxidation of the thin fuel cladding that houses the fuel pellets and accumulated fission products.
In the course of molten salt reactor operation, other undesirable fission products also build up in the fuel. These include oxygen, which gives rise to particulate deposits of solid metal oxides, and highly corrosive sulphur and metals. These also require removal.
Techniques for extracting these various contaminants from the molten salt fuel were originally developed at Oak Ridge. Typical processing temperatures are in the region of 400C to 600C and involve the use of highly reactive chemicals such as hydrogen, hydrogen fluoride and hydrofluoric acid. This creates a highly hazardous environment.
These chemicals are safely used on tonnage scales throughout various industrial processes. With LFTR these chemicals do not enter the waste stream, but are part of a closed cycle in which the reactants are continuously regenerated and reused.
Further ‘pyro-processing’ techniques that are highly applicable to MSRs were developed at the Argonne National Laboratory in the context of its Integral Fast Reactor (IFR) programme. These involve high temperature ‘pyrometallurgy’ and electro-refining.
Some proposed integral fast reactor (IFR) reprocessing techniques can benefit from use of molten salt chemistry, but such reprocessing is not needed for LFTR. The LFTR fuel cycle relies on much simpler fluorination, reduction, distillation and electrolysis processes.
Note that these technologies could be used to produce very high purity streams of fissile uranium and plutonium well above weapons grade as currently defined (see A3.2 and A3.3 below).
LFTRs do not “produce” any net gain in fissile material let alone a stream of divertible fissile material. LFTRs are “isobreeders”, meaning they simply maintain equilibrium between consumption and production of 233U. Any attempted diversion would disrupt this balance and lead to reactor shutdown. At no point in the LFTR fuel cycle is there ever a stream of fissile material leaving the reactor containment or an increase in the amount of total fissile material in the system. In short, LFTR is configured such that neutrons from consumption of 233U in the core salt produce an equal amount of 233U via neutron capture by thorium in the blanket salt. The new 233U is then moved in a closed system from the blanket salt to the core salt where it is fissioned and new thorium is added to blanket salt.
The continuous purification of the molten salt inevitably creates a waste stream of fission products in various combinations and in mixtures of reagents and waste chemicals arising from the process. The safe handling and disposal of these wastes, while minimising radioactive releases to the environment, presents further serious challenges in radio-chemical engineering.
Yes, all fission produces fission products. Fortunately, LFTR’s liquid fuel form allows for complete fuel burn-up and extraction and commercialization of a number of the fission products, removing both from the waste stream. The remainder can be vitrified in an inert glassy form suitable for storage until the fission products have safely stabilized to background radiation levels. There is obviously process development needed for a modern LFTR demonstration and further refinement and scaling of these processes to be performed between demonstration and large-scale commercialization. Fission product handling presents tractable challenges with reasonable engineering solutions.
A2.3 Safety in operation
Proponents of the LFTR claim important safety advantages for the technology:
LFTRs are unable to suffer reactor core meltdown, as occurred at Chernobyl and Fukushima. If the core temperature rises too high, the liquid fuel expands and the fission chain reaction slows down. Also the removal of the ‘neutron sink’ isotope 135Xe can be halted so as to slow down the fission chain reaction. As a fail-safe, a salt plug is included in the bottom of the reactor which will melt at a set temperature and allow the fuel to drain into a holding tank where fission will halt.
the LFTR operates at near-atmospheric temperature is therefore less susceptible to explosive pressure release venting fuel and fission products to secondary containment or the atmosphere.
These claims are broadly accurate. But while LFTRs do indeed reduce certain risks, other new risks appear. There is the risk of materials failure in the reactor / liquid fuel containment as the alloys become brittle under neutron irradiation, or suffer cracking and surface damage in the high-temperature, intensely corrosive reactor environment.
The certain risks that are reduced by LFTR are significant and cannot be overstated. Engineering new materials for safe performance even in the harsh environment within a LFTR is within the ability of skilled materials engineers. That is also the purpose of a demonstration reactor, to determine the suitability of various materials for safe deployment. Regulatory bodies will assure themselves of the suitability of every material used in a licensed reactor.
Any accidental release of the hot fluoride salt fuel could be highly damaging owing to the fuel’s highly corrosive chemistry, and cause radioactive releases to the environment. Similar accidents could also take place in the continuous fuel reprocessing system, which will use highly reactive and potentially explosive chemicals such as hydrogen, fluorine and hydrofluoric acid, all at very high temperatures.
Actually, the LFTR salts are extremely stable and will not react with air or water like liquid metals would. Many of the problematic fission products, such as cesium, remain chemically bound and entrained within the salt. More importantly, any leakage of salt from the reactor vessel would be collected by the containment structure and directed safely via gravity to storage tanks. There is no stored energy term, e.g., high pressure or chemical reactivity that would drive any outward release of salt. It is not clear what extreme hypothetical circumstances the author imagines would lead to displacement of low-pressure salt from both the reactor vessel and containment structure.
Fission product extraction will obviously require careful operation of well- engineered salt handling systems.
Due to the intensely radioactive nature of some the isotopes that need to be handled during reprocessing, there is no scope for direct human intervention in the fuel reprocessing system or the reactor itself in the case of failure. All personnel will need to be shielded and will only be able to intervene via remote handling systems or robots, themselves subject to potential failure. In the event of an accident, there might be little alternative but to abandon the reactor, possibly for an extended period of time, until radioactivity declined to sub-lethal levels permitting human access.
Remote operation of processes with adequate shielding is a common challenge with nuclear energy. However, systems and processes have been and can be engineered to suitably safe levels. Many of the fission product extraction processes will not require direct personnel involvement. Liquids are highly amendable to continuous bulk processing in highly-automated systems. The removal of the need for personnel involvement through automated liquid processing is more rightly counted a plus that a minus on LFTR’s score card. Moreover, these processes are not conducted in the open air but within a suitable containment structure designed to retain any such release. By eliminating the stored energies of high pressure operation and chemical reactivity of the fuel characteristic of legacy reactors, LFTR makes the handling of fuel and fission products dramatically easier and safer. Without these driving forces, any incidental or accidental releases within the containment structures will remain fully contained.
It is therefore hard to sustain with any certainty the idea that LFTRs are intrinsically safe. Indeed considerable dangers appear to be attached to LFTRs and their routine operation.
The regulatory agencies are tasked with determination of safety of reactor designs and evaluation of a demonstration reactor is a key step in that determination. LFTR will be engineered for safety and demonstrated as safe before it is deployed commercially.
Appendix 3 – Nuclear weapons proliferation
A3.1 General considerations
As already noted, thorium reactors work by breeding 233U, a fissile isotope of uranium. It has been stated that thorium reactors present no nuclear weapons proliferation hazard because they do not breed plutonium like conventional uranium reactors. However, there are a number of stages of the thorium fuel cycle in which fissile material for weapons could be diverted.
The more accurate characterization of LFTR’s non-proliferation advantage is that the thorium fuel cycle are undesirable for weapons efforts and that LFTR’s fuel cycle and operating environment greatly disincentivise any attempt at diversion.
First, the thorium fuel cycle needs to be initiated by externally supplied fissile material, whether uranium (235U or 233U) or plutonium (239Pu). Accordingly there is the risk that some of this externally-supplied fissile material could be diverted into weapons.
As discussed
earlier, LFTR does require an initial fissile charge, after which no further
preparation, storage or transport of fissile material is required. This is in
stark contrast to current reactors that require constant preparation,
transport, storage and resupply of fissile material. This should be counted a
significant advantage. Adequate security can be ensured for this single
sensitive delivery of the intialy charge required per LFTR.
Second, the 233U that is bred in thorium reactors is
highly weaponisable. Such a bomb was exploded in the Nevada desert in 1955 as
part of Operation Teapot, a series of 14 nuclear bomb tests conducted by the US
government. In one of these tests, the Military Equipment Test or ‘MET shot’,
engineers replaced the 235U core of the uranium / plutonium bomb with 233U. The
bomb successfully detonated and the principle that 233U can be used to make
nuclear bombs, with fearsome destructive potential, was firmly established.
233U is fissile and
so was tested along with other fissile materials and found impractical for
weapons purposes. While the Teapot test details remain classified, it is
reported that the 233U was added to a conventional plutonium weapon and that
the yield of the detonation was so disappointingly diminished that there have
been no further attempts. Indeed, not one of the tens of thousands of warheads
in the world’s arsenal is based on the thorium fuel cycle.
It has been
suggested that the inevitable presence of 232U as a contaminant of 233U – as we
have noted, a powerful gamma emitter though its decay products – renders 233U
unusable as a bomb-making material, due to health damage to handlers and
machinists, and disruption to electronics. Given that a 233U bomb has already
been assembled and detonated using 1950′s technology, this is clearly not an
insuperable problem. Pyroprocessing technologies could also be used to produce
highly concentrated 233U untainted by 232U or any other uranium isotope from thorium
fuel (see A3.2 below).
It is far more
likely that would-be weapons builders would resort to enrichment of natural
uranium than attempting to storm an operating LFTR at 600 C with newly formed
fission products requiring many tons of shielding for transportation if the
fuel salt could even be remotely extracted in any reasonable timeframe. There
is no reason that LFTR’s fuel would be found anywhere other than in the
high-temperature, high-radiation reactor containment or at the high-security
facility for preparation of the initial fissile charge. There is no 233U in the
LFTR waste stream. Accordingly, adequate security measures at the initial
charge preparation facility and at each securely sited LFTR, along with LFTR’s
high-temperature high-radiation operation itself, should sufficiently mitigate
against any unlikely threat of diversion. Moreover, any 233U produced in the
LFTR fuel cycle does inherently contain 232U and the corresponding hard gamma
emissions are a further deterrent to handling. Lastly, these hard gamma
emissions are readily detectable by regulators and security forces.
One means that
has been proposed to prevent the generation of sufficiently pure 233U to build
a nuclear bomb is to add to the thorium fuel a significant percentage of natural
or depleted uranium, rich in non-fissile 238U. It is impossible to chemically
separate 233U from 238U since they are both the same element.
As a last resort,
238U can be kept at the ready for rapid denaturing of the fuel salts if there
is ever a need to render the fuel useless for all purposes. Only liquid fuels
can allow for such instantaneous denaturing. LFTR is not configured to use
depleted uranium and 238U is excluded from the LFTR fuel cycle.
However as the
NNL notes, “Attempts to lower the fissile content of uranium by adding U-238
are considered to offer only weak protection, as the U-233 could be separated
in a centrifuge cascade in the same way that U-235 is separated from U-238 in
the standard uranium fuel cycle.” Indeed, owing to the greater difference in
atomic mass the centrifuge separation would operate more efficiently.
If one already had
uranium enrichment capabilities, why would they add the additional daunting
task of 233U diversion when there are more readily available sources of natural
uranium for enrichment? The endless conjuring of unlikely hypothetical
scenarios eventually begins to border on the ridiculous.
The presence of
238U in the fuel would also create another hazard. One of the advantages of
using thorium fuel is its low level of conversion to long-lived transuranics.
But if the fuel contains 238U as an anti-proliferation measure, then it will
absorb neutrons (in the process degrading the ‘neutron economy’) and form
plutonium. Next, ‘pyroprocessing’ technologies could be used to extract any
plutonium from the molten salt fuel, potentially producing 239Pu at very high
concentrations well above ordinary weapons-grade (see A3.2, below).
Some have proposed
a form of molten salt reactor employing a denatured fuel to further alleviate
proliferation concerns. This would increase the total fuel inventory
significantly and add to the complexity of the waste stream reprocessing,
including formation of transuranics. These are not part of the LFTR design.
LFTRs would exclude
238U from the fuel cycle to minimize formation of transuranics, specifically
239P.
A3.2 Weapons
grade uranium (233U)
The process
whereby fissile 233U is bred in the thorium fuel cycle involves an intermediary
stage, with the protactinium isotope 233Pa. The 233Pa undergoes beta decay to
233U with a half life of 27 days. But the 233Pa may first absorb a second
thermal neutron to make 234Pa. This isotope may either decay to the undesirable
uranium isotope 232U, or the non-fissile uranium isotope 234U.
This makes it
beneficial to remove the protactinium from the molten salt fuel before it can
intercept a second neutron. Left to itself, away from the neutron flux, the
233Pa decays over a period of months to produce very pure 233U. Furthermore the
removal of the protactinium leaves more neutrons to maintain the reactor’s
neutron economy, maintaining both the thorium fuel cycle and the fission chain
reaction.
Oak Ridge
demonstrated several effective methods of protactinium removal, for example,
precipitation by addition of thorium oxide to the molten salt [Removal of
protactinium from molten fluoride breeder blanket mixtures, C. J. Barton and H.
H. Stone, 1966]. This technique precipitated 233Pa when present at just 0.1
parts per billion.
But the authors
note that the protactinium is highly radioactive and hazardous to handle even
on the milligram scale, due to its short half life of 27 days, and its high
combined beta and gamma energy of 570 KeV. For this reason, Oak Ridge used a
mixture of 233Pa and much less radioactive 231Pa for its experiments, in which
the 231Pa was ~100,000 times more abundant than the 233Pa. This indicates that
the handling of the highly radioactive 233Pa extracted from molten salt fuel
would need to be done entirely by remote handling with any personnel shielded
from radiation.
This method of
producing very pure fissile 233U, while highly desirable as regards reactor
operation, represents a significant weapons proliferation hazard. Purities of
233U well above accepted weapons grade (85% for 235U) would be achievable.
LFTR does not
necessarily require protactinium removal, although this would improve the
neutron efficiency overall. Intermediate protactinium removal improves neutron
efficiency, however, LFTR can operate while leaving protactinium in the fuel
salt.
A3.3 Weapons grade plutonium (239Pu)
A similar
approach could also be used to produce weapons-grade plutonium from molten salt
fuel rich in 238U. In the normal operation of a uranium solid fuel reactor the
238U captures a neutron and then undergoes two beta decays to form 239Pu.
However, the 239Pu itself has a high probability of capturing further neutrons
to make 240Pu. Hence the plutonium in spent fuel typically comprises under 80%
239Pu and most of the remainder is 240Pu with some 241Pu.
240Pu is highly
undesirable in a plutonium weapon since it can trigger premature fission giving
rise to a low-yield explosion known as a ‘fizzle’. As a result weapons grade
plutonium contains a maximum of 7% 240Pu. So in order to make weapons grade
plutonium the fuel is only left for a short time in the reactor before
reprocessing.
The same result
could be served in a molten salt reactor by including a high level of 238U in
the fuel, and extracting the 239Pu as it is formed. It would make sense to use
the pyro-processing technology developed by the Argonne National Laboratory for
its Integrated Fast Reactor – a high temperature (500C) electrolysis process
using molten salts. This could be employed to concentrate the plutonium,
together with some residual uranium and any minor actinides from the molten
salt fuel [Proliferation-Proof Uranium/ Plutonium Fuel Cycles, By G. Kessler].
Standard aqueous methods such as PUREX could then be employed to purify the
plutonium. This approach should be capable of producing plutonium with a very
high 239Pu content well above normal weapons-grade.
Again, LFTR cannot
use 238U in place of thorium and LFTR does not produce any increase in overall
fissile material and any attempt at diversion of the fissile derived from
thorium to replenish the fuel would halt the isobreeding cycle, precluding
further production of fissile material. LFTRs are designed to iosbreed 233U and
could not be used to breed plutonium.
The likelihood of a
hypothetical outcome must always be considered alongside the hypothetical
severity of that outcome to arrive at any reasonable quantification of the real
risk. In each of the hypothetical scenarios posited above, the outcome could no
doubt be severe, but the likelihood can be made so remote as to render the
actual risk extremely low.
The authors receive
full marks for their due diligence on appreciation of LFTR’s benefits and for
their creativity in conjuring hypothetical extremes. The extreme hypothetical
outcomes proposed above, however, involve significant deviations from the
design and operation of LFTRs and the thorium fuel cycle. On balance, this
discussion will likely prove helpful to the public appreciation and discussion
of thorium and LFTR and their promising roles in our energy future.
END
Most likely not.
Source: http://energyfromthorium.com/tickell-rebuttal-jul-2012/ (improved and reformatted)
