A new lease for nuclear?

Graphite core of the ORNL molten salt reactorI’ve focused on nuclear energy for the past three posts, and my appraisal is bleak. It looks cheaper than other options, but only if you ignore costs such as construction overruns, accidents, waste storage, and the inherent risk in long-term energy demand forecasting.  Chernobyl and Fukushima prove that the technology was and continues to be far from foolproof. And between the need for public financing and inevitable swings in public opinion, cost overruns are inescapable. Is that the end of the story? Is there any hope that nuclear energy can overcome these problems and recapture the public imagination the way it did back in the glory days of the 1950s?

There is hope. But we must change the fuel and change how it is burned.

Uranium is rare. Mining uranium ore is difficult and costly, as the mineral deposits are often deep underground. Separating the uranium from the rest of the ore leaves tailings ponds which are a huge environmental liability. The uranium must be enriched to separate out the usable isotope; the rest is wasted. The enriched uranium is then used as fuel, and all of it eventually ends up as waste. Some of that waste can be used to produce nuclear weapons. The waste will remain highly radioactive and deadly for tens of thousands of years.

It takes about 100,000 tonnes of low-grade ore to produce 250 tonnes of natural uranium, which then yields only 35 tonnes of enriched uranium. That’s enough fuel to produce 1GW of power, and 100% of the spent fuel ends up as waste. The nuclear industry produces 10,000m3 of high-level, long-lived waste each year.

However, uranium isn’t the only game in town. Thorium has the necessary properties to be a nuclear fuel. It is three times more abundant than uranium. It is already being produced and discarded as waste in the process of extracting rare earth metals. In any case it is often found in monazite sand deposits on the surface, so mining is much easier than it is with uranium. Thorium is usable as-is, so it does not require enrichment the way that uranium does. Thorium can be used as nuclear fuel in liquid form, so there is no need to form the fuel into rods or pellets as is required with uranium. A given amount of power can be produced with 3% of the amount of thorium compared to uranium. The waste cannot easily yield material for nuclear weapons, only 17% of it needs to be stored for the long term, and that for the comparatively brief period of 300 years.

It takes less than 20 tonnes of monazite to yield one tonne of thorium. That’s enough to produce 1GW of power. Of the resulting one tonne of spent fuel, only 17% requires long-term storage. The technical and anthropological challenge of storing waste for 300 years is anything but trivial, but it is not nearly as daunting as planning a dump that has to last longer than the span of recorded human history.

That’s the fuel. Now let’s look at the engine.

In a typical uranium-fuelled reactor, fuel rods in the reactor core undergo fission reactions to produce heat. Control rods, which stifle the fission reaction by absorbing neutrons, are used to modulate the amount of heat being produced. Circulating water carries the heat away to turn a turbine, thereby generating electricity.

The big problem with these types of plants is that the coolant is water. If the water gets too hot it becomes steam, expanding in volume a thousand fold. Because the coolant water is in direct contact with the core, it is radioactive, and must be contained in the event of a leak. That means that the containment vessel around the reactor must be a thousand times larger than the reactor itself. That is why commercial nuclear power plants are so massive.

One of the most promising thorium-based reactor designs is the Liquid Fluoride Thorium Reactor, or LFTR. In the LFTR, the fuel is dissolved in the working fluid itself, and that working fluid is a molten salt rather than water. The system operates at atmospheric pressure, and never approaches the boiling point of the fluid. As a result, the containment structure only needs to be slightly larger than the reactor itself, allowing for very small plants. One experimental reactor produced 2.5MW of power and was designed to be small enough to power an aircraft.

A particularly compelling design feature of the LFTR is its safety. At the bottom of the reactor is a plug made of a material which melts at ambient temperature. The plug must be constantly cooled; if the power goes out, the cooling stops, the plug melts, and the working fluid – which includes all of the fuel – drains out of the reactor and into a reservoir where the fission chain reactions stop.

In summary, a thorium-fuelled molten salt reactor such as the LFTR offers some very promising advantages over traditional uranium-fuelled, water-cooled reactors. The fuel comes from a raw material that is cheap, abundant, and high-yielding. A given amount of fuel generates much more energy, and produces much less waste. That waste is less susceptible to covert diversion toward nuclear weapons programs, and does not need to be stored for an unimaginable time frame.

The reactors can be made much smaller and therefore much faster and cheaper, rendering public sector financing unnecessary. Build-out plans can also be adjusted over time if electricity demand turns out to be different from what was predicted.

There are just two problems. Uranium is an elephant, and thorium is a flea on its backside. And LFTRs are pie-in-the-sky.

According to the European Nuclear Society, there were 435 nuclear reactors in operation worldwide as of February 2012 with a total output of 368GW. Of these, eight were thorium-fuelled, with a cumulative power of less than one GW. All eight are in India, where abundant thorium deposits have made the fuel very attractive. For all intents and purposes, all of the worldwide industry expertise and experience is in uranium and not thorium. Changing the industry over to the thorium fuel cycle would be a mammoth undertaking.

For all its apparent advantages, the LFTR is theoretical. The only one ever built, at the Oak Ridge National Laboratory, was shut down 43 years ago. Those involved have long since retired and their knowledge has been lost. Any new LFTR initiative will start from scratch. Some state-funded initiatives are doing just that, as is at least one private company (Flibe Energy), but they will need a great deal of time and money before they will have anything to show for it.

Proponents of LFTRs have their work cut out for them. Thorium may well be all that they claim. However, they face an uphill battle. Uranium-based fission did not live up to its potential. Neither did thermonuclear fusion, nor cold fusion. The history of nuclear energy is littered with broken promises. I won’t be a bit surprised if thorium, for all its possibilities, never gets the chance to prove its case in the court of public opinion.

Many thanks to my readers Steve Dyck and Dr. Alexander Cannara for inspiring this week’s post.

Advertisements

8 thoughts on “A new lease for nuclear?”

  1. Really fun to read your post, and I think you provide a balanced view. You did a great job of simplifying the science.

    That said, if someone wanted to harvest U235 to make a bomb, I think all they would need to do is turn off the cooling system on the bottom of the LFTR. The process of enriching to weapons grade would be easy (not that I have the equipment). Let me know if I have this wrong.

    That said, I continue to be a proponent for a least researching Liquid Thorium reactors.

    Like

    1. The problem is, that LFTR produces both U233 and U232. U232 has decay products the are hard gamma emitters. There are also proposed ideas to have a U238 scuttle system in place to make any U235/U233 produced hard to extract. You would need to perform isotopic separation to do such, and any group of doing said operation could most likely make their own reactor to breed bomb worthy materials.

      Like

      1. designing a reactor was easy.I think our genarel goal in the research community is to get to process temperatures. Any excess temperature can be added through a number of means (e.g. heat pump, electrical, chemical, etc). There is a thermodynamic penalty that can be successfully mitigated based off of configuration with each of these designs. Why fight the material restrictions on the reactor side? Find a non-nuclear solution, just do a patent search and you will see the sate of the art.Even with salt reactors, getting above 650C is going to be a challenge from a materials selection and reactor design stand point. By the way 650C is still not hot enough for what we need for process heat. I’m not sure where you are getting 704C from. I sat in a project proposal meeting with a slew of really smart people from around the country talking about a MSR design. Those really smart people saying even 650C would be a challenge. We need about 800 C for process heat.We will probably see thorium commercially used in light water reactors before seeing MSR’s. Think Shippingport redux. The cost of the plant for a LWR is a fraction of that of a higher temperature reactor, especially LWR SMR’s. Unless you come up with a way to break the process heat temperature barrier, the infrastructure investment that we have in LWR technology will be difficult to displace. LWR are technically, regulatory, and commercially well understood. Not to mention our nation posses little to no lithium reserves and there is a complete lack of industrial lithium enrichment facilities. FLiBe is a nifty coolant with amazing properties, but not without its problems (extreme costs, supply chain vulnerabilities, and consumption requirements).As for IFR, the electrorefiner development continued after the termination of EBR-II in the mid 1990 s. A google search of engineering scale electrorefiner will be fruitful. The scale up from engineering to commercial scale is small and throughput can be achieved through replication. It doesn’t take much to reprocess the fuel of a few reactors. BTW, thorium is not without reprocessing technical difficulty either. Thorium is not an easy substance to work with chemically.I am an energy yes kind of dude. I am not poo-pooing thorium. It is a great fuel source and it does have its role, especially in thermal spectra reactors. Uranium fast reactors also have their place too. The point I am trying to get through to you is, unless we can get nuclear heat to 800C in the next 15 years, you might as well plan on 30-40 before seeing commercial applications. That is true for IFR, MSR, NGNP, LFTR.Once you start trying to crack the process heat role, it gets more difficult, and the regulatory philosophy on any system that can affect reactivity or come in contact with fission products significantly complicates implementation. Exxon is not going to invest $10 billion in a liquid fuels plant that they could build for $5 billion and not have to deal with the NRC.We need every scrap of energy that we can get our hands on. This is not a thorium uranium, LFTR-IFR fight. We need all of those technologies. Try and see how they integrate into the entire picture, not just electricity generation, but also the material supply, regulatory, and process heat as well.Thorium is incredible. The path from U-233 to any significant MA concentration is long and tortuous. From that fact alone Th fuel cycles have 1/1000th the MA concentration of Uranium cycles. Thus their radiotoxicity is on par with IFR. It is from that standpoint alone that we will see Th commercially implemented. Reply

        Like

  2. Steve,

    LFTR’s breed U-233 (although U-235 may be necessary to start the reactor it would most likely by 20% or less enriched, proliferation resistant), so other than the start up, there should be no U-235 in the reactor. But U-233 is a usable bomb material and for argument sake lets say it is as good as or better than enriched U-235 as bomb material.

    However, in the process of breeding U-233 from Th-232, you also produce small amounts of U-232. U-232 has a decay chain of hard gamma emitters and alpha emitters. The presence of U-232 in U-233 makes it impossible to handle by humans, reeks havoc on electronics, and shows up readily on radiation detectors. Separating U-232 from U-233 is almost impossible due to their 1 neutron weight difference, so you can’t “enrich” the U-233. These facts cause U-233 produced in a LFTR to be one of the most difficult ways to produce fuel for nuclear bombs. There is a reason no one has really gone this route to produce bomb material.

    By the way, LFTR’s cooling system is circulating the fuel salt. This stuff is highly radioactive, it isn’t as easy as “turning off the coolant” to get to the uranium in the salt.

    Like

    1. The ORNL Molten Salt Reactor Experiment was successfully pecmletod in five years (1960-1965) and then operated for 4 years (2 of those 4 years as a Thorium Molten Salt Reactor or LFTR).That was an experiment, not a commercial product, and all of this occurred well before the regulatory changes following the TMI accident. Sorry, but 1970 was a long time ago. Those days, like 8-track tapes and $0.30 per gallon gas, are never coming back. The estimate provided of a twenty-five year development time only reflects the impacts historically low fission reactor R&D budgets Well, that’s not going away anytime soon. It’s a political impossibility from both ends. The Left these days don’t want to fund anything practical if it is associated with the word nuclear. They only want to fund basic science, which won’t get anything new built anytime soon. The Right just wants to cut budgets. and of the impacts of NRC regulatory obstruction for new commercial nuclear reactors.Well, this is a chicken and egg problem. You can’t get the NRC to work on new technology unless you have a very willing, very credible customer, and you can’t get a real customer unless you can assure them that you can get past the regulatory roadblock.If you want the DOE to build yet another experimental reactor, then good luck. I’m not sure that it will do much. History has shown that experimental reactors don’t necessarily lead to commercial designs.The NGNP project was supposed to build a full-scale prototype of a new reactor design (very similar to a commercial version, but with extra equipment to gather valuable engineering data) in Idaho, and the ups and downs of budget decisions by Congress has run the project into the ground. You can’t depend on the National Labs. Reply

      Like

  3. The roadblock to building bombs starting from LEU is the enrichment capability. The US and others are very active to try to prevent non-weapons states from building the capability to enrich to weapons grade uranium – with varying degrees of success. Blocking nations from taking a nuclear power plant off line and then enriching its contents is done by blocking access to enrichment technology.
    In my view, blocking enrichment technology is more difficult the more enrichment services have a legitimate market. One advantage of LFTR is that it can be designed to be a iso-breeder – that is a machine that makes just enough of its own fissile to keep itself running. This would reduce the demand for enrichment services – helping to reduce proliferation in the short term and in the long term (once we stop expanding nuclear power) it eliminates the demand for enrichment services. That should make it much easier to block access to this technology.

    Like

    1. designing a reactor was easy.I think our aeenrgl goal in the research community is to get to process temperatures. Any excess temperature can be added through a number of means (e.g. heat pump, electrical, chemical, etc). There is a thermodynamic penalty that can be successfully mitigated based off of configuration with each of these designs. Why fight the material restrictions on the reactor side? Find a non-nuclear solution, just do a patent search and you will see the sate of the art.Even with salt reactors, getting above 650C is going to be a challenge from a materials selection and reactor design stand point. By the way 650C is still not hot enough for what we need for process heat. I’m not sure where you are getting 704C from. I sat in a project proposal meeting with a slew of really smart people from around the country talking about a MSR design. Those really smart people saying even 650C would be a challenge. We need about 800 C for process heat.We will probably see thorium commercially used in light water reactors before seeing MSR’s. Think Shippingport redux. The cost of the plant for a LWR is a fraction of that of a higher temperature reactor, especially LWR SMR’s. Unless you come up with a way to break the process heat temperature barrier, the infrastructure investment that we have in LWR technology will be difficult to displace. LWR are technically, regulatory, and commercially well understood. Not to mention our nation posses little to no lithium reserves and there is a complete lack of industrial lithium enrichment facilities. FLiBe is a nifty coolant with amazing properties, but not without its problems (extreme costs, supply chain vulnerabilities, and consumption requirements).As for IFR, the electrorefiner development continued after the termination of EBR-II in the mid 1990 s. A google search of engineering scale electrorefiner will be fruitful. The scale up from engineering to commercial scale is small and throughput can be achieved through replication. It doesn’t take much to reprocess the fuel of a few reactors. BTW, thorium is not without reprocessing technical difficulty either. Thorium is not an easy substance to work with chemically.I am an energy yes kind of dude. I am not poo-pooing thorium. It is a great fuel source and it does have its role, especially in thermal spectra reactors. Uranium fast reactors also have their place too. The point I am trying to get through to you is, unless we can get nuclear heat to 800C in the next 15 years, you might as well plan on 30-40 before seeing commercial applications. That is true for IFR, MSR, NGNP, LFTR.Once you start trying to crack the process heat role, it gets more difficult, and the regulatory philosophy on any system that can affect reactivity or come in contact with fission products significantly complicates implementation. Exxon is not going to invest $10 billion in a liquid fuels plant that they could build for $5 billion and not have to deal with the NRC.We need every scrap of energy that we can get our hands on. This is not a thorium uranium, LFTR-IFR fight. We need all of those technologies. Try and see how they integrate into the entire picture, not just electricity generation, but also the material supply, regulatory, and process heat as well.Thorium is incredible. The path from U-233 to any significant MA concentration is long and tortuous. From that fact alone Th fuel cycles have 1/1000th the MA concentration of Uranium cycles. Thus their radiotoxicity is on par with IFR. It is from that standpoint alone that we will see Th commercially implemented. Reply

      Like

  4. Charles,Is there any real hope for construction of the LFTR someitme in our lifetimes? It seems like this nation if not the world is going full throttle with light water reactors while the possibility of construction of lftr on a meaningful scale seems rather remote. Hopefully I am wrong in my analysis and am willing to be corrected. The way you describe lftr seems like it has much potential to help an energy starved planet but that technology appears to be side tracked, at least for the time being.

    Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s