Nuclear Energy: Gas-Cooled Reactors

In a small article series about nuclear reactors, I’d like to look into the different types of reactors that are currently under active development. There is renewed interest in new and revamped designs for various applications and I will look into current activities in the commercial space. This series starts with a look at gas-cooled reactors.

The operating principle of a gas-cooled reactors can be simplified to: fuel-elements are arranged in a a critical configuration and gas is forced to circulate along the fuel elements to cool them and transfer the heat to do useful tasks, like turning a turbine or generating steam. With gases used as a coolants, there is no phase change in the reactor and no two-phase flows can occur, which simplifies the design process. Because gases have in general really poor heat transfer properties, the systems have to be operated at fairly high pressures in the neighborhood of 7 MPa to increase the density of the gas enough to cool the fuel elements sufficiently.

Historically, the first generation of gas-cooled reactors was deployed relatively early, from the 50s onward. France and the UK worked on gas-cooled reactors. Both countries used magnesium-alloys which, did not corrode in the CO2 they used as coolant and had fairly beneficial neutronic properties.
These reactors were graphite-moderated and fueled by natural uranium Italy and Spain also built and operated such reactors.
As already mentioned previously, reactors of this first generation were – and in the case of North Korea still are – used to produce plutonium for weapon programs.
The initial generation of reactors was improved and the resulting Advanced Gas-cooled Reactor design is still in use in the UK. These reactors abandoned the magnesium-cladding and use steel as cladding material instead. This necessitates the usage of slightly-enriched uranium.
These types of reactors could not compete economically with PWRs, once low-enriched uranium become more widely available and fell out of favor. But they weren’t forgotten.

Generation IV systems

The Generation IV International Forum (GIF) selected the very-high temperature reactor (VHTR) as a candidate technology to reach the goals of the GIF. According to their definition, these reactors have to be graphite-moderated and helium-cooled. I don’t know why it is relevant to specify the coolant or the moderator. For the coolant I assume it’s because helium itself doesn’t get activated. In the unlikely event of an accidental release of large amounts of coolant, it would be inert to the environment. But since radioactive dust particles might also be present in the released coolant, I can’t judge how much of an actual risk gets eliminated by this particular choice of coolant. I am even less clear on the moderator. I think all solid, high-temperature moderators would do comparably well, while there is of course by far the most experience with graphite moderators. The new generation of VHTRs in envisioned to use TRISO particles as fuel. These fuel elements constitute a major part of the improved safety case and enable the operation at higher temperatures. Each fuel kernel has its own fission product retention and containment system.
Another, also gas-cooled Generation IV reactor type, is the Gas-cooled Fast Reactor (GFR). It’s a high-temperature, helium-cooled reactor with a fast neutron spectrum. It doesn’t have a moderator. This concept is supposed to benefit from the research activities of sodium-cooled fast reactors, which uses the same fuel reprocessing technology, and of the VHTR, which uses the same coolant and probably a lot of the same structural materials, power conversion systems and reactor structures.
Core design and safety analysis are areas of research that have to be conducted for GFR specifically.

Hydrogen-production

The initially stated target temperature for these GenIV systems was over 1000°C, but it seems that material issues and a change of the commercial landscape have move the goal posts back to 700-900°C.
The temperature of 1000°C was initially targeted, because it enables the usage of the direct conversion of heat to hydrogen, for example by General Atomics’ sulfur-iodine process.

This cycle doesn’t split water directly, like in a electrolyzer, but rather uses three chemical reactions to do it in steps. One reaction splits water and binds the oxygen and hydrogen to an iodine or sulfur compound. If enough heat of sufficiently high temperature is applied, these intermediary compounds can be cracked and converted back to their initial state, but also to free molecular hydrogen and oxygen. The net result is that water was split by applying heat and you have regenerated your reactants to repeat the cycle. The cycle is depicted graphically here:

All compounds are either liquid or gaseous, which is advantageous for continuous operation. The presence of highly corrosive chemicals at high temperatures naturally creates material issues.
This thermo-chemical cycle poses serious demands on the temperature level of the heat source, but does not put a requirement on the type of the heat source. In other words: if another reactor system can produce outlet temperatures of around 1000°C, it should be considered as a candidate heat source, irregardless of the type of coolant.

Reactor systems under development

HTR-PM

China has been conducting research on the HTR-10, which is a TRISO-fueled pebble bed reactor. This design, originally developed in Germany and South Africa, has proven itself to be passively safe in relevant accident scenarios. However, in large scale tests, it has suffered from large amounts of radioactive dust created by wear on the pebbles. The HTR-PM is a scaled-up version of the HTR-10. It’s a 250MWth reactor. In a recent talk, Prof. Li Fu said that these reactors have the same footprint as light water reactors of the same power capacity. 6 HTR-PM would produce very similar steam conditions to some of China’s modern coal boilers (567°C, 13.25 MPa) and could work as a replacement for some of these highly polluting fossil power stations in population dense areas. All six reactor cores and steam generators are housed in a containment building and connected to a single steam turbine. Below is a picture of this 655MWe configuration.

Construction has been underway for a 2-core demonstrator since 2012, making China the forerunner of the development of this new generation of gas-cooled reactors. The reactor outlet temperature is 750°C.

TCR

The Transformational Challenge Rector (TCR) is a research project at Oak Ridge National Labs aiming at 3D-printing a helium-cooled reactor. It tries to demonstrate different new materials and manufacturing techniques, like the 3D-printed silicon carbide core. While this reactor does not seem to fit the description of a VHTR, because it is not graphite-moderated, I include it in this list, because it is using another solid, high-temperature moderator, yttrium hydride, which I think is the intent of the GIF definition. It also demonstrates TRISO particles that contain uranium nitride fuel kernels. It is not supposed to generate a new commercial reactor, but rather to demonstrate the new manufacturing techniques and embedded instrumentation for real-time analytics as a basis for new regulatory approaches.
Their first demonstrator will not be used to produce electricity, but rather as a heat source.

X-Energy Xe-100

X-Energy is developing a pebble bed helium-cooled reactor. It’s based on HTGR technology and their staff seems to have been drawn form South African talent originally working on the technology.
It’s fueled by some 220 000 graphite pebbles loaded with TRISO-X fuel particles. The core consists of a high-temperature tolerant graphite core structure. The reactor is designed for a 60-year operational life. The pebble bed design enables online refueling, has a good safety case and a high burn-up of 160 GWd/tHM.
The reactor is sized to produce 200MWth at 750°C outlet temperature and 7 MPa pressure. This is fed through a heat exchanger to produce steam at 565°C and 16.5 MPa. This is fed into a turbine to produce 75MWe.
The reactor can be scaled into a “four-pack” 300 MWe power plant.
It’s designed to bring nuclear power to countries, utilities and communities that were not able to afford or integrate a nuclear power system into their energy infrastructure.
X-Energy is producing their own TRISO fuel particles commercially.
Kam Ghaffarian, the founder of X-Energy, has also founded Axiom Space, which tries to commercialize the ISS and Intuitive Machines, which aims at build a space craft for landing on the moon.
NASA has contracted X-Energy to develop nuclear thermal propulsion systems, which are scheduled for a near-Earth technology demonstration in the 2025 timeframe. It seems like NASA’s intend is to enable a manned Mars mission by 2030.
Another envisioned application for the X-Energy systems is to provide lunar surface power, possibly for the future moon tourists, X-Energy’s sister companies are working to create.
Overall, the design and operating specifications of this reactor are quite similar to the HTR-PM.

Ultra Safe Nuclear Comapany MMR

Ultra Safe Nuclear Company (USNC) is developing the Micro Modular Reactor (MMR). USNC and Ontario Power Generation (OPG) have formed a joint venture to build, own and operate a proposed project at the Chalk River Laboratories site. In a sense, this makes this reactor the forerunner of the GenIV developments in the Western world.

Sporting a power output of 15MWth, 5MWe, the reactor is really small. The reactor system is factory built and shipped in standard ISO containers. Multiple modules can be linked together, which enables energy systems ranging from 5 MWe to 50 MWe or up to 150MWth or a combination of both. The usable temperature is designed to be 630°C. The core design has a rather low power density and a high heat capacity, which makes transients slow and predictable. The core is cooled by helium as primary coolant, which transfers heat to a molten salt secondary loop. This loop is a heat storage, which allows for an increased flexibility in the supply of electricity and process heat.
The system is supposed to be deployed in remote communities, which rely on diesel-electric power, to relief the people from the tremendous costs imposed on them by this extremely costly form of power generation. I am not aware of any plans to create an economically competitive electric energy system for more common locations.
USNC states that they are working on technologies for ship reactors, mobile emergency power and process heat comparable to the price of natural gas. At least in Asia and Europe.

The core consists of hexagonal graphite blocks containing stacks of USNC’s own “Fully Ceramic Microencapsulated” (FCM) fuel. This fuel consists of standard TRISO particles, that are encased in a fully dense silicon carbide matrix. In contrast to other TRISO fuel elements, this eliminates the outer graphite matrix, which determines the shape of the fuel elements. Using silicon carbide should make the fuel even more robust. USNC states resistance to temperatures of more than 2000°C. This makes this fuel into a candidate for space applications, where reliable power is needed for nuclear electric propulsion, in-situ resource utilization, mining, reprocessing of materials and general life support. This fuel could also enable nuclear thermal rockets with high specific impulses, which would be quite handy to unlock solar system transport.

U-Battery

U-Battery is an advanced reactor designed to produce low-carbon, cost effective and be a reliable power source of power and heat for energy intensive industries and remote locations. The conceptual design was developed at the University of Manchester and Delft. The project was initiated by Urenco and is now being developed in partnership with Jacobs and Kinectrics to realize the first deployment by 2028, possibly at a Urenco site.

The reactor is powered by accident tolerant TRISO fuel. The combination of fuel, reactor design and size creates an inherently safe design and reduces the area of the emergency planning zones, which should enable the usage very close physically to the consumers.
U-Battery has hexagonal graphite blocks containing the TRISO fuel and a helium-cooled first loop, which transfers heat to a secondary nitrogen loop, which in turn runs a gas turbine. The power output is supposed to be 10MWth and 4MWe at 710°C.

The consortium is working on licensing in the UK and in Canada.

StarCore

StarCore develops – or at least developed – a 36MWth, 20MWe high-temperature gas reactor, which is supposed to be buried in 57m deep silos, made of double-walled ultra high strength concrete. Above the ground is a 100m x 30m building housing the turbines and access tunnels to the silos.
The reactor has a 5 year refuelling schedule with uninterrupted power provided throughout the refuelling. The reactors units are sized to be delivered by truck. It operates at 6 MPa, is helium-cooled and uses TRISO fuel compacts supplied by BWXT. The fuel compacts are in a StarCore specific form: truncated cuboctahedrons, which look like a cube with its corners cut of. This enables a dense fuel packing, optimized gas channels and a stable, self-supporting matrix in the core. This design is called rectlinear core.
StarCore was one of the candidate designs under consideration for siting a small modular reactor unit at one of the CNL’s campuses.
While their website seems abandoned, there were recently news about the company’s activities, so I guess they are still active.

HolosGen

One of the most radical designs is HolosGen. This company integrates the nuclear fuel into a closed-loop Brayton turbine to form independent modules.
The module compresses helium via an electrically driven compressor, heats it with TRISO particles, which are filled into channels in a graphite block, directly feeds it into the turbine, cools it with an organic Rankine cycle and then runs it back through the compressor to repeat the cycle.

The ORC acts as a secondary heat conversion system, but is not critical for the safety case of the nuclear system. Other heat sinks would also work.

Another radically new feature of the design is the usage of multiple modules, which are each subcritical on the own. Each of the integrated “nuclear turbines” has a too high neutron leakage to become critical on its own. Several of these modules, normally 4, are positioned by a fast acting mechanical system – akin to the landing gear of airplanes – called Automatic Module Positioning System (AMPS), close enough together, so that the leakage neutrons of each module enter adjacent modules to make them critical.
If the system should ever lose power, the AMPS will be forced by gravity back into a position, where the entire configuration is non-critical. The AMPS could also be used to control the heat production of the modules by controlling the criticality of the independently subcritical cores via their relative positions.

The company is currently working on two turbine sizes. 4 of the smaller ones neatly fit into an ISO container an can produce up to 13MWe. One of the turbines can produce up to 3MWe. In this case neutron reflectors have to be positioned around the unit to make it critical.

he larger subcritical module, called the Holos Titan, occupies an ISO container on its own. 4 of these turbines can be combined to create a 81MWe power plant, which, including a secondary ORC power conversion system and auxiliary systems, fits into 12 standard containers.

While the smaller generators are aimed at remote communities and disaster relief, the larger version is supposed to be competitive with commercial power generation systems. Operating, above ground reactor installation would have to be housed in prefabricated shield building with walls built of concrete, rebar and boron. A small stack would be used to dilute activated argon. A perimeter fence around the reactor at 50m distance would keep dose rates well below NRC regulations for the general public.
In the paper “The Holos Reactor: A Distributable Power Generator with Transportable Subcritical Power Module” by HolosGen founder Claudio Filippone et al., the following estimates for LCOE are given, which, if they were to materialize, would be quite game-changing:

All of the used technologies have a technology readiness level of 8 individually, but have never been combined in this way before.
Rod Adams has a great interview on his show with the founders of HolosGen here.
Interestingly enough, the basic principle will work with helium, but also other gases, such as nitrogen or even supercritical CO2. They are also agnostic about the reactor specifics, as long as it produces reliable heat. In a sense, HolosGen is more about the development of the power conversion system, than it is about advancing nuclear engineering itself.
The proximity of the turbines spinning at exceedingly high velocities in close proximity to the fuel elements seems to pose some regulatory challenges.

Adams Nuclear Engines

A company, that’s long been defunct, still warrants an honorable mention here. Rod Adams, host of the Atomic Show, was working on a company for nuclear power systems. It featured TRISO fuel, an annular fuel core, where the coolant, which is nitrogen, enters at the outside and flows to the center to be fed into a Brayton gas turbine. The idea was that TRISO offers a superior safety case and the nitrogen coolant is close enough to air to enable the usage of fairly common turbines. The compressor was envisioned to be driven electrically. The decoupling of the turbine and compressor, while making the system more complicated and possibly costly, offers the ability to operate the system at high thermal conversion efficiencies for a wide-range of operating conditions, which is supposed to be net positive. The same principle design could be used for reactors of sizes between 1 and 100 MWe. It’s described here.
Recently Rod introduced the idea of using heavy nitrogen as coolant. Heavy nitrogen is far less likely to absorb neutrons than regular nitrogen. In fact the neutron absorption of regular nitrogen is high enough to use it as a secondary means of shutting the CO2-cooled AGRs down. It also eliminates some of the activation issues related to the creation of C-14 from nitrogen and might also contribute to a negative temperature coefficient of the overall system. On a per volume basis, heavy nitrogen offers a 40% greater heat capacity than helium. The mass of that volume is 7 times as big though, which will affect the electricity needed to run the compressor compared to helium. Nitrogen has to be enriched, which adds to the price of the system. But given that is eliminates the need for specialized helium turbines, the overall effect on the costs of electricity might still be positive.

Energy Multiplier Module

The last design on this list is the only fast reactor type on here. The Energy Mulitplier Module (EM²) designed by General Atomics, is a 265MWe helium-cooled fast reactor with a core outlet temperature of about 850°C. It converts fertile material in situ to fissile material and burns it. The core and the fuel elements are manufactured from General Atomics proprietary SiGA material, which is a silicon carbide composite.
The helium drives a Brayton cycle and a vertically mounted high-speed asynchronous generator.

There is some work on the possibility of using thorium as fertile material in this reactor.
The reactor and power plant are designed to be modular for an easier construction process. The power plant design incorporates redundant Direct Reactor Auxiliary Cooling Systems (DRACS) that are capable of 100% core heat removal via natural circulation, without any actuation, even under “station blackout” conditions.
These usage of a virtually melt-resistant construction material and a passive cooling system is the basis for the system’s safety case. It doesn’t seem like GA is actively working on this reactor.

Summary

There is a wide variety of designs for gas-cooled reactors and commercial developers. All of them try to create inherently safe systems that mostly rely on TRISO fuel to make their safety case. As you can see from the pictures, hardly any concept incorporates a protective building around the reactor. Instead, the “built in” containment of each TRISO particle is relied upon. It remains to be seen, if regulators agree to this choice. While TRISO seems to be great in terms of safety, it’s probably the most expensive nuclear fuel yet invented. It remains to be proven that the overall system and energy costs can be realized economically. The availability of uranium enriched to more than 5% U-235 as well as the recycling of spent TRISO fuel and large amounts of irradiated graphite pose questions yet unanswered.

Several companies are actively developing nuclear systems for space applications, like nuclear thermal and nuclear electric propulsion and are even looking into deploying such nuclear power systems on the moon.
The characteristics of gas-cooled nuclear systems enable their usage in mobile and remote locations for process heat and electricity at rather low power levels, which seems to be the basis of the business case for quite a few companies. All of these applications are in areas, where “normal” energy systems can’t operate or the cost of alternative energy systems is really high.

Astonishingly few companies seem to have the goal of producing electricity competitively on a utility scale. It doesn’t seem like there is commercial interest in increasing the output temperatures to the levels necessary for hydrogen production. Even GA’s design, utilizing the most advanced material, only barely reaches the minimal temperatures chemically necessary for driving the sulfur-iodine cycle.