Many engineers may have felt that the premature shutdowns of Fort St Vrain (320 MWe) and THTR (300 MWe) in 1989 put an end to high temperature gas cooled reactor development. However the concept has several attractive features and, currently, at least four HTGR development programmes are at an advanced stage. Current thinking is moving away from large output electrical generating plants and interest has also spread to parts of the world far removed from those countries which pioneered the HTGR.


The original idea came from Harwell in the UK and a zero-power physics reactor, Zenith, was built at a nuclear facility at Winfrith. In 1959, several OECD (Organisation for Economic Co-operation and Development) countries joined together to build the experimental reactor known as Dragon which operated successfully from 1966 to 1975. Dragon had a thermal output of 20 MW and achieved a gas outlet temperature of 750°C. It was intended to demonstrate the concept and to test different types of fuel assemblies but did not produce electricity. Dragon fuel assemblies contained fuel particles, approximately 1 mm in diameter, consisting of a kernel of fissile material and an outer coating. During the time Dragon was operating, new techniques for producing impervious coatings were developed. Typically these consisted of layers of pyrolytically deposited carbon and impervious silicon carbide which effectively sealed in all fission products. Tests carried out in Dragon showed that these particles operated very well and it was unnecessary to provide elaborate systems to remove fission products from the helium coolant. They can withstand operating temperatures of 1200 – 1300°C and continue to retain fission products at much higher temperatures. They have remained the basis for all HTGRs since then.

After Dragon, HTGR development split into two streams depending on how the fuel particles are compacted into fuel assemblies. In the pebble bed reactors, the particles, together with other moderator material, are compacted into spherical balls about 6 cm in diameter. The balls are loosely stacked within the reactor vessel and cooled by helium. Balls can be removed from the bottom of the reactor and after checking for damage and burn-up, they are returned to the reactor or discharged. The first reactor of this type was AVR, built at Jülich in West Germany. This 46 MWt (15 MWe) reactor operated from 1967 – 1988 and achieved gas outlet temperatures of 950°C.

The other stream consists of reactors in which the fuel particles are incorporated into replaceable graphite blocks of various configurations which have passages for the helium coolant. Typically these blocks are hexagonal prisms about 800 mm high and 360 mm across the flats. The forerunner of this type was the American Peach Bottom reactor which operated from 1967 – 1974 although it used pin-type fuel rather than blocks. It operated at the same time as Dragon and advanced the fuel technology by incorporating a buffer layer between the fuel kernel and the silicon carbide layer. It had a thermal rating of 115 MW and produced 40 MWe with a gas outlet temperature of 715°C.

These demonstration reactors were followed by two prototype 300 MWe class reactors, the pebble bed THTR in Germany which operated from 1985 to 1989 and Fort St Vrain in the US, a prismatic fuel reactor which operated from 1976 to 1989. Both plants were shutdown before the end of their design lifetimes for reasons which were not directly related to the HTGR concept. After that interest shifted away from large power reactors to designs which exploit other characteristics such as the HTGR’s high gas temperature, its inherent safety features and the possibility of using direct cycle gas turbines.


At present, four HTGRs are under construction or in the detailed design stage.

HTTR – High Temperature Test Reactor

Japan’s HTTR has a 30 MWt rating and no generating plant. Initially all heat will be rejected to the atmosphere but later, when the characteristics of the plant have been established, high temperature helium will be taken off from a 10 MWt He/He heat exchanger and used to test various chemical applications including the methane-steam reforming process mentioned above.

This reactor (described previously in the October 1997 issue of NEI) uses the prismatic fuel design and the reactor together with the He/steam and He/He heat exchangers is enclosed in a steel containment vessel.

Construction was completed last year with fuel loading to be completed in September. The initial core loading consisted of installing only the outer ring of fuel blocks, leaving the centre filled with dummy fuel, so that low power tests can be done with an annular core configuration. The centre blocks will then be filled in to give the full size core. These tests will provide useful data for designing future annular cores. Initially tests will be done with a gas outlet temperature of 850°C and this will later be increased to the design value of 950°C.

HTR-10 (China)

The HTR-10 is a 10 MWt demonstration reactor under construction at the Institute of Nuclear Energy Technology (INET) of Tsinghua University, north west of Beijing. Like HTTR, its aims are to demonstrate that modular HTGRs can be built and operated safely and can produce high temperature gas for industrial processes and electricity.

The HTR-10 is a pebble bed reactor and so it will be useful to share and compare data obtained from HTR-10 and the prismatic-fuel reactor HTTR. The buildings for HTR-10 have been constructed and equipment is currently being installed. The reactor is expected to go critical in 1999.

The fuel consists of 60 mm diameter spheres and is very similar to that of earlier pebble bed reactors. Equipment for prototype scale production was transferred from the German Nukem facility to INET. The complete core will contain 27000 of these spheres. In other respects, the design of the plant differs considerably from earlier German plants. The primary circuit consists of two steel pressure vessels, one containing the reactor and the other a single heat exchanger unit. They are arranged side by side in concrete containments and are connected by a single concentric gas duct.

The heat exchanger is of a unique design in which the outer annulus contains a number of helically tubed steam generating units. The space in the centre will be used later to install a helium-gas heat exchanger. In the first stage, HTR-10 will be operated with a gas inlet temperature of 250°C and an outlet temperature of 700°C. Heat will be removed by the steam generators and used to drive a small turbine. In the second stage, a 5 MWt He/nitrogen heat exchanger will be added and the temperatures increased to 300°C inlet and 900°C outlet. Gas at 850°C will be taken from the heat exchanger for process heat applications.

Decay heat removal is done using the completely passive method mentioned earlier. Heat is conducted through the core and radiated from the pressure vessel wall to the surface of the concrete housing. The concrete surfaces are covered with water cooled panels in which the water is circulated by natural convection to air coolers where the heat is discharged to atmosphere. These panels are mainly intended to protect the concrete and pressure vessel from overheating and the ultimate safety of the reactor does not depend on this cooling system.

There are plans to test HTR-10 as a heat source for the same methane-steam reforming process that will be investigated with HTTR. Eventually, however, it may be possible to use this type of reactor for the liquefaction or gasification of coal. China has large reserves of coal in the north west of the country and alternative ways of transporting this source of energy to distant consumers would be beneficial in relieving the burden on rail and river transportation.

Gas turbine – modular helium reactor (GT-MHR)

The GT-MHR is an international joint project currently underway in Russia. After the Three Mile Island accident, emphasis in the US changed to modular designs and GA developed the MHT-GR design as part of a US Department of Energy (DOE) programme. The design used small modules to take advantage of the emergency core cooling feature mentioned earlier. Large plant outputs were to be achieved by grouping together several of these modules. At this stage it was expected that conventional steam cycle generating equipment would be used but later the interest switched to direct cycle gas turbines and the design became known as GT-MHR. In December 1994, the DOE stopped funding new reactor designs including the GT-MHR.

Subsequently, GA proposed that the GT-MHR plant should initially be used to burn military plutonium. In February 1995, GA and the Russian Ministry of Atomic Energy, MINATOM, signed a memorandum of understanding to develop the GT-MHR. GA has the role of project manager and MINATOM coordinates the design and development work being done by several Russian organisations. Framatome joined the project in January 1996 and Japan’s Fuji Electric joined in April 1997. The CEA of France will also participate in safety analyses. The project is currently being financed by MINATOM and the participating companies. The US government is not contributing but, of course, the project has benefited from the results of the earlier DOE sponsored programme.

The project is to be implemented in 4 steps. A conceptual design was produced in step 1 and this was completed in October 1997. In step 2 a detailed design is being produced and a programme of research and development has been started. This work should take 5 years. During step 3, a prototype is to be built and tested in Russia and in step 4 a commercial version will be made available for marketing worldwide.

The joint development with Russia has several advantages. Development and construction cost are considerably reduced. By using Russian engineers, research facilities and construction companies, costs are about one third of the costs needed to develop and deploy the GT-MHR for plutonium disposal in the US. Also Tomsk 7 (now renamed Seversk), offers a secure site for the prototype which would be able to replace the power currently generated by plutonium production reactors. Another, non-technical advantage is that it provides peaceful employment for scientists and engineers who were previously employed in the Russian military projects.

The design is for a 600 MW thermal, 285 MWe modular plant. Each module consists of two pressure vessels mounted side by side in a concrete containment 35 m below ground level. One vessel contains the reactor and the other the power conversion equipment. They are interconnected by a concentric duct. Hot gas leaves the reactor at 850°C and 70 bar pressure and passes through the turbine, exiting at a pressure of 26 bar. It then passes through a “recuperator” (a regenerative heat exchanger) and a water cooled heat exchanger before entering the two stage compressor equipped with an intercooler. The compressor returns the helium to the reactor through the other side of the recuperator where the temperature is increased to 490°C. This scheme is expected to give a thermal efficiency of close to 50%.

As would be expected from GA, the core is a prismatic design. The centre consists of 61 columns of graphite reflector blocks, containing no fuel. These are surrounded by 102 columns of fuel bearing blocks which form the annular core and these in turn are surrounded by replaceable reflector blocks. This arrangement allows the power output to be increased without losing the passive shutdown heat removal characteristic. The fuel uses pure plutonium with erbium burnable poison. As no uranium is used to dilute the plutonium, no secondary plutonium is formed. Consequently, the GT-MHR can consume 90% of the loaded Pu-239 in a single pass, which is much higher than other reactor types (for a LWR the comparable figure is 50%). The GT-MHR also produces more power per ton of weapons grade plutonium than other reactors assuming that the spent fuel is not recycled. If three 4-module plants were constructed they could dispose of 50 t of weapons grade plutonium in 25 – 30 years. The cost of the three 4-module plants was estimated to be $3.12 billion in 1996 (less than $1000/kW).

The reactor vessel is equipped with a water cooled heat exchanger to remove shut down heat. However the ultimate safety does not depend on this system; as mentioned before, if coolant circulation is lost, the reactor can be safely cooled by radiation from the reactor vessel to the containment cavity walls. These are covered with water cooled panels to absorb the heat but if these in turn should fail the heat would be absorbed by the concrete itself. The reactor can withstand loss of coolant circulation and loss of coolant inventory accidents without operator intervention or use of active safety systems.

South African Pebble Bed Modular Reactor (PBMR)

The PBMR-SA project is being carried out by ESKOM, the South African state-owned utility. ESKOM operates two PWRs at its Koeberg site. The ambitious aims of this project are to achieve a $1000/kW construction cost and a 2 cents/kWh generating cost using a modular pebble bed reactor with a direct cycle gas turbine generator. The modules are to be approximately 100 MWe and a typical power station would have 10 modules. The first such plant is expected to be built at ESKOM’s Koeberg site.

The design work is being done by South African companies including the SA Atomic Energy Corporation and Integrators of Systems Technology. A nuclear reactor consulting team has also been assembled from German companies and experts. The conceptual design was carried out from March 1996 to March 1997 and the plan was approved by the government in February 1998. ESKOM made a policy decision to proceed in March.

The plan calls for the detailed design and licence application to be completed this year. After this, a prototype could be constructed between 1999 and 2001 and tests to evaluate its performance between 2001 and 2003. ESKOM is taking the risk of constructing the prototype without going through extensive rig tests. This is justified by the fact that the technologies are already well advanced and the cost of the prototype could be cheaper than the rigs.

Each module consists of three steel pressure vessels in a concrete building, one for the reactor and two for the power conversion equipment. Two high speed (14,000 rpm) turbo-compressors and an intercooler are in one vessel and the power generator, recuperator and pre-cooler are in the other vessel. Reactor gas inlet conditions are about 558°C at a pressure of 70 bar with an outlet temperature of 900°C. Plant efficiency is expected to be about 45%.

ESKOM thinks that the design objectives can be achieved because very little safety grade equipment is required due to the inherent safety features of the reactor. This will enable them to investigate suppliers worldwide to find the most suitable products and, in fact, most components can be made in South Africa. Serial production of the modules will reduce costs and construction times could be as short as 24 months due to the small size of the modules.

Other Design Studies

In addition to the four projects described above, HTGR technology has been the basis for a number of design studies. One of these, the INCOGEN Pre-feasibility study was supported by the Netherlands Ministry of Economic Affairs. INCOGEN stands for Inherently Safe Nuclear Cogeneration. The study was made between 1996 and 1997 by a number of consultant and research organisations in the Netherlands (ECN, IRI, KEMA, Stork Nucon and Romawa) and the report was issued in October 1997.

The study is based on the concept that a small high temperature reactor could be shown to be inherently safe to such a degree that it could be installed close to an industrial facility where it could supply both power and process heat. It would then perform the same function as the diesel or gas turbine cogeneration plants in use today. Although similar to modular designs, the nuclear cogen plants would be dispersed rather than concentrated in a power station. To make this possible, the design has to be cheap, demonstrably safe and very easy to operate and maintain. It must therefore be as simple as possible.

The INCOGEN study was based on a HTR design produced by the American company LPI (Longmark Power International). This is a 40 MW thermal pebble bed reactor with a single shaft gas turbine and heat exchangers which provide process heat. The generator produces 16.8 MWe, process heat in the range of 40 – 150°C, accounts for 18.2 MW and the remaining heat between 40°C and ambient is rejected to the environment.

The reactor is based on the original German pebble-bed design principles and uses the same 60 mm balls. However to keep operations simple, the continuous recycling of the balls has been dispensed with and, instead, the little-by-little (sometimes called PAP from the French peu-à-peu) system has been adopted. In this system, no balls are discharged while the plant is operating and a small number of balls are added each day to compensate for burnup.

This process could continue for a long time, 10 years in the INCOGEN study, before the reactor vessel becomes full. The INCOGEN core would be 2.5 m in diameter and initially would be filled to a height of 1.05 m with balls containing 10% enriched uranium. Balls with an enrichment of 19.75% would be added gradually until the height reached 4.5 m, when the reactor would be shut down for refuelling. An alternative method would be to load a pebble bed core with fuel containing sufficient burnable poison to last for say 3 years. The whole core could then be replaced in one operation.

Although the INCOGEN study has finished, further studies are going ahead in the Netherlands. The report however, cautions that such a plant, although feasible, may not be economical in regions which have access to cheap natural gas.

Strategic characteristics of HTGR

• Industrial applications
Gas temperatures of 950°C are high enough to be used for many chemical processes. The one which will be tried first is the reforming of steam and natural gas to produce hydrogen. At a later stage, other chemical processes and the gasification of coal or even steel making could be possible. The Japan Atomic Energy Research Institute (JAERI) is currently working on a method of decomposing water into hydrogen and oxygen at high temperatures using iodine and sulphur dioxide as reactants which are continuously recycled. The process, which originated at GA, is still in the laboratory stage but some promising results have been obtained. HTGR temperatures would be suitable for such a process which would be a way of making hydrogen fuel without producing carbon dioxide as a by-product.

• Inherent safety
The inherent safety characteristics of this reactor arise from the fact that the core is composed of graphite and ceramic materials which can withstand very high temperatures. If the output of the plant is not too large and the power density is reasonably low, shutdown and decay heat can be removed by conduction through the core and radiation from the pressure vessel walls. For such a reactor, maintaining coolant circulation is no longer an issue and there is no need for engineered core cooling systems. The limiting power output of this type of reactor can be increased by using annular cores.

• Direct cycle gas turbines
The advantages of direct cycle gas turbines are self-evident. The whole secondary steam cycle is eliminated together with the steam generators which have been the cause of so many headaches in other types of reactors.