Technology development for ITER

Seven major technology projects to support the ITER design activities

November 29, 1999

The next major step in the world strategy for fusion development is an experimental fusion reactor, the International Thermonuclear Experimental Reactor ITER. In 1995, to support the ITER design activities, seven major technology projects were started. Two model coils, a prototype vacuum vessel sector, blanket modules, divertor components and remote handling devices for blanket and divertor, were established to demonstrate manufacturing feasibility and to confirm the performance of the main ITER components.

The cost of these projects - about 400 million US Dollars up to July 1998 - incorporated a large part of the already existing fusion technology programmes worldwide, leading to their strong focus on a common goal.

ITER has been developed in an international framework, involving the world's leading fusion programmes - the European Union (including Canada), Japan, the Russian Federation (including Kazachstan) and - till 1998 - the USA. Its objective is to demonstrate the scientific and technological feasibility of fusion. ITER would achieve this by producing an energy-delivering deuterium-tritium plasma and by demonstrating technologies essential to a fusion power station.

While manufacturing of most of these prototypes and models is now successfully completed, the testing programmes are still continuing. As the results already confirm the ITER design, the construction of the device could have started after completion of the ITER Final Design Report in 1998. However, due to financial difficulties of the ITER parties, it was decided to prolong the design phase till 2001 for examination of reduced-cost options for ITER. While the detailed technical objectives had to be reduced, the main programmatic objectives should be maintained. This smaller device having similar technology needs, it will make use of most of the existing design solutions and the favourable results of the ITER technology projects.

Full size model – relative to the 1998 design – of one half sector of the ITER plasma vessel, made in Japan.

Vacuum Vessel Sector
With a height of 15 metres the ITER plasma vessel - a ring-shaped stainless steel shell, containing the plasma - is about three times the linear dimensions of the largest existing plasma vessel and also much more complex. Therefore only a model at full scale could confirm its feasibility and verify the mechanical characteristics of the design. The vessel is double-walled for stability reasons and fitted with 60 ports giving access to plasma heating, diagnostics, vacuum pumping, and remote handling. Outer and inner wall are connected by supporting ribs, and the interspace is filled with water and an array of steel plate inserts for neutron shielding. A full size model of one sector - 1/20 of the ITER vessel - has been built to provide information about critical issues of fabrication such as welding distortions, achievable tolerances as well as non-destructive inspection techniques.

The model has been manufactured by the Japanese Home Team, with Hitachi and Toshiba having each built half sectors. This distributed manufacturing offered opportunities to test and compare different weld schemes. The Russian Home Team has built a port extension and shipped it to Japan for integration with the sector model. With assembly and testing of the vessel sector model being completed successfully, the viability of design and fabrication techniques has been confirmed. In particular, it proved possible to weld together the two sector halves by a fully remotized welding and cutting system developed by the US Home Team. The tolerances achieved - about ± 5 mm - were well below the required values. This should enable several manufacturers worldwide to fabricate the vessel, thus widening the manufacturing base and allowing costs to be minimised. The vessel model will be further used for remote cutting and welding tests to simulate assembly and changing of vessel segments.

Model Magnet Coils
ITER will be a superconducting tokamak. The magnetic field cage to confine the plasma will be built up by 29 magnet coils of three different types, the largest superconducting coils ever built. To reach the desired magnetic field strength of up to 13 Tesla (at the coil) an advanced superconducting material - niobium-tin - is used. For superconductivity - and consequently strongly reduced electrical power demands - the coils will be cooled down by liquid Helium to a temperature of 4.5 Kelvin, near absolute zero. Two model coils of smaller size than the original ITER coils are sufficient to verify the conductor performance and to demonstrate the major steps of manufacturing - starting from a single superconducting strand up to the complete coil: a sub-size model coil for the Central Solenoid, which in ITER will drive the plasma current, and a model coil for one of the 20 main field coils.

The Central Solenoid Model Coil
The Central Solenoid model coil is to produce a magnetic field comparable to that of the full size coils. Having an outer diameter of 3.6 and a height of 2.8 metres, it consists of two modules nested inside each other. An inner module, made in the USA, is placed into an outer module, made in Japan. The superconducting niobium-tin strands used were fabricated in Europe, Japan, and the USA. For conductor tests, three separate smaller insert coils, which fit within a bore of the main coil, will also be tested. They are made from different conductors fabricated in Japan and Russia. In total 10 institutes, 11 industries for superconducting strand and cable manufacturing, 3 industries for jacket and conductor fabrication, and 7 for winding and coil fabrication, participated in the effort. Interface issues, intercontinental transportation, quality assurance, and custom clearance problems, have all been solved successfully. Thus not only technological but also valuable administrative experience in handling a complex international collaboration was gained.

Building the model coils required significant advances on present technology, but the main manufacturing issues are now all successfully solved. These include the production of a substantial quantity of niobium-tin, the enclosing of a cable of this strand in a massive metal jacket to provide support against magnetic forces, and bending the conductor accurately to the winding shape. As the superconducting niobium-tin compound is brittle, the wires are to be fabricated initially containing both compounds separated in a copper matrix. They react together only after a 200 hour heat treatment at 650 °C. This must be performed in an oxygen-free atmosphere after all cabling and conductor bending operations are completed, but before any temperature sensitive coil components are added. After the heat treatment in an oxygen-free atmosphere the conductor was insulated and impregnated with epoxy resin. Finally, the different layers have been connected. The assembly of the two modules has recently taken place in the Test Facility in Naka, provided by the Japanese Home Team. When the model coil and its different inserts have demonstrated the basic operation requirements, they will subsequently be operated beyond the design conditions. The results will be used to optimise the magnet design, especially conductors and insulation system.

Main Field Model Coil
While the model coil for the Central Solenoid forms the testbed for the conductor, the sub-size model for the 20 main field coils is mainly intended to demonstrate the more complex winding technique used for these coils. Critical manufacturing steps need to be defined and the performance and reliability of each component integrated in the magnet demonstrated. The European Home Team in Garching has the lead responsibility for the project in collaboration with the Joint Central Team in Naka.

For manufacturing the subsize coil - about 4 metres high and 3 metres wide - a different mechanical support compared to the Central Solenoid model was used. The niobium-tin cable was jacketed in a thin-walled conduit, bent, heat-treated, and insulated. For mechanical support it was enclosed in a spiral grove on both sides of a massive radial plate. Five of these plates were then stacked together and again supported by a steel case. The coil is now to be tested in the TOSKA facility at Forschungszentrum Karlsruhe, Germany, under realistic magnetic loads.

Blanket Modules
The blanket, which covers the inside of the plasma vessel, collects the power of the high-energy neutrons produced in the fusion process and shields plasma vessel, magnet coils and outer parts of the device from the neutrons. (In a later ITER operation phase this shielding blanket may be replaced by a breeding blanket providing a large proportion of the tritium needed.) To ease remote maintenance, the blanket consists of single modules. With a weight of about 4.5 tonnes each they will be inserted into the plasma vessel through vessel ports by a special remote handling device. There the modules will be fixed on the vessel wall and connected to water cooling pipes, both stainless steel and pressurised water providing the shielding against the neutrons.

To assess the manufacturing feasibility and to develop bolting, welding and cutting tools for use in remote removal, several small scale mockups and two full-scale blanket modules and their attachment systems were fabricated. To demonstrate the performance, representative parts of the components are being tested under relevant conditions. In addition, confirmation of the design choices is being obtained by accompanying research on materials, joining techniques and neutronics. All four Parties are involved in this work. The European Home Team has lead responsibility for implementing the project, in collaboration with the Japanese and Russian Home Teams and the Joint Central Team at Garching.

The results up to now show that the ITER blanket modules can be manufactured. The main materials and joint techniques have been tested, options selected and qualified for ITER operation. Two full scale massive multi-layered modules were build in Japan and Europe. Made of stainless steel as structural material, they are covered by a first wall - to diffuse electro-magnetic radiation from the plasma to the coolant - made from copper alloy with steel channels for water cooling. Facing the plasma directly is protection material such as beryllium, carbon or tungsten. For manufacturing the complex shaped prototypes two different techniques - conventional forge drilling as well as "powder hot isostatic pressing" - have been used. In this advanced manufacturing technique steel powder is baked together into the desired shape and, by the same process, subsequently connected to copper and beryllium. With both methods the desired tolerances of less than 2 mm have been achieved. Future work will focus on the completion and testing of the prototype modules, and on assembly tests - such as module insertion, bolting and unbolting, cutting, and rewelding.

Blanket Remote Handling Project
Um beschädigte Module austauschen und - in einer späteren ITER-Experimentierphase - das Tritium-Brutblanket in die ITER-Anlage einbauen zu können, müssen die einzelnen Module auswechselbar sein. Wegen der hohen Strahlung wurde dazu ein fernbedient arbeitendes System entwickelt - vier auf einer Schiene im Plasmagefäß laufende Roboterfahrzeuge. Vollständig fernbedient werden die Schienenteile zunächst durch vier Öffnungen in das Plasmagefäß eingeführt, dort zusammengebaut und verankert. Die Fahrzeuge, die jeweils mit einem sechs Meter langen Greifarm ausgerüstet sind, handhaben die schweren Blanket-Module mit hoher Genauigkeit. Der Greifer kann die Module an der Stützwand befestigen und wieder lösen, Kühlwasserleitungen zusammenschweißen und trennen sowie mit einer Videokamera das Gefäß inspizieren.

Full-scale testing and verification has been successfully carried out on a test platform set up at Naka in Japan, involving also contributions from Europe. Simulating the full scale structure of a complete half of the ITER in-vessel region, it comprises module and port handling equipment, auxiliary remote handling tools and a blanket mock-up structure. The heavy modules could be positioned with the high accuracy of one millimetre. A sophisticated test programme - with purposely installed misfunctions - is now to follow.

Divertor Project
The divertor is an important component of a fusion device. It exhausts heating power, plasma impurities, and the fusion reaction product, helium, from the plasma. For this purpose a specially formed magnetic field guides the plasma boundary to water-cooled divertor plates at the bottom of the plasma vessel. The ITER divertor technology project was set up to demonstrate the feasibility of full-scale armoured plasma facing components with the necessary heat flux handling capability of up to 20 MW/m2. This includes development of the materials and joints, their qualification by high heat flux tests, as well as characterisation of their erosion properties.

To ease remote replacement the ITER divertor will be build up from 60 modular "cassettes" able to accomodate a stationary power load of up to 300 Megawatts. Aside from several small scale models the project culminated in construction of a full scale mockup of part of the divertor cassette body and its characteristic plasma facing components. All four parties - Europe, Japan, the Russian Federation and the USA - have contributed to this development.

At the start of the project the technology for the ITER divertor high heat-flux components did not exist. The programme in the meantime has fully met its objective: the technology needed to fabricate full-scale armoured components has been developed. In a broad effort a variety of concepts for materials, geometry, radiation resistance, and manufacturing techniques were assessed. Materials now are chosen and the basic design is confirmed. Joints between armour and heat sink (such as between carbon or tungsten and copper) proved to be possible. The components are able to sustain thermo-hydraulic and electro-mechanical loads, whilst utilising the most cost effective and reliable manufacturing processes. Full size vacuum compatible castings for the cassette body have been demonstrated and shown to have satisfactory mechanical properties. In high heat flux tests the reliability of the plasma facing components has been demonstrated. Various geometries and materials tested could meet the ITER requirements and withstand also abnormal heat-loads up to 20 MW/m2 for more than 1000 cycles without failure. Also erosion as well as fatigue lifetimes meet the ITER requirements.

Divertor Remote Handling
Quick and reliable replacement of the highly stressed divertor components are crucial for high availability of ITER. This project therefore was established to demonstrate the feasibility of the rapid remote replacement of the divertor cassettes, and their refurbishment - i.e. the replacement of plasma facing components - in hot cells. For this purpose a full scale mockup of part of the lower vacuum vessel and ports was constructed, along with all the various remote handling equipment and several dummy cassettes. In addition, a refurbishment platform to demonstrate the remote replaceability of plasma facing components on the cassette was constructed. Both platforms were set up at Brasimone (Italy, Europe). The project involves also contributions from Canada and Japan.

The divertor test platform is used to simulate at full scale all handling operations inside the vacuum vessel - such as locking and securing the supports, making water pipe connections, and assembling electrical connectors, including also removal and replacement the 25 tonnes divertor cassettes through vessel ports. The tests already have confirmed the maintenance concept. Further investigations are planned to reduce costs, and to improve the man-machine interface. The divertor refurbishment platform - now fully operational - is simulating the most critical operations to be undertaken in the hot cell. The assembly and disassembly of high heat flux components is simulated with prototype tools. Tests so far show that an armour mockup can be installed on the cassette with the required accuracy. Further work is needed to streamline procedures and to shorten the refurbishment time.

Die Reparatur-Plattform ist inzwischen vollständig betriebsbereit. Hier werden die kritischen Arbeitsschritte in der Heißen Zelle simuliert: Mit Prototyp-Werkzeugen wird das Auswechseln der hoch-hitzebelasteten Bauteile getestet. Die Montage der Divertorplatten auf der Kassette konnte inzwischen bereits mit der nötigen Genauigkeit vorgeführt werden. Weitere Untersuchungen sollen insbesondere dazu dienen, die Reparaturzeiten zu verringern.

The ITER Design Activity
The ITER Final Design Report (July 1998) provided the first ever comprehensive design of a full-scale fusion reactor experiment based on well established physics and technology. However, though the cost of the device of 6.7 billion Euro had been maintained at a comparable level throughout the 10 year design phase, due to changes in the economic climate the ITER parties were unable to proceed to the construction of the device. In 1998 the USA withdrew from the project, but completed their technology contributions till mid-1999. The remaining ITER parties decided to examine reduced-cost options for ITER by reducing the technical goals and technical margins while maintaining the overall programmatic objective.

Under the same technical assumptions, a device which would achieve an energy gain of at least 10 at a direct capital cost of approximately 50 % of the 1998 ITER design would achieve this. Such a device would make cost-effective use of existing design solutions. The remaining ITER Parties therefore decided to extend the design phase for three years, to design the reduced cost device, as well as to negotiate a construction and siting agreement. Work on ITER since mid-1998 has focussed on the key issues involved in the revised technical objectives, and on understanding the choices and trade-offs that need to be made between performance and cost reductions. This process of convergence of the design to a single option is nearly finished. The Engineering Design Activities will be completed by mid-2001. About ten years after granting of a construction licence ITER could produce the first plasma.

Isabella Milch

Go to Editor View