Status Report on the Toroidal Field Coils for the ITER Project

IEEE TRANSACTIONS ON APPLIED SUPERCONDUCTIVITY, VOL. 20, NO. 3, JUNE 2010 381 Status Report on the Toroidal Field Coils for the ITER Project F. Savary, A. Bonito-Oliva, R. Gallix, J. Knaster, N. Koizumi, N. Mitchell, H. Nakajima, K. Okuno, and C. Sborchia Abstract—The magnet system for ITER comprises 18 Toroidal cable-in-conduit superconductor, Field (TF) Coils using 3 which operate at 4.5 K in supercritical helium. The procurement of the TF Coils and Structures is amongst the first which have been launched following the creation of the ITER Organization (IO). It is organized in 4 phases. A Procurement Design Readiness Review held in April 2008 confirmed the readiness of the design to proceed with Phases I and II. Procurement Arrangements (PA) were signed with the European and Japanese Domestic Agencies (DA) respectively in June and November 2008. After a brief description of the TF Coils and Structures, the paper gives an overview of the PA showing the milestones towards series production. The procurement strategy of both DA involved is described, in particular the first step which covers pre-production activities: qualification of raw materials, manufacturing trials, mock-ups and full-scale prototype radial plates, impregnation tests and, possibly, winding trials. The work carried out by IO is also presented: optimization of the cover plate welding to satisfy the allowable stress criteria while minimizing the associated distortions, qualification of blends of cyanate ester with epoxy resin for the impregnation of the winding packs and design of the coil terminal region including integration of the needed instrumentation. Nb Sn Index Terms—Cover plate welding, cyanate ester, ITER toroidal field coils and structures, procurement arrangement, TFC terminal region. I. INTRODUCTION Fig. 1. A pair of TF Coils with their terminal regions, OIS, IOIS, ILIS, slots for the poloidal shear keys or IIS, gravity support and interface areas for the PFC supports. Central Solenoid (CS), 6 Poloidal Field Coils (PFC) and 18 Correction Coils (CC). II. DESCRIPTION OF THE TF COILS AND STRUCTURES HE construction phase of the ITER Project started following the signature of the ITER Agreement on November 21st, 2006. After the Headquarters Agreement establishing the legal status of the ITER Organization (IO), which entered into force on April 9th 2008, each Member has established its Domestic Agency (DA) [1]. The IO is located at the construction site in Cadarache-France. The magnet system for ITER includes 18 Toroidal Field Coils (TFC) using cable-in-conduit superconductor (CICC) operating at 4.5 K thanks to supercritical helium flowing in the central cooling channel of the conductor. The magnet system also includes a T Manuscript received October 19, 2009. First published March 01, 2010; current version published May 28, 2010. F. Savary, R. Gallix, J. Knaster, and N. Mitchell are with the ITER Organization, Saint-Paul-Lez-Durance 13108, France (e-mail: [email protected]). A. Bonito-Oliva and C. Sborchia are with Fusion for Energy (F4E), Barcelona 08019, Spain (e-mail: [email protected]). N. Koizumi, H. Nakajima, and K. Okuno are with the Japan Atomic Energy Agency (JAEA), Naka-shi, Ibaraki-ken 311-0193, Japan (e-mail: koizumi. [email protected]). Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/TASC.2010.2040914 The 18 TF Coils are ‘D’ shaped as shown in Fig. 1 and consist of a Winding Pack (WP) enclosed in a structural steel case, the Toroidal Field Coil Case (TFCC), made up of 4 sub-assemblies as shown in Fig. 2. The WP is a bonded structure of 7 Double Pancakes (DP), each made up of a radial plate (RP) housing the reacted CICC, closed by cover plates and wrapped with ground insulation. A WP is made of 5 regular DP containing each 760 m of conductor and 2 side DP containing each 415 m of conductor. It has an external ground insulation of 7 mm thickness. A cross section of the WP is shown in Fig. 3. The coil terminal region, which protrudes from the TFCC at its lower curved part, includes the 2 conductor terminal joints, 6 DP joints and the helium feeder manifolds. The ITER magnet system embodies mechanical structures to provide support against gravity and the much larger magnetic forces. The TFCC and the TFC Structures (TFCS) together with the gravity supports (GS) carry the entire magnet system, i.e. the TFC, the PFC, the CC and the CS. They also support the vacuum vessel thermal shield. The intercoil structures are shown in Fig. 1. They include: the Outer Intercoil Structures (OIS), the Intermediate Outer Intercoil Structures (IOIS), the poloidal shear keys or Inner Intercoil Structures (IIS) and the Inner Leg Intercoil Structures (ILIS). 1051-8223/$26.00 © 2010 IEEE 382 IEEE TRANSACTIONS ON APPLIED SUPERCONDUCTIVITY, VOL. 20, NO. 3, JUNE 2010 III. OVERVIEW OF THE PROCUREMENT ARRANGEMENTS The TFC and TFCS are procured in-kind. The 18 TFC plus one spare unit will be delivered by the Japanese and the European Domestic Agencies (DA). A Procurement Arrangement (PA) was signed in June 2008 with Fusion for Energy (F4E), the European DA located in Barcelona-Spain, for 10 TFC and in November 2008 with the Japan Atomic Energy Agency (JAEA), the Japanese DA located in Naka-Japan, for 9 TFC. Another PA was signed in November 2008 with JAEA for the supply of the structures for 9 TFC. It will be amended to cover the structures for the 10 remaining TFC. The PAs are split in 4 phases as follows. A. Phase I: Work Definition for Phase II and Call for Tender Fig. 2. The 4 elements of the TFCC, marked-up AU-BU for the U-shaped sub-assemblies and AP-BP for the closure plates, and the WP with its terminal region. During this phase, the DA identifies its strategy for procuring the TFC, carries out a call for tender and places contracts with industry. B. Phase II: Process Optimization and Qualification Fig. 3. Cross-section of the WP showing the inboard and the outboard legs with 5 regular DP and 2 side DP. TABLE I MAIN PARAMETERS OF THE TFC This phase includes: 1) selection, characterization and qualification of materials; 2) industrialization and cost optimization of the fabrication processes, including detail design of the main tooling, procurement and qualification of the critical tooling and finally, some design adjustment to improve manufacturability; 3) preparation of manufacturing drawings; 4) manufacture of a full-scale dummy double pancake to qualify the tooling, the processes and the final electrical design performance; manufacture of mock-ups for the development and qualification of the TFCC welds; 5) definition of Manufacturing Plans for the TFC and TFCS as a result of items (2) and (4). C. Phase III: First-of-Series Coils This phase covers the fabrication and testing of a WP, a TFCC for the casing operation, the casing operation, i.e. the insertion of the WP into the TFCC, and the finishing operations for the first-of-series coils, as well as a full set of TFCS intercoil components. The European and Japanese DA will both produce a first-of-series coil. D. Phase IV: Series Production This phase comprises the production of the 16 (18 minus 1 per each DA) remaining TFC plus the spare unit, including the WP, the TFCC for the casing operation and the TFCS intercoil components for the assembly at the ITER site. IV. PROCUREMENT STRATEGY OF THE DOMESTIC AGENCIES A. The European Strategy by F4E The main parameters of the TFC are given in Table I. The TFCS also include 2 sets of 3 pre-compression rings (PCR) to put the 18 TFC assembly in hoop compression and thus, avoid breathing and wear of the poloidal shear keys and key slots. The 18 TFC have identical WP but the TFCC attachments and interfaces required to support the CS, PFC and CC are not identical for all TFC and 3 TFC include Rogowski loops to measure the plasma current, resulting in 6 TFC variants. Because the construction of the TFC requires a combination of large capacity and broad range of capability, which are very difficult to find on the market, F4E split its scope of supply in three more accessible and specific work packages: the RP, the WP and the casing operation. This strategy is expected to bring higher level of competition and to identify best suppliers for each specific work package. The RP work package is split in 2 phases: the first covering 2 prototype RP (1 side and 1 regular) through an open call for SAVARY et al.: STATUS REPORT ON THE TOROIDAL FIELD COILS FOR THE ITER PROJECT tender aiming at 2 separate contracts and the second covering the series production through an additional call for tender. The 2 prototypes are made intentionally with different technologies to explore different routes and finally, the series production will be carried out with the technology demonstrating better quality and being most cost effective. 383 TABLE II ITER STRUCTURAL DESIGN CRITERIA—ALLOWABLE STRESSES AT 4 K B. The Japanese Strategy by JAEA Contrary to F4E, JAEA is seeking coordination of the work through a single main contractor to reduce interfaces and contain sub-contracting costs. This should also allow applying a uniform overall QA program and a coordinated approach to the schedule. The whole process covering Phases II, III and IV as described in the PA will be likely split in 3 contracts covering both TFC and TFCS. The first contract will cover a major part of Phase II as stipulated in the PA including: 1) trials and mock-ups to set-up the manufacturing procedure for the radial/cover plates and fabrication of a full-scale radial plate including a set of cover plates, 2) complete manufacturing plans, 3) a cost estimate covering all the required deliverables, i.e. the fabrication of the radial plates, of the double pancakes, of the WP, of the entire TFCS and the casing operation for the 9 TFC, including the design, procurement, installation and commissioning of the necessary tooling, 4) the design work for all the tooling necessary for a complete production line. The second contract will cover the remaining part of Phase II including a radial plate and a set of cover plates as deliverables of the first contract to be completed to a full-scale dummy double pancake and Phase III. The third contract will cover the series production, i.e. Phase IV. The last 2 calls for tender will be fully open and worldwide. V. ENGINEERING ACTIVITIES BY THE ITER ORGANIZATION While the DA are launching their call for tender and placing their first contracts in industry, the IO is proceeding with specific engineering activities aiming at completing the specification drawings and confirming the overall manufacturability of the TFC. The most relevant activities are described in the following sections. A. Optimization of the Cover Plate Welding The CICC is locked in the RP groove by means of a cover plate (CP). The CP has holes at 20 cm spacing to enable resin flowing and wetting the conductor insulation during the vacuum pressure impregnation (VPI) process. The CP are fixed to the RP by laser welding with a continuous seam of at least 1 mm depth in the outboard leg and at least 2 mm depth in the inboard leg region to satisfy both stress and stiffness requirements under the operation loads. Analyses were carried out to reduce the welding depth from 5 mm down to the values mentioned above in order to minimize the out-of-plane deflections of the DP induced by the welding operation. The structural design criteria applicable are summarized in Table II for the material of the RP and CP. For the inboard leg, which is more critical as it withstands the highest loads, material of class C2 is necessary whereas for the outboard leg, material of the lower class C4 is acceptable. The results of the stress TABLE III STRESS INTENSITY IN THE WELD Fig. 4. Stress distribution in the DP at the inboard leg for a 2 mm deep weld. analysis are summarized in Table III. A snapshot showing the stress distribution in the DP at the inboard leg for a 2 mm deep weld is shown in Fig. 4. For the analyses, we assumed that the gap between the RP groove and the conductor is not filled with resin, which is a conservative approach. As a next step, a code is being built to estimate the weldinduced out-of-plane distortions as a function of the welding sequence and of the main welding parameters. B. Qualification of Cyanate Ester Blends for the Insulation System The fabrication of the WP requires four different insulation and impregnation processes: 1) conductor or turn insulation, 2) DP insulation, 3) WP ground insulation, 4) terminal region insulation. The turn, DP and WP insulation will be impregnated through vacuum pressure impregnation (VPI) with either pure cyanate ester (CE) or a blend of CE with epoxy resin at a volume ratio of 60% CE and 40% diglycidyl ether of bisphenol F (DGEBF) to assure the long term integrity of the insulation system under the expected neutron fluence throughout the tokamak life time. The insulation layout consists of several layers of half overlapped glass-fiber/polyimide tapes of 0.175 mm thickness. Voltage simulations have shown that in case of failure, the maximum voltage between the coil terminals of 2 adjacent coils will be 14 kV if a grounding fault is combined with a fast discharge unit failing to open. In normal operating conditions, this maximum voltage is 7.3 kV. Typical experimental values of the 384 IEEE TRANSACTIONS ON APPLIED SUPERCONDUCTIVITY, VOL. 20, NO. 3, JUNE 2010 Fig. 5. Layout of the DP and WP insulation. dielectric strength of VPI glass/polyimide insulation are comprised between 80 and 90 kV/mm [2], which give comfortable margin with respect to the above mentioned values. The integrated neutron fluence seen by the TFC during the excannot be withpected 20 years of operation stood reliably by commercially available low viscosity epoxy resins [3]. CE is known to present radiation hardness significantly higher than the neutron fluence mentioned above. Moreover, it has suitable viscosity and pot life. For these reasons, R&D programs were implemented in Europe and Japan in the recent years to qualify ad’hoc blend of epoxy resin-CE as thermoset. The degradation of insulation under irradiation by different sources and doses is correlated with the CE content, but occurs at levels above the neutron fluence expected in ITER. The blending ratio of 40% CE and 60% epoxy has been chosen to satisfy stoechiometric laws to enhance the cross linking between the molecules of different nature during the curing process. The lack of experience in VPI with CE as thermoset was overcome in 2008 with the fabrication at ASG Superconductors of a full-scale WP mock-up under a Task Agreement with F4E. Seven steel plates of 1 m length and about 110 mm thickness (corresponding to that of a real DP) were wrapped with the insulation layout shown in Fig. 5, stacked together, wrapped according to the same layout as the WP ground insulation and successfully impregnated and cured in autoclave. Specimens prepared from this mock-up to check ultimate tensile strength, inter laminar shear stress and tension fatigue showed a perfect correlation with data relating to previous tests [3], [4]. Despite the radiation hardness of the glass-fiber/CE blend composites, the existing test data show a significant dispersion in the results due to the different methods used to prepare the samples and the different nature of the ionizing radiation utilized for all the tests performed around the world. CE blends of 2 suppliers which participated to the R&D program (Huntsman International LLC and Composite Technology Development Inc) are under qualification by the Institute of Atomic and Subatomic Physics, ATI, in Vienna Austria. Samples were prepared from composite sheets of 4 mm thickness fabricated by wrapping half-overlapped glass/polyimide tapes around an aluminum sheet coated with a mould release agent. The new resins developed by Huntsman and CTD in response to the specification provided by the IO, F4E and JAEA were used to impregnate the samples. The samples will be irradiated in a . TRIGA reactor in Vienna to a neutron fluence of Fig. 6. Layout of the TFC terminal region. Then, the degradation of their mechanical properties will be verified and compared to those of non irradiated samples. C. Design of the TFC Terminal Region The TFC terminal region, shown in Fig. 6, is quite complex as it integrates different types of sub-systems and interfaces in a very limited space: 6 DP to DP joints, 2 terminal joints between DP and feeder bus bar, insulation breaks, cryogenic instrumentation, helium inlet and outlet manifolds, current-limiting resistors and high voltage cables for the DP joints. These items assure liaison with the power supply, the cryoplant, the protection system and the control system of the tokamak. The instrumentation includes 3 absolute pressure cryogenic sensors, 11 redundant temperature sensors, 2 helium flow meters and 44 voltage taps. An electro-mechanical study of the terminal area predicts a relative vertical movement of the WP with respect to the TFCC of about 3 mm during operation. As the TFC terminal region is clamped to the lower part of the TFCC, a sliding surface made of Vespel is incorporated in between the impregnated DP to DP joints and their support cover to accommodate this relative movement. VI. CONCLUSION Significant progress was achieved between 2008 and the first half of 2009 for the TFC of the ITER Project. Two PA were signed by the IO with the European and Japanese DA and first contracts were signed by both DA with contractors for the qualification phase preceding the series production. In parallel, the IO is proceeding with specific engineering activities aiming at completing the design work and confirming the overall manufacturability of the TFC. REFERENCES [1] N. Holtkamp, “The status of the ITER design,” in 22nd AIEI Fusion Energy Conf., Geneva, Switzerland, Oct. 2008. [2] K. Humer et al., “Dielectric strength of irradiated fiber reinforced plastics,” Phys. C, vol. 354, pp. 143–147, 2001. [3] J. Knaster et al., “The electrical insulation of the TF coils of ITER,” Fusion Sci. Technol., vol. 56-2, pp. 666–672. [4] R. Prokopec et al., “Mechanical characterization of the ITER insulation mock-up after reactor irradiation,” in ICMC 2009 Proc., Tucson.