Home Search Collections Journals About Contact us My IOPscience

The ITER project construction status

This content has been downloaded from IOPscience. Please scroll down to see the full text.

2015 Nucl. Fusion 55 104023

(http://iopscience.iop.org/0029-5515/55/10/104023)

View the table of contents for this issue, or go to the journal homepage for more

Download details:

IP Address: 139.80.123.50
This content was downloaded on 03/10/2015 at 08:13

Please note that terms and conditions apply.

 International Atomic Energy Agency Nuclear Fusion

Nucl. Fusion 55 (2015) 104023 (20pp) doi:10.1088/0029-5515/55/10/104023

The ITER project construction status

O. Motojima
ITER Organization, Route de Vinon-sur-Verdon, CS 90 046, 13067 St Paul-lez-Durance, France

E-mail: [email protected]

Received 1 June 2015, revised 8 June 2015

Accepted for publication 29 June 2015
Published 1 October 2015

Abstract
The pace of the ITER project in St Paul-lez-Durance, France is accelerating rapidly into its
peak construction phase. With the completion of the B2 slab in August 2014, which will
support about 400 000 metric tons of the tokamak complex structures and components, the
construction is advancing on a daily basis. Magnet, vacuum vessel, cryostat, thermal shield,
first wall and divertor structures are under construction or in prototype phase in the ITER
member states of China, Europe, India, Japan, Korea, Russia, and the United States. Each
of these member states has its own domestic agency (DA) to manage their procurements
of components for ITER. Plant systems engineering is being transformed to fully integrate
the tokamak and its auxiliary systems in preparation for the assembly and operations
phase. CODAC, diagnostics, and the three main heating and current drive systems are also
progressing, including the construction of the neutral beam test facility building in Padua,
Italy. The conceptual design of the Chinese test blanket module system for ITER has been
completed and those of the EU are well under way. Significant progress has been made
addressing several outstanding physics issues including disruption load characterization,
prediction, avoidance, and mitigation, first wall and divertor shaping, edge pedestal and SOL
plasma stability, fuelling and plasma behaviour during confinement transients and W impurity
transport. Further development of the ITER Research Plan has included a definition of the
required plant configuration for 1st plasma and subsequent phases of ITER operation as
well as the major plasma commissioning activities and the needs of the accompanying R&D
program to ITER construction by the ITER parties.

Keywords: ITER, tokamak, fusion

(Some figures may appear in colour only in the online journal)

1. Introduction such as the case of the difficult CS cable issue which was
solved successfully in 2012, it is clear that the ITER Project
The ITER project is exploring many important horizons of has passed the point of no return. The in-kind manufacturing
Fusion Science and Technology; ITER is a necessary step on is making progress all over the world, and the first large com-
the way to a commercial fusion reactor. ITER will demon- ponent already arrived at the ITER site on 4 September 2014.
strate the availability and integration of science, technology, This is also highlighted by the ITER Organization (IO) and
and safety features for a fusion reactor. The self-sustained the European domestic agency (EU-DA) joint achievement of
D–T (fusion fuel) burning plasma in ITER will generate 500 the concrete pouring of the B2 slab, which was successfully
MW, which is ten times more power than it receives. The started on 10 July 2014, immediately after receiving the letter
ITER Project is an international megaproject, which is in the from the Nuclear Safety Authority (ASN) to release the hold-
process of creating a new collaborative culture and pioneering point, and it was completed on 27 August. Therefore, the only
standards for solving the world’s energy and environmental way is to go forward together. This fact should be understood
problems, and for contributing to world peace. The construc- and shared by all the stakeholders.
tion phase of the ITER Project has advanced considerably Presently, the seven domestic agencies (DAs) of the ITER
during the two years since the last Fusion Energy Conference Organization (IO) from China, EU, India, Japan, South
in San Diego. Since no show stopper has appeared up to now, Korea, Russia, and the United States have committed to

0029-5515/15/104023+20$33.00 1 © 2015 IAEA, Vienna Printed in the UK

Nucl. Fusion 55 (2015) 104023 O. Motojima

in-kind procurement of 90% of the total credit value and to steel elements that make up the superstructure of the Assembly
more than 70% of the number of Procurement Arrangements. Hall began to arrive on site at the end of September 2014.
Construction is accelerating and the appearance of the site is Preparatory work for other construction contracts has also
changing on a daily basis. As the ITER project advances in the commenced on the site including those for the Site Services
construction phase, there is increased coordination of plant Building and general infrastructure. Construction of the ITER
systems engineering, tokamak construction, and building and Headquarters Extension was completed and handed over to
site infrastructure. The recent engineering progress and con- the IO on 30 September 2014. Work has commenced on the
tribution to world fusion research are reported in this paper. access road to the planned 10 000 m2 storage warehouse, and
the designs and permits have been submitted to the relevant
local authorities. It is expected that this facility will be opera-
2. Safety, buildings and site preparation tional by the end of 2015.
The cryostat workshop, currently the tallest building on
On 9 November 2012, ITER was awarded with the Decree the site, is complete. Facilities for the full site workforce have
of Authorization of creation of a nuclear facility, becoming also been completed and work on two component storage
nuclear facility number 174 in France: INB-174. Following zones is underway.
the French regulation, the Technical Prescription accom- Two test convoys for the transport of bulky and very heavy
panying the ITER decree was issued one year later through components have made the journey from the port at Fos-
the 2-ASN Decision 2013-DC-0379 dated 12 November sur-Mer to the ITER site over three nights. These tested the
2013 establishing the prescriptions applicable to the ITER infrastructure modifications and improvements carried out by
Organization for the design and construction of the licensed the French Authorities. In the process, some minor improve-
nuclear facility INB No 174 called ITER. In 2013, The ITER ments were carried out and this massive exercise was a great
installation passed successfully the stress test considering the success, paving the way for the delivery of some of the largest
accident at the Fukushima nuclear power plant in Japan. The ITER components.
construction of the slab of the tokamak complex was autho-
rized by the ASN in July 2014. Several nuclear inspections
from the Nuclear Safety Authorities took place regarding the 3. Magnets
construction of the buildings, the fabrication of the vacuum
3.1. Superconducting strands and conductors
vessel, and the prototype of the fast discharge units. These
inspections have underlined that the nuclear operator, the By August 2014, 535 tons of Nb3Sn strand, representing 90%
ITER organization, must continue its effort to control the of the ITER requirement, have been produced [1]. Also, 242.5
activities of construction and fabrication. tons of NbTi strand have been manufactured and 88% of the
Construction on the site is carried out under the responsi- ITER requirement was achieved. The strands were twisted
bility of the European Union. For the main Tokamak Complex into cables and the conductors were produced by inserting
buildings (Diagnostic, Tokamak and Tritium buildings), the the cable into the jacket made of austenitic stainless steel. All
seismic isolation pit, which includes the 1.5 m thick lower conductors were fabricated in unit length. The production per-
basemat and the excavation support structure walls, and the formance of the unit length (number of unit lengths produced/
493 elastomeric anti-seismic bearings were completed in early number of unit lengths required) is 81% for TF conductors
2013. Concerning the basemat of the Tokamak Complex ‘B2 (Nb3Sn), 22% for PF conductors (NbTi), 14% for CS conduc-
slab’, which sits on the anti-seismic bearings, the year 2013 tors (Nb3Sn) and 56% for correction coil conductors (NbTi).
was mainly dedicated to the consolidation of the construction The production of the strands and conductors continues.
design. The first concrete section, out of 15 sections in total, The qualification tests of the CS conductor that was manu-
was poured in December 2013 and the last concrete pour, for factured in Japan were conducted at the SULTAN facility in
the supporting structure of the Tokamak machine, was made Switzerland and all samples satisfied the ITER requirements.
at the end of August 2014, completing then 15 000 m3 of
reinforced concrete (figure 1). The concrete used for the B2
3.2. TF coils
slab was qualified with specific tests to verify its confinement
requirement and with a dedicated on-site full scale mock-up The Nb3Sn conductors are wound in a D shape and set in the
of a portion of the slab, representative of the areas with very radial plate after the heat treatment that generates the Nb3Sn
dense reinforcements and embedded plates. This mock-up phase. The TF coil winding tooling facilities and the electric fur-
was also used for the qualification of the work procedure for naces have been constructed in the EU and Japan. The winding
concrete implementation and vibration activities. machine can control the bending of the TF conductor with high
All of the critical underground trenches, dedicated to house precision and the furnace can keep the same temperature in a
the electrical and I&C cables and the cooling water system wide space (rough size of the inside is 12 m  ×  16 m  ×  3 m,
pipes, were completed in July 2014, thereby allowing the for the EU furnace). The heat treatment condition depends on the
installation of the construction cranes around the Tokamak conductor but the heat treatment temperature is around 600 °C
complex building. Work has now started to prepare for the and the holding time at the flat top is over 100 h. One full size
construction of the remaining levels of the Tokamak Complex dummy double pancake winding was already heat treated
and Assembly Hall. New cranes are being installed and the in Japan (figure 2). The full size radial plates for the dummy

2
Nucl. Fusion 55 (2015) 104023 O. Motojima

Figure 1. Photo of the last concrete pour of the B2 slab on 27 August 2014.

Figure 2. Heat treatment facility in Japan. Heat treatment of the first double pancake for JA-DA series production of the TF coil was
completed.

double pancake were welded and machined to form the D double pancakes by the end of 2014 and to further assemble the
shape, and the radial plates for the first real TF coil in Japan are winding pack in the course of 2015. In addition, the winding
being assembled. (Each TF coil has seven double pancakes.) activity for the second TF coil in the production sequence has
The EU has already qualified the manufacturing process of the started. Regarding the TF coil structure, full size mock-ups and
double pancake which includes the winding of the conductor, the first series segments of the TF coil case have been machined
the heat treatment, the fabrication of helium inlets and termi- and welded in Japan and Korea to assess and confirm the con-
nals, the conductor insulation, and the cover plate welding trol of welding deformation. The qualification investigations of
(figures 3 and 4). The first vacuum pressure impregnation of the welding processes and non-destructive inspection methods
the dummy pancake with radiation hard cyanate ester–epoxy have been implemented in Japan. Pre-compression ring mate-
blend resin has been completed. By taking this process, the rials were specified and manufactured, and the mechanical tests
space between the conductor and the groove wall in the radial were conducted in the EU. A mock-up of the gravity support
plate will be filled with resin, which becomes the final turn shown in figure 5 was constructed in China and the mechanical
insulation. In parallel with the above activities, the fabrication test at room temperature was carried out.
of the first winding pack for the real TF coil has started in the
EU. More than ten conductor windings were manufactured 3.3. PF coils
and all helium inlets and joints were fabricated successfully.
The heat treatments of these double pancakes are in progress The PF coil system consists of 6 coils, PF-1 to PF-6. PF-2 to
and are expected to achieve the first set of seven heat treated PF-5 will be constructed in the PF coil building on the ITER

3
Nucl. Fusion 55 (2015) 104023 O. Motojima

Figure 3. EU TF coils. Double pancake prototype being loaded inside the impregnation facility in La Spezia, Italy by the supplier.
(Consortium ASG-Elytt-Iberdrola).

Figure 4. Turn over tooling for radial plate. (EU, SIMIC).

Figure 5. Gravity support mock-up (CN) together with a 3D drawing showing the gravity support at the base of each TF magnet.

site by the EU, because these coils are too large to transport. fabrication have been prepared and two have already started
The upper and lower coils, PF-1 and PF-6, have small enough including the design optimization for fabrication and the design
diameter to be transported and will be manufactured in Russia of the tooling. The call for tender of the site and infrastructure
and China, respectively. Six independent contracts for PF coil contract was launched in June 2014 and placed in November

4
Nucl. Fusion 55 (2015) 104023 O. Motojima

2014. Component qualification started in June 2014. A 3   ×   3 magnet operation. A magnet coil feeder consists of a dry box
conductor mock-up with dummy conductors and the turn insu- (DB), a combined coil terminal box/S-bend box (CTB/SBB),
lation samples were fabricated. The mechanical and electrical a cryostat feedthrough (CFT) with a vacuum barrier (VB)
tests of these samples were carried out at room temperature and an in-cryostat feeder (ICF), including a terminal transi-
and at 77 K. To accelerate the PF-6 schedule, a collaboration tion box for structural support to the coil terminal joint con-
agreement for PF-6 was settled between F4E and ASIPP in nections. A pair of high temperature superconducting current
China in October 2013 with CN-DA support. The preparation leads (HTSCLs) is installed in the CTB/SBB to transfer the
activities for the PF-6 manufacturing building and the tooling electric current from the room temperature bus bar to the cryo-
facilities are well under way at ASIPP. genic magnet. Three pairs of prototype current leads are being
constructed for full current cryogenic qualifications, which
was carried out in mid-2015. Qualifications of the full size
3.4. CS coils
prototype vacuum box and the thermal shield for the CTB/
The final design review (FDR) of the CS coils was held in SBB have been completed. These components will be used as
October 2013. The CS coil system consists of 6 CS modules, the test bed for qualification of cryogenic components such as
each with 40 pancakes, which will be provided by the United the HTSCLs and the superconducting joints. Manufacturing
States. The CS module manufacturing facilities include two of full size prototype CFT and ICFs are on-going with accep-
winding lines, a stacking station, a heat treatment oven, an tance tests carried out in late 2014. Qualifications of a few key
insulation wrapping station, an impregnation station and a cold components like the seismic bellows for the CFT connection
test facility. The pancakes will be stacked and all joints and the to the cryostat, the cold mass support, and the vacuum barrier
necessary works for each module will be performed before the in the CFT are in progress.
heat treatment for Nb3Sn phase generation. A large electric fur-
nace will be installed for the heat treatment of each module, and
4. Cryostat, vacuum vessel, thermal shields, and
a vacuum pressure impregnation facility will be used to make
in-vessel coils
the resin insulation. The CS coil system includes the struc-
tural components, the lower key blocks, the tie plates and the 4.1. Cryostat
upper key blocks. The contract of the lower key blocks with a
US-company was placed in June 2014. The call for tender of the The cryostat is a multifunctional system that provides vacuum
tie plates will be completed this year and that for the upper key insulation for the superconducting magnets operating at 4.5
blocks will be carried out in 2015. Cold tests of the CS modules K and for the thermal shields at 80 K [2]. It also has a struc-
are required to qualify the CS modules. Cryogenic facilities for tural function and is designed to withstand gravity and seismic
the cold tests at 4 K will be installed at the coil manufacturer. events through its lower support structure. It is a very large
The FDR of the CS module assembly tooling was completed in vacuum containment vessel with a volume of ~16 000 m3
September 2014. The assembly tooling includes the assembly evacuated to a base pressure of 10−4 Pa.
platform, the rotating fixture and the lifting fixture. A single procurement arrangement was signed with the
IN DA to fabricate and assemble the cryostat. Manufacturing
activities began in 2013, including the manufacturing designs
3.5. Correction coils
of the four main sections and fabrication of a full-scale mock-
There are three correction coil (CC) positions including the up of one base section sector (figure 6). A temporary work-
top CCs (TCC), side CCs (SCC) and bottom CCs (BCC). shop has been built at the ITER site to assemble the parts that
Since the TCC and the BCC are symmetric, there are two will be shipped from India. The detailed designs of the lower
types of coils, B/TCC and SCC. The winding facility for the supports, rectangular bellows, and torus cryopump housing,
CCs has been constructed and commissioned in China. The are being finalized.
prototype of the double pancakes for B/TCC and SCC were
wound and the winding process was qualified. The vacuum 4.2. Vacuum vessel
pressure impregnation of a dummy B/TCC was carried out
and the functional tests including a 10 000 cycled bending test The manufacture of vacuum vessel (VV) components has
was completed in 2014. The prototype coil cases for the B/ started by four DAs (figure 7) [2]. Since the VV provides the
TCC and the SCC were machined and welded. To perform primary confinement boundary for the ITER machine, the
the high level welding on 3D weld lines of the coil case, auto- design, manufacture and assembly of the VV has to comply
matic laser welding was used for the closure welds. All tooling with the requirements of the French nuclear pressure vessel
has been commissioned and the qualification of the manufac- regulations (ESPN) and the RCC-MR code. In particular, the
turing process is ongoing. The first forged 316LN plate for the manufacturing design and procedures have to be reviewed
coil case was produced in China in 2014. and approved by the Agreed Notified Body (ANB). The engi-
neering design was approved by the ANB in 2012 after a
series of design changes and now the detailed manufacturing
3.6. Feeders
design of the first VV sector is complete, while for the other
The magnet system is connected with 26 coil feeders, 3 struc- eight sectors and ports, it is still in progress. Seven sectors will
ture cooling feeders, and 2 instrumentation feeders to support be fabricated by Europe, with Korea manufacturing the two
5
Nucl. Fusion 55 (2015) 104023 O. Motojima

Figure 6. Manufacture of the cryostat base pedestal ring and sandwich structure (courtesy of IN DA). A 3D drawing of the base pedestal of
the cryostat is also shown for reference.

procedures, and fabrication of mock-ups are on-going, while

welding and non-destructive inspection procedures are being
qualified. The manufacturing design of the irregular sectors
is on-going and will be completed in 2015. The upper port
manufacture began in May 2014 after signature of the fab-
rication contract with the German company MDT. This con-
tract is managed by the Efremov Institute. Progress consists in
the finalization of the manufacturing design, procurement of
batches of forgings and plate material and qualification of the
forming and welding procedures, also with the use of relevant
mock-ups. Full scale fabrication of the IWS elements for sec-
tors#5 and #6 has started by the Indian company Avasarala,
while the manufacturing design of the blocks for other sectors
is being finalized. The procurement of plates, bolts, fasteners,
Figure 7. 3D View of the vacuum vessel and its ports. The colours and washers for the IWS blocks is almost complete.
correspond to the European contribution (green), the Korean
contribution (blue), and the Russian contribution (red).
4.3. Thermal shields
other sectors and the equatorial and lower ports. Russia will
supply the upper ports and India the in-wall shielding (IWS), The vacuum vessel and cryostat thermal shields (TS) consist
consisting of neutron shielding and ferromagnetic inserts of silver plated panels cooled to 80 K by He gas (figure 9).
located inside the VV double wall. Following the manufacture Korea is responsible for the procurement, which was awarded
of full-scale mock-ups of two segments and ports (figure 8), to the company SFA in February 2013, following detailed
cutting and forming of the plates of the first VV sector#6 engineering design and R&D of the shields since 2007.
began at Hyundai Heavy Industry (HHI) in Korea in February Several design changes were required during the manufac-
2012. Welding of the sector#6 segments and fabrication of turing design stage due to interfaces with other tokamak
elements of the lower and equatorial ports are in progress. The components and updated assembly procedures. Since fitting
manufacturing design of the second Korean sector#1 is com- the thermal shields requires strict geometrical tolerances and
plete and cutting/forming activities have started. Procurement complex assembly procedures, Korea has undertaken sub-
of the base plates and forgings is complete. The European stantial mock-up manufacturing tests to verify the assembly
consortium AMW (Ansaldo Nucleare, Mangiarotti and procedures and the necessary metrology. Two mock-ups for
Walter Tosto) has started manufacture of its first sector#5 in different kinds of sector field joints were made and their
February 2014 with the first cutting and forming. Procurement assembly feasibility checked. The complex shape of cooling
of the base plates and forgings, finalization of manufacturing tube routing for the VVTS lower port is possible through

6
Nucl. Fusion 55 (2015) 104023 O. Motojima

Figure 8. Full-scale mock-up of PS2 vacuum vessel segment by Hyundai Heavy Industries.

Due to the high radiation and temperature environment, chal-

lenging thermo-mechanical analyses and design solutions
are required. Mineral insulated water-cooled copper conduc-
tors will be used to resist the high radiation level. Substantial
R&D work has been managed by ITER in close collabora-
tion with the Chinese Academy of Sciences (ASIPP) and
the Princeton Plasma Physics Laboratory (PPPL). Prototype
coils were made by ASIPP in April 2014 (figure 11). The
aim was to develop suitable manufacturing procedures and
techniques necessary to fabricate the ELM and VS coils,
and to qualify electrical and mechanical test procedures to
meet the acceptance criteria. An extensive set of analyses to
evaluate the effects of the high thermal and electromagnetic
loads on the IVCs has been carried out. The baseline design
for the ELM coils relies on brazing of the support brackets
Figure 9. 3D view of vacuum vessel and cryostat thermal shield. to the conduit, both made of Inconel 625. Since this process
has not proven successful, alternative designs with square
a novel bending method. The design of the manifolds for stainless steel conductors welded together are being inves-
the coolant supply to the TS has also been performed and tigated. The procurement of the IVCs and their conductors
its structural integrity verified. All of the manufacturing will be done via direct contracts from the ITER Organization
methods, except silver coating, were tested with the fabrica- and preparation has already started. The calls for tender are
tion of a full-scale VVTS 10 degree section (figure 10) before expected to be launched in early 2016 after completion of
starting TS manufacturing, which began at the end of 2014. the design and R&D activities and a final design review. It is
In parallel, all of the VV and Cryostat TS plate material has expected that the supply contracts will be signed by the end
been procured. SFA has also set up a brand new dedicated of 2016.
facility for the manufacture of the ITER TS.

4.4. In-vessel coils 5. Blanket shield and divertor

Following the ITER Council decision in November 2013, two The Blanket system and the divertor cover the majority of the
types of in-vessel coils (IVCs), namely ELM Coils to miti- ITER plasma facing surface with a total area of about 830
gate edge localized modes and VS Coils to provide vertical m2 [4]. Being at the frontline with the plasma, their design
stabilization of the plasma, have been incorporated in the has to address a number of challenging requirements, which
ITER baseline design (figure 11) [3]. The IVCs are mounted include: (1) the surface heat flux, (2) the electromagnetic
to the VV inner wall, in close proximity to the plasma, just loads, (3) extremely tight tolerances towards the plasma and
behind the Blanket Shield Modules (BSM). Fitting the coil the interfacing systems, (4) plasma–wall interactions (plasma
systems in between the BSM and the VV leads to difficult disruptions, vertical displacement events, ELMs, particle flux,
integration with diagnostics and cooling water manifolds. etc), (5) compatibility with ultra-high vacuum, (6) capability

7
Nucl. Fusion 55 (2015) 104023 O. Motojima

Figure 10. Equatorial section of a 10 degree vacuum vessel thermal shield mock-up.

Figure 11. 3D layout of ELM and vertical stability coils and a photo of a prototype ELM coil assembly at ASIPP.

to remove up to 850 MW of power, (7) accommodation of the shutdown, and (4) to provide passage for and accommodate
various and sometimes conflicting requirements (shielding, interface requirements of the plasma diagnostics. The Blanket
assembly, remote handling, integration with other in-vessel design and analysis effort is conducted within the Blanket
components), (8) respect of the highest quality standards Integrated Product Team, which includes resources from the
and (9) the required high heat flux technologies and complex IO and the DAs.
welded structures [4]. The Blanket System consists of Blanket Modules (BM)
comprising two major components: a replaceable plasma-
facing first wall (FW) panel and a semi-permanent shield
5.1. Blanket system
block (SB). There are 440 BMs, arranged in 18 poloidal and 18
The main functions of the blanket system are: (1) to exhaust (inboard)/36 (outboard) toroidal rows. The BMs are mechan-
the majority of the plasma power, (2) to contribute to neutron ically mounted to the vacuum vessel (VV) via a system of
shielding of the superconducting coils, (3) to provide limiting connections, which includes the flexible attachments (to react
surfaces that define the plasma boundary during startup and to the radial loads), the pads and keys (to react to poloidal

8
Nucl. Fusion 55 (2015) 104023 O. Motojima

Figure 12. The ITER blanket system and procurement sharing.

Figure 13. Blanket FW mock-up manufactured in China (left) and Russia (right) (Courtesy of CN DA and the RF DA).

and toroidal loads and moments), and the electrical straps

(to route any induced currents to the VV via a well-defined
path). Figure 12 shows an example BM with the procurement
sharing among DAs.
The blanket design effort culminated in April 2013 with the
FDR, which was held in July 2013. Since then, two procure-
ment arrangements (PAs) have been signed with the Chinese
and Korean DAs for the supply of the SBs, and one with the
Russian DA for the FW panels. A fourth PA was signed in
December 2014 with the Russian DA for the manufacturing
of the Blanket connections. The remaining PAs will be signed
in 2015 and 2016.
The FW panels consist of a plasma-facing material made of
beryllium, which is joined onto a copper alloy (CuCrZr) heat
sink (figure 13), which is in turn actively cooled by pressur-
ized water. There are two principal FW panel designs, accom-
modating a heat flux of 2 MW m−2 (Normal Heat Flux panels)
or of 4.7 MW m−2 (enhanced heat flux panels). Mock-ups and
semi-prototypes have been manufactured by all the procuring
DAs and have been subject to high heat flux tests in excess of Figure 14. Blanket shield block prototype manufactured in Korea.
Courtesy of KO-DA.
the design requirements. The FW panels are shaped to mini-
mize the plasma heat load on edges caused by ports, diag- the signature of the procurement arrangement, full-scale pro-
nostic openings, assembly gaps between the FW panels and totypes were manufactured to validate the technology and to
FW panel radial misalignments (see also section 10). demonstrate the achievement of the required tolerances.
The SBs are made of forged stainless steel (316 L(N)-IG) in
which a number of cooling channels are obtained by drilling
5.2. Divertor
(figure 14). Their main function is to provide most of the
shielding, to support the FW panels and to provide the connec- The aim of the Divertor is to minimize the impurity content
tions for the supply of cooling water to the FW panel. Before in the plasma and, in a burning plasma, to exhaust the He ash

9
Nucl. Fusion 55 (2015) 104023 O. Motojima

Figure 15. Cassette assembly of the divertor and its procurement sharing.

produced by fusion reactions by intercepting plasma flowing monoblocks due to assembly tolerances. A simple toroidal
on open magnetic field lines, neutralizing the plasma at the chamfering has been initially proposed, which minimizes the
divertor targets and allowing neutral particles to be pumped loss of wetted surface area for steady state power handling.
away by the vacuum system. The price to pay is that the However, this is a difficult and uncertain area and a final deci-
divertor must withstand extremely high steady state heat loads sion has still not been taken, pending further studies within the
(up to 20 MW m−2), by far the highest steady loads of any of physics community (section 10).
the ITER plasma-facing components (PFCs).
The divertor consists of 54 cassette assemblies making
a toroidal ring at the bottom of the plasma chamber. Each 6. Plant systems
cassette assembly is formed by a cassette body, which sus-
tains three main PFCs, the inner and outer target, and the The first plant system components were delivered to the ITER
dome. Figure 15 shows the main layout of the divertor and site in September 2014, twelve high voltage surge arrestors for
the procurement sharing among the DAs. In September the steady state electrical network, which arrived on schedule
2011, the IO proposed to begin ITER operation with full- from the US DA from New York. Prototype ac/dc convertors
tungsten PFCs to minimize costs and gain early operational have been manufactured in both China and Korea and electro-
experience with tungsten during the non-active phases of mechanical components have been demonstrated in Russia.
the machine. An engineering task force was established to All the electrical distribution procurement arrangements have
develop the full-tungsten divertor option within the two- been signed with China, the EU and the United States. In their
year time frame set by the ITER Council following the IO turn, the Parties concerned have signed more than 50% of the
proposal. In parallel, a major effort was launched by the IO contracts with suppliers.
throughout the Physics Community (in particular through A system is a set of interacting or interdependent compo-
the International Tokamak Physics Activity (ITPA), to nents forming an integrated whole. A system has intercon-
study a number of key outstanding issues relating to nectivity, performs processes (to fulfil functions), and exhibits
physics operation and plasma–wall interactions [5] (see a behaviour (reaction to events). Hence systems (instead of
also section 10). products or components) must consider all interactions/inte-
A full-tungsten Divertor Final Design Review was held gration aspects including functions and external constraints.
in June 2013 and closed in October 2013. As a consequence Considering ITER as an integrated system means to look
of favourable progress in the demonstration of tungsten high first at the capacity of ITER to react simultaneously to satisfy
heat flux handling technology in both the DAs responsible for operational performance as well as safety performance. The
supply of the outer and inner divertor PFCs, and encouraging ‘external constraints’ fall into two mains categories: (1) inte-
results from the physics program, the ITER Council, at their gration (EMC, space management, functional performance,
November 2013 meeting, formally mandated the IO to begin safety analyses), and (2) requirements coming from the other
ITER operation with a full tungsten divertor. One main design project phases (procurement or manufacturing, assembly,
change from the previous carbon/tungsten baseline was the integrated commissioning, operation, radiological shielding).
introduction of the shaping of the baffle region (upper part) Safety is a constraint and a main function.
of the outer vertical target to protect leading edges during The IO emphasizes Systems Engineering activities to
downward vertical displacement events. A second key issue assure proper integration and develop an integrated functional
is the question of individual monoblock surface shaping in analysis. Processes have been established and implemented
the highest heat flux areas. Such shaping is required to pro- to determine the functional maturity of systems, not just indi-
tect against leading edges arising between neighbouring vidual components, and to characterize interfaces to enhance

10
Nucl. Fusion 55 (2015) 104023 O. Motojima

systems engineering. Key deliverables from the systems engi- have been made in user support and training to further assist
neering improvement activities include: the plant system suppliers. High performance networks cov-
ering time synchronization, deterministic real-time commu-
• Production of design plans for each PBS (Plant nication, and massive scientific data streaming have been
Breakdown Structure), as well as for integration among designed and implemented. The detailed identification and
systems and safety analysis specification of machine protection and nuclear safety func-
• Completion of plant system intelligent P&IDs (Piping tions are advancing. CODAC is also in production of on-site
and Instrumentation Diagrams)
monitoring of the construction site power and the building
• Implementation of a phased approach to conduct transient
management system in the poloidal field coil fabrication
design analysis (initially considering each system and its
building as well as at various test facilities around the world.
clients, and thereafter all systems progressively coupled
with their clients)
• Implementation of a functional analysis seminar and 8.2. Diagnostics
assistance in developing a distributed knowledge
More than 85% of the diagnostic system conceptual design
• Formal issuance of building General Arrangement
reviews (CDRs) have been completed and FDRs of the upper
Drawings as 2D drawings
and equatorial port plugs have been completed. Several key
• Improved provision of central CAD services and s­ pecialist
in-vessel diagnostics and especially diagnostic engineering
nuclear calculations support
infrastructure are needed early in the assembly schedule.
• Systematic revision for system maturity of each system
These include many magnetic diagnostics and the wiring loom
against transversal requirements (i.e. shielding, safety
infrastructure. The requirements for these are stringent both
performance, functional performance, in service
in terms of the measurement needs and the engineering con-
inspection and maintenance, confinement function, fire
straints. These items are currently being accelerated and given
protection)
the highest priority to ensure no impact on the ITER construc-
• Development of commissioning sequence and commis-
tion schedule. A strong push is also underway to ensure that
sioning procedures for single system and suppliers and
the key engineering infrastructure for the port-based diagnos-
single system with supplier and clients
tics is fully prepared as soon as possible to allow the smooth
• Development of transversal commissioning procedures
development and integration of these diagnostics. A brief look
for integration among systems.
at some of the different areas is covered below.
Today the IO is facing the urgent issue of integrating,
across the IO and DAs, the function of Systems Engineering 8.2.1. General considerations. A number of diagnostics
in the design and fabrication process. in ITER will measure light in UV–visible–IR range provid-
ing time-and spatially resolved information about plasma
parameters and plasma surface interactions. The first mirrors
7. Preparations for assembly and installation will be exposed to plasma and fluxes of impurities creating
deposits on first mirrors (FM) that will lead to their degrada-
Following completion of the buildings, the IO will have respon- tion. Significant efforts were spent on R&D to develop a FM
sibility for the assembly of the Tokamak and the installation cleaning system. One of the most promising methods for FM
of associated plant systems. Detailed engineering studies have cleaning is based on RF discharge. Currently this technique
been undertaken to ensure the constructability of the systems is being implemented in the design of all concerned optical
and structures, and the IO is now undertaking detailed plan- systems. In addition, the expected stray light level for various
ning and preparation for the works execution phase, including plasma conditions is being assessed together with benchmark-
definition of its contract strategy. Qualification of welding ing from existing tokamaks and development of techniques
processes, design of tools and detailed planning is underway. to mitigate reflections. All optical diagnostics will use win-
dows to transmit light through vacuum and the safety bound-
8. CODAC, diagnostics, heating and current drive ary that makes them Protection Important Components (PIC).
systems A standard design for windows was developed at the IO tak-
ing into account experience at the leading existing machine
8.1. CODAC (JET). Almost all diagnostics will use PICs (including not
only windows but gas, electrical and vacuum feedthroughs,
The Central Interlock and Central Safety System preliminary torus vacuum extension tubes, etc). A careful assessment of
designs have been approved and part 1 of the CODAC FDR various risks and loads at each design step as well as develop-
has been completed including the design for the network ment of mitigation strategies for them is very important, and
infrastructure and plant control. The main challenge remains forms a part of the safety culture that is being implemented in
the integration of hundreds of local plant control systems sup- collaboration with the DAs.
plied by the DAs. Key components to achieve integration are a
set of hardware, called the I&C integration kit, and a software 8.2.2. Neutronics. For diagnostics, neutronics is a big design
framework, called the CODAC core system, currently being driver. Many aspects of the detectors and their infrastruc-
shipped by the IO to the plant system suppliers. Investments ture have to withstand highly energetic neutrons. The need

11
Nucl. Fusion 55 (2015) 104023 O. Motojima

Figure 16. Integrating many diagnostics into ITER is achieved through port plugs. The picture shows an example of the integration of port
plug 11. The structures firstly collect the data close to the tokamak and then couple the signals out of the vacuum chamber where they are
coupled to detectors or relayed further using waveguides or fibre optics.

Figure 17. Photo of the PRIMA facility under construction in Padua, Italy.

to provide shielding to reduce the doses to workers in areas and prototyping for integrated solutions for cabling, con-
where maintenance is foreseen has guided the evolution of the nectors and feedthroughs is also underway. Inner vessel
diagnostic port plugs, developing single and double labyrinth sensors (coils and loops) and their infrastructure are all in
configurations. Studies for the material selection for the neu- the detailed design stage. In particular, the saddle loops are
tron shielding of the Port Plugs showed boron carbide as the approaching final design and pre-qualification of manu-
most promising candidate. For the diagnostics bioshield, spe- facturers is underway. Pickup coils have key technology
cial concrete compositions will make it possible to meet the elements in R&D. In particular, the structural integrity
required doses in the port cells. For the VV mounted diagnos- of the chosen low-temperature co-fired ceramic (LTCC)
tics with cutouts in the blankets, which are the primary shield- composite pickup coil technology has been initially veri-
ing components, studies are being carried out to minimize fied through dedicated irradiation experiments. Coils
their impact on the heating of the VV and ­superconducting based on reactor-proven mineral-insulated cable (MIC) are
coils. used wherever possible; R&D to select the manufacturing
method of compact solid (brazed) coils is underway. The
8.2.3. Magnetics and infrastructure. The ­continuous Rogowski remote handling compatibility of in-vessel pickup coils has
coils, embedded in the toroidal field coils (TFCs), have been confirmed. High frequency coils (and all other in-ves-
been through final design and are in early manufacturing sel components) must withstand stray electron cyclotron
and final qualification phases. Outer vessel sensors (coils, heating (ECH) power reaching them. These scenarios have
loops, fibres and steady-state sensors) are in the final design now been calculated, and the design of sensors adapted to
stage. All early installation items have been prototyped and meet them, for example, by using metal film on insulator
tested as part of the design activities. The detailed design ECH shields.

12
Nucl. Fusion 55 (2015) 104023 O. Motojima

8.2.4. Diagnostic port plugs. The diagnostic generic equato- in the manufacturing process. They have also achieved reli-
rial port plug (GEPP) and diagnostic generic upper port plug able operation of 97% for 40 pulses of 1.0 MW and 1000 s.
(GUPP) designs are meant to be common to all (or most) Such performance would result in  >18 MW delivered to the
equatorial and upper port-based diagnostic systems. The port plasma after a 3600 s plasma pulse, assuming similar reli-
plug structure must contribute to the nuclear shielding, or ability from the 24 RF sources, which is significant progress.
plugging, of the port and further contain circulated water to Considerable progress has also been made on the USDA
allow cooling during operation and heating during bake-out. procured EC transmission line (TL) component designs.
Figure 16 shows the general view of the assembled equato- The design has been revised to minimize the amount of
rial port plug which consists of the GEPP structure, diagnostic higher order modes (HOM) generated in the waveguide and
shield modules (DSMs) and diagnostic first walls (DFWs). prototyping has been carried out. Presently manufactured
The generic port plug structures are designed to survive samples have increased straightness with a radius of curva-
throughout the lifetime of ITER for 20 years, 30 000 dis- ture exceeding the desired 1.2 km. They have developed a
charges, and 3000 disruptions. The system must provide both new miter bend design and coupling design and are devel-
adequate cooling to deal with nuclear heating and adequate oping a Monte Carlo code to highlight the major sources
bake-out heating to raise the entire port plug structure to a of HOMs and to concentrate on component development
minimum temperature of 200 °C in 48 h. A significant feature to minimize those losses. The Ion Cyclotron system is also
of the generic port plug structure design is the use of gun- advancing in the PDR phase. The IO is working closely
drilled coolant channels to ease the inspection process and with F4E and a European consortium CYCLE to develop
minimize the risk of stress corrosion cracking phenomena the ITER antenna. Prototypes of RF power amplifier tubes
during the 20 year lifetime. that are being procured by the INDA are currently in the
The generic port plug structure Final Design Reviews factory acceptance tests. The design of the IC TL, a US-DA
were passed in late 2013. Currently, ITER is preparing for the procurement components (and their experimental validation
common manufacturing of the equatorial and upper port plug under 6 MW hour long pulses is also advancing. Design
structures. review of early-procurement TL items to be used for the
commissioning of the RF sources is planned for late 2015.
8.3. Heating and current drive
9. Test blanket module systems
The heating systems of ITER include the neutral beams, elec-
tron cyclotron, and ion cyclotron systems. The lower hybrid ITER provides a unique opportunity to test tritium breeding
system, not reported on here, is also being designed as a pos- technologies needed for the development of fusion as an
sible future upgrade. Work on the NB systems is progressing energy source. So, one of the main engineering perfor-
well with the focus on the Neutral Beam Test Facility, mance goals of ITER is to test and validate design concepts
called PRIMA (Padua Research on Injectors with Megavolt of tritium breeding blankets relevant to a power producing
Acceleration) under construction in Padua, Italy (figure 17). reactor. Various mock-ups of breeding blanket concepts, cur-
PRIMA will consist of two subsidiary test facilities at rently developed by the ITER members (IM), will be tested
Consorzio RFX in Padua: the test bed for Megavolt ITER in the framework of the ITER Test Blanket Modules (TBM)
Injector Concept and Advancement (MITICA), which dif- Program [6].
fers from the actual injector only in the diagnostic capa- The TBM program in ITER foresees the operation of six
bility and the connection to the ITER torus by the ‘front different Test Blanket Systems (TBSs) utilizing three of the
end components’, and an ion source test bed ‘Source for ITER equatorial ports. Each TBS includes a TBM, directly
Production of Ions of Deuterium Extracted from an RF facing the plasma, involving either flowing Pb–16Li eutectic
plasma’ (SPIDER), which has to be capable of testing a full alloy and/or ceramic pebbles of ternary lithium ceramics as
size ITER ion source at full ITER parameters. The facilities tritium breeding material. The coolants used are high-tem-
of the PRIMA site are being installed from October 2014 and perature, high-pressure helium, water, and/or Pb–16Li. Each
the first components of the SPIDER facility arrived at the TBM has several of its own associated ancillary systems
end of 2014. All of the build-to-print MITICA components (cooling systems, tritium-extraction system, measurement
have reached FDR stage and all tenders will be launched systems and others).
by the end of 2015. Progress on the other components of The six TBSs to be tested in the initial phases of ITER
the heating neutral beams is also well advanced. All of the operation are the following:
build-to-print components have reached final design stage
and the DD components have completed preliminary design • In port #2: the helium-cooled ceramic breeder TBS
review (PDR). The European test facility ELISE, located (HCCB-TBS), and the lithium–lead ceramic breeder TBS
at IPP-Garching, Germany, has produced very successful (LLCB–TBS), proposed respectively by China and India.
results after just over 1.5 years of operation operating in • In port #16: the helium-cooled lithium–lead TBS
hydrogen very close to the ITER requirements. The electron (HCLL-TBS), and helium-cooled pebble bed TBS
cyclotron (EC) system is also progressing very well. The (HCPB-TBS), both proposed by the EU.
RFDA has reported 1.0 MW and 1000 s operation from the • In port #18: the water-cooled ceramic breeder TBS
3rd prototype gyrotrons, which also demonstrates reliability (WCCB-TBS), and the helium-cooled ceramic reflector

13
Nucl. Fusion 55 (2015) 104023 O. Motojima

Figure 18. Schematic view of the three equatorial ports with six Test Blanket Systems, along with their cooling systems and tritium circuit
connections.

TBS (HCCR-TBS) proposed respectively by Japan and six TBSs have been performed in 2014. The CDRs for all
Korea. the TBSs are planned to be completed in 2015. The manu-
facturing assessment of the TBM frames and dummy TBMs
The United States and the Russian Federation are also con-
is on-going to be ready for the PDR that is planned at the
tributing to the TBM Program by performing R&D on specific
beginning of 2016.
areas of breeding blanket technologies applicable to the above
The use of RAFM steels for the TBMs structures could
TBSs.
generate some perturbation to the plasma confinement. Recent
Following the various ITER plasma operation phases,
experiments on DIII-D with an electromagnetic mock-up of a
several versions of each TBM are planned to be tested with
pair of TBMs have shown that either external or internal cor-
different testing objectives. To simulate the correct dem-
rection coils, similar to those being designed for ITER, can
onstration fusion power plant (DEMO) conditions, these
correct the intrinsic plus TBM mock-up error fields at high
versions are made by the same structural materials as fore-
performance to avoid significant degradation in plasma per-
seen in DEMO, which are reduced-activation ferritic mar-
formance [7].
tensitic steels (RAFMs), and make use of the ‘engineering
The overall objective of the TBM Program is to acquire
scaling’ strategy that allows to have very important infor-
all relevant data and information concerning a given tritium
mation even with the relatively less severe ITER operation
breeding blanket (BB) to validate the applied codes for the
conditions.
relevant analyses and to be able to design, manufacture, and
The locations of the various sub-systems of the six TBSs
operate a BB system in DEMO and in future fusion power
within the tokamak building and the tritium building are shown
reactors, provided that data on long-term irradiation effects
in figure 18. They are designed and procured by the corre-
and failure modes are obtained in parallel in other facili-
sponding DAs. Two TBMs and their associated TBM shields
ties. Since several BB designs are tested simultaneously in
are inserted in a water-cooled steel frame (TBM frame) and
ITER, the TBM program could also determine the figure of
located in a port plug (PP) inside the appropriate TBM equato-
merit of the various designs prior to DEMO-BB design and
rial port. The IO is charged with the definition of the required
manufacturing.
maintenance operation and the design and procurement of
the associated maintenance tools and equipment. Moreover,
the IO is also charged with the design and procurement of 10. ITER physics issues
the TBM Frames for the three TBM PPs and of the design
and procurement of six dummy TBMs that are required to Significant progress has been made addressing several out-
replace the TBMs and their shields in case the TBMs are not standing physics issues including disruption load character-
available. ization, prediction, avoidance, and mitigation [8–12], first wall
The TBM program has made substantial progress in the and divertor shaping [13, 14], edge pedestal and SOL plasma
last couple of years. From the management side, the signature stability [15–17], fuelling and plasma behaviour during
of six TBM arrangements with each of the DAs involved were confinement transients [17–19] and W impurity transport
achieved. [20–22]. The ITER research plan [23] has been further devel-
From the technical side, the designs of the six Test oped including a definition of the required plant configura-
Blanket Systems (TBSs) and of the associated maintenance tion for 1st plasma and subsequent phases of ITER operation
tools and equipment have reached the conceptual level and as well as the major commissioning activities with plasma
the corresponding analyses have been performed. The CDRs and the needs of the accompanying R&D program to ITER
of the CN-HCCB TBS and of the connection pipes for all construction by the ITER Parties. The conceptual design of

14
Nucl. Fusion 55 (2015) 104023 O. Motojima

Figure 19. Power flux distribution on an inner midplane first wall panel top surface for a full bore limiter plasma near the end of the
start-up phase of an ITER discharge assuming a double exponential parallel heat flux profile in the SOL (Ip = 5 MA, power into the SOL
~5 MW). Left: the case of the original first wall panel toroidal shaping (optimized for a single exponential heat flux profile). Right: a
panel with shape optimized for the double exponential. Note that the maximum steady state power handling capacity of these panels is
~4.5 MW m−2 (see section 5.1).

the plasma control system (PCS) has been approved and the loads and radiation asymmetries. Further R&D is required to
preliminary design for 1st plasma and early plasma operation demonstrate a reliable technique to mitigate runaway elec-
has begun with a consortium of EU and US plasma control trons that can be extrapolated to ITER conditions. In view of
experts. the very demanding requirements and the remaining uncer-
tainties for disruption/RE mitigation, the DMS design should
remain very flexible and include both techniques of shattered
10.1. Disruption prediction, avoidance, and mitigation
pellets and massive gas injection.
The magnetic and thermal energy stored in a burning plasma
in ITER will significantly exceed that in present devices. The 10.2. First wall and divertor component shaping
rapid release of this energy during a disruption could cause
surface melting of PFCs and the resulting high electromag- As mentioned in section 5, the ITER FW panels are toroidally
netic loads could, in some cases, approach design limits. shaped to ensure that leading edges are not presented to the
Because of the high plasma current in ITER, possible high plasma, either in the case of start-up plasmas in limiter con-
energy runaway electrons (RE) following a disruption could figuration or in the case of main chamber far scrape-off layer
carry substantial energy that could eventually be released to (SOL) heat fluxes during divertor operation. The original
very localized areas of the PFCs, resulting in localized melting shaping design was based on the assumption of a single expo-
of the first wall. So, there is an R&D program on existing nential radial profile of parallel heat flux in the SOL plasma.
devices to learn how to avoid, predict, and best mitigate dis- It has recently become clear, however, following a multi-
ruption and RE loads. The US-DA has taken responsibility for machine effort initiated by the IO [13], that the very narrow
designing disruption and RE mitigation systems (DMS) [10] gradient in the near SOL heat flux found in diverted H-mode
for ITER that will inject large amounts of high-Z impurities plasmas at high power [24], is also present in simple Ohmic
into the plasma on the timescale of tens of milliseconds to and L-mode inner wall limiter plasmas [25]. The limiter SOL
dissipate plasma thermal and magnetic energy and distribute parallel heat flux can thus in general be described in current
it over a large surface area of the PFCs by radiation. devices by a double exponential profile and this has important
Disruption avoidance is a critical strategy for efficient consequences for FW power handling on ITER during limiter
operation of ITER and increasing such R&D efforts on start-up/ramp-down if the same physics applies at the ITER
existing devices will yield valuable experience to develop scale.
avoidance strategies for ITER. Reliable disruption prediction To quantify this, key parameters and scalings have been
is the basis for a high mitigation success rate and it is essential extracted from the new limiter SOL heat flux database and
to develop a physics based and transferable multi-parameter used, in conjunction with 3D field line tracing to derive a new
threshold predictor in parallel to the data-driven approaches to toroidal shape for the ITER inner wall FW panels which miti-
quickly achieve a high prediction rate from the very beginning gates the effect of the narrow SOL feature, should it be found
of ITER operation. The ongoing design of the ITER disrup- on ITER, but has no significant impact on the power handling if
tion and RE mitigation system is expected to be finalized in the phenomenon does not occur. This is illustrated in figure 19
2017. This system will be essential from the start of the first in terms of power flux distributions on the plasma-facing sur-
experimental campaign, particularly to avoid RE generation. face of an inner midplane ITER FW panel (see figure 12 for a
Experiments have shown reliable mitigation of electromag- 3D image of an individual BM) [13]. If a double exponential
netic loads using massive gas injection of neon or argon, but, heat profile is imposed on the panel with the original toroidal
uncertainties remain with respect to the mitigation of thermal profile optimized for a single exponential, the peak steady
15
Nucl. Fusion 55 (2015) 104023 O. Motojima

Figure 20. (a) Mode structure of an n = 40 ballooning in the ITER pedestal and SOL with λ p = 1.2 m. (b) Growth rate of the n = 40
ballooning mode as a function of the inner radius of the simulation domain.

state heat flux significantly exceeds the design power han- element in the bulk tungsten outer divertor [26], which yielded
dling capacity. In contrast, a toroidal profile designed for a favourable conclusions for ITER in terms of transient induced
double exponential spreads the power much more effectively tungsten melting and melt splashing, but found major discrep-
and power handling stays well below power handling limits. ancies between the expected and measured/inferred heat loads
Section 5.2 described the intensive program of engineering/ on the misaligned edge. These findings put into question the
materials and physics R&D which has been pursued by the basic assumptions used in the ITER modelling [14] to estab-
IO, DA partners and the physics community, notably the ITPA lish the optimum monoblock shaping. A final decision on the
over the past 3 years to establish the basis for the recent ITER appropriate shaping is to be taken by the IO in early 2016. In
Council decision to begin ITER operations with a full W the meantime, a very extensive effort is being made within the
divertor. On the physics side, some of the key issues concern R&D community, coordinated by the ITPA, to better under-
material surface damage due to unmitigated transients (ELMs stand the physics of edge heat loading at glancing angles, high
and disruptions), the consequences of very high numbers of parallel heat flux density and strong magnetic fields.
mitigated ELM transients, the potential for surface degrada-
tion due to non-active operation with pure helium plasma,
10.3. Edge pedestal and SOL plasma stability
the transport of tungsten eroded by ELMs (see section 10.3),
and the physics of power loading on misaligned edges in the Application to ITER and validation with existing experi-
high heat flux areas of the divertor targets. This latter issue ments of the non-linear MHD simulations with the JOREK
is the final outstanding element of the full tungsten divertor code have advanced considerably since the last IAEA con-
design, since it determines the design of the monoblock sur- ference [15–17]. Application of the JOREK code to evaluate
face shaping, which was not completely resolved in time for the MHD stability of the H-mode scrape-off layer plasma has
the 2013 Final Design Review of the Divertor (section 5.2). shown that the SOL plasma is MHD stable down to separatrix
A very detailed study of expected power loading on scale lengths of ~1 mm, as shown in figure 20 [17], and thus
both shaped and unshaped monoblock surfaces in the ITER that MHD stability is unlikely to limit the scale-length of the
divertor has recently been completed [14], approaching the pressure and heat flux in the SOL in ITER Q = 10 plasmas
problem both from a geometric (optical, or guiding centre) down to λ p ~1 mm. Whether this will be the case or not would
viewpoint and taking into account charged particle orbits in a thus depend entirely on the mechanisms determining heat
ballistic Monte Carlo treatment. The study demonstrates that SOL transport near the plasma separatrix in H-modes [15].
the toroidal chamfer proposed in the divertor design review Validation of JOREK predictions for ELM control by pellets
will adequately protect leading edges from heat load, but that, with JET experimental results has been carried out including
as a result, steady state power handling can become an issue, not only the ELM triggering phase but also the controlled
particularly in the case of slow transient (e.g. divertor plasma ELM crash itself [17]. The good agreement between JOREK
reattachment) events. Moreover, the large Larmor orbits of predictions and JET measurements for controlled ELMs dem-
ions in ELM perturbations allow particles to ‘skip’ over edges onstrates that previous studies regarding the requirements for
protected against steady state power flux so that the reduced ELM control by pellet injection in ITER are appropriate to
projected area (due to the chamfer) increases the surface provide a physics basis for the design of the corresponding
power deposition due to ELMs. The issue has been made more pellet injectors for ITER. Finally, modelling studies to iden-
complex by the experiment performed at JET in 2013 inves- tify the physics processes behind alternative solutions to
tigating ELM induced melting of a deliberately misaligned those in the ITER Baseline for ELM control (pellet triggering

16
Nucl. Fusion 55 (2015) 104023 O. Motojima

Figure 21. (a) The density profile as a function of the normalized poloidal flux before and in the presence of the n = 1 kink-peeling mode.
(b) The normalized density at the ψ N = 0.95 surface as a function of time, showing the typical non-sinusoidal behaviour characteristic of an
EHO perturbation.

and ELM control by 3D fields) have been carried out. These together with the anomalous inwards pinch expected in the
have been focused on the QH-mode regime developed at the core ITER plasma lead to the performance of gas fuelled 15
DIII-D that has the potential for high density/high confine- MA DT H-modes to be much higher than previously esti-
ment H-mode operation without ELMs in ITER. The JOREK mated with values as high as Q DT ~7 being achievable. For the
modelling studies have revealed that in this regime the peri- achievement of Q DT ~ 10 pellet fuelling is required in agree-
odic expulsion of energy and particles by ELMs is replaced ment with previous evaluations.
by a continuous particle expulsion driven by a saturated low Access and exit to/from high Q DT phases in ITER are
n kink-peeling mode, as shown in figure 21, which has many most challenging from the plasma control point of view. In
of the features of the experimentally observed edge harmonic order to develop an adequate control strategy for such phases
oscillation (EHO) [16]. The next step in this line of work for an appropriate description of plasma behaviour is required
ITER, which is in progress, is to investigate if such saturated for ITER. This has been the focus of modelling studies by
low n kink-peeling modes will develop in the expected ITER the JINTRAC code suite whose modelling assumptions
pedestal plasma conditions. have been compared with those required to model JET
experimental results [19, 20]. These studies have shown that
10.4. Fuelling and plasma behaviour during confinement appropriate adjustment of the fuelling following the increase
transients of additional heating after the L-mode phase at 15 MA is
required to ensure a robust access to burning plasma condi-
Extensive modelling studies have been carried out to evaluate tions in ITER because of the influence of plasma density on
the requirements for fuelling of ITER plasmas and of plasma the H-mode threshold and on the alpha power increase in
behaviour in confinement transients in various phases of its that phase. For robust H-mode Q DT ~ 10 it is required that
operational scenarios and operational phases [18–20]. Studies the density is maintained by gas fuelling at a constant level
of fuelling requirements including both the confined plasma of ~ 0.4n GW for longer than typically 7 s after the increase of
as well as the SOL plasma have confirmed that fuelling effi- additional heating and then raised by pellet fuelling to ~nGW
ciency of the recycling flux in ITER is very low [18]. This in a time interval longer than 20 s as shown in figure 22(a).
leads to the maximum plasma density achievable by gas fuel- A faster density evolution prevents the build-up of alpha
ling to be determined by divertor detachment and core plasma heating power in this phase and leads to the return of the
transport which, for ITER low collisionality plasmas, has a plasma to L-mode as the plasma density is raised to ~nGW.
significant inwards convection component. Estimates for Studies of the exit phase from Q DT ~ 10 H-mode condi-
ITER show that this maximum density for L-mode plasmas tions have shown that previous evaluations for ITER based
might be lower than that required for stationary NBI injec- on simplistic transport assumptions (instantaneous transi-
tion due to limitations to the shine-through loads on the first tion from H-mode to L-mode when additional heating is
wall for plasma currents under 10–15 MA [18]. Thus, if NBI switch off) are unphysically pessimistic for ITER, which is
is required in such conditions these L-mode plasmas might in agreement with measurements during ITER-like experi-
have to be fuelled by pellets instead of gas puffing. On the ments at JET [20]. Due to the dominance of alpha heating
contrary, the same modelling studies find that the fusion in these ITER plasmas and the core plasma density being
­performance of gas fuelled H-mode DT plasmas in ITER is determined by pellet fuelling the degradation of confinement
significantly higher than previously expected. The larger edge following the additional heating switch off in ITER takes
power flow and high plasma current, leading to a narrow λ q, place in timescales much longer than previously assumed (6
and detachment to occur and very high separatrix densities s instead of 1.8 s) as shown in figure 22(b). This leads to the
nsep > 5× 1019 m−3, and the low neutral source in the pedestal power fluxes deposited on PFCs to be a factor of ~3 lower

17
Nucl. Fusion 55 (2015) 104023 O. Motojima

(a) (b)
Figure 22. (a) Operational space for the achievement of a transition to high QDT ~ 10 15 MA/5.3 T H-mode in ITER for PAUX = 53 MW
in terms of the nominal density rise time and delay of the ramp in density with respect to the start of the high heating power have for two
models for the anomalous core transport in ITER (GLF 23 and Bohm–gyroBohm). (b) Plasma parameter evolution following the sudden
switch-off of additional heating and pellet fuelling at 530 s for QDT =10 H-mode plasma in ITER. From top to bottom: plasma energy,
alpha heating, core and pedestal density and temperature are displayed. The time at which the H–L transition takes place is indicated by
a vertical dashed line; the time at which the plasma energy decreases by e−1 from its value during the stationary burning plasma phase is
indicated by the intersection of the horizontal dashed line with the time trace of the plasma energy.

Figure 23. (a) Evolution of the plasma temperature at the ITER outer divertor around the H-mode transition in ITER showing that gas
fuelling is effective to maintain the divertor temperature at low values so that the W sputtering source is minimized. (b) Plasma parameter
evolution following the increase of additional heating at 83.5 s for access to a QDT = 10 H-mode plasma in ITER. From top to bottom:
plasma energy, alpha heating, auxiliary heating and W radiated power, pedestal temperature and density and resulting W pinch velocity in
this phase (positive pinch velocity implies poor W penetration in the plasma).

than previously assumed for this H-mode exit phase and to a into account this leads to the neoclassical pinch velocity to be
much better control of the plasma position during that phase directed outwards for W in the pedestal and to a very low con-
than previous estimates. tamination of the plasma between controlled ELMs for ITER
Q DT = 10 stationary phases. The same physics basis regarding
edge W transport has been applied to optimize W control in
10.5. W impurity transport
the H-mode access phase in ITER, as shown in figure 23. In
Major progress has taken place regarding the modelling of the initial phase after the increase of the additional heating
W behaviour in the edge and pedestal plasmas of ITER since after the L-mode phase at 15 MA the plasma is fuelled with
the last IAEA conference [20–22]. Experiments in ASDEX- gas puffing. This is appropriate to decrease the divertor plasma
Upgrade have shown that impurity transport in the pedestal can temperature to values under 10 eV thus ensuring a low W
be accurately described by neoclassical transport. When such a sputtering source during this phase (figure 23(a)). After the
hypothesis is applied to ITER Q DT = 10 and the need of a high pedestal temperature has built-up after the L–H transition,
separatrix density for divertor power load control are taken which ensures large temperature screening in the pedestal

18
Nucl. Fusion 55 (2015) 104023 O. Motojima

33 MW + 20 MW off-axis ECRH
33 MW + 20 MW on-axis ECRH 2.0 33 MW + 20 MW off-axis ECRH
14 33 MW + 20 MW on-axis ICRH 50:50 e-i (He3) 33 MW + 20 MW on-axis ECRH
33 MW + 20 MW on-axis ICRH 50:50 e-i (He3)
12
1.5
10

nWnormalized
ni(1019m-3)
8
1.0
6

4 0.5
2

0 0.0
0 1 2 0 1 2

r (m) r (m)
Figure 24. Results of ASTRA simulations for ITER 15 MA/5.3 T QDT = 10 plasma with different heating schemes (33 MA of NBI  +  
20 MW of off axis or on axis ECRH and 33 MA of NBI  +  20 MW of on axis ICRH with He3 minority to provide maximum ion heating):
(a) ion density profile versus minor radius, (b) normalized W density profile versus minor radius.

plasma for W and low plasma contamination, the plasma requirements for each of the stages of ITER operation in prep-
density is increased by pellet fuelling leading to a slow den- aration of a revised ITER schedule up to DT operation that
sity increase (figure 23(b)), which maintains the favourable will be presented to the ITER Council in 2015. The research
W transport to ensure poor plasma contamination during this plan has been developed in stages to define requirements and
phase all the way up to the ITER Q DT = 10 stationary phase. an initial commissioning and operations plan for 1st plasma,
It is important to note that this optimum fuelling strategy that up to 3.5 MA plasma current operation, up to 7.5 MA opera-
ensures core plasma W control in the H-mode access phase to tion at 2.65 T in the H/He non-activated phase, H-mode opera-
Q DT = 10 in ITER is the same that ensures robust access to tion in He plasmas, 15 MA operation at 5.3 T in H/He, D
Q DT = 10, as shown in section 10.4, and thus provides a fully operation, and DT operation. Commissioning of disruption
self-consistent scenario to Q DT = 10 in ITER. detection, avoidance and mitigation will make a significant
Regarding core W transport, the solid development of its demand on the experimental time, requiring possibly 10–15%
physics basis from recent experimental results in the JET of the operational time during the non-active phase. Timely
ITER-like wall program has enabled to carry out physics- availability of the heating and current drive systems is essen-
based simulations of the expected W transport behaviour in tial, as is the provision of an adequate diagnostic capability to
ITER. Simulations with ASTRA and JINTRAC show that dif- support protection and control functions (but even diagnostics
fusive anomalous transport dominates W transport across the required for physics understanding will play a crucial role if,
cross section of the plasma except in the very central region as might be anticipated, ITER plasmas do not behave exactly
(r /a ⩽ 0.3) where neoclassical W transport dominates, which as predicted). Fuelling and ancillary control systems (such as
is in good agreement with the JET experiments and modelling in-vessel coils) are also essential to the efficient implementa-
quoted above. Due to the low fuelling efficiency provided by tion of the program and a rapid transition to DT operation,
the 1 MeV NBI heating in ITER the scale lengths of density which relies on an effective integration of these systems both
and temperature in that core region are such that no significant to support the development of the necessary plasma scenarios
W peaking is expected in ITER reference Q DT = 10 plasma and understanding of plasma behaviour at the ITER scale, and
conditions, as shown in figure 24. This is despite the overall to provide a comprehensive control and protection capability
peaked density profiles resulting from anomalous transport to allow the ambitious experimental program to be imple-
for the DT ions expected in ITER in figure 24. This favour- mented while minimizing the risk to the ITER tokamak. The
able physics picture regarding the lack of strong W accumula- more detailed integrated commissioning plan does not result
tion in ITER has been demonstrated experimentally in Alcator in additional delays to reaching high performance DT opera-
C-Mod experiments operating with ITER-like H-mode tion, but will depend on meeting an aggressive construction
plasmas (in terms of core heating/fuelling and plasma den- and system implementation schedule.
sity peaking). In addition, the ITER Baseline heating schemes
have been proven to provide significant capabilities for W
concentration control when applied to the plasma centre at 11. Conclusion
the 20 MW power level because of their much more peaked
deposition profile than the dominant plasma heating by alpha It is very important to implement an integrated ITER plant con-
heating (Pa = 100 MW) as shown in figure 24. struction time schedule assuring management of the quality
of the installation, consistency with the design and IO design
10.6. ITER Research Plan
authority, on time resolution of in-field non-conformities, cost
control on a day-to-day basis, begin commissioning activi-
The ITER Research Plan [23] has been further devel- ties upon system turn over and plant delivery to the oper-
oped including a more detailed commissioning plan and ating organization team upon commissioning completion.

19
Nucl. Fusion 55 (2015) 104023 O. Motojima

In such a large project like ITER, the financial constraints require [3] Encheva A. et al 2014 [FIP/1–5] paper presented at the 25th
cost efficiency and cost savings. Therefore, it is an important and IAEA Fusion Energy Conf. (St Petersburg, Russia)
unique opportunity to assure an interesting future for the life of [4] Merola M. et al 2014 SOFT-28 Conf. (San Sebastian, Spain,
29 September–3 October 2014)
the ITER project. On the other hand, the ITER plant completion [5] Pitts R.A. et al 2013 J. Nucl. Mater. 438 S48–56
generates industrial business opportunities for the international [6] Giancarli L. et al 2012 Overview of the ITER TBM program
industrial system in high technology products, systems, and ser- Fusion Eng. Des. 87 395–402
vices. A culture of cost efficiency has to become a critical effort [7] Lanctot M.J. et al 2014 A path to stable low torque
without exceptions for ITER’s success. Industrial participation, plasma operation in ITER with test blanket modules
25th IAEA Fusion Energy Conf. (St Petersburg,
together with the ITER Organization, in supporting and devel- Russia)
oping cost efficient products, becomes strategic as well as neces- [8] Lehnen M. et al 2014 [EX/P3-18] paper presented at the
sary. There is no technical barrier to assure the future of ITER 25th IAEA Fusion Energy Conf. (St Petersburg, Russia)
technology. To be successful, the ITER IO  +  DAs and the indus- [9] Aleynikov P. et al 2014 [TH/P3-38] paper presented at the
trial system have to work hard in an integrated way, looking 25th IAEA Fusion Energy Conf. (St Petersburg, Russia)
[10] Baylor L.R. et al 2014 [FIP/2-1] paper presented at the
continuously at cost savings in engineering/procurement and 25th IAEA Fusion Energy Conf. 2014 (St. Petersburg, Russia)
installation, time schedule saving, and the quality of the job. [11] Martín-Solís J.R. et al 2014 Nucl. Fusion
It is my pleasant duty to thank all seven Members, DAs and 54 083027
the IO as the UIT for their large effort and contribution towards [12] Martín-Solís J.R. et al 2014 [TH/P3-43] paper presented
the progress of the ITER Project. Over the past two years, the at the 25th IAEA Fusion Energy Conf. (St Petersburg,
Russia)
ITER Project has made the transition from design to building [13] Kocan M. 2015 Nucl. Fusion 55 033019
up the infrastructure by the IO and seven DAs. Design com- [14] Gunn J.P. et al 2014 [FIP/1–2] paper presented at the 25th
pletion is leading to actual manufacturing of in-kind compo- IAEA Fusion Energy Conf. (St Petersburg, Russia)
nents. Our aim and responsibility is to make progress in ITER [15] Loarte A. et al 2015 J. Nucl. Mater. 463 401
construction and to keep the project risks in terms of scope, [16] Liu F. et al 2015 Nucl. Fusion. 55 113002
[17] Huijsmans G.T A. et al 2014 [TH/6-1Ra] paper presented
cost, and schedule acceptable to the stakeholders. at the 25th IAEA Fusion Energy Conf. (St. Petersburg,
Russia)
[18] Romanelli M. et al 2014 [PPC/P3-22] paper presented at the
Acknowledgments 25th IAEA Fusion Energy Conf. (St. Petersburg, Russia)
[19] Loarte A. et al 2014 Nucl. Fusion 54 123014
I would like to acknowledge Joseph Snipes’ great contribution [20] Koechl F. et al 2014 [TH/P3-24] paper presented at the
25th IAEA Fusion Energy Conf. (St. Petersburg,
to accomplish this review paper together with the excellent IO Russia)
engineers and physicists who are every day successfully meet- [21] Hogeweij G.M.D. et al 2014 [EX/P3-17] paper presented
ing the challenge of the world’s difficult job, the ITER project. at the 25th IAEA Fusion Energy Conf. (St. Petersburg,
Russia)
[22] Loarte A. 2015 Phys. Plasmas 22 056117
References [23] Campbell D.J. 2012 [ITR/P1-18] paper presented at the
24th IAEA FEC (San Diego, CA)
[24] Eich T. et al 2013 Nucl. Fusion 53 093031
[1] Devred A. et al 2014 Challenges and status of ITER conductor [25] Goldston R.G. 2015 J. Nucl. Mater. 463 397
production Supercond. Sci. Technol. 27 044001 [26] Matthews G. et al 2014 [EX/4-1] paper presented at the
[2] Sborchia C. et al 2014 [FIP/1-6Ra] paper presented at the 25th 25th IAEA Fusion Energy Conf. (St. Petersburg,
IAEA Fusion Energy Conf. (St Petersburg, Russia) Russia)

20