P

GCa

b

a

AA

KFIPESHHAPS

1

psria

btffi

0d

Fusion Engineering and Design 84 (2009) 1562–1569

Contents lists available at ScienceDirect

Fusion Engineering and Design

journa l homepage: www.e lsev ier .com/ locate / fusengdes

lasma scenarios, equilibrium configurations and control in the design of FAST

. Ramogidaa,∗, G. Calabroa, V. Cocilovoa, F. Crisanti a, A. Cucchiaroa, M. Marinuccia, A. Pizzutoa,. Ritaa, F. Zoncaa, R. Albaneseb, G. Artaserseb, F. Mavigliab, M. Matteib

Associazione Euratom-ENEA sulla Fusione, Via Enrico Fermi 45, I-00044 Frascati, RM, ItalyAssociazione Euratom-ENEA-CREATE sulla Fusione, Via Claudio 21, I-80125 Napoli, Italy

r t i c l e i n f o

rticle history:vailable online 21 March 2009

eywords:ASTTERlasmaquilibriumcenario-modeybriddvanced Tokamakosition controlhape control

a b s t r a c t

The Fusion Advanced Studies Torus (FAST) conceptual study has been proposed [A. Pizzuto on behalf ofthe Italian Association, The Fusion Advanced Studies Torus (FAST): a proposal for an ITER Satellite facilityin support of the development of fusion energy, in: Proceedings of 22nd IAEA Fusion Energy Conference,Geneva, Switzerland, October 13–18, 2008; Nucl. Fusion, submitted for publication] as possible EuropeanITER Satellite facility with the aim of preparing ITER operation scenarios and helping DEMO design andR&D. Insights into ITER regimes of operation in deuterium plasmas can be obtained from investigationsof non linear dynamics that are relevant for the understanding of alpha particle behaviours in burningplasmas by using fast ions accelerated by heating and current drive systems.

FAST equilibrium configurations have been designed in order to reproduce those of ITER with scaledplasma current, but still suitable to fulfil plasma conditions for studying burning plasma physics issues inan integrated framework. In this paper we report the plasma scenarios that can be studied on FAST, withemphasis on the aspect of its flexibility in terms of both performance and physics that can be investigated.All plasma equilibria satisfy the following constraints: (a) minimum distance of 3 energy e-folding length(assumed to be 1 cm on the equatorial plane) between plasma and first wall to avoid interaction betweenplasma and main chamber; (b) maximum current density in the poloidal field coils, transiently, up to

2

around 30 MA/m . The discharge duration is always limited by the heating of the toroidal field coils thatare inertially cooled by helium gas at 30 K. The location of the poloidal field coils has been optimizedin order to: minimize the magnetic energy; produce enough magnetic flux (up to 35 Wb stored) forthe formation and sustainment of each scenario; produce a good field null at the plasma break-down(BP/BT < 2 × 10−4 at low field, i.e. BT = 4 T and ET = 2 V/m for at least 40 ms).

Plasma position and shape control studies will also be presented. The optimization of the passive shelll stab

position slows the vertica

. Introduction

Fusion Advanced Studies Torus (FAST) has been proposed as aossible option for a European ITER Satellite facility [2], aimed atupporting the preparation of operation scenarios and the explo-ation of technologies relevant to DEMO physics and technologyssues in a wider (dimensionless) parameter space than JT-60SAnd with characteristic values closer to ITER.

FAST has been conceived to contribute drawing the maximumenefit from ITER before as well as in parallel with ITER exploita-

ion in a time window lasting significantly longer than currentlyoreseen for any existing European devices and within reasonablenancial constraints.

∗ Corresponding author.E-mail address: gramogid@frascati.enea.it (G. Ramogida).

920-3796/$ – see front matter © 2009 Elsevier B.V. All rights reserved.oi:10.1016/j.fusengdes.2009.02.026

ility growth time down to 100 ms.© 2009 Elsevier B.V. All rights reserved.

The R&D objectives in fusion physics, technology and engineer-ing have been structured by the Fusion Facilities Review Panel inseven interrelated missions [2] along the path from ITER towardsDEMO and further: FAST has been designed to address several dif-ferent aspects of the first five of these interconnected objectives ina integral fashion. FAST will be able to explore Fast Particle physicsissues (mission 1), to investigate general aspects of ITER relevantPlasma Operations (mission 2), to look into the physics of largeheat loads on divertor plates (mission 3), to investigate AdvancedTokamak (AT) scenarios (mission 4) and to promote the validationof numerical simulation codes to predict ITER fusion and burningplasma performance (mission 5).

As the contribution to mission 1 objectives on burning plasma

achievement in ITER, FAST will be able to investigate non lineardynamics that are relevant for the understanding of alpha parti-cle behaviours in burning plasmas by using fast ions accelerated byheating and current drive systems, working with deuterium plas-mas in a dimensionless parameter range closer to the ITER one

G. Ramogida et al. / Fusion Engineering

FQti

ttqpettepNi[fiptepowowcrttr

aatf

wttif

coils to allow plasma shaping flexibility, manufacture easiness andefficient cooling, six external poloidal field coils, the vacuum ves-sel (VV) with its internal components and the mechanical supportstructure.

ig. 1. FAST physics operational space in the (Q/(Q + 5), a/RLarmor) plane. Since/(Q + 5) = N/Nc ∝ T5/2 and T controls edge physics conditions as well as PWI [5,6],

his Fig. emphasizes one of the fundamental aspects of burning plasma physicsntegration, which is the very original motivation of FAST.

han that of existing machines [1], as depicted in Fig. 1. Althoughhe use of the Q factor to define the physics operational domain isuestionable for a machine dedicated to operate with deuteriumlasmas and not with a deuterium–tritium mixture, it has beenmphasized in [3,4] that �SD/�E ∝ T5/2/N, with N = nT�E and �SD/�Ehe ratio of alpha particle (fast ions) collisional slowing down timeo the energy confinement time. Thus, for fixed �SD/�E in order tonsure similarity of ˇfast/ˇ (ˇfast/ˇ is the ratio of fast ion to thermallasma kinetic energy densities) and of electron–ion equipartition,∝ T5/2 and it is reasonable to fix T when defining operation scenar-

os of a burning plasma relevant experiment since Q/(Q + 5) = N/Nc

3,4], with Nc the critical triple product at ignition. Meanwhile,xing T corresponds to controlling edge physics conditions andlasma wall interactions (PWI) [5,6]. Thus, although the choice ofhe y-axis in Fig. 1 may be questionable and is not the most gen-ral one, it captures one of the fundamental aspects of burninglasma physics integration, which is the very original motivationf FAST. Meanwhile (see Section 4), the fact that FAST can operateith characteristic dimensionless (both thermal and fast) particle

rbits similar to those of ITER ensures that FAST transport physicsill indeed be relevant since it will reproduce micro- to meso-scale

ross-couplings typical of burning plasma conditions [7–11] and theatio between energy confinement time and electron–ion equipar-ition time will be comparable to that of ITER; thus, the access tohe high performance regimes will occur at dominant and DEMOelevant electron heating.

FAST will be able to contribute, as other machines, to severalspects related to mission 2 issues on reliable operations: plasmand ELM control, assisted break-down development, assessment ofhe toroidal field ripple (TFR) effects, power coupling studies in aast particles operational space closest to that of ITER.

The high magnetic field in FAST together with its compactness

ill make possible to obtain a very high power flux P/R, greater

han ITER and approaching the DEMO target value, allowing FASTo test, in relevant conditions, technical approaches to mission 3ssues, related to first wall and divertor power handling, such asull-tungsten wall/divertor and liquid lithium divertor solutions.

and Design 84 (2009) 1562–1569 1563

Even if it is not a superconducting machine, FAST will be indeedcapable to approach steady state conditions of interest for mission4 objectives, thanks to its availability of heating and current drivepower sufficient to access a full non inductive current drive scenariowith high bootstrap current fraction.

The contribution of FAST to extend the validation of predic-tive codes to wider parameter regimes, as far as mission 5, will bealso relevant, to fill the gap in operational space foreseen betweenITER and JT-60SA. The various physics that can be addressed inFAST to verify and validate numerical codes and theoretical modelsare described in Refs. [7–12] and in the extended versions of Refs.[13,14].

2. The FAST machine

The requirement for plasma behaviour sufficiently close to ITERsets stringent constraints to FAST features that must be accom-plished:

(1) plasma current, Ip, from 2 MA (corresponding to full non induc-tive current drive scenario) up to 8 MA (corresponding tomaximized performance scenario);

(2) auxiliary heating systems able to accelerate the plasma ions toenergies in the range 0.5–1 MeV;

(3) plasma major radius of about 1.8 m and minor radius around0.65 m;

(4) pulse duration from 20 s for the reference H-mode scenarioup to 160 s (∼40 resistive times �res) for the longest AdvancedTokamak scenario at 3 MA/3.5 T.

These features have been satisfied in the current FAST design ofa compact (Ro = 1.82 m, a = 0.64 m, triangularity ı = 0.4) and cost-effective machine able to investigate, at the same time and inintegrated way, non linear dynamics effects in the fast particlebehaviours [1], plasma wall interaction under ITER relevant powerload [13], ITER relevant operational issues and Advanced Tokamakregimes up to fully non inductive plasma current driven scenarios.

FAST load assembly is shown in Fig. 2: it consists of 18 toroidalfield coils (TFC), a central solenoid (CS) vertically segmented in six

Fig. 2. The load assembly of FAST.

1564 G. Ramogida et al. / Fusion Engineering and Design 84 (2009) 1562–1569

tpaatThtii

F(iT[afgaa[ttrbef

mmvtfd

smflm

wmfl

ch

3. Plasma equilibrium configuration

The location and the dimension of the poloidal field coils, shownin Fig. 4, have been optimized with the constrains to minimize the

Fig. 3. Coils temperature evolution for the longest (AT2) scenario.

The copper coils are kept at cryogenic temperatures (30 K) beforehe pulse starts, to lower their adiabatic heating during the plasmaulse increasing the duration of the AT scenarios: the final temper-ture at the end of the pulse is never expected to exceed 85 K forny poloidal coil and 150 K for the inner legs of the TFCs; in Fig. 3he coil temperature tim evolution is reported for the AT2 scenario.he cooling of the magnet system after the pulse is guaranteed byelium gas flow. Each TFC is contained by a stainless steel belt fittedo the outside zone of the coil. Two pre-compressed rings situatedn the upper–lower zone keep the whole toroidal magnet structuren wedged configuration.

A preliminary 3D magnetostatic analysis [15] showed that inAST the toroidal field ripple can be kept at an acceptable valuebelow 0.3% on the plasma separatrix) by mean of ferromagneticnserts located inside the VV on the outboard side, in front of theF coils [15]. A future proper optimization of the size of this inserts16] and of the plasma shape should allow to limit the TFR withincceptable values at all the toroidal field configurations, rangingrom 3.5 to 8.5 T. An alternative active system to reduce the TFRuaranteeing the greatest flexibility of the machine operation withny value of the magnetic field has been studied, by using of smallctive control coils located between the outer branches of the TFC15]. A preliminary analysis showed that this system could reducehe maximum TFR on the plasma separatrix well below 0.3% in allhe toroidal field configurations, feeding the active coils with a cur-ent never exceeding 1/14 of the TFC current. Further analyses wille nevertheless required to assess the stray fields, the EM loads, thelectrical feed-through, the power supply and cooling requirementsor this solution.

The vacuum vessel, segmented by 20◦ modules, is a single shellade of Inconel 625 with 30 mm maximum thickness, in order toinimize the flux consumption during the plasma start-up. The

ertical, oblique and equatorial access ports are made of 20 mmhick stainless steel. The operating temperature of the vessel rangesrom room temperature to above 100 ◦C. A suitable water loop isedicated to regulate the vessel temperatures.

A copper shell is inserted inside the vacuum vessel aimed atlowing down the growth rate of the vertical instability to ≈13 s−1,aking the control of the plasma vertical position easier. To avoid

ux shielding during plasma break-down the shell is toroidally seg-ented.The first wall (FW) consists of a bundle of tubes armoured

ith ∼4 mm plasma spray tungsten. The divertor is made with theonoblock technology, which has been tested in high value heat

ux range.FAST is equipped with three auxiliary heating systems: ion

yclotron resonant heating (ICRH), electron cyclotron resonanteating (ECRH) and lower hybrid (LH). The injection of 30 MW ICRH

Fig. 4. FAST poloidal field coils system and field null region during the plasma break-down.

accelerate the plasma ions to energies in the range of 0.7–0.8 MeVin H-mode scenario (6.5 MA/7.5 T) [14]; 6 MW of LH have beenadopted to actively control the current profile and 4 MW of ECRHare devoted to MHD control.

Fig. 5. FAST reference H-mode equilibrium.

G. Ramogida et al. / Fusion Engineering and Design 84 (2009) 1562–1569 1565

Table 1FAST plasma parameters for the analysed configurations (the H-mode extreme configuration is foreseen in a second phase of the machine, with the additional NNBI system.

FAST H-mode reference H-mode extreme Hybrid AT AT2 AT full NICD

Ip (MA) 6.5 8 5 3 3 2q95 3 2.6 4 5 3 5BT (T) 7.5 8.5 7.5 6 3.5 3.5H98 1 1 1.3 1.5 1.5 1.5〈n20〉 (m−3) 2 5 3 1.2 1.1 1Pth H (MW) 14–18 22–35 18–23 8.5–12 8.5–12 5–7ˇN 1.3 1.8 2.0 1.9 3.2 3.4�E (s) 0.4 0.65 0.5 0.25 0.18 0.13�res (s) 5.5 5 3 3 5–6 2–5T0 (keV) 13.0 9.0 8.5 13 13 7.5Q 0.65 2.5 0.9 0.19 0.14 0.06tdischarge (s) 20 10 20 70 170 140tflat-top (s) 13 2 15 60 160 130INI/Ip (%) 15 15 30 60 80 100Padd.heat (MW) 30 40 30 30 40 40Pheat

ICRH (MW) 30 30 30 30 30 30Pheat

PP

msdBtl

(Iosa

pfTficTtsiac[

4

psnpcsp

eiefowo

〈ı〉 = 0.4, plasma volume > 20 m3. An overview of the main possi-ble configurations is given in Table 1, where the H-mode extremescenario is foreseen only in a second phase of the machine life, withthe additional NNBI system.

Table 2Main parameters of the reference H-mode equilibrium.

I (MA) 6.5

LH (MW) 0 0heatECRH (MW) 0 0heatNNBI (MW) – 10

agnetic energy, to have enough flux to allow any selected plasmacenario and to have a quite good field null during the plasma break-own, shown in Fig. 4: a very large central exapolar region withP/BT < 2 × 10−4 at low field, BT = 4 T. The toroidal electric field forhe break-down has been imposed to be 2 V/m for a time lasting ateast 40 ms.

The poloidal field circuit can provide the necessary flux∼35 Wb stored) to build-up an X-point plasma equilibrium withp = 6.5 MA and to sustain a scenario with a high current flat-topf around 13 s (the discharge lasting around 22 s) in a high den-ity (n = 2 × 1020 m−3) H-mode configuration, assuming 32 MA/m2

s maximum current density in the PFC.All plasma equilibria guarantee a minimum distance between

lasma and first wall greater than 3�E, where �E is the energy e-olding length assumed to be about 1 cm on the equatorial plane.he maximum transiently allowable current density in the poloidaleld coils is around 32 MA/m2. Within these constraints, suffi-ient flexibility is maintained to allow different plasma shapes.he divertor region has been designed to provide enough spaceo scan substantially the plasma triangularity, to allow sometrike point sweeping and to have an efficient pumping capabil-ty. The FAST divertor concept is well described in [17]. The crucialspects of the thermal loads on the divertor plates and of theore plasma purity for the proposed scenarios are discussed in13].

. FAST flexibility in physics and performances

FAST is a very flexible device able to reproduce, with scaledlasma current, the three ITER main equilibrium configurations:tandard H-mode with broad pressure profile, hybrid mode witharrower pressure profile and Advanced Tokamak scenario witheaked pressure profile. Although FAST will work at lower plasmaurrent than ITER, it is still suitable to fulfil plasma conditions fortudying operation problems, plasma wall interaction and burninglasma physics issues in an integrated framework.

The achievement of these capabilities in a compact, cost-ffective device, has been obtained by designing a machine workingn a dimensionless parameter range close to ITER, with similar

quilibrium profiles, dominant electron heating and plasma per-ormances in the fusion parameter space with Q ≥ 1. The choicef Q as indicator of equivalent “fusion performance” in a machineorking with deuterium captures one of the fundamental aspects

f burning plasma physics integration, which is the very original

0 0 6 60 0 4 4

– – – –

motivation of FAST, although it involves some questionable issues(see Section 1). A similarity scaling of FAST dimensionless param-eters based on those of ITER implies an electron temperature Te

∼10 keV and effective fast particles perpendicular temperature Tf∼ 600 keV for the 3He minority ions accelerated by ICRH in a Dplasma. This condition is reached, at the proper ˇN, with less than30 MW of ICRH additional power, as currently foreseen in FAST.The requirement to preserve the fast ion induced fluctuation spec-trum, both in mode number �∗

fast,FAST ∼ �∗fast,ITER and in frequency

(normalized to Alfvén frequency) (ωfast/ωA)FAST∼(ωfast/ωA)ITER, aresatisfied by Ip ∼6.5 MA and (n/nGW)FAST∼0.5(n/nGW)ITER. Refs. [1,14]present detailed discussions of the self-consistency and physicsintegration of FAST plasma scenarios, based on the fact that sim-ilarity scalings can be equally and consistently derived from bothfast particle dynamics as well as thermal plasma properties. Theduration of the plasma flat-top is greater than the resistive diffu-sion time in all scenarios and the ratio �tflat-top/�RES is about halfof that in ITER.

The design of FAST allows reaching ITER and DEMO relevant PWIregimes with large power load P/R ∼22 MW/m and expected ELMscomparable in size with those tolerable by ITER, �WELM ∼1.5 MJ.The external heating power in all scenarios is provided by the ICRHsystem (30 MW) and, for the Advanced Tokamak scenarios, by theLH system (6 MW) plus the ECRH system working in second har-monic (4 MW). For the most extreme H-mode scenario (Ip = 8 MA,BT = 8.5 T), additional power by a suitable NNBI system (10 MW) hasbeen assumed, which can be accommodated in equatorial ports ofthe machine, which have been designed for this scope. In all casesthe configuration has been designed to have always the same geo-metrical plasma features, shown in Fig. 5: major radius R = 1.82 m,minor radius a = 0.64 m, elongation k = 1.7, average triangularity

p

ˇN 1.3P0/〈P〉 2.35q95 3qaxis 1.09Volume (m3) 22.8

1566 G. Ramogida et al. / Fusion Engineering and Design 84 (2009) 1562–1569

F

5

ttBa

orrt

with 6.5 MA plasma current.The plasma parameters obtained for this scenario have been

fully validated in order to determine the transport and the con-finement of fast ions (∼0.7 MeV) produced by the 3He minorityions accelerated by 30 MW of ICRH [14]. Predictive simulations of

ig. 6. Evolution of the plasma configurations in the reference H-mode scenario.

. H-mode scenarios

Two possible H-mode scenarios have been extensively inves-igated: a reference H-mode (Ip = 6.5 MA, BT = 7.5 T) to be used inhe extensive integrated studies and an extreme H-mode (Ip = 8 MA,T = 8.5 T) corresponding to the highest achievable performance byssuming the availability of an additional 10 MW NNBI power input.

The main parameters of the reference H-mode equilibrium,btained by using MAXFEA [18] and FIXFREE [19] codes, are

eported in Table 2. The time evolution of the equilibrium configu-ations is shown in Fig. 6 and the time evolution of the PFC currentshat is reported in Fig. 7.

Fig. 7. Evolution of the PFC currents in the reference H-mode scenario.

Fig. 8. Predictive JETTO simulations: plasma current density (Jp), safety factor (q),electron (Te) and ion temperature (Ti) profiles during the high ˇ phase.

After the break-down the plasma current rises at Ip = 2 MA in1.5 s, maintaining a circular shape, then the plasma begin to elon-gate while its current keep raising for the next 3 s, when the finalX-point plasma shaped is achieved with Ip = 4.5 MA. The plasma cur-rent achieves its target value Ip = 6.5 MA at t = 7 s. At t = 7.5 s it isassumed to apply the full additional heating, causing an increase ofthe internal kinetic energy on a time scale (about 1 s) longer than theplasma energy confinement time. During this strong increase of ˇNthe plasma boundary it is assumed to be preserved by using a shapecontrol technique like the extreme shape controller (XSC) used inJET [20]. A long experimental flat-top (up to 13 s, correspondingabout two resistive decay time) is then possible at maximum ˇN

Fig. 9. FAST reference equilibrium configuration and passive structures (vessel andstabilizing shell) in CREATE NL model.

ering

taTtsd

qh∼tw

6

cw

wfii∼Nsf

G. Ramogida et al. / Fusion Engine

his scenario have been performed by means of JETTO code, usingsemi-empirical mixed Bohm/gyro-Bohm transport model [21,22],ORIC code [23] and SSQLFP code allowing the fully simulation ofhe scenario dynamic evolution. The plasma density current, theafety factor, the electron and ion temperature profiles obtaineduring the high ˇ phase are shown in Fig. 8.

The extreme H-mode scenario (Ip = 8 MA, BT = 8.5 T, safety factor95 ∼2.6), corresponding to the highest achievable performance,as been also studied. In this transient (�flat-top = 2 s, �E ∼0.7 s, �RES5 s) scenario, the plasma density has been assumed to be close to

he Greenwald limit, so a quite large species coupling is foreseenith Te = Ti.

. Hybrid and Advanced Tokamak scenarios

The hybrid scenario would allow to reach an equivalent Q ∼ 1onsidering an enhanced confinement factor equal to 1.3H98, mean-hile, ˇN = 2 and n/nGW = 0.8.

FAST design permits to accomplish Advanced Tokamak scenariosith quite different features: moderate ˇN ∼2 with large toroidaleld BT = 6 T (AT scenario), ˇN ∼3.2 greater than the MHD stabil-

ty with lower toroidal field BT = 3.5 T (AT2 scenario), very large ˇN3.4 with fully non inductive driven plasma current Ip = 2 MA (fullICD scenario with n = 1 × 1020 m−3) consisting of a 60–70% boot-

trap fraction and a 30–40% driven by the LH system (3.7 or 5 GHz)raction of the total plasma current.

Fig. 10. Position and shape controller performances for a

and Design 84 (2009) 1562–1569 1567

For all the Advanced Tokamak cases, the plasma boundary shapeis essentially the same of the reference H-mode scenario, reportedin Fig. 5 and the q profile is assumed to be slightly reversed withqaxis > 2 and qmin < 2 with a peaked pressure profile, as it should beexpected in an Advanced Tokamak scenario.

The poloidal circuit can sustain the discharge for a long timein the two scenarios with residual inductive current, assuming theplasma residual resistivity about 60–100 mV (tflat-top ∼60 s in ATand tflat-top ∼160 s in AT2). In the fully non inductive scenario amaximum tflat-top ∼160 s is foreseen as constrained by the toroidalmagnet heating. In all these cases the discharge last quite longerthan the resistive time (up to 40 times).

The LH current drive system at 3.7 or 5 GHz (6 MW) will generatea significant fraction of the total plasma current for the AT scenarios,guaranteeing the access, control and sustenance of the required cur-rent profiles. A study of the LH penetration and absorption has beenperformed in a parameter range typical of FAST AT scenarios [1,14].In the reference AT scenario the LH driven current is 0.65 MA, cor-responding to 22% of the total plasma current, while 38% is drivenby the bootstrap current. According to simulations this is enoughto produce a negligible evolution of the current profile during the

whole discharge.

An analysis of the global MHD stability for the long pulse ATscenarios has been performed using the MARS code [24] in order toinvestigate the possibility of stabilizing resistive wall modes (RWM)[1]. The ECRH system (4 MW) will provide enough power for MHD

minor disruption simulation by CREATE NL model.

1 ering

c(u

7

imptmb

vatuavthtcls

otpdcwpa

cpeTbvio

vgna

avip

tttc

rlasdt

[

568 G. Ramogida et al. / Fusion Engine

ontrol. Besides the stabilization of neoclassical tearing modesNTM) in long pulse AT scenarios, the ECRH system on FAST is alsosed for electron heating and current drive tasks at lower densities.

. Plasma position and shape control

The control of the plasma position and shape is a crucialssue as in every compact, elongated and high performance Toka-

ak as FAST. The capability of the poloidal field coil system, asresently designed, to provide an effective vertical stabilization ofhe plasma has been investigated using the CREATE NL response

odel [25], assuming axisymmetric deformable plasma describedy few global parameters.

The plasma chamber has been schematized as an Inconel 625essel, 25 mm thick, with a resistivity equals to 1.29 �� m at oper-ting temperature: the torus resistance results 62.6 ��, neglectinghe 3D effects of the ports. A stabilizing copper shell inside the vac-um vessel has been designed, optimizing its thickness (26 mm)nd location to provide a slowing down of the growth rate of theertical instability around 13 s−1 with a safe stability margin equalso 0.973. To avoid flux shielding during plasma break-down the shellas been toroidally segmented, providing the up–down connec-ion by the poloidal path around the ports, so the net total toroidalurrent flowing in it is zero. Fig. 9 describes the FAST reference equi-ibrium, modelled by CREATE NL code, together with the passivetabilizing structures: vessel and copper shell.

Preliminary analyses have been performed to study the controlf the plasma current, shape and position during the flat-top ofhe reference H-mode plasma scenario. The structure of the pro-osed controller consists in a feedback loop which controls theerivative of the vertical position (using CS2U-CS2L and PF3–PF4oil imbalance current) and a slower multivariable feedback loop,hich controls the plasma current, shape and position. The twoairs of coils selected for the vertical control will be fed by up–downnti-symmetric currents provided by a dedicated power supply.

The voltage for the vertical stabilization system is supplied by aonverter driven by simple saddle network controller driven by thelasma vertical velocity. The vertical diagnostics has been mod-lled as a first order low pass filter with a time constant of 3 ms.he closed loop system guarantees that, in the presence of a distur-ance, the plasma vertical velocity goes to zero, while the plasmaertical position of the current centroid is not recovered. The stabil-ty of the vertical velocity loop is guaranteed with a phase marginf about 65◦ and a gain margin of 5 dB.

The current and shape controller structure uses as controlledariables, besides plasma current, six linear combinations of 39aps (between the plasma separatrix and the plasma facing compo-ents), strike points and X-point descriptors, obtained using a SVDpproach.

The power supply system has been modelled, in a conservativepproach as a pure time delay of 10 ms: under this assumption theertical stabilization controller and the power supplies voltage lim-ts have been designed so as to guarantee a settling time for thelasma velocity of about 800 ms.

As far as the current/shape disturbances rejection (recovery ofhe gaps within 1 cm), the controller has been evaluated simulatinghe system response to a 1 cm plasma Vertical Displacement Event,o a 100 kA step in the plasma current and to a minor disruptionharacterized by a 20% fall in internal inductance and poloidal beta.

In all these cases, the recovery is guaranteed by the cur-ent/shape controller as presently designed, with a settling time

ess than 2 s. The maximum power required for this stabilization isbout 14 MW, in the range of the capabilities of the designed PFCystem, with the most demanding disturbances, VDE and minorisruption. The voltage and current required for the stabilization,ogether with the time evolution of the plasma position, velocity,

and Design 84 (2009) 1562–1569

current and gaps, are showed in Fig. 10 in the case of the minordisruption simulation.

8. Conclusions

The Fusion Advanced Studies Torus has been designed to pro-vide a European ITER Satellite facility able to explore Fast Particlephysics, to investigate ITER relevant Plasma Operations issues, tostudy the physics and test the technologies required to deal withlarge heat loads on ITER and DEMO plasma facing components, toinvestigate long lasting Advanced Tokamak regimes up to fully noninductive scenarios, to validate numerical simulation codes predic-tions of ITER fusion and burning plasma performance. FAST willbe able thus to address most of the EFDA strategic missions andto support the preparation of ITER operation scenarios by usingfast ions accelerated by heating and current drive systems, workingwith deuterium plasmas in a dimensionless parameter range closeto that of ITER.

FAST equilibrium configurations permit the preparation of ITERscenarios in a compact, cost-effective device still suitable to fulfilplasma conditions required to study burning plasma physics issuesin an integrated framework. The FAST flexibility in terms of bothperformance and physics that can be investigated is emphasizedby the variety of plasma scenarios that can be studied, from theextreme high performance H-mode to the full not inductive currentdriven scenario.

The feasibility of a proper plasma position and shape controlwith the current poloidal field system design has been also intro-duced, showing the possibility of guaranteeing a wide stabilityregion and of rejecting undesired shape modification.

References

[1] A. Pizzuto on behalf of the Italian Association, The Fusion Advanced StudiesTorus (FAST): a proposal for an ITER Satellite facility in support of the develop-ment of fusion energy, in: Proceedings of 22nd IAEA Fusion Energy Conference,Geneva, Switzerland, October 13–18, 2008; Nucl. Fusion, submitted for publi-cation.

[2] Report of the Fusion Facilities Review Panel, R&D Needs and Required Facilitiesfor the Development of Fusion as an Energy Source, Bruxelles, Belgium, October2008.

[3] F. Romanelli, A. Coletti, C. Gormezano, F. Lucci, A. Pizzuto, G.B. Righetti, et al.,Fusion Sci. Technol. 45 (2004) 483.

[4] The FAST Team, FAST Conceptual Study, Technical Report ENEA/FPN-FAST-RT-07/001, Frascati, Italy, 2008.

[5] K. Lackner, Comments Plasma Phys. Control. Fusion 15 (1994) 359.[6] P.J. Catto, D.A. Knoll, S.I. Krasheninnikov, Phys. Plasmas 3 (1996) 3191.[7] F. Zonca, S. Briguglio, L. Chen, G. Fogaccia, T.S. Hahm, A.V. Milovanov, et al.,

Plasma Phys. Control. Fusion 48 (2006) B15.[8] L. Chen, F. Zonca, Nucl. Fusion 47 (2007) S727.[9] F. Zonca, Int. J. Mod. Phys. A 23 (2008) 1165.

[10] F. Zonca, L. Chen, in: P.K. Shukla, B. Eliasson, L. Stenflo (Eds.), Frontiers in ModernPlasma Physics, CP1061, AIP, 2008, pp. 34–44.

[11] F. Zonca, L. Chen, in: O. Sauter, X. Garbet, E. Sindoni (Eds.), Theory of Fusionplasmas, CP1069, AIP, 2008, pp. 355–360.

12] L. Chen, Plasma Phys. Control. Fusion 50 (2008) 124001.[13] G. Maddaluno, R. Zagórski, V. Pericoli Ridolfini, M.L. Apicella, G. Calabrò, F.

Crisanti, et al., Edge Plasma Issues of the Tokamak FAST (Fusion Advanced Stud-ies Torus) in Reactor Relevant Conditions, in: Proceedings of 22nd IAEA FusionEnergy Conference, Geneva, Switzerland, October 13–18, 2008; Nucl. Fusion,submitted for publication.

[14] A. Cardinali, S. Briguglio, G. Calabrò, F. Crisanti, C. Di Troia, G. Fogaccia et al.,Minority Ions Acceleration by ICRH: a tool for investigating Burning PlasmaPhysics, in: Proceedings of 22nd IAEA Fusion Energy Conference, Geneva,Switzerland, October 13–18, 2008; Nucl. Fusion, submitted for publication.

[15] G. Calabrò, V. Cocilovo, F. Crisanti, A. Cucchiaro, F. Lucca, A. Marin, et al., Toroidalfield ripple reduction studies for ITER and FAST, Fusion Eng. Des., this issue,doi:10.1016/j.fusengdes.2009.01.030.

[16] M. Roccella, F. Lucca, R. Roccella, A. Pizzuto, G. Ramogida, A. Portone, et al.,Fusion Eng. Des. 82 (2007) 709–715.

[17] A. Cucchiaro, R. Albanese, G. Ambrosino, G. Brolatti, G. Calabrò, V. Cocilovoet al., Load Assembly Design of the FAST Machine, in: Proceedings of22nd IAEA Fusion Energy Conference, Geneva, Switzerland, October 13–18,2008.

[18] P. Barabaschi, The MAXFEA code, ITER Plasma Control Technical Meeting, Naka,Japan, 1993.

ering

[[

[

G. Ramogida et al. / Fusion Engine

19] F. Alladio, F. Crisanti, Nucl. Fusion 26 (1986) 1143.20] R. Albanese, G. Ambrosino, M. Ariola, A. Cenedese, F. Crisanti, G. De Tommasi,

et al., Fusion Eng. Des. 74 (2005) 627–632.21] G. Cenacchi, A. Taroni, JETTO: a free boundary plasma transport code (basic

version), Technical Report JET-IR(88)03, Culham, UK, 1988.

[

[[[

and Design 84 (2009) 1562–1569 1569

22] G. Vlad, M. Marinucci, F. Romanelli, A. Cherubini, M. Erba, V.V. Parail, et al., Nucl.Fusion 38 (1998) 557–570.

23] M. Brambilla, Plasma Phys. Control. Fusion 41 (1999) 1.24] A. Bondeson, G. Vlad, H. Lutjens, H. Phys. Fluids B 4 (1992) 1899.25] R. Albanese, F. Villone, Nucl. Fusion 38 (1998) 723.