the aries-at advanced tokamak, advanced …aries.ucsd.edu/lib/report/journal/fed/06-fed-80...f idaho...

Fusion Engineering and Design 80 (2006) 3–23

The ARIES-AT advanced tokamak,Advanced technology fusion power plant

Farrokh Najmabadia,∗, The ARIES Team: A. Abdoub, L. Brombergc, T. Brownd,V.C. Chane, M.C. Chue, F. Dahlgrend, L. El-Guebalyb, P. Heitzenroederd,D. Hendersonb, H.E. St. Johne, C.E. Kesseld, L.L. Laoe, G.R. Longhurstf,

S. Malangc, T.K. Maua, B.J. Merrillf , R.L. Miller a, E. Mogahedb,R.L. Mooref, T. Petriee, D.A. Pettif , P. Politzere, A.R. Raffraya,

D. Steinerg, I. Sviatoslavskyb, P. Syndere, G.M. Syaeblere, A.D. Turnbulle,M.S. Tillacka, L.M. Waganerh, X. Wanga, P. Weste, P. Wilsonb,

a University of California, Center for Energy Research, San Diego, 460 EBU-II, La Jolla, CA 92093-0417, USAb University of Wisconsin, Fusion Technology Institute, 1500 Engineering Drive, Madison, WI 53706-1687, USA

c Massachusetts Institute of Technology, Plasma & Fusion Center, Cambridge, MA 02139-4307, USAd Princeton Plasma Physics Laboratory, Princeton, NJ 08543, USA

e General Atomics, P.O. Box 85608, San Diego, CA 92186, USAf Idaho National Engineering and Environmental Laboratory, Fusion Safety Program, P.O. Box 1625, Idaho Falls, ID 83415, USA

g Rensselaer Polytechnic Institute, Troy, NY 12180, USAh The Boeing Company, P.O. Box 516, St. Louis, MO 63166, USA

ancedevings ofesma. A

sertthe

hen

Abstract

The ARIES-AT study was initiated to assess the potential of high-performance tokamak plasmas together with advtechnology in a fusion power plant and to identifying physics and technology areas with the highest leverage for achiattractive and competitive fusion power in order to guide fusion R&D. The 1000-MWe ARIES-AT design has a major radiu5.2 m, a minor radius of 1.3 m, a toroidalβ of 9.2% (βN = 5.4) and an on-axis field of 5.6 T. The plasma current is 13 MA and thcurrent-drive power is 35 MW. The ARIES-AT design uses the same physics basis as ARIES-RS, a reversed-shear pladistinct difference between ARIES-RS and ARIES-AT plasmas is the higher plasma elongation of ARIES-AT (κx = 2.2) whichis the result of a “thinner” blanket leading to a large increase in plasmaβ to 9.2% (compared to 5% for ARIES-RS) with onlya slightly higherβN. ARIES-AT blanket is a simple, low-pressure design consisting of SiC composite boxes with a SiC infor flow distribution that does not carry any structural load. The breeding coolant (Pb–17Li) enters the fusion core frombottom, and cools the first wall while traveling in the poloidal direction to the top of the blanket module. The coolant t

∗ Corresponding author. Tel.: +1 858 534 7869; fax: +1 858 822 2120.E-mail address: [email protected] (F. Najmabadi).

0920-3796/$ – see front matter © 2005 Elsevier B.V. All rights reserved.doi:10.1016/j.fusengdes.2005.11.003

4 F. Najmabadi et al. / Fusion Engineering and Design 80 (2006) 3–23

returns through the blanket channel at a low speed and is superheated to∼1100◦C. As most of the fusion power is depositeddirectly into the breeding coolant, this method leads to a high coolant outlet temperature while keeping the temperature of theSiC structure as well as interface between SiC structure and Pb–17Li to about 1000◦C. This blanket is well matched to anadvanced Brayton power cycle, leading to an overall thermal efficiency of∼59%. The very low afterheat in SiC compositesresults in exceptional safety and waste disposal characteristics. All of the fusion core components qualify for shallow land burialunder U.S. regulations (furthermore,∼90% of components qualify as Class-A waste, the lowest level). The ARIES-AT studyshows that the combination of advanced tokamak modes and advanced technology leads to an attractive fusion power plant withexcellent safety and environmental characteristics and with a cost of electricity (4.7 ¢/kWh), which is competitive with thoseprojected for other sources of energy.© 2005 Elsevier B.V. All rights reserved.

Keywords: Fusion power plant; Tokamak; Reversed shear plasma

1. Introduction

The ARIES program research aims at establishingthe economic, safety, and environmental potential offusion power plants, and at identifying physics andtechnology areas with the highest leverage for achiev-ing attractive and competitive fusion power in order toguide fusion R&D. The ARIES Team is a U.S. nationaleffort with participation from national laboratories,universities and the industry, and with strong inter-national collaborations. The Team performs detailedphysics and engineering analyses using the most cur-rent and detailed models available, and then uses theresults to perform optimization and trade studies via acost-based systems code.

Progress in tokamak physics during the past decadeFig. 1. The ARIES-AT fusion power core.

-

l

has been remarkable. Our vision of tokamak powerplants in the 1980s, large devices operating in thepulsed mode, has been replaced by high-performance,advanced tokamak modes today. During this period, theARIES Team has studied a variety of tokamak powerplants. Continuation of research has allowed us to applylessons learned from each ARIES design to the next andexplore the impact of different extrapolations in physicsand technology on the overall attractiveness of the toka-mak concept. The ARIES-AT is the latest tokamakpower plant study undertaken by the ARIES program

for fusion power plants and describes the key R&D topics. A complete discussion of ARIES-AT systems canbe found in the accompanying papers in this speciaissue[1–8]. Figs. 1 and 2show the cross-section ofARIES-AT.

2. Design directions

and represents the latest in the evolution of our thinkingon the potential of the tokamak concept as an attractivefusion power plant. In Section2, we will explore thed thel igns.S ly-s ntt nts

The ARIES research aims at establishing the eco-nomic, safety, and environmental potential of fusionpower plants. A set of top-level goals and requirementsfor fusion power has been developed in collabora-tions with representatives from US electric utilities andindustries[9] and has been framed in a quantitativeway [10,11] to help establish the minimum necessaryfeatures of a fusion power plant that would lead to

esign directions for ARIES-AT as derived fromessons that we have learned from previous desections3–5present overviews of our detailed anais and trade-off studies. Section6 examines the exteo which ARIES-AT has met the top-level requireme

F. Najmabadi et al. / Fusion Engineering and Design 80 (2006) 3–23 5

Fig. 2. Cross-section of ARIES-AT fusion power core.

its likely introduction into the US and world elec-tricity market. These goals can be divided into threegeneral categories: (1) power plant should have aneconomically competitive life-cycle cost of electricity,(2) it should gain public acceptance by having excel-lent safety and environmental characteristics, and (3) itshould have operational reliability and high availabil-ity. Similar to all ARIES designs, ARIES-AT researchwas directed toward achieving these goals (as describedbelow) while maintaining a balance between attractive-ness (as measured by the customer requirements) andcredibility (as measured by the extent of extrapolationfrom the present database).

2.1. Plasma physics

The plasma regime of operation is optimized mainlyto reduce the plant cost (goal 1 above) through (a)reducing the recirculating power needed to maintain theplasma, (b) reducing the plant size by increasing plasmapower density, and (c) reducing the cost of plasma sup-port technologies such as magnets.

In the 1980s, tokamak fusion power plants werebased on large devices operating in the pulsed-plasmamode. While the benefits of steady-state operation wereclearly understood at the time, no credible steady-state tokamak plasma operation was identified. Thisis due to the fact that the non-inductive current drive,necessary for sustaining plasma current, is quite inef-ficient, leading to a very high recirculating powerfraction (approaching unity). Operation with a highbootstrap current fraction as the approach to steady-state operation was proposed by the ARIES-I[12]and SSTR[13] studies simultaneously and indepen-dently in 1990. In order to reduce the current-drivepower, the plasma current is reduced while the boot-strap fraction is maximized. In this manner, the amountof plasma current that should be driven by exter-nal means and the associated current-drive powerare reduced and the power plant recirculating powerfraction becomes acceptable (∼10–30%). In the first-stability regime (i.e., discharges with a monotonicqprofile and stable to kink modes without a conduct-ing wall) this can be accomplished by operating witha moderately high plasma aspect ratio (A = 1/ε ∼ 4.5)to reduce the plasma current (I ∼ 10 MA) at a rel-atively high poloidal beta (βp ∼ 0.6). This modeof operation, however, has a low value of plasmaβ ∼ 2% becauseβp (which determines the boot-strap fraction) is related to the achievable plasmaβ

through

ε

wp -i ).I en-s tht nsitya iledM t ab ithaM ag-n out∼ ana evero ger

βpβ

ε=

(βN

20

)2

S (1)

hereS = (1 +κ2)/2 is the plasma shape factor,κ thelasma elongation, andβN =β a I/B is the normal

zedβ (typically βN < 3.5 in the first stability regimen effect, there is a trade-off between the power dity (lower β) and recirculating power fraction wihe optimum system having a moderate power dend a moderate recirculating power fraction. DetaHD and current-drive analysis have shown thaootstrap fraction of∼60% can be achieved but wslightly lowerβN ∼ 3 (and a low plasmaβ ∼ 2%).ost of the driven current is located near the metic axis, requiring a current-drive power of ab100–200 MW delivered to the plasma. There ismple experimental database for this regime, howperation in discharges with durations much lon


than the current diffusion time as well as in a burningplasma are needed.

A comparison of major parameters of a 1000-MWesteady-state and pulsed-plasma tokamak power plantsusing the same physics and technology basis was gen-erated as part of the Starlite study[10]. Our resultsindicate that both pulsed-plasma and steady-state firststability devices optimize in the same physics regime:low current, highA, and moderate bootstrap currentfraction. Because of the long burn time in a pulsed-plasma power plant, the plasma is essentially in steadystate. As such, the physics needs of pulsed-plasmaand steady-state first-stability devices are identical (theonly difference, the physics of non-inductive currentdrive, is well established). Generally, the low recir-culating power fraction of pulsed-plasma operation ismore than offset by the lower power density, the largersize, and the expensive PF system. As such, pulsedpower plants tend to be more expensive than steady-state ones—a pulsed-plasma power plant is inferior toa steady-state one based on similar physics and tech-nology extrapolations.

Starting from a steady-state, first-stability device,power plant economics can be improved by increas-ing the fusion power density and decreasing the recir-culating power fraction. It should be noted that theeconomics improvement with increasing fusion powerdensity or decreasing recirculating power fraction “sat-urates” after a certain limit. This is due to the fact that

at low power density, any reduction in the plasma size(increase in power density and wall loading) propor-tionally decreases the volume of fusion core (blan-ket/shield/coils) that surrounds the plasma (and thesystem cost). However, when the plasma size is com-parable and/or smaller than the blanket and shieldthickness, the fusion core volume (and its cost) doesnot change appreciably with reduced plasma size andincreased power density[14]. For 1000-MWe powerplants, economic improvements with increased wallloading saturates at an average neutron wall load of∼4 MW/m2. As a reference, a first-stability steady-state device with a 16-T magnet technology achieves aneutron wall loading of∼1.5 MW/m2. Similarly, thereis no economic benefit when the recirculating powerfraction falls below a few percent.

High-field magnets can be used to increase thefusion power density of a first-stability device as isshown inTable 1. ARIES-I featured a 19-T cryogenictoroidal-field system. Because the fusion power den-sity scales asβ2B4, the impact of increased toroidal-field strength is quite dramatic. The increase in fusionpower density leads to a substantially smaller deviceand a slightly lower current-drive power, both factorscombine to help improve power plant economics eventhough the magnet system is more expensive. Alterna-tively, one could seek plasma regimes with higherβNandβ that have a large bootstrap current fraction. In1980s, second-stability operation with a high plasmaβ

TM nts

High-fi

M 6.75P 4.5P 1.8β 2 (3.0P 19O 11A 2.5I 1.9P 10B 0.57C 202R 0.28T 0.49C 8.2

able 1ajor parameters of 1000-MWe advanced tokamak power pla

First stability

ARIES-I

ajor radius (m) 8.0lasma aspect ratio 4.5lasma elongation,κ 1.8(%) (βN) 2 (2.9)eak field at the coil (T) 16n-axis field (T) 9.0verage wall load (MW/m2) 1.5

TER89-P multiplier 1.7lasma current (MA) 12.6ootstrap current fraction 0.57urrent-driver power (MW) 237ecirculating power fraction 0.29hermal efficiency 0.46ost of electricity (¢/kWh) 10

Reverse-shear

eld: ARIES-I ARIES-RS ARIES-AT

5.5 5.24.0 4.0

1.9 2.1) 5 (4.8) 9.2 (5.4)

16 11.58 5.8

4 3.32.3 2.011.3 130.88 0.91

80 350.17 0.14

0.46 0.597.5 5


was the focus of the theoretical research. However, theARIES-II/IV research[15] showed that high-β second-stability operation leads to bootstrap current overdriveand misalignment of the bootstrap current profile withthe equilibrium current profile, resulting in a largecurrent-drive power. Therefore, second-stability oper-ation results in only marginal improvements in powerplant attractiveness.

Reversed-shear[16] plasmas were proposed in mid1990s. The benefits of this configuration are that itachieves both highβN andβ; obtains large bootstrapfractions with very good current profile alignment; andfeatures an internal transport barrier necessary to sus-tain the peaked pressure profiles that is consistent withthe highβ and high bootstrap current. The negativecentral magnetic shear is responsible for stability to bal-looning modes. A conducting wall, however, is neces-sary for stabilization of external kink modes—and theresistive-wall modes should be stabilized with plasmarotation and/or feedback coils.

Reversed-shear regime of operation was analyzedin the ARIES-RS study[11]. A large database of stableMHD equilibria was generated and their non-inductivecurrent drive needs were calculated. This databasewas then used by the ARIES Systems Code to arriveat the optimum power plant configuration. Analysesshowed that[11] to zeroth order, the cost of elec-tricity is insensitive to the plasma aspect ratio in therange of 2.5–4.5 (the lower plasmaβ at the higherAis compensated by a higher toroidal-field strength ona atioo ringc

ka-m en at ncet per-i beep aly-sR asee onsl asa er-e

( ofof

current-drive power was necessary to drive a rela-tively “small” plasma current mid-plasma and nearthe plasma edge. In ARIES-AT, effort was made toeliminate the need for non-inductive current drivein mid-plasma. As a result, while the bootstrapfraction was only increased slightly in ARIES-ATto 91%, the current-drive power was reduced by afactor of∼2.

(2) Typically the plasma profile optimization (e.g.,MHD stability, bootstrap current, and current-driveanalyses) is performed using fixed boundary equi-libria. Plasma boundaries are forced to match the95% flux surface boundary from the free-boundaryequilibrium. For ARIES-AT, plasma boundariesused in the optimization are forced to coincide withthe 99% flux surface from the free-boundary equi-librium. This has resulted in an increase inβN asthe plasma shaping is stronger as one approachesthe separatrix. This approach also provides a strictself-consistency between the free-boundary equi-libria generated by the poloidal-field coils and thefixed-boundary equilibria used for MHD stabilityand current-drive analyses.

(3) A distinct difference between ARIES-RS andARIES-AT plasmas is the higher plasma elonga-tion of ARIES-AT (κ = 2.1) which is the result ofthe ARIES-AT “thinner” blanket. In ARIES toka-mak designs, passive vertical stabilization shellsare located behind the blanket. As such, the blan-

heeto

r

for-

heandis-r-esasmaome oflf-

xis and a lower current-drive power). An aspect rf 4 was chosen for ARIES-RS based on engineeonsiderations.

Based on the results from our previous ARIES toak studies, the reversed-shear regime was chos

he reference tokamak discharge for ARIES-AT. Sihe ARIES-RS study, extensive theoretical and exmental studies of reversed-shear plasmas haveerformed. As such, considerably more physics anis was performed for ARIES-AT (see Section3 andefs. [1,2]) and, as a result, a more credible cxists for the ARIES-AT discharge. In addition, less

earned from the ARIES-RS plasma optimization wpplied to the ARIES-AT research. The major diffnces between ARIES-RS and ARIES-AT are:

1) ARIES-RS has a bootstrap current fraction∼88%. Unfortunately, a substantial amount

s

n

ket thickness helps set the plasma elongation. Thigher plasma elongation of ARIES-AT and thcorresponding larger plasma triangularity leda large increase in plasmaβ to 9.2% (comparedto 5% for ARIES-RS) with only a slightly higheβN = 5.4 (βN = 4.8 for ARIES-RS).

(4) More in-depth analysis has been performedARIES-AT (compared to ARIES-RS). For example, in the MHD area, we have considered teffects of edge density/temperature gradientsanalyzed the stability requirements for the restive wall modes (RWM), the neoclassical teaing modes (NTM), and the edge localized mod(ELM). Modern transport analysis tools, suchGLF23, were used to study consistency of plasprofiles. Radiative and conductive heat losses frthe plasma and edge regions under a wide rangimpurity doping were considered to arrive at a seconsistent solution for the divertor system.


The major plasma properties of ARIES-AT are alsoshown inTable 1. As in the ARES-RS study, we foundthat the cost of electricity is insensitive to the plasmaaspect ratio in the range of 2.5–4.5. An aspect ratio of 4was chosen for ARIES-AT based on engineering con-siderations (more space, more uniform load on compo-nents, etc.). An important observation fromTable 1isthat ARIES-RS with aβ = 5% and a 16-T Nb3Sn super-conductor achieved an average neutron wall loading of4 MW/m2. As discussed above, the power plant eco-nomic performance is insensitive to increased powerdensity above this value. This can be seen in the ARIES-AT design point (Table 1) where the increased plasmaβ

is used to reduce the toroidal field requirement insteadof increasing the power density and reducing the sys-tem size.

2.2. Fusion technologies

Fusion power technologies (first wall, blanket,shield, and power conversion system) play an impor-tant role in the attractiveness of a power plant. Notonly must these components perform the vital func-tion of tritium breeding and power recovery, but thechoice of materials and configurations associated withtheir designs also have a major impact on the safetyand environmental characteristics of the power plant.In addition, the economics of the power plant is directlytied to the performance of fusion power technologies,i.e., thermal conversion efficiency. Three classes ofl der-a els,v pro-g ed ont

ate-r inA ono saryf owi te-r ets inw ES-I re of9 har-a twos ationi all,

resulted in small separations between the SiC compos-ite tube sheets, making manufacturing of such a blanketdifficult. Second, the coolant outlet temperature wastoo “low” for a reasonable thermal conversion effi-ciency from the Brayton cycle envisioned at the time.As such, a Rankine cycle was used in ARIES-I leadingto a thermal conversion efficiency of∼46%. The per-formance of the ARIES-I blanket was disappointing inthis regard as the high coolant temperature did not leadto a high thermal conversion efficiency.

Self-cooled liquid breeder blankets are inherentlysimpler than blankets with solid breeders as thefusion neutron energy is deposited directly in thebreeder/coolant. ARIES-RS featured a self-cooleddesign with liquid lithium as the coolant and breederand vanadium alloy as the structural material. The high-temperature capability of vanadium alloys allowed acoolant outlet temperature of∼700◦C and a thermalconversion of efficiency of 49% (better than ARIES-Idue to a lower pumping power). Vanadium is a low-activation material and all components of ARIES-RSqualified for shallow land burial under NRC regula-tions. Lithium/vanadium blankets, however, have twomain disadvantages: (1) the vanadium structure mustbe coated with an electrical insulator in order to havea reasonable MHD pressure drop and pumping power,and such an insulating coating has not yet been demon-strated; (2) care should be taken in design and opera-tion in order to avoid safety concerns associated withlithium fires.

baseo ed,h at-i lsw er-m sedo turef xi-ma ano ema et.T ikeg eel.P tings ot-t the

ow-activation structural materials are under consition for fusion applications: ferritic/martensitic steanadium alloys, and SiC composites. The ARIESram has studied a number of blanket designs bas

hese materials.Use of SiC composites as advanced structural m

ial for fusion power plants was first proposedRIES-I [12]. SiC composites offer a combinatif high-performance at high temperature (neces

or high thermal efficiency) as well as a very lnduced activation. In ARIES-I, solid breeder maials are located between SiC composite tube shehich flows the high-pressure He coolant. The ARIdesign achieved a high coolant outlet temperatu00◦C and excellent safety and environmental ccteristics. The ARIES-I blanket, however, hadhortcomings. First, the large nuclear heat genern the solid breeder, especially close to the first w

Ferritic/martensitic steels have the largest dataf all fusion low-activation materials. It was perceivowever, that the relatively “low” maximum oper

ng temperature (∼550◦C) of ferritic/martensitic steeould lead to a blanket with an unacceptably low thal conversion efficiency. This perception was ban the assumption that the coolant outlet tempera

rom a blanket would always be lower than the maum structure temperature. ARIES-ST[17] proposedn innovative dual-coolant ferritic steel blanket withutlet coolant temperature of∼700◦C (higher than thaximum steel structure temperature of 550◦C) andthermal efficiency similar to those of Li/V blankhe ARIES-ST first wall and blanket have a box-leometry and are made of ferritic/martensitic stb–17Li is used as breeder and coolant, circulalowly in the blanket: it enters the blanket at the bom in the first row of blanket boxes, traverses to


top and return through the last three rows. Becausethe fusion neutron energy is directly deposited in thePb–17Li breeder/coolant, it reaches an exit tempera-ture of 700◦C. SiC inserts are used inside each boxto help keep the heat conducted to the structural mate-rial from the coolant at a low level and maintain theferritic/martensitic steel below its maximum operatingtemperature. The walls of the steel box, including thefirst wall, are cooled by high-pressure helium (flowinginternally). Helium coolant exits the first wall/blanketat 500◦C and is then superheated by Pb–17Li coolantto ∼700◦C in a heat exchanger. Recent interest in gas-turbine power generators has resulted in large R&Defforts in high-efficiency gas cycles. In particular, thedevelopment of high-efficiency recuperators in a Bray-ton cycle provides the possibility of high efficiencyeven with a relatively low coolant temperature[18].ARIES-ST utilizes such a Brayton cycle and achievesa thermal conversion efficiency of 46%.

Use of liquid breeder and the development of mod-ern Brayton cycles directly address the shortcomingsof the ARIES-I SiC blanket. As such, the ARIES-ATstudy revisited the use of SiC composite as blanketstructural material (Section4and Refs.[3–5]) utilizingthe same innovation as those in ARIES-ST for rais-ing the coolant temperature. The ARIES-AT blanket isa simple, low-pressure design consisting of SiC com-posite boxes. Each box contains a SiC insert for flowdistribution that does not carry any structural load. Thecoolant enters the fusion core from the bottom, andc c-t henr eeda np hodl pingt rfacebT tonp y of∼

sev-e m-p fieldc ra-t lityo sim-p hus,

reduce the cost. As such, we considered HTS as thebase line for ARIES-AT. Since, the ARIES-AT designpoint optimized at a maximum field of 11 T, we didnot utilize the high-temperature capability of HTS. Assuch, Nb3Sn superconductors can equally be used (infact, with some cost penalty in increasing the size ofthe fusion core, NbTi superconductor will also be fea-sible). Section4.4and Ref.[6] summarize our findingson the impact of HTS on the attractiveness of fusionpower plants.

3. Plasma physics

Plasma physics analyses of ARIES-AT were per-formed in an iterative manner with systems analysisand engineering design and with an increased levelof detail and self-consistency. The ARIES-AT plasmaoptimization was started by developing a databaseof fixed-boundary equilibria for which the total par-allel current and density and pressure profiles wereprescribed. In each case, the bootstrap current pro-file was calculated and monitored for alignment withthe prescribed current density profile. This approachallowed us to efficiently scan the MHD stability ofthese equilibria. Current-drive calculations were per-formed for selected optimized equilibria. The resul-tant plasma equilibria and current-drive database wasthen used by the systems code to arrive at the initialARIES-AT design point. More detailed analysis wasp ATd ilib-r filea rnalc bridc strapc e ofe venc ity.T nsa rrivea int.I xed-b thef uxs encyb y thep ria

ools the first wall while traveling in the poloidal direion to the top of the blanket module. The coolant teturns through the blanket channel at very low spnd is superheated to∼1100◦C. As most of the fusioower is deposited directly into the coolant, this met

eads to a high coolant outlet temperature while keehe SiC structure temperature as well as the inteetween SiC structure and Pb–17Li to about 1000◦C.his blanket is well matched to an advanced Brayower cycle, leading to an overall thermal efficienc59%.High-temperature superconductors (HTS) offer

ral attractive features for fusion applications coared to cryogenic ones: (a) a higher magneticapability, (b) operation at liquid nitrogen tempeure and simplifying cryo-plant, and (c) the possibif using advanced manufacturing techniques tolify the manufacture of superconductor coils and, t

erformed at this stage around this initial ARIES-esign point. Self-consistent fixed-boundary equia were developed in which only the pressure prond the parallel current density profiles from exteurrent-drive sources (i.e., fast-wave and lower-hyurrent-drive systems) were prescribed. The booturrent-density profile was calculated at each stagquilibrium and was added to the externally driurrent to provide the total parallel current denshrough iteration with the current-drive calculationd system analysis, this approach was used to at the final self-consistent ARIES-AT design po

n this analysis, plasma boundaries used in the fioundary equilibrium calculations were taken from

ree-boundary equilibrium at the 99% poloidal flurface. This approach provides strict self-consistetween the free-boundary equilibria generated boloidal-field coils and the fixed-boundary equilib


used for MHD stability and current-drive analyses. Inaddition, this approach can have a strong impact onthe calculated stableβ values since the plasma shapingbecomes stronger as one approaches the separatrix[2].

The robustness of the ARIES-AT equilibrium wasstudied by considering the impact of a variety of den-sity/temperature gradients and current values at theplasma edge. Stability against the resistive wall modes(RWM), the neoclassical tearing modes (NTM), and theedge localized modes (ELM) were also analyzed. Thetransport properties were investigated using the GLF23transport model to ensure self-consistency of assumeddensity/temperature profiles. This section summarizesthe physics analysis in support of ARIES-AT. Moredetails can be found in Refs.[1,2]. The ARIES-ATequilibria are shown inFig. 3 and the major plasmaparameters are summarized inTable 2.

3.1. MHD stability

Studies were performed to determine the impact ofplasma elongation, triangularity, pressure and currentprofiles, and the kink-stabilization shell position onplasma performance.Fig. 4 shows the maximum nor-

Table 2Major plasma parameters of ARIES-AT

Major radius (m) 5.2Minor Radius (m) 1.3Plasma current (MA) 12.8On-axis field (T) 5.8Plasma elongation,κx 2.20Plasma triangularity,δx 0.90Toroidalβa (%) 9.2Normalized beta,βN

a 5.4Poloidal beta,βp 2.28On-axis safety factor,q0 3.50Minimum safety factor,qmin 2.40Plasma-edge safety factor,qe 3.70Plasma-edge safety factor,q* 1.85Internal inductance,li (3) 0.29Location of kink-stabilization shellb 0.33Bootstrap current fraction 0.91Current-driver power (MW) 35Electron and ion temperature (keV) 18Temperature peaking factor (T0/〈T〉) 1.72Electron density (m−3) 2.15Total ion density (m−3) 1.71Density peaking factor (n0/〈n〉) 1.34ITER89-P multiplier 2.0

a ARIES-AT operates at 90% of maximum theoreticalβ limit.b Distance from plasma edge normalized to minor radius.

F tor profile, parallel current density profiles, poloidal flux contours, and toroidalc

ig. 3. The reference ARIES-AT plasma equilibrium: safety facurrent density profiles.


Fig. 4. Normalized beta values for ballooning modes as function ofplasma elongation and triangularity. Note that there is an optimumtriangularity for each value of plasma elongation.

malized beta (βN) values for high-n ballooning modesas functions of plasma elongation and triangularity. Ineach case, the plasma pressure profile was optimizedto provide a high stableβ and a large well-alignedbootstrap current fraction simultaneously (larger sta-ble βN values are possible but at the expense of amuch larger current-drive power).Fig. 4shows thatβNincreases with plasma elongation only if the triangular-ity is increased simultaneously—i.e., there is an opti-mum triangularity for each value of plasma elongation.

The low-n external kink stability limits of the aboveequilibria were analyzed and the marginal wall locationwas found for toroidal mode number up to 6 (n = 6–4modes require a closer wall location). Our balloon-ing and kink stability analyses led to an elongation ofκx = 2.2 and a triangularity ofδx = 0.9 for ARIES-ATunder the requirement that the kink stabilization shellbe located behind the blanket (seeFig. 2) at 0.33 timesthe minor radius. Note that this distance is measuredbetween the shell position and the 99% flux surface(approximately the plasma separatrix).

Increasing the plasma elongation improvesβ, butmakes the plasma more unstable to axisymmetricmodes. The maximum elongation is limited by thepassive-stabilization conducting structure and the feed-

back control system. It is desirable to utilize thekink-stabilization shell for passive stabilization ofaxisymmetric modes. The conducting structure isdesigned to provide a stability safety margin,fs = 1.2(fs = 1 +τg/τL/R, where τg is the vertical instabilitygrowth time andτL/R is the longest up–down asym-metric time constant of the structure). The conductingstructure is made of tungsten, 4-cm-thick, and oper-ating at 1100◦C so direct cooling is not required.Feedback control simulations for the plasma verticalposition were performed using the Tokamak Simula-tion Code[19]. A peak feedback power of 30 MVA isneeded for 1 cm random vertical displacement (notethat∼85% of this power is reactive).

As the kink-stabilization shell is not an ideal wall,resistive wall modes (RWM) can grow although thegrowth time is reduced to values below the shell timeconstant. While physics of RWM is not fully under-stood, it is known that plasma rotation and/or feedbackcontrol coils can help stabilize these modes. Stabi-lization of RWM in the ARIES-AT plasma by plasmarotation was evaluated using the MARS stability code[20]. The critical rotation frequency for stabilization ofn = 1 RWM modes is found to between 0.07 and 0.08 ofthe Alfven transit frequency and the stability windowfor the shell location is between 0.30 and 0.45 of theplasma radius. Forn = 2 mode, the rotation requiredis substantially reduced. Our analysis indicates thatneutral beam injection is not a practical mean for gen-erating the necessary plasma rotation in ARIES-AT[1].A n-s alllb bys ee r-r

o-c d. InA res-o eb iteda n-t esst pro-fi e int nt ofm sys-

nother possible approach is utilization of “flux coerving intelligent coils” located at the resistive wocation (“intelligent shell feedback” scheme)[1]. Sta-ilization of RWM by control coils were examinedcaling the DIII-D C-coils[21]. Six saddle coils armployed (seeFig. 2) each carrying a maximum cuent of 50 kA-turns[1].

The stability of ARIES-AT with respect to nelassical tearing modes has also been considereRIES-AT, ELMs may seed perturbations to excitenantq = m/n = 5/2 mode atρ = 0.92 location. The largootstrap current can produce a large island if excnd allowed to grow[1]. However, current profile co

rol with radially localized current drive could supprhe island. It seems that a combination of currentle modification (to make the free energy availablhe current profile more negative) and replacemeissing bootstrap (by the external current-drive


tem) is necessary and most effective. Design of the RFsystem for this purpose, however, has been left for thefuture.

The GLF23 drift-wave based model[22] was usedto predict the transport properties of the ARIES-ATplasma and to analyze the consistency of the assumedplasma profiles with transport processes. The GLF23model predict global energy confinement timeswhich exceeds the ARIES-AT design requirementif the density profile is as peaked as assumed. Theenergy confinement improvement is primarily due toShafranov shift stabilization of drift waves. Moreover,the computed pressure profile is close to the onefound for optimum MHD stability. Results from thedetailed GLF23 simulation of the ARIES-AT plasmaare reported in Ref.[1].

3.2. Current drive

The ARIES-RS plasma required three current-drivesystems: ICRF fast waves for central current drive;high-frequency fast wave (HFFW) for driving currentin mid plasma; and lower hybrid for edge plasma cur-rent drive. The HFFW drive was necessary as the lowerhybrid waves could not penetrate the inner parts ofthe plasma. As a result, a large current-drive powerwas necessary to drive a relatively “small” plasma cur-rent at mid-plasma. In ARIES-AT, effort was made toeliminate the need for non-inductive current drive inmid-plasma by optimizing plasma profiles. As a result,w htlyi asr t-d wase

erep hee ncepa ray-t ep ar-i mac isu nc ma( xisc es.

Fig. 5. Profiles of bootstrap (B), diamagnetic (D), and RF-driven(C) current densities. The sum of these components (T) matcheswell with the reference equilibrium current density profile (E).

The current-drive system details are given inTable 3.In order to minimize intrusion in the fusion powercore, launchers with high power density capabilitieswere utilized: folded wave guides for fast wave andactive/passive waveguide grill for lower hybrid. All RFlaunchers fits in one blanket module and cover less than0.5% of the first wall area.

Using neutral beams (NBI) for driving current inthe outer region of the plasma was considered as it wasperceived that such a NBI system can also help gen-erate the necessary plasma rotation. A 120-keV NBIsystem at a pivot angle of 70◦ can penetrate the outerregion of the plasma and drive the necessary current.The needed NBI power is∼50 MW. Unfortunately,analysis indicated that this NBI system will not gen-erate the necessary plasma rotation and this option wasnot pursued.

3.3. Divertor physics

Because of the high power density of ARIES-AT, theheat fluxes on the divertor structures can be high. Note

Table 3The ARIES-AT current-drive system

System Frequency N|| Power(MW)

Driven current(MA)

ICRF fast wave 96 MHz 2 4.7 0.15

Lower hybrid 3.6 GHz 1.65 3.1 0.15

hile the bootstrap fraction was only increased slign ARIES-AT to 0.91, the current-drive power weduced by a factor of∼2. More importantly, a currenrive system with its associated launcher systemliminated.

A large number of current-drive calculations werformed in support of plasma optimizations. Tffect of impurities injected in the plasma to enhalasma radiation (see Section3.3) was taken intoccount. Calculations were performed using the

racing code Curray[23]. Fig. 5 shows the referenclasma equilibrium current density profile with the v

ous current-drive components. About 91% of plasurrent is self-driven. Lower hybrid current drivetilized to drive the bulk (1.1 MA) of the RF-driveurrent which is located in the outer part of the plaslarger than 80% of minor radius). A small on-aomponent (0.15 MA) is driven by ICRF fast wav

3.6 GHz 2.0 4.4 0.23.6 GHz 2.5 8.2 0.33.6 GHz 3.5 8.9 0.22.5 GHz 5.0 12.4 0.15


that in general, the heat flux capability of the first wallwould be substantially smaller than that of the divertorbecause of the long cooling path of the inboard firstwall. (In the ARIES-AT design, the heat flux capabilityof the first wall and divertor are, respectively,∼0.45 and∼5 MW/m2.) As such, “radiative mantle” scenarios inwhich almost all of the plasma energy is radiated moreor less uniformly on the first wall usually lead to anunacceptable solution. A combination of radiation inthe core and edge/divertor plasmas is required.

To get an “upper” bound estimate of heat fluxes onthe divertor, we considered a case with no impurityradiation (∼30% of plasma energy is radiated throughbremsstrahlung). Divertor plates were assumed to havean inclination of 10◦. Assuming a ratio of 8:1 for thepower flow into the outboard scare-off layer comparedto the inboard andλq|| = 1.2 cm, the heat fluxes on theinner and outer divertor plates were estimated at 3.3 and13.7 MW/m2, respectively. Thus, some form of impu-rity injection is required to reduce the peak heat flux onthe outboard divertor. (Note thatλq|| is calculated basedon formulation in Ref.[24] with ARIES-AT parame-ters which givesλq|| of roughly 1.2 cm for L-mode and2.1 cm for H-mode.)

Because of the relatively low heat flux limit of theARIES-AT first wall, the radiation fraction in the coreshould be below 36%. In this case,∼43% of powerin the divertor should be radiated. The MIST[25] andSTRAHL [26] codes are used to model impurity trans-port by injection of Ne, Ar, or Kr, and the resultant radi-a ileda nrT arebl trolo w-e

4

truc-t etd lowi lantt malc en-

tal goals (because of the very low induced activation).Since SiC composites were first proposed in ARIES-I,substantial R&D have been performed and new designideas have been developed. While the ARIES-I outletcoolant temperature was high (900◦C), the best possi-ble thermal cycle available at the time (a Rankine cycle)led to a thermal conversion efficiency of 49%. Recentadvances in low-cost recuperators make Brayton cyclesvery attractive even at a relatively low coolant temper-ature (700◦C) as was shown in the ARIES-ST study[17]. At the higher temperatures made possible withSiC-based blanket (∼1100◦C), the Brayton cycle effi-ciency approaches 59%[18] making these types ofblanket very attractive for fusion application. Sepa-rately, the ARIES-I design used solid breeders, whichled to a complicated nested shell design. Use of a liquidbreeder such as Pb–17Li simplifies the blanket consid-erably. Both directions were pursued in ARIES-AT.

Attempts were made to simplify manufacturing ofthe fusion core by utilizing a low-pressure design withsimple geometry for blanket modules. In addition,the blanket and shield were designed in the form ofintegrated sectors to ensure rapid replacement of com-ponents (maximizing plant availability). Waste mini-mization and additional cost savings were also achievedby radially subdividing the blanket modules into twozones: a replaceable zone and a life of plant zone.

4.1. First wall, blanket, and shield

oft an-k terst thep le.T nnela -ST,t oledw lan-k theb -e tureo eels

ssi-b im-p lyP esire

tion in the ARIES-AT core and edge plasmas. Detanalysis can be found in Ref.[1]. In all cases, the desigequirements could be met with small increase inZeff.he results[1] indicate that higher Z rare gasesetter as they lead to higher radiation at lowerZeff and

ower fuel dilution. The dynamics of feedback conf the radiated power using impurity injection, hover, have not been investigated.

. Fusion power core engineering

Silicon-carbide composite is selected as the sural material of the ARIES-AT first wall and blankue to its high-temperature capability and its very

nduced activation. This material helps the power po meet its economic goals (through a high theronversion efficiency) and its safety and environm

The ARIES-AT blanket concept is similar to thathe dual-coolant ARIES-ST. The first wall and blet is a box-like structure. The breeding coolant enhe fusion core from the bottom and travels inoloidal direction to the top of the blanket moduhe coolant then returns through the blanket chat very low speed and is superheated. In ARIES

he ferritic/martensitic structure was separately coith He gas. The separate cooling circuit for the bet box structure as well as the SiC insert allowedreeding coolant in ARIES-ST to reach 700◦C, considrably higher than the maximum allowable temperaf ∼550◦C assumed for the ferritic/martensitic sttructure.

While the same dual-coolant arrangement is pole for a blanket with SiC composite structures, a sler design was chosen for ARIES-AT utilizing onb–17Li as breeding coolant because of: (1) the d


Fig. 6. The ARIES-AT first wall and blanket. (a) Each sector comprised of 12 modules arranged in two rows, (b) cross-section of each sector,and (c) internals of a blanket module highlighting the insert and the positioning ribs.


to have a simple low-pressure SiC composite structurewith relative ease in manufacturing; (2) the desire to uti-lize only one coolant, and (3) the low nuclear heating ofSiC composite as compared to a high nuclear heatingin the Pb–17Li reducing the need for direct cooling ofthe box. However, a dual-coolant arrangement wouldallow for a much higher first wall heat load, if needed.

The ARIES-AT first wall and blanket is a simple,low-pressure design consisting of SiC composite boxesas shown inFig. 6. Each module includes an insertwith position ribs which is free flowing and does notcarry structural loads. These ribs divide the annularregion between the box and the insert into a number ofchannels through which the coolant first flows at highvelocity to cool the walls of the box (from the bottom totop in the poloidal direction). The coolant turns at thetop of the module and returns slowly through the largeinner channel (insert). The breeding coolant exits thefirst wall and blanket box at the bottom at high temper-ature. This flow arrangement allows Pb–17Li coolantto exit at 1100◦C while maintaining the SiC compositestructure and the SiC/Pb–17Li interface temperaturesbelow 1000◦C. The wall of the box facing the plasma(first wall) consists of a 4-mm SiC structure on a whicha 1-mm CVD SiC armor layer is deposited.Table 4summarizes the power flow and major thermal param-eters of the ARIES-AT blanket.

For stress analysis, a 1-MPa inlet pressure isassumed for the coolant to adequately account forboth the pressure drop in the blanket (∼0.25 MPa) and

TM

P

O

25

2.11

the hydrostatic pressure of a∼6 m Pb–17Li column(∼0.5 MPa). ANSYS analyses[3] indicate that the pri-mary stress at the first wall is quite low (∼60 MPa) andthe maximum thermal stress is 114 MPa resulting in amodest combined stress of 174 MPa, well within the190-MPa limit adopted for SiC composites.

The nuclear performance of the ARIES-AT fusioncore was optimized in order to achieve a TBR of1.1 (3D) while minimizing the radial build and wastestream. The blanket in the outboard is divided intotwo regions; a 30-cm-thick region which is replace-able and a 35-cm-thick region which is a life-of-plantcomponent (seeFig. 2). On the inboard side, thereis only one 30-cm-thick replaceable blanket region.The blanket is surrounded by hot shield zones (24-cm-thick in the inboard and 15-cm-thick in the outboard)which are also life-of-plant components. The vacuumvessel which is made of ferritic/martensitic steel andcooled by water also helps shield the magnets. Thissegmentation strategy has led to a reduction of∼50%in rad-waste compared to ARIES-RS (see Section6).

The ARIES-AT blanket with its high coolant outlettemperature is well matched to an advanced Braytonpower cycle, leading to an overall thermal efficiency of∼59%. The Brayton cycle considered includes three-stage compression with two intercoolers and a highefficiency recuperator. Recent advances in manufactur-ing of low-cost, high efficiency recuperators are the keyin achieving the high cycle efficiency[18]. The powercycle requires using He as the secondary fluid. A heate s noty

4

ato rsi les.E ard,o lds;S andP atesr of2 ud-i ickS Thec lantc s for

able 4ajor parameters of ARIES-AT first wall and blanket

ower flowsFusion power (MW) 1719Multiplied neutron power (MW) 1375Total transport power (MW) 369Core radiation (MW) 103Average neutron wall load (MW/m2) 3.2Peak neutron wall load (MW/m2) 4.8

utboard blanket parameters (Region I)Number of sectors 16Number of modules per sector 12Power core coolant inlet temperature (◦C) 654Blanket outlet temperature (◦C) 1100Blanket pressure drop (MPa) 0.Maximum SiC temperature (◦C) 996Maximum SiC/Pb–17Li interface temperature (◦C) 994Average Pb–17Li velocity in the first wall (m/s) 4.Average Pb–17Li velocity in the inner channel (m/s) 0

xchanger between Pb–17Li and He coolants haet been developed but, in principle, is possible.

.2. Divertor

The ARIES-AT divertor system is similar to thf ARIES-RS[27]. Each of the 16 ARIES-AT secto

ncludes a pair of top and bottom divertor moduach divertor module consists of three plates (inboutboard, and dome), a structural ring, and manifoiC composite is used as structural materialb–17Li as coolant. Each divertor plate incorpor

ectangular coolant channels with a toroidal pitch.1 cm and a total radial dimension of 3.2 cm, incl

ng a 3.5-mm-thick W armor over a 0.5-mm-thiC composite wall on the plasma-facing side.oolant flows in the poloidal direction in these coohannels which serve as inlet and outlet header


the 2-mm-thick channel on the plasma-facing sidewhere the Pb–17Li flows toroidally over a shortdistance (∼2 cm) to cool the high heat flux region.This configuration is capable of handling a peak heatload of 5 MW/m2 [3]. This divertor design was chosenbecause of the desire to integrate the divertor coolingcircuit into the blanket circuit, to utilize only onecoolant for both blanket and divertor, and to minimizethe number of inlet and outlet coolant headers.Higher heat flux capabilities are possible by usinghigh-pressure He as the coolant and utilizing W as theplate structural material (instead of SiC composite).

4.3. Configuration and maintenance

Achieving a high plant availability goal significantlyinfluenced the overall configuration of ARIES-AT andthe major design decisions. As for ARIES-RS, a sec-tor removal scheme has been utilized for ARIES-AT.An important aspect of this approach is the integra-tion of each sector (seeFig. 7). The first wall, blanket,

parts of the shield, divertors, and stability shells form anintegral unit. This integrated sector construction elimi-nates in-vessel maintenance operations and providesa very strong continuous structure which is able tosupport large loads (such as disruption forces). In addi-tion, there is only one inlet-outlet header to each sectorminimizing the number of connects/disconnects duringmaintenance. Maintenance is performed by extracting acomplete sector and replacing it with a refurbished sec-tor from the hot cell. Removal of the sector is performedthrough large ports (seeFigs. 1 and 8) using a rail sys-tem together with transporter casks[7]. Port doors onthe back of the shield and the cryostat prevent the spreadof radioactive dust to the containment building.

Extracting a complete sector requires positioningof the poloidal-field (PF) coils in regions above andbelow the maintenance doors. The outer leg of thetoroidal-field (TF) coils must be sufficiently large toprovide adequate clearance between the coils to extracta complete sector. This has the benefits of reducingthe toroidal field ripple considerably. An importantrequirement for sector maintenance is that all replace-able components should have a nearly identical powercore lifetime. Otherwise, the plant availability will suf-fer due to many plant shutdowns to replace smallercomponents.

4.4. Magnet systems

At temperatures greater than 20 K, the only HTSm ica-t rial,B larlya COt orm ityc -c hesr ulart en-d thisr uctori tly,t pur-p sibletl d atp
Fig. 7. An integrated ARIES-AT fusion core sector.
aterial that appears promising for fusion applions at present is YBCO tapes. Another HTS mateSSCO, has great performance capability, particut high field but at cryogenic temperature. The YB

hick film superconductor is highly anisotropic. Fagnetic field parallel to the film, its current dens

apability is much larger than that of Nb3Sn superonductors. Its current-carrying capability diminisapidly, however. as the field strength perpendico the tape is increased. For ARIES-AT, the perpicular field strength has been kept below 5 T foreason. The actual performance of the superconds a strong function of the film thickness; presenhinner films have a higher current density. For theose of this study, it is assumed that it would be pos

o make a relatively thick film (∼20–30�m) of longength with a performance similar to that observeresent for thinner films.


Fig. 8. Top view of ARIES-AT maintenance scheme. An integrated sector is removed horizontally.

A major benefit of HTS is the simplificationof magnet design. High-temperature superconductorsdo not suffer from flux jumping when operated attemperatures above 10–20 K because of the largeincrease in the thermal capacity of the metals atthese temperature compared to 4 K. As a result, onlya small amount of stabilizer is needed which sim-plifies the magnet design and increases its averagecurrent density. In addition, it may be possible to man-ufacture the coil by directly depositing the supercon-ductor on structural plates (with intermediate layersof insulator and stabilizer) by epitaxial techniques.This should lead to considerable savings in the mag-net cost[6]. The HTS may be more damage-resistantcompared to LTS (data to date indicate that irra-diation damage limits are not lower than those forthe LTS material; data at higher doses are not yetavailable). As a whole, HTS have the potential ofsubstantially relaxing the design constraints due tolower irradiation damage, smaller stabilizer, and lowernuclear and AC heating at the cryogenic tempera-ture. Magnet operation at 20 K will also substan-tially simplify the cryo-plants (“dry” magnets may bepossible).

The ARIES-AT magnets are shown inFig. 9.Because of the modest value of the ARIES-AT toroidal-field strength, the stresses are quite low[6]. A majorconstraint of YBCO superconductor is that the con-

ductor strain should be less than 0.2%. As a result,the ARIES-AT magnet support structure is “thicker”and more massive than those of cryogenic supercon-ductors. It should be noted that since the ARIES-ATdesign point optimized at a maximum field of 11 T, wedid not utilize the high-temperature capability of HTS.As such, Nb3Sn superconductors can equally be used(in fact, with some cost penalty in increasing the sizeof the fusion core, NbTi superconductor would also befeasible).

Fig. 9. The ARIES-AT toroidal-field coil assembly.


5. Safety and licensing

Safety and environment goals for ARIES designsinclude: (a) no need for evacuation plan and (b) no high-level rad-waste and minimization of low-level waste. Inaddition, we have tried to ensure that achieving thesegoals are transparent (e.g., by minimizing the sourceterm) in order to simplify the licensing process.

Compared to previous ARIES designs, there is agreater depth in the safety analysis for ARIES-AT. Forexample, a large number of off-normal events were con-sidered. In addition to loss-of-flow (LOFA) and loss-of-coolant (LOCA) events, we considered an ex-vesselloss of coolant, an in-vessel off-normal event that mobi-lize in-vessel inventories (e.g., tritium and tokamakdust) and bypass primary confinement such as a loss-of-vacuum and an in-vessel loss of coolant with bypass.

Fig. 10shows the specific activity and specific decayheat in various outboard components of ARIES-AT as

F oardc

a function of time after shutdown. Note that the highinitial decay heat of the first wall drops to levels belowferritic/martensitic steel components within 20 min to1 h after shutdown. This is indicative of the very low-level of induced activity in SiC and the very smallamount of total decay heat in SiC structure. As a result,analyses show[6] that following a LOCA or LOFA,the temperature of the first wall and blanket in ARIES-AT drops below the normal operation temperature andthere is no temperature excursion. It is interesting tonote that within 1 h after shutdown, the decay heat ofSiC structure in the first wall and blanket falls belowthat of the Pb–17Li coolant. As such, component tem-peratures are estimated to be higher during a LOFAthan a LOCA.

The major radiological inventories in ARIES-AT aretritium and activation products in fusion core compo-nents (plasma facing components, structural material,and coolant). Overall, no accident has been identifiedthat can lead to mobilization of radioactive materialin the ARIES-AT first wall and blanket. Therefore,inventories that may be mobilized during an accidentare tritium, W dust, and203Hg and210Po isotopes inthe Pb–17Li coolant. Detailed analyses[5] indicatethat ARIES-AT can meet the no-evacuation limit evenassuming a conservative on-line removal of Bi from thecoolant. For example, the tritium inventory in ARIES-AT is estimated at 745 g, of which∼680 g is estimatedto be due to plasma interaction and/or co-depositedlayer in the first wall. The worst case scenario con-s asse tot iong

Tc uire-mA ardc ore,t ledt toA -e ion( werpc iatedwF

ig. 10. (A and B) Specific activity and decay heat on the outbomponents of ARIES-AT.

idered (an in-vessel LOCA together with a by-pvent) would lead to the release of 7.6 g of tritiumhe environment far below 150 g limit for no evacuatoal.

Activation analyses indicate that all ARIES-Aomponents meet the Class-C low-level waste reqents by a wide margin. In fact,∼90% of ARIES-T waste meets Class-A limits (lowest hazlassification under US regulations). Furthermhe segmentation of ARIES-AT fusion core haso a reduction of∼50% in rad-waste comparedRIES-RS. A total of∼1270 m3 of waste is genrated after 40 full-power year (FPY) of operatassuming that the coolant is reused in other polants). Of this amount,∼1000 m3 is for lifetimeomponents (end of service). The waste associth the replaceable component is∼30 m3 every 4PY.


6. Design evaluation and R&D needs

Similar to all ARIES designs, the ARIES-ATresearch has aimed at arriving at a balance betweenattractiveness (as measured by customer needs) andcredibility (as measured by the degree of extrapola-tion from present database). These two dimensions areexplored in this section.

6.1. Design evaluation

The top-level goals and requirements for fusionpower can be divided into three general categories: (1)

cost, (2) safety and environmental features, and (3) reli-ability, maintainability, and availability.

Cost of electricity—The ARIES-AT cost of electric-ity of 4.7 ¢/kWh is competitive with current sources ofenergy and is smallest of modern fusion power plantstudies.Table 5provides a detailed breakdown of theARIES-AT costs. Further reduction in the cost of elec-tricity can be achieved by utilizing devices with highernet electric output (seeFig. 11).Safety and environmental features—ARIES-ATmeets the no-evacuation criteria with minimum need

Table 5ARIES-AT power-plant economic parameters (1992 $)

Account no. Account title Million dollars

20. Land and land rights 10.621. Structures and site facilities 253.522. Reactor plant equipment 761.022.1.1. FW/blanket/reflector 64.322.1.2. Shield 69.422.1.3. Magnets 126.722.1.4. Supplemental-heating/CD systems 37.122.1.5. Primary structure and support 26.922.1.6. Reactor vacuum systems (unless integral elsewhere) 98.822.1.7. Power supply, switching and energy storage 50.722.1.8. Impurity control 4.122.1.9. Direct energy conversion system 0.022.1.10. ECRH breakdown system 4.022.1. Reactor equipment 482.022.2. Main heat transfer and transport systems 126.023. Turbine plant equipment 243.02 ent2 quipm229 ing con9 and e9 ing and9 g and999 uction9

U e)kWe)

C

4. Electric plant equipm5. Miscellaneous plant e6. Heat rejection system7. Special materials0. Direct cost (not includ1. Construction services2. Home office engineer3. Field office engineerin4. Owner’s cost6. Project contingency7. Interest during constr9. Total cost

nit costs Unit direct cost ($/kWUnit overnight cost ($/
Unit total cost ($/kWe)
ost of electricity Capital return (mill/kWh)O&M (2.10%) (mill/kWh)Component replacement (miDecommissioning (mill/kWh)Fuel (D) cost (mill/kWh)Total cost of electricity, COE

98.5ent 47.4

23.383.8

tingency) 1521.1quipment 171.9

services 79.1services 79.1

277.8311.8

(IDC) 403.22844.0

15212441
2844
36.876.87

ll/kWh) 3.510.250.03

(mill/kWh) 47.53


Fig. 11. Cost of electricity of ARIES-AT type power plant as a func-tion of net electric output showing the nuclear economy of scale. Theaverage neutron wall load is kept fixed at 3.3 MW/m2 (ARIES-ATvalue). The ARIES-AT design point is shown by a square.

for the confinement of radioactive material because ofthe low radioactive source term. In addition, ARIES-AT components meet the Class-C low-level wasterequirements by a wide margin (∼90% of ARIES-ATwaste meet Class-A limits). Furthermore, the segmen-tation of ARIES-AT fusion core has led to a reductionof ∼50% in rad-waste compared to ARIES-RS.Reliability, maintainability, and availability—Detailed modeling of sector removal maintenancescenario has shown that schedule maintenance canbe performed in 1 month. An availability of >90%was predicted for ARIES-AT[7,8]. (A plant factorof 85%, however, was used in ARIES-AT costing.)Special attention was given to manufacturability indesigning ARIES-AT components.

Compared to ARIES-RS, these improvements inmeeting customer requirements have been achievedmainly through using SiC composites as structuralmaterial and achieving a high thermal conversion effi-ciency. In addition, SiC composite has allowed theblanket to be thinner, moving the passive stabilizationcloser to the plasma resulting in a larger plasma elon-gation and plasmaβ.

As discussed in Section2, the decrease in the costof electricity with decreasing power plant size satu-rates above a certain value. This can be clearly seenin Fig. 12; the COE increases only by∼0.15 ¢/kWh(∼3%) if the major radius is increased from 5.2 to6 m (about 15%). As a result, the increased plasmaβ the

Fig. 12. Cost of electricity of 1000-MWe ARIES-AT type powerplant as a function of neutron wall loading demonstrating insensi-tivity of COE to power plant size. The ARIES-AT design point isshown by a square.

machine size; rather it was used to reduce the toroidal-field strength. In addition, because of this insensitiv-ity to size, ARIES-AT is more robust to variations inachievable physics or engineering parameters (e.g., alower value ofβ and/or a higher divertor heat loadcan be easily accommodated by utilizing a larger sizedevice with minor cost penalties).

6.2. R&D needs

A major aim of the ARIES-AT study was to identifyphysics and technology areas with the highest leveragefor achieving attractive and competitive fusion powerin order to guide fusion R&D. Refs.[1–8] describethe detailed analysis of ARIES-AT components andinclude a comprehensive list of R&D topics. Some ofthe more critical issues are listed below (the ordering isnot intended to imply any priority). It should be notedthat the ARIES-AT physics assumptions are similar tothose of ARIES-RS.

6.2.1. Demonstration of a stable and controllableoperating point

The benefits of the reversed-shear plasma operatingmode are that it achieves both highβN and highβ, it obtains large bootstrap current fractions withvery good current profile alignment, and it appearsto provide the transport suppression necessary tosustain the pressure profile that is consistent withthe higherβ and high bootstrap current. Significant
in ARIES-AT has not been utilized to decrease


progress in understanding this regime of operation hasbeen made in recent years and higher performancedischarges has been obtained. Still, demonstration andcontrol of these discharges in steady-state conditionsand with a burning plasma remain a central issue. Itshould be demonstrated that plasma parameters (e.g.,density, temperature, position, current,. . .) and theirprofiles can be controlled precisely as large variationin fusion power would cause significant problems forthe designs of fusion core components.

6.2.2. Divertor and edge physicsDivertors remain one of the most difficult physics

and engineering challenges. Radiative modes whichresult in dispersing most of the plasma transport poweron the first wall and divertor plates can help keep theheat load on the divertor plates to a manageable level. Itappears that a combination of radiation in the core andin the divertor would be necessary. The higher densityof an L-mode type edge can help to reduce the heatloads.

6.2.3. Disruption avoidanceOne of the most serious problems with the tokamak

is the possibility of a major plasma disruption requir-ing plant shutdown. Besides the potential for damage tothe surrounding structures and the possibility to initiateevents leading to chemical and radioactivity release,unplanned shutdowns are an unacceptable operatingcharacteristic for a power plant. Aside from the impacto esw en-t uldn

dis-r uumv in-v bero d-i veryl onp tiona

6c

hath w

cost of electricity and high level of safety and environ-mental attractiveness. A large amount of developmentis still needed before silicon-carbide composites canbe qualified as structure material for fusion applica-tions. It should be noted that major advances has beenmade in SiC composites during the last 10 years[28].The latest results indicate that irradiated “third gener-ation” composites show no degradation of propertiesup to 10 dpa. This improved performance is due todevelopment of stoichiometric crystalline SiC fiber andadvanced fiber/matrix interphases. Note that the irra-diation lifetime of SiC composite is set probably byburn-up rate (a 3% burn-up rate is assumed for ARIES-AT, corresponding to∼15 MW/m2). For comparison,the irradiation lifetime of ferritic steel is estimated at150–200 dpa.

6.2.5. Compatibility of Pb–17Li with SiCcomposites

The maximum interface temperature betweenPb–17Li and SiC composites have been set at 1000◦Cin ARIES-AT. Tests at ISPRA have shown thatthe two materials are compatible at 800◦C understagnant conditions[29]. Test in flowing Pb–17Liis needed to better establish operating temperaturelimits.

6.2.6. High-temperature heat exchangers(Pb–17Li and He)

A high temperature heat exchanger betweenP uti-l yclee elop-m mon-s er areN wnt

6p

for-m etherw osea uxc thee s thatt rtora ent

n plant availability, it is likely that extended outagould follow an unplanned plasma shutdown to id

ify the source of the problem and ensure that it coot happen again.

As the SiC composite is a semiconductor, theuption forces appear on the sturdy shield and vacessel of ARIES-AT. However, thermal load onessel components is still large and only a small numf disruptions (∼20) can be tolerated. Further stu

es are needed to determine operating points withow probability of disruption (less than 1 disruptier year) and/or reliable active measures of disrupvoidance.

.2.4. Development and qualification of SiComposites

Utilization of SiC composite is the key feature tas allowed ARIES-AT to achieve its relatively lo

b–17Li coolant and He should be developed toize the Brayton cycle and achieve a high thermal cfficiency. This was considered a reasonable devent step, although the technology needs to be de

trated. Possible materials for such a heat exchangb or SiC composite (if Pb–17Li and SiC are sho

o be compatible at 1100◦C).

.2.7. High heat flux removal combined with higherformance

The simultaneous requirements of high perance under extreme heat and particle loads, togith extended lifetime, maintainability, and safety pdifficult problem for divertors and high heat fl

omponents. The lack of predictive capabilities fordge plasma exacerbates the problem. It appear

ungsten is the only acceptable material for divermor (and possibly divertor structure). Developm


of divertor systems with capability of removing highheat flux at steady state as high-grade heat is a majorfeasibility issue.

Acknowledgements

Institutions involved in the ARIES-AT study, inaddition to UC San Diego, are: The Boeing Company,General Atomics, Idaho National Engineering Labora-tory, Massachusetts Institute of Technology, PrincetonPlasma Physics Laboratory, Rensselaer PolytechnicInstitute, and the University of Wisconsin–Madison.The work at University of California San Diego wassupported by the United States Department of Energy,Office of Fusion Energy DE-FC03-95ER54299.

References

[1] S.C. Jardin, C.E. Kessel, T.K. Mau, R.L. Miller, F. Najmabadi,V.S. Chan, et al., Physics basis for the advanced tokamak fusionpower plant, ARIES-AT, Fus. Eng. Des. 80 (2006) 25–62.

[2] C.E. Kessel, T.K. Mau, S.C. Jardin, F. Najmabadi, Plasma pro-file and shape optimization for the advanced tokamak powerplant, ARIES-AT, Fus. Eng. Des. 80 (2006) 63–77.

[3] A.R. Raffray, L. El-Guebaly, S. Malang, I. Sviatoslavsky, M.S.Tillack, X. Wang, and The ARIES Team, Advanced power coresystem for the ARIES-AT power plant, Fus. Eng. Des. 80 (2006)79–98.

[4] L. El-Guebaly, The ARIES Team, Nuclear performance assess-

.and

The. 80

sys-180.l-IES-. Eng.

n,erport

[ e,s for005,al-

[11] F. Najmabadi, C.G. Bathke, M.C. Billone, J.P. Blanchard, L.Bromberg, E. Chin, et al., Overview of the ARIES-RS reversed-shear tokamak power plant study, Fusion Eng. Design 38 (1997)3–25 (also, see other articles in that issue).

[12] F. Najmabadi, R.W. Conn, P.I.H. Cooke, S.P. Grotz, M.Z.Hasan, E. Ibrahim, et al., The ARIES-I tokamak fusion reactorstudy—the final report, UCLA-PPG-1323, 1991 (available athttp://aries.ucsd.edu/ARIES/DOCS/final-report.shtml);R.W. Conn, F. Najmabadi, ARIES-I: a steady-state, first-stability tokamak reactor with enhanced safety and envi-ronmental features, Nucl. Fusion Suppl. 3 (1991) 659–670.

[13] Y. Seki, M. Kikuchi, S. Nishio, A. Oikawa, T. Ando, Y. Ohara, etal., Concept study of the steady-state tokamak reactor (SSTR),Japan Atomic Energy Research Institute Report JAERI-M-91-081, 1991.

[14] T. Kammash, Fusion reactor physics—principles and technol-ogy, Ann. Arbor Sci. (1977), Chapter 13.

[15] F. Najmabadi, R.W. Conn, Directions for attractive tokamakreactors: the ARIES-II and ARIES-IV second-stability designs,Nucl. Fusion Suppl. 3 (1993) 295–420.

[16] C.E. Kessel, J. Manickam, G. Rewoldt, W.M. Tang, Improvedplasma performance in tokamaks with negative magnetic shear,Phys. Rev. Lett. 72 (1994) 1212;T. Ozeki, et al., Profile control for stable high-βp tokamaks withlarge bootstrap current, in: Proceedings of 14th IAEA Interna-tional Conference on Fusion Energy, Wursberg, Germany.

[17] F. Najmabadi, the ARIES Team, Overview of the ARIES-STspherical torus power plant study, Fusion Eng. Des. 65 (2003)143–164 (also, see other articles in that issue).

[18] R. Schleicher, A.R. Raffray, C.P. Wong, An assessment of theBrayton cycle for high performance power plant, Fusion Tech-nol. 39 (2001) 823–827.

[19] S.C. Jardin, N. Pomphrey, J. Delucia, Dynamic modeling oftransport and position control of tokamaks, J. Comp. Phys. 66

[ on-995)

[ er,itiesoils,

[ M.port

[ ur-actor

Heat-els,

[ nn,uxater.

[ as,

ment of ARIES-AT, Fus. Eng. Des. 80 (2006) 99–110.[5] D.A. Petti, B.J. Merrill, R.L. Moore, R.L. Moore, G.R

Longhurst, L. El-Guebaly, et al., ARIES-AT safety designanalysis, Fus. Eng. Des. 80 (2006) 111–137.

[6] F. Dahlgren, T. Brown, P. Heitzenroeder, L. Bromberg,ARIES Team, ARIES-AT magnet systems, Fus. Eng. Des(2006) 139–160.

[7] L.M. Waganer, The ARIES Team, ARIES-AT maintenancetem definition and analysis, Fus. Eng. Des. 80 (2006) 161–

[8] L.M. Waganer, F. Najmabadi, M. Tillack, X. Wang, L. EGuebaly, and The ARIES Team, Design approach of the ARAT power core and vacuum vessel cost assessment, FusDes. 80 (2006) 181–200.

[9] J. Kaslow, M. Brown, R. Hirsch, R. Izzo, J. McCanD. McCloud, et al., Criteria for practical fusion powsystems—report from the EPRI Fusion Planet, EPRI ReBR-104469, 1994.

10] F. Najmabadi, M.S. Tillack, R.L. Miller, T.-K. Mau, B.J. LeX. Wang, et al., The Starlite study: assessment of optiontokamak power plants, UC San Diego Report UCSD-ENG-1997 (available athttp://aries.ucsd.edu/ARIES/DOCS/finreport.shtml).

(1986) 481.20] M.S. Chu, J.M. Greene, T.H. Jensen, R.L. Miller, A. B

deson, R.W. Johnson, M.E. Mauel, Phys. Plasmas 2 (12236.

21] C. Cates, M. Shilov, M.E. Mauel, G.A. Navratil, D. MaurS. Mukherjee, et al., Suppression of resistive wall instabilwith distributed independently controlled active feedback cPhys. Plasmas 7 (2000) 3133.

22] R.E. Waltz, G.M. Staebler, W. Dorland, G.W. Hammett,Kotschenreuther, J.A. Konings, A gyro-Landau-fluid transmodel, Phys. Plasmas 4 (1997) 2482.

23] T.K. Mau, S.C. Chiu, R.W. Harvey, Modeling of fast wave crent drive in standard and second-stability bootstrapped replasmas, in: EPS Topical Conference on Radio frequencying and Current Drive of Fusion Devices, vol. 16E, Bruss1992, p. 181.

24] A. Loarte, S. Bosch, A. Chankin, S. Clement, A. HerrmaD. Hill, et al., Multi-machine scaling of divertor peak heat fland width for L-mode and H-mode discharges, J. Nucl. M266-269 (1999) 587.

25] R.A. Hulse, Numerical studies of impurities in fusion plasmNucl. Technol./Fusion 3 (1983) 259.

http://aries.ucsd.edu/aries/docs/final-report.shtml

http://aries.ucsd.edu/aries/docs/final-report.shtml


[26] K. Behringer, Description of the impurity transport codeSTRAHL, JET Report JET-R(87)08, JET Joint Undertaking,Culham, 1987.

[27] C.P.C. Wong, E. Chin, T.W. Petrie, E.E. Reis, M.S. Tillack, X.Wang, et al., ARIES-RS divertor system section and analysis,Fusion Eng. Des. 38 (1997) 115–137.

[28] A.R. Raffray, R. Jones, G. Aiello, M. Billone, L. Giancarli, H.Golfier, et al., Design and material issues for high performanceSiCf /SiC-based fusion power cores, Fusion Eng. Des. 55 (2001)55–95.

[29] S. Malang, Forschungszentrum Karlsruhe, private communica-tions, 2001.

the aries-at advanced tokamak, advanced …aries.ucsd.edu/lib/report/journal/fed/06-fed-80...f idaho...

Documents