Fusion Engineering and Design 58–59 (2001) 629–634

The EU advanced lead lithium blanket conceptusing SiCf/SiC flow channel inserts as electrical

and thermal insulators

P. Norajitra a,*, L. Buhler a, U. Fischer a, K. Kleefeldt a, S. Malang a,G. Reimann a, H. Schnauder a, L. Giancarli b, H. Golfier b, Y. Poitevin b,

J.F. Salavy b

a Forschungszentrum Karlsruhe, P.O. Box 3640, D-76021, Karlsruhe, Germanyb CEA/Saclay, DRN/DMT/SERMA, 91191, Gif-sur-Y�ette, France

Abstract

Preparatory work on the EU advanced dual coolant (A-DC) blanket concept using SiCf/SiC flow channel insertsas electrical and thermal insulators has been carried out at the Forschungszentrum Karlsruhe in co-operation withCEA (SiCf/SiC composite-related issues) as a conceptual design proposal to the EU fusion power plant study plannedto be launched in 2001 within the framework of the EU fusion programme having the main objective of specifyingthe characteristics of an attractive and viable commercial D–T fusion power plant. The basic principles and the studymethod for the A-DC blanket concept are presented in this report. The results of this study show that the A-DCblanket concept has a high potential for further development due to its high thermal efficiency and its simple conceptsolution. © 2001 Elsevier Science B.V. All rights reserved.

Keywords: Dual coolant; SiCf/SiC; EU fusion; A-DC blanket

www.elsevier.com/locate/fusengdes

1. Introduction

The A-DC blanket concept [1] (Fig. 1) with aself-cooled lead–lithium breeding zone and he-lium cooled steel structure is based on the use ofoxide dispersion-strengthened (ODS) steel as firstwall (FW) material and SiCf/SiC flow channelinserts (FCIs) as electrical and thermal insulators

as proposed in ARIES-ST study [2]. As a result,there is no longer any need for insulating coatingson the steel structure inside the liquid-metal flowchannels (for MHD reasons). Furthermore, it al-lows a high exit temperature of the lead–lithium,which gives the potential for a high efficiency ofthe power conversion system. In this case, theSiCf/SiC FCIs have no structural function (i.e. nopressure difference between the inside and outsideof the FCI walls is expected due to the pressureequalisation slot). Therefore, neither a highstrength nor a high thermal conductivity (i.e. astrong thermal barrier is favoured) of this mate-

* Corresponding author. Tel.: +49-7247-82-3673; fax:+49-7247-82-7673.

E-mail address: prachai.norajitra@imf.fzk.de (P. Norajitra).

0920-3796/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved.

PII: S0920 -3796 (01 )00506 -3

P. Norajitra et al. / Fusion Engineering and Design 58–59 (2001) 629–634630

rial is required. Moreover, the use of ODS ferriticsteel as FW material allows an about 100 Khigher FW temperature limit [3,8] in comparisonwith the MANET steel (Tmax-MANET-FW�550 °C)used in the layout [4] of the earlier EU DCLblanket concept for DEMO. For these reasons ahigher neutron wall load (resulting in higherpower densities and surface heat load) as well asan enhancement of the thermal efficiency could beexpected. The goal of this study is to evaluate thepotential of this A-DC blanket concept with re-gard to a high thermal efficiency and blanketperformance limitations in terms of maximumneutron wall load (NWL), max. surface heat load(SHL), max. power density, maximum structuraltemperature, availability, and lifetime.

2. Main features

Fig. 1 shows a cut-out of an A-DC outboardblanket in the torus equatorial zone with a ra-dial– toroidal–poloidal dimension of about 1×1.4×12 m3. The whole blanket structure is madeof ferritic steel. The U-shaped FW having a frontlayer thickness of 4 mm (total thickness: 44 mm),

together with the radial–poloidal and poloidal–toroidal steel grids and the shielding structure,makes up a stiff segment box. The steel gridsforming rectangular Pb–17Li coolant channels(about 335×240 mm2) are rigidly welded to theexternal structure. As recognised by temperaturecalculations, the maximum FW temperature be-yond 550 °C (maximum allow. e.g. for EURO-FER) occurs within an about 2 mm thin layer onthe plasma facing FW surface. This allows aconstructional solution that avoids the fabricationof the whole ODS structure (at present, onlydiffusion welding is recommended for ODS joints[5,6]) by choosing EUROFER as base materialfor the whole structure which contains a thinODS layer plated onto the plasma facing FWsurface. The blanket box is not expected to with-stand the maximum He coolant pressure becausethe use of a passive accident management scheme(pressure suppression system) is foreseen.

Within the FW and side wall, He coolant chan-nels (radial–poloidal dimension: �30×20 mm2,24 mm poloidal channel pitch and 4 mm thickwebs) are passed by He gas in alternating direc-tions. The counter-flowing He gas contributes toachieving a uniform temperature distribution in

Fig. 1. Isometric view of an A-DC outboard blanket segment at torus centre.

P. Norajitra et al. / Fusion Engineering and Design 58–59 (2001) 629–634 631

Table 1Main data for the A-DC reference case

Blanket

Average/maximum NWL [MW/m2] 4.0/5.00.68/0.88Average/maximum SHL [MW/m2]

Energy multiplication 1.15

Blanket coolantsHelium

300/480Inl./outl. Temperature [°C]Pressure [MPa] 14

40Maximum He velocity [m/s]5aPumping power [% of electrical power]

Pb–17LiInl./outl. Temperature [°C] 460/700

0.3Pumping power [% of electrical power]

Power con�ersion systemSecondary heliumInl./outl.temp. [°C] 280/650

18Pressure [MPa]Thermal efficiency (power converstion 0.46

system)Net efficiency (blanket cycle) 0.44

a About 2% for the base case with an average NWL of 2.22MW/m2.

shaped coolant channels (to keep the maximuminterface temperature between the steel grids andthe stagnant Pb–17Li under 500 °C for corrosionreasons) and heated up to 480 °C before leavingthe blanket top.

The Pb–17Li, besides having the function of atritium breeder and neutron multiplier, acts as acoolant as well. The Pb–17Li outlet temperatureis determined by compatibility issues and, forefficiency reasons, needs to be maximised. Be-cause of the high magnetic field present in theblanket region (�7 T), the Pb–17Li needs to beelectrically insulated from the steel wall despitethe relatively low flow velocity (maximum 0.5m/s). This is achieved by adding SiCf/SiC FCIs(about 5 mm thick) which also act as thermalinsulators (in order to maximise the Pb–17Litemperature without exceeding the allowable steeltemperature). In this concept the most importantspecific requirements for SiCf/SiC (no structuralfunctions) are a low electrical conductivity andlow thermal conductivity, together with a compat-ibility with high Pb–17Li temperatures. The Pb–17Li enters the top end of the blanket at 460 °C,flows downwards through the four channels in thefront zone, and is heated to 630 °C at the blanketbottom. It is then diverted into the two rearchannel zones containing eight channels alto-gether, where it flows back to the top at lowervelocity and leaves the blanket at 700 °C.

3. Calculations

The methods of the study are described in detailin [1]. In the first step, detailed neutronic calcula-tions (tritium breeding ratio, overall blanketpower, and power densities), MHD calculations(velocity profile and MHD pressure loss in thePb–17Li flow), thermohydraulic calculations(mass flow rates, temperature rises, heat transfers,pressure losses and temperature distributions)were performed for the base case using the reactorparameters specified by UKAEA Culham with anaverage NWL of 2.22 MW/m2. After fixing of thetemperature constraints, i.e. 650 °C for the FWdue to creep rupture strength and 500 °C at theinterface between Pb–17Li and the steel for cor-

the external structure and to minimising the ther-mal stresses. Besides, separation of the He supplysystems 1 and 2 facilitates emergency coolingupon failure of a coolant system. A relatively highHe pressure of 14 MPa was chosen in order tominimise the pressure losses and to keep the Hepumping power for the reference case with arelatively high performance under a reasonablelimit of about 5% relative to the extracted power(Table 1). In case of lower performance as in thebase case (see next point) as well as under theDEMO layout conditions [4], a He pressure of 8MPa would be sufficient to cope with pressurelosses. The He coolant gas enters the upper blan-ket segment at 300 °C and is routed four times onits way downwards in a meandering flow throughthe segment walls containing the parallel coolantchannels by means of a header system. After thelast passage, the He temperature at the loweroutlet of the segment walls attains a value of440 °C. On its flow upwards, it is passed throughthe channel arrays in the steel grids with U-

P. Norajitra et al. / Fusion Engineering and Design 58–59 (2001) 629–634632

rosion reasons, temperature calculations were car-ried out in the second step for a higher blanketperformance by linear extrapolation of the powerdensities and the SHL of the base case. Theevaluation of the maximum allowable power lim-its for the FW and the blanket breeding zoneleads to the definition of the A-DC reference casewith a considerably higher power density than thebase case with the corresponding thermomechani-cal calculations for check-up of the feasibility.Finally, the thermal efficiency and assessment ofthe availability and lifetime were performed forthe A-DC reference case.

Since the material data for ODS steel are notyet available, the comparable physical propertiesof the ferritic steel T91 were taken [1], e.g. thethermal conductivity � (400–600 °C)�29 W/(mK), thermal expansion coefficient ��12.0×10−6

l/K and specific heat cp�750 J/(kg K). Two-di-mension stress calculations are carried out assum-ing generalised plane strain boundary conditions.For stress evaluation according to the RCC-MRcode, the Sm values of T91 are assumed on thetemperature level shifted upwards by 100 K [3,8],

e.g. Sm,t-ODS=174/146/101 MPa at 500/600/700 °C, respectively, with t�104 h. For Pb–17Lithe same data as applied for the DEMO calcula-tions [4] are used, e.g. � (600 °C)�19 W/(m K)and cp (600 °C)=187 J/(kg K).

4. Short summary of the results of the study

Two temperature constraints for the FW (creeprupture strength) and the Pb–17Li breeding zone(corrosion) were taken into account. The latterwas found to be decisive for the power limitationof this blanket concept. As a result of this study,a reference case with an average NWL of 4 MW/m2 (corresponds to 0.9 MW/m2 maximum SHL)was determined (Table 1), which is 80% higherthan that in the base case. Moreover, the use ofODS steel leads to a relatively large safety marginfor the FW (e.g. against peaking factor uncer-tainty) allowing a maximum SHL of up to 1.5MW/m2. Detailed thermomechanical calculationsfor the reference case show that all temperaturesand stresses in the FW structure (Tmax=

Fig. 2. Temperature distribution in the first wall at the torus centre for the reference case.

P. Norajitra et al. / Fusion Engineering and Design 58–59 (2001) 629–634 633

Fig. 3. Von Mises primary plus secondary stresses in the first wall at the torus centre for the reference case.

617 °C�allowable 650 °C Fig. 2; maximum to-tal stress �max=413 MPa�allowable 418 MPa,Fig. 3) and the SiCf/SiC FCIs �max=93 MPa�allowable 140 MPa) are well below the allowablelimits. Assuming a three-stage Brayton gas tur-bine cycle for the power conversion system, a netefficiency of 44% was obtained for the blanketcycle. This value could even be increased to about46% by integrating the divertor cycle into theblanket cycle in order to exploit the divertor heat.For the steel structure assumed to play a decisiverole for the blanket lifetime (since no structuralfunctions of SiCf/SiC FCIs), the lifetime predic-tion assuming an allowable displacement damagefor ODS of 150 dpa (a value achieved for similarsteels in fission reactors tests without intolerablylarge swelling) yields an average neutron flux of15.1 MWa/m2 (corresponds to e.g. 6.8 FPY at2.22 MW/m2 or 3.78 FPY at 4.0 MW/m2).

5. Conclusion

The advanced dual coolant blanket concept ischaracterised by its simple construction, simplefunction as well as by its high thermal efficiency.

The overall results of this study show that thisA-DC blanket concept has a high potential forfurther development due to its high thermal effi-ciency and its simple concept solution. Limitingthe ODS steel to a few millimetres thick layer atthe FW allows to utilise the high temperaturestrength of this material without the fabricationdifficulties typical for ODS. Employing flow chan-nel inserts as electrical and thermal insulatorsallows to achieve a liquid-metal exit temperatureof 700 °C at a maximum temperature of the steelstructure of 550 °C, resulting in a thermal effi-ciency of about 45%. Open issues are mainly thefabrication of SiCf/SiC flow channel inserts andtheir compatibility with flowing Pb–17Li (up tonow a compatibility at 800 °C was experimentallyconfirmed in [7]) as well as their irradiation be-haviour. Nevertheless, the expected degradationof the thermal conductivity of SiCf/SiC caused byirradiation is even favourable for this concept.

References

[1] P. Norajitra, L. Buhler, U. Fischer, K. Kleefeldt, S.Malang, G. Reimann, H. chnauder, G. Aiello, L. Gian-

P. Norajitra et al. / Fusion Engineering and Design 58–59 (2001) 629–634634

carli, H. Golfier, Y. Poitevin, J.F. Salavy, The secondadvanced lead lithium blanket concept using ODS steel asstructural material and SiCf/SiC flow channel inserts aselectrical and thermal insulators, FZKA 6385, 1999.

[2] D.K. Sze et al., The ARIES-ST blanket design, proceed-ings of the 13th topical meeting on technology of fusionenergy, Nashville. Tennessee (June 1998), to be publishedin Fusion Technology.

[3] S. Malang, M. Billone, How well can we predict thebehaviour of the different materials in solid breeder blan-kets?, Ninth International Conferrence on Fusion ReactorMaterials (ICFRM9), Colorado Springs, Colorado, USA,October 1999, pp. 10–15.

[4] P. Nomjitra, Thermohydraulics design and thermomechan-ics analysis of two european breeder blanket concepts forDEMO; FZKA 5580, 1995.

[5] K. Shinozaki, H. Kuroki, Y. Nakao, Bonding phenomenaand mechanical properties of liquid phase bonded joints inFe-base ODS alloy, MA956, Welding Int. 10 (10) (1996)778–787.

[6] M. Rees, R.C. Hurst, J.D. Parker, Diffusion bondingof ferritic oxide dispersion strenghtened alloys to austeni-tic superalloys, J. Mater. Sci. 31 (17) (1996) 4493–4501.

[7] R.J. Fordham (Ed.), High Temperature Corrosion ofTechnical Ceramics, Elsevier Applied Sceince, London,1990.

[8] K. Ehrlich, R. Lindau, A. Moslang, M. Schirra, EURO-FER 97 and development strategy for RAFM steels inEurope, IEA-Working Group Meeting on Ferriticilvlarten-sitic Steels, Tokyo, November 2–3, 2000, proceedings to bepublished.