CFETR的等離子體配置的優(yōu)化設(shè)計外文文獻(xiàn)

Contents lists available at ScienceDirect Fusion Engineering and Design journal homepage Optimization design for plasma confi guration at the CFETR H Lia b G Q Lic X J Liuc J P Qianc Y Huangc Z P Luoc Y Guoc X F Liuc L Liuc J X Zhengc S L Liuc X Gaoa c CFETR Design Team aAdvanced Energy Research Center Shenzhen University Shenzhen 518060 China bKey Laboratory of Optoelectronic Devices and Systems of Ministry of Education and Guangdong Province College of Optoelectronic Engineering Shenzhen University Shenzhen 518060 China cInstitute of Plasma Physics Chinese Academy of Sciences Hefei 230031 China A R T I C L E I N F O Keywords CFETR Equilibrium Snowfl ake Divertor A B S T R A C T Equilibrium is essential in the design of the Chinese Fusion Engineering Test Reactor CFETR It is based on the principle underlying various physical and engineering issues such as vertical instability core plasma physics divertor physics and design etc Built on the 0 D parameters with elongation 2 0 using the equilibrium code TEQ series of equilibria with diff erent lower triangularity range of 0 3 0 8 were constructed Through iterations with the divertor and blanket structures we optimized the plasma shape including in particular the position of the X point hence obtaining the fi rst acceptable single null plasma with 0 418 This optimal ITER like equilibrium has in turn served as the starting point for various other design works Besides the standard divertor confi guration the quasi snowfl ake plus QSF divertor confi gurations with diff erent total plasma currents have also been produced which aimed at controlling excessive heat fl ux load on the divertor target The results indicate that the poloidal fi eld PF coil system has the capacity to sustain the shape of the QSF however the PF coil system will continue to be optimized with particular attention of the divertor coil 1 Introduction The conceptual design study for CFETR has been widely carried out since 2010 1 CFETR was originally planned in two phases Phase I to demonstrate steady state operation and a full cycle of tritium self sustained with a fusion power of 200MW and major minus radius R 5 7 m a 1 6 m Phase II to generate a fusion power over 1GW to validate the feasibility of commercial electric fusion power plant after a major upgrade with a larger device size 2 Considering that the cost of magnets takes up large portion of the direct cost of a fusion reactor 3 and that CFETR should be upgraded to a large size in Phase II a new tokamak featuring major and minor radii of 7 2m and 2 2m respec tively was brought forward to perform a detailed physical and en gineering design in 2018 in order to achieve the goals of Phase II while saving the cost of construction Based on the zero order parameters one of the most fundamental tasks of CFETR design is a lower single null diverted SND equilibrium The basic plasma confi guration is the starting point of several physical and engineering study issues such as vertical instability core plasma physics divertor physics and design etc Thus it is critical to fi nd an acceptable equilibrium that leads to feasible physical and engineering solutions The shape or separatrix of the plasma is strongly associated with the structures of the divertor and blanket The spatial arrangement of the blanket modules needs to refer to the separatrices and the design of the divertor also needs detailed information regarding the magnetic fi eld topology in the divertor area 4 5 In contrast the blanket and divertor would constitute some en gineering and physical constraints to the establishment of equilibrium The optimization of the equilibrium is mainly focused on optimizing the shape or the last closed fl ux surface LCFS of the plasma including fi nding appropriate values of the elongation and triangularity and modifying divertor magnetic structure The analyses above will help to determine the divertor structure fi rst wall FW shape etc All these calculations were performed within the framework of the specifi ed poloidal fi eld PF coil system 6 The PF system of CFETR consists of eight central solenoid CS coils 6PF coils and one extra divertor confi guration DC coil These superconducting PF coils can provide a maximum fl ux of 400 VS and feedback control of the plasma current position and shape Table 1 shows the position size turns and current limits of the PF coils 7 Power exhaust on fusion reactors is also a critical challenge In previous work 8 9 as one solution to resolve this challenge a series of snowfl ake SF diverted equilibrium for CFETR were established with https doi org 10 1016 j fusengdes 2019 111447 Received 10 September 2019 Received in revised form 29 November 2019 Accepted 19 December 2019 Corresponding author E mail address ligq G Q Li Fusion Engineering and Design 152 2020 111447 Available online 23 December 2019 0920 3796 2019 Elsevier B V All rights reserved T two additional DC coils DC1 DC2 e g exact SF confi guration and its two derivatives i e snowfl ake plus SF and snowfl ake minus SF The exact snowfl ake SF confi guration is characterized by a second order null 10 The gradient of the poloidal magnetic fi eld in the vi cinity of the second order null decreases leading to larger fl ux expan sion and longer connection length The SF and SF have an additional one order null but have similar geometric eff ects to that of the exact SF Compared with previous design there is only one divertor coil DC1 in current reactor because the engineering design concluded that the port for the divertor module maintenance needs large space It would result in a huge infl uence on the capability of sustaining advanced diverted confi gurations and the currents in some PF and CS coils increase dra matically Therefore it is necessary to verify the capability of PF coils system that sustain an SF like confi guration This paper mainly presents the progress of establishing the fi rst acceptable equilibria with the latest parameters for CFETR The opti mization procedures for SND confi guration are presented in Section 2 Section 3 presents the feasibility study of SF like confi gurations Fi nally Section 4 contains a summary and outlook All equilibria in this article are taken from TEQ code 11 2 Determining the plasma confi guration The plasma shape can be described by elongation and triangu larity A series of equilibrium with diff erent were provided to the divertor and blanket design groups and then the feedback was returned as references or constraints 2 1 Elongation The plasma shape can be described by the elongation and tri angularity Plasma elongation is a zero order parameter that aff ects the plasma performance and the global structure of the machine Recall that large elongation is good for plasma confi nement and MHD stability 12 making it easier to achieve higher plasma temperature and den sity higher bootstrap current fraction and higher fusion power Hence large elongation can reduce the size and cost of the machine The fi rst step in the machine design is the 0 D design using the system code In our 0 D design the elongation at 95 poloidal fl ux surface is 95 2 0 2 5 13 Note that 95appears in the scaling laws and in the theoretical study However in the engineering design we are concerned with the elongation at the separatrix X since the separatrix shape determines the shapes of FW and the blanket In our engineering design the separatrix elongation is set as X 2 0 with the corre sponding 95 surface elongation as 95 1 84 which is less than the value for 0 D design The elongation of ITER at separatrix and 95 poloidal fl ux surface are 1 85 and 1 71 respectively Hence the elongation is slightly greater than that of ITER which can facilitate the achievement of good performance However vertical instability is a challenge for large elongation plasma thereby setting an elongation limit to the plasma shape The general parameter dependence study 14 15 shows that the maximum elongation depends on the inverse aspect ratio triangularity wall distance b a plasma poloidal beta p and capability of the vertical displacement VD control system W For the parameters in today s tokamaks and ITER the maximum elongation 95 is generally above 2 For CFETR the parameters are 0 31 0 4 b a 1 1 and p 1 1 which are moderate parameters The capability of the VD control system relies on the engineering design Actually the design of the control system is ongoing and the engineering design is attempting to meet our physics design requirement If we refer to ITER design W 1 22 then all the CFETR parameters are moderate and the maximum elongation 95 2 0 should be achievable Thus our design of 95 1 84 is acceptable In addition specifi c simulation studies have been carried out for the VD control of CFETR In 2015 a preliminary study 16 was conducted using the TOKSYS code The passive structures and in vessel coils ICs were used to control the VD Recently more realistic simulation was performed by considering the eff ects of blanket structure and the pos sible position of the IC From the experience of dealing with the internal structures in EAST the blanket structures are modeled as three layer metal wall 6 The growth rates of vertical instability were simulated using DINA TSC TOKSYS codes respectively and the dependence on the wall resistivity was calculated 7 The results show that growth rates of 12 6 37 s 1are achieved without ICs Based on the results above and the ICs which are still under physics and engineering design it should be able to stabilize the VD 2 2 Triangularity and divertor magnetic geometry It is also benefi cial to employ higher triangularity to enable H mode operation with good core confi nement and global MHD stability 17 18 As reported earlier 19 20 the triangularity has a strong infl uence on the pedestal height and edge pressure gradient profi les i e higher pedestal height and stiff core profi les lead to an increase in plasma stored energy hence plasma performance is better and plasma beta can be more readily obtained with a large At the beginning of the design the lower triangularity lis set to 0 8 average 0 57 whose primary X point is at high magnetic fi eld HFS Then with self consistent core current and pressure profi les a preliminary H mode equilibrium characterized by Ip 13 78MA BT 6 53 T R 7 2 m a 2 2m q95 5 2 2 0 l 0 8 was established as shown in Fig 1 a The pressure pedestal was obtained by EPED model 21 The is interconnected with the divertor geometry and shape of the FW As the inner divertor leg must keep enough length to maintain highly radiative divertor there would not be enough space left at the inboard for the shielding and bracing structures with a large l 0 8 In order to get an acceptable a series of equilibria with a range of lower triangularity 0 3 0 8 are constructed With these equilibria the di vertor physics design group calculated the heat fl ux distribution by SOLPS code 22 while considering many issues e g pumping im purity screen inclined angle etc After complicated iteration the po sition of the primary X point which had been moved outwards hor izontally was preliminarily determined and the ldecreased to 0 52 Considering that the heat load on the outer divertor target is at least three times more than that on the inner target 23 the length of the outer divertor leg was extended as long as possible inside the vacuum vessel A long outer leg has some benefi ts for the reduction of the heat load 24 25 In the long leg cases the ion and electron temperature at the outside target were decreased signifi cantly and the volume of ra diation was increased resulting in the enhancement of the volumetric radiation power loss As shown in Fig 1 b the length of the outer and inner divertor was raised to 1 6 m and 2 8 m respectively However Table 1 PF coil parameters CoilsR m Z m R m Z m TurnLimit KA Turn CS1U1 701 02512 0573860 CS2U1 703 07512 0573860 CS3U1 705 12512 0573860 CS4U1 707 17512 0573860 CS4L1 70 7 17512 0573860 CS3 L1 70 5 12512 0573860 CS2L1 70 3 07512 0573860 CS1L1 70 1 02512 0573860 PF1U4 609 801 11 544835 PF2U13 208 001 11 122545 PF3U15 733 151 11 122545 PF1L4 60 9 801 11 544850 PF2L15 30 6 901 11 122545 PF3L15 73 3 151 11 122545 DC17 10 10 001 11 122550 H Li et al Fusion Engineering and Design 152 2020 111447 2 after iteration between the divertor and shielding structures there was still a space limitation in which the space for shielding and support structures in the vicinity of the divertor was insuffi cient To address this problem the primary X point was moved horizontally outward by 20 cm A new proper divertor design was determined using this optimal position of X point The plasma and divertor confi gurations are plotted in Fig 2 and the minimum radial coordinate of the last closed fl ux surface LCFS in the poloidal plane was fi xed as 5000mm A concept termed dRsep was used to quantify the position of the secondary separatrix dRsep represents the distance between the two separatrices at the outer midplane OMP 26 We need to keep a large dRsep to minimize particles and heat transport across the secondary separatrix and prevent the intersection of the secondary separatrix with FW Since most particles move along the fi eld lines in scrape off layer SOL region the shape of the secondary separatrix has a direct impact on the distribution of the wall heat loads If the secondary separatrix touches the FW in particular at the midplane it is probable that the refl ected and sputtered particles can aff ect the core plasma as the hitting position is closer to the core than other positions Due to the blanket must maintain a minimum thickness for the purpose of at taining tritium self suffi ciency and shielding a large dRsep also can result in a shortage of space for the arrangement of the upper and in ward blanket The minimum radial coordinate of the FW at high fi eld side HFS was determined by the blanket design group and the co ordinate was 4860 mm After scanning a series of equilibria with a range of dRsep 4 10cm the optimal dRsep was confi rmed as 6cm ensuring that the secondary separatrix did not touch the FW With re ferencetothenewsecondaryseparatrix themaximumradial Fig 1 CFETR SND a CFETR SND equilibrium with a large lower triangularity l 0 80 0 57 the X point is very close to the HFS b Version of CFETR SND with a smaller l 0 52 0 418 which ensures enough space for the arrangement of the blanket and long leg divertor The numbers indicate the currents on the PF coils in units of kA turn H Li et al Fusion Engineering and Design 152 2020 111447 3 coordinate of the FW at the low fi eld side LFS increased to 200mm The modifi ed FW and the optimal plasma are shown in Fig 2 The main parameters of the optimized baseline SND equilibrium are listed in Table 2 3 Quasi snowfl ake plus QSF diverted confi guration 3 1 Optimizing the QSF confi guration CFETR was designed to generate fusion power 1GW Under the present operating scenarios the exhausted power PSOL into the SOL is expected to be 180 220MW According to the scaling law the power decay length 27 mmBqPR 0 73 TSOLgeoq 0 78 95 1 20 10 the value of CFETR is approximately 2 1mm This implies that the wetted area on the di vertor plate is less than 3m2 hence the divertor heat load will exceed the material tolerance limit by 10MW m2 resulting in serious sputtering of divertor materials Though impurity seeding is considered an eff ective way to control power handling in the divertor region by enhancing the radiation loss power the impurity at the boundary will partly be transported into the core plasma and degrade the core energy confi nement 28 Another approach to solve this problem is based on employing advanced diverted plasma confi gurations such as X divertor XD 29 super X divertor SXD 30 and snowfl ake divertor SFD 10 XD enhances fl ux expansion by producing an additional x point near the strike point whereas SXD improves upon this by additionally moving the plates to the largest radius inside the toroidal fi eld TF coils SFD is characterized by a second order null and the main se paratrix divides the poloidal plane into six sectors this confi guration is known as exact SF which is topologically unstable 31 Its two deri vatives i e SF and SF which possess a one order additional X point are more stable SF whose additional X point is in SOL has four strike points meaning that there are two other transport channels in the di vertor area It will be a challenge to design a divertor that is compatible with SND and SF SF has an additional X point located in the private fl ux region PFR In the previous design of CFETR it was easy to achieve SF and the PF coils could provide about 30 VS for the fl attop phase 8 However in the current large device the currents in some PF coils will extremely exceed their limits due to the greater interaction distance between the PF coils and the plasma as well as due to the higher plasma current and removal of DC2 as shown in Fig 3 Due to the limits of the coil currents and of space inside the vacuum vessel it is also diffi cult to achieve an XD or SXD for CFETR at the LFS The values of currents in CS4L PF2 3L will increase substantially There is also not suffi cient space to accommodate a long led divertor for the XD or SXD Therefore the distancedxptbetween the main X point x1 and the additional X point x2 should be increased to reduce the coil currents Hence an alternative confi guration termed quasi snowfl ake plus QSF was given more attention QSF is similar to SF but in this case thex2is outside the PFR Thedxpt has a strong infl uence on the currents in the PF coils and the magnetic topology in the vicinity of the X points In other words it is needed to fi nd an optimal value ofx2that can reduce the coil currents while sustaining the favorable characteristics of the QSF plasma e g large fl ux expansion and long connection length For quantitative as sessment of the eff ect ofx2 at diff erent positions polar coordinates