- Email: [email protected]
Overview of magnetic control in ITER
G Model FUSION-7396; No. of Pages 5
ARTICLE IN PRESS Fusion Engineering and Design xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Fusion Engineering and Design journal homepage: www.elsevier.com/locate/fusengdes
Overview of magnetic control in ITER L. Zabeo a,∗ , G. Ambrosino b , M. Cavinato c , Y. Gribov a , A. Kavin d , V. Lukash e , M. Mattei f , A. Pironti f , J.A. Snipes a , G. Vayakis a , A. Winter a a
ITER Organization, Route de Vinon sur Verdon, 13115 St. Paul Lez Durance, France CREATE/Universitá di Napoli Federico II, Dip. Ingegneria Elettrica e delle Tecnologie dell’informazione, Naples, Italy Fusion for Energy (F4E), Josep Pla 2, Torres Diagonal Litoral - B3, 08019 Barcelona, Spain d D.V. Efremov Scientific Research Institute, 196641 St. Petersburg, Russia e Kurchatov Institute, Moscow, Russia f CREATE/Seconda Universitá di Napoli, Dip. Ingegneria Industriale e dell’informazione, Naples, Italy b c
a r t i c l e
i n f o
Article history: Received 24 May 2013 Received in revised form 28 February 2014 Accepted 11 March 2014 Available online xxx Keyword: Plasma control
a b s t r a c t ITER is targeting Q = 10 with 500 MW of fusion power. To meet this target, the plasma needs to be controlled and shaped for a period of hundreds of seconds, avoiding contact with internal components, and acting against instabilities that could result in the loss of control of the plasma and in its disruptive termination. Axisymmetric magnetic control is a well-understood area being the basic control for any tokamak device. ITER adds more stringent constraints to the control primarily due to machine protection and engineering limits. The limits on the actuators by means of the maximum current and voltage at the coils and the few hundred ms time response of the vacuum vessel requires optimization of the control strategies and the validation of the capabilities of the machine in controlling the designed scenarios. Scenarios have been optimized with realistic control strategies able to guarantee robust control against plasma behavior and engineering limits due to recent changes in the ITER design. Technological issues such as performance changes associated with the optimization of the final design of the central solenoid, control of fast transitions like H to L mode to avoid plasma-wall contact, and optimization of the plasma ramp-down have been modeled to demonstrate the successful operability of ITER and compatibility with the latest refinements in the magnetic system design. Validation and optimization of the scenarios refining the operational space available for ITER and associated control strategies will be proposed. The present capabilities of magnetic control will be assessed and the remaining critical aspects that still need to be refined will be presented. The paper will also demonstrate the capabilities of the diagnostic system for magnetic control as a basic element for control. In fact, the noisy environment (affecting primarily vertical stability), the non-axisymmetric elements in the machine structure (affecting the accuracy of the identification of the plasma boundary), and the strong component of eddy current at the start-up (resulting in a poor S/N ratio for plasma reconstruction for Ip < 2 MA requiring a robust plasma control) make the ITER magnetic diagnostic system a demanding part of the magnetic control and investment protection systems. Finally the paper will illustrate the identified roles of magnetic control in the PCS (plasma control system) as formally defined in the recent first step of the design and development of the system. © 2014 ITER Organization. Published by Elsevier B.V. All rights reserved.
1. Introduction
∗ Corresponding author. Tel.: +33 4 42 17 65 24; fax: +33 4 42 17 65 00. E-mail addresses: [email protected] (L. Zabeo), [email protected] (G. Ambrosino), [email protected] (M. Cavinato), [email protected] (Y. Gribov), [email protected] (A. Kavin), lukash@nfi.kiae.ru (V. Lukash), [email protected] (M. Mattei), [email protected] (A. Pironti), [email protected] (J.A. Snipes), [email protected] (G. Vayakis), [email protected] (A. Winter).
Axisymmetric magnetic control for a tokamak device is the basic control for maintaining the plasma properly shaped (performance) and positioned in the vacuum vessel (avoid close proximity with the internal components at high plasma performance). In the case of ITER [1], magnetic control will also play an important role in machine protection because of the high plasma energy and electromagnetic forces.
http://dx.doi.org/10.1016/j.fusengdes.2014.03.051 0920-3796/© 2014 ITER Organization. Published by Elsevier B.V. All rights reserved.
Please cite this article in press as: L. Zabeo, et al., Overview of magnetic control in ITER, Fusion Eng. Des. (2014), http://dx.doi.org/10.1016/j.fusengdes.2014.03.051
G Model FUSION-7396; No. of Pages 5
ARTICLE IN PRESS
2
L. Zabeo et al. / Fusion Engineering and Design xxx (2014) xxx–xxx
Fig. 1. Layout of the magnetic actuators in ITER. Poloidal field (PF) coils, central solenoid (CS) coils and vertical stabilization systems (VS1 and VS3).
It is a common approach allocating three main actions to the magnetic control areas such as driving the plasma current, control the shape and the position of the plasma [2]. Plasma initiation is also usually included in the magnetic control area by means of requiring the control of the magnetic actuators. The paper will illustrate the present understanding of the magnetic control in ITER with attention also to the uncertainties and open questions still under investigation. In the paper, space will be given to both the plasma scenarios development and the controller design. 2. Plasma scenario optimization A wide set of analysis have been performed so far with the primary purpose to validate the capabilities of the ITER machine in achieving the desired plasma performance while still remaining inside the technical limits of the device. Those limitations are on the performance of the actuators (e.g. maximum current and voltage at the coils) and the machine structural integrity (e.g. maximum thermal load on first wall or electromagnetic forces at the coils). Scenarios have been simulated in all the phases from the plasma breakdown, ramp-up, flat-top to ramp-down showing that the present design of ITER is able to pursue its goals. Naturally, in the process of the development of the scenarios, modifications have been made and simulations repeated each time a change in the performance of the actuators or an additional restriction coming from one of the plant system of the machine was identified. Alternatives and adjustments to the scenarios have been tested in order to verify the largest set of possible conditions that ITER would encounter. Input for the design includes also specific plasma events and conditions that could compromise the performance of ITER and in some cases jeopardize the controllability of the plasma with the risk of damaging first wall components. In the following, the main activities in the respect of the scenario design will be reported with a particular emphasis on the magnetic control aspects (Fig. 1). 2.1. Plasma initiation Plasma initiation in ITER defines the interval of the start of discharge of the central solenoid (CS) followed by the gas breakdown and impurity burn-through till plasma current of about 0.5 MA, when the control of the plasma position is activated. Breakdown
occurs under specific conditions [3] such as the applied voltage history, the quality of the field null region and the vertical field evolution. Additionally, a radio frequency source from an EC system is envisaged for ITER [4]. The magnetic system needs to prove the ability to realize a high quality field null region (<∼2 mT) simultaneous with the maximum toroidal electric field (>∼0.3 V/m), followed by the correct evolution of the magnetic field to provide stable radial and vertical plasma equilibrium for the rising plasma current. Usually performed with pre-programmed actions, optimization in the magnetic distribution will be probably required by means of a feedback control. Optimization is also desired in order to maximize the flux at the breakdown leaving more flux for sustaining the long in ITER. The analyses have been carried out by means of simulations with different codes. Transport and equilibrium codes have been used together with the purpose to simulate the full conditions for the plasma formation in a realistic way. A first set of simulations makes use of the TRANSMAK [5] code for approaching the conditions for plasma breakdown and the current ramp-up till about 0.5 MA [6]. In parallel a different set of simulations has also been proposed [7]. That optimizes the set of voltages to apply to the coils prior the breakdown. Then the plasma equilibrium was calculated by using the CREATE-NL [8]. Several representative scenarios of plasma initiation have been studied such as inboard and outboard with partly (∼60 Wb) or fully (∼118 Wb) charged CS and ‘first plasma’ scenario. Some important results can be summarized as follows: • The operating space for Ohmic breakdown is extremely limited. Electron cyclotron assistance is required to guarantee a reliable breakdown and impurity burn-through. • There is a rather large region of low field (<3 mT in the area with minor radius ∼1.6 m) in which one or more field null points can be found at the breakdown time. Non-axisymmetric eddy currents, ferromagnetic elements closely located to the plasma (ferromagnetic inserts and test blanket modules) and measurement errors in the coil currents could reduce precision in the control of magnetic configuration with the possibility of undesired breakdown. • The flux loss at breakdown is in the range of 8 Wb for all the cases with an external voltage at the center of the plasma formation region between 12 V and 14 V. An Ip current ramp rate of about 1 MA/s can be achieved up to Ip = 0.5 MA for all the cases. • PF3–PF5 voltages are fully saturated in most of the cases. CS voltages are used at the maximum of their capabilities leaving a reduced margin of control for the plasma rise. • Forces and fields in coils remain within the imposed limits although for fully charged CS cases some of them are very close to limits (vertical forces on the coils PF1 and PF6, max field on the CS conductors). • The present limitation of the grid is on the order of 500 MW, where the analyses show a variation of the required peak power during the breakdown between 400 MW and 800 MW. Optimization is still possible to remain inside the limit. 2.2. From plasma rump-up to termination The baseline plasma scenarios [1] have been designed and simulated. Two major sets of analysis have been performed in order to provide a representative operational space for ITER within the specifications of the machine. In here we will focus on the axisymmetric magnetic control aspects without disputing the physics results, assuming that the physics target has been achieved. We will report the main results on the 15MA–DT scenario, main goal for ITER, although many others have been simulated.
Please cite this article in press as: L. Zabeo, et al., Overview of magnetic control in ITER, Fusion Eng. Des. (2014), http://dx.doi.org/10.1016/j.fusengdes.2014.03.051
G Model FUSION-7396; No. of Pages 5
ARTICLE IN PRESS L. Zabeo et al. / Fusion Engineering and Design xxx (2014) xxx–xxx
2.2.1. DINA simulation of plasma scenarios Simulations of the whole scenarios have been performed using the DINA code [9] from the start of the discharge till the end of the plasma. Some results are reported in [10], [11]. In DINA, the simulations start with the resistances of the switching network units (SNU) and waveforms of the feed-forward voltages in all CS and PF coils as taken from the corresponding TRANSMAK simulation up to the first available equilibrium (around 0.1 MA). When the plasma current increases to about 0.4 MA the feed-forward voltages were corrected by feedback voltages produced by the system controlling the plasma vertical and radial positions. During the simulation the profile of plasma current is calculated in a self-consistent way. The simulations assumed an early X-point formation (at 3.5 MA) during the plasma current ramp-up, the plasma current rampdown in a divertor configuration, and vertical stabilization with external coils (vertical stabilization system VS1, varying differential current in the coils PF2 to PF5) and/or in-vessel (system VS3, varying the current in the in-vessel coils connected in series for generating of radial magnetic field). The simulations take into account the dynamic characteristics of the power supplies and the limits on the coil voltages and currents. The current limits were taken into account in the control algorithm for limiting also the value of the magnetic field on the coil PF6 during the current ramp-up at low li . One of the engineering constraints imposed on the PF system is the limit on vertical electromagnetic forces acting on the coils and in particular on the CS modules. The most critical appears to be the vertical downward force acting on the CSL3 module. This vertical force depends on the current in this module. One proposed solution to cope with the limit is a separated feedback loop on the current in CS3L applied via a feed-forward current although in this case CS3L does not participate in the feedback control of the plasma current, position and shape. The majority of the DINA simulations were performed with the plasma vertical stabilization by VS1 system only by using a new more robust controller with higher control margin, allowing plasma vertical stabilization with a RMS value of low frequency noise in the vertical speed diagnostic signal up to 0.6 m/s (1 kHz bandwidth). That choice generated a significant oscillation and an increase in the AC losses [12] in the PF, coils requiring additional studies and controller optimization. Fewer scenarios were simulated using the in-vessel coils with a higher noise up to 1.2 m/s on the vertical speed and with two different controllers (VS3 only or combination of VS3 and VS1). One of the main results indicates that the use of only VS3 could end up in an excessive temperature of the conductors while the combined set would perform inside the limits. It should be noted, that different designs of the plasma current and shape controllers are required depending on the version of the plasma vertical stabilization system. Based on the power supply and coil capabilities, these analyses indicate that the fastest plasma current ramp-up to 15 MA can be performed in 50 s where the fastest ramp-down can be performed in about 65 s. The longest ohmic plasma current ramp-down lasts about 300 s (with auxiliary heating even up to 1400 s obtained for an extended burning plasma scenario simulation [11]). 2.2.2. CREATE-NL and JINTRAC simulation of plasma scenarios Simulation of ITER plasma scenarios from plasma current rampup till plasma termination was also performed with a suite comprising the 1.5D code JINTRAC [13] and the free-boundary equilibrium evolution code CREATE-NL. Some results are reported in [14]. Results on a set of the most representative scenarios have demonstrated the capabilities of the ITER PF and CS systems to meet the target confirming the previous analyses. The results are not exactly the same because of the different control solutions and
3
scenario development choices but the order of magnitude is the same. The 15MA-DT inductive scenario has been designed and simulated adopting different solutions. Standard ramp-up of about 80 s followed by a ∼500 s of flat-top and with the longest ramp-down of about 200 s is achievable with the ITER PF/CS. There is a low voltage margin in the PF system with cases of saturation during breakdown, initial ramp-up and L–H/H–L transitions [15]. A beneficial H–L approach would be delaying the event until lower plasma current (around 10 MA) during the ramp-down solution. During the flat-top, there is a tendency of an increase on the vertical forces in the PF5 and CS modules In addition, for long pulses, the CS currents are fully consumed leaving a small control margin at the final stage of the discharge, if needed. Space for optimization is nevertheless available on both scenario definition and control strategy. As for the previous simulations, the use of VS3 is strongly suggested not only to guarantee a higher controllable vertical displacement (up to ∼16 cm) but also to improve control during H–L transitions and ramp-down. Additional analyses are in progress with the purpose of providing higher performance vertical controllers using VS1, VS3 or a combination of the two systems. It is important to mention that the number of scenarios investigated is much wider than the single 15 MA scenario presented here. Although the overall conditions for plasma magnetic control may be less demanding, they are extremely important for quantifying the operational space of ITER. For example an additional scenario of interest is the 7.5 MA half field with a flat-top duration of about 1000 s. This case shows that the ITER magnetic system is able to guarantee the necessary conditions. The ramp-up phase is fundamental for allowing the plasma to reach the needed conditions, such as li , ˇp or density while avoiding conditions that could lead to instabilities. The ramp-down as well is a critical phase, not only for the need to guarantee a proper control of the plasma to avoid proximity with the first wall, but also to provide a set of optional termination scenarios based on the overall conditions, including failure modes. Those two phases, of course, depend on the plasma scenario at the flat-top requiring dedicated development. 2.3. Fast events ITER needs to be able to respond properly to fast plasma events such as ELMs [16], minor vertical displacements and confinement transitions (H–L and L–H). The slow response time of the vessel and the limits in the actuators make the control complex. Moreover, those events must be controlled in ITER because of the significant impact on in-vessel components (i.e. disruptions and high thermal load on the wall). Dedicated control solutions need to be identified and implemented in the control system. Discrete ELMs at 10 MA and minor disruptions at 15 MA during the flat-top seem to be controllable with the present ITER design and with a minor optimization of the controllers. With the in-vessel coils (VS3), a vertical displacement of up to ∼16 cm appears achievable, although only a few centimeter displacements is permitted with the external coils (VS1) alone. Instead, an H to L transition at 15 MA is difficult to control, causing voltage saturation. For avoiding plasma contact with the wall in the case of unexpected H to L mode transition, minimum value of the innermost plasma-wall gap needs to be increased depending on how close to the engineering limit is the current of CS1 module [17]. Feed-forward associated with feedback control could also help in the transition as demonstrated in [18]. However the feasibility of the technique depends on the capability of predicting the H–L transition at least a few seconds in advance, which still needs to be demonstrated.
Please cite this article in press as: L. Zabeo, et al., Overview of magnetic control in ITER, Fusion Eng. Des. (2014), http://dx.doi.org/10.1016/j.fusengdes.2014.03.051
G Model FUSION-7396; No. of Pages 5
ARTICLE IN PRESS
4
L. Zabeo et al. / Fusion Engineering and Design xxx (2014) xxx–xxx
Fig. 2. Reference control schema for plasma current, position and shape.
3. Controllers, diagnostics and actuators The results presented clearly indicate the importance of the design of optimized controllers able to perform inside the stringent ITER constraints. The design needs to carefully take into account the machine characteristics (i.e. double layer thick vessel with a ∼300 ms response time), actuator performance (i.e. voltage and current limits and coils forces), and the diagnostics details (i.e. noise level). 3.1. Controllers The standard control schema (Fig. 2 ) used in many of the previous simulations makes use of two feedback loops operating on different timescales. The fast loop performs the plasma vertical stabilization and the slow loop the plasma current, position and shape controls. More complex solutions have been proposed for both the loops making use of different optimization techniques aimed at improving robustness and performance to the control. The two available systems for the vertical stabilization, VS1 and VS3, provide different performance if used alone or in combination. Performance needs to be estimated in all cases to define the operational space under all possible conditions. The internal coils will operate in a very critical environment with a risk of long term failure. As already mentioned, the plasma needs to be maintained far from the vessel and prompt action is required when abrupt movements occur in case of rapid transient (drop or step) in the plasma. Developments are planned with the goal of providing ITER with a set of controllers to be used in simulations and later during operation. Evolution in the control area will, nevertheless, be required during the operational phase when the real plasma behavior and environment conditions will be available.
been used for the scenario design. A low frequency noise in the signal is expected for ITER. Studies performed in JET and ASDEX-U [20] extrapolated for ITER, indicate that the expected plasma noise can be assumed band-limited at 1 kHz white noise with a standard deviation between 0.2 and 0.6 m/s. Those values could limit the performance of VS1 where a higher noise level is still acceptable if VS3 is used. In order to investigate more realistically the vertical control issue, an estimator for the vertical velocity of the plasma has been designed starting from the magnetic measurements. The performance of the vertical control system has been tested also making use of this estimator proving positive results [20]. Plasma shape: the position and shape of the plasma is detected by means of a set of gaps (distance plasma-wall in pre-defined positions) that can vary in number from one controller to another. An extended set of studies reports generally positive responses on the achievable performance of the magnetic system with an accuracy within the ITER requirements [21]. Unfortunately, especially for long pulses the drift in the integrators for the measurements introduces a significant error in the estimation. Those cases could be critical for control although the low plasma current and energy during those phases make the problem less dangerous for the machine integrity. The effect of the non-axisymmetric components of the magnetic field due to distributed non-axisymmetric sources could also be a problem reducing the accuracy of the measurements. A final assessment on the capability of detecting those corrupting components with sufficient accuracy so to correct the parameters for control has not been given yet although preliminary results have been provided. The high level of eddy currents together with noise on the measurements could pose some limitations in the accuracy in the plasma reconstruction during the first phase of the discharge for Ip < 2 MA. An estimation of the eddy current distribution is probably needed although the decay time should be short in comparison to the control of the shape. A final evaluation still needs to be provided.
3.3. Actuators The power supply system in ITER (Fig. 3 ) has been already established within the constraint posed, in particular, by the structural limits of the superconducting coils. Those limitations concern not only the maximum voltage and current but also the maximum allowable electromagnetic forces as due to the cross coupling between coils. All the scenarios have been designed trying to keep those forces far from the limits although some cases show a small margin. Not being space for new a design with more powerful power supply or coil systems, it the control system mandate to optimize their usage.
4. PCS roles 3.2. Diagnostics The primary diagnostic dedicated to the axisymmetric magnetic control is the set of magnetic measurements from which the plasma current, plasma shape and position can be calculated. A large number of magnetic measurements will be installed in ITER [19]. The high coverage of the machine and the redundancy in the detectors seem to be sufficient for ITER purposes. Nonetheless achievable accuracy in the plasma geometry and the level of noise, partially still unknown, could make the control of the plasma difficult. Plasma current: the requirements seem to be achievable with the present set of diagnostics and with a good redundancy. Vertical velocity: simple models for the vertical velocity for the vertical stabilization with the injection of an artificial noise have
Axisymmetric magnetic control is part of a wide set of controls that will be installed in the plasma control system in ITER [22], [23]. Recently a first step in the design of the PCS – plasma control system has concluded. The requirements and functionalities for the system have been identified for each of the control areas. The requirements for the diagnostics and the actuators have been also partially discussed. Axisymmetric magnetic control has been extensively discussed as well although it is probably the most advanced in development. The functional breakdown given to axisymmetric control includes also support functionalities and additional control loops as support for the magnetic control. Plasma boundary reconstruction, magnetic field distribution identification together with current
Please cite this article in press as: L. Zabeo, et al., Overview of magnetic control in ITER, Fusion Eng. Des. (2014), http://dx.doi.org/10.1016/j.fusengdes.2014.03.051
G Model FUSION-7396; No. of Pages 5
ARTICLE IN PRESS L. Zabeo et al. / Fusion Engineering and Design xxx (2014) xxx–xxx
5
Fig. 3. Power supply system design for PF, CS, VS1 and CC (correction coils).
loop control and saturation avoidance were some of the supporting elements widely discussed during the review. The next milestone for the PCS is to provide a first possible solution for the implementation of the system together with a possible implementation for the algorithms and control system architecture as it is required for the first plasma and early operation phase. In the axisymmetric magnetic control area the important targets would be the basic control of the actuators with the possibility to optimize the field distribution for the null point maybe with a feedback loop as well. 5. Future work and conclusions The paper has presented the actual understanding of the axisymmetric magnetic control in ITER. Although in some conditions ITER will probably work with small margin of control, the general considerations are positive confirming the achievable ITER targets with the actual design. Studies and analysis need still to be performed so to clarify and confirm critical situations that could increase complexity in the control of the plasma. Acknowledgments The authors would like to thank all the people from the IO and Domestic Agencies for their support and for reports and analyses essential for the success of the ITER project. The views and opinions expressed herein do not necessarily reflect those of the ITER Organization. References [1] Progress in ITER physics basis. Nucl. Fusion, 2007, 47 (Special Issue). [2] M. Ariola, Pironti A., Magnetic Control of Tokamak Plasmas, Springer-Verlag, London, 2008. [3] Y. Gribov, D. Humphreys, K. Kajiwara, E.A. Lazarus, J.B. Lister, T. Ozeki, et al., Plasma operation and control, Nucl. Fusion 47 (2007), Chapter 8. [4] B. Lloyd, P.G. Carolan, C.D. Warrick, ECRH-assisted start-up in ITER, Plasma Phys. Control. Fusion 38 (1996) 1627.
[5] V.A. Belyakov, Plasma initiation stage analysis in tokamaks with TRANSMAK code Plasma Device and Operations 11 (3) (2003). [6] RA-DA, Study of plasma initiation using TRANSMAK code: ITER design “Baseline 2010”, ITER private communication, December 2010. [7] EU-DA, Task on the study of plasma start-up, ITER private communication, December 2011. [8] R. Albanese, M. Mattei, G. Calabrò, F. Villone, ISEM, 2003, pp. 404–405. [9] R.R. Khayrutdinov, V.E. Lukash, Studies of plasma equilibrium and transport in a tokamak fusion device with the inverse-variable technique, J. Comp. Phys. 109 (1993) 193–201. [10] V.E. Lukash, A.A. Kavin, Y.V. Gribov, R.R. Khayrutdinov, A.B. Mineev, Simulation of ITER plasma scenarios starting from initial discharge of central solenoid, in: 38th EPS Conference on Plasma Physics, 27 June–1 July 2011, Strasbourg, France, 2011, P2.109. [11] V.E. Lukash, A.A. Kavin, Y.V. Gribov, R.R. Khayrutdinov, A.B. Mineev, Study of burn duration in ITER with reduced internal stress in the bottom module of the central solenoid, in: 39th Conference & 16th International Congress on Plasma Physics, 2–6 July 2012, Stockholm, Sweden, 2012, P5.068. [12] L. Bottura, P. Bruzzone, J.B. Lister, C. Marinucci, A. Portone, Computation of AC losses in the ITER magnets during fast field transients, IEEE Trans. Appl. Superc. 17 (2) (2007) 2438–2441. [13] S. Weisen, JINTRAC-JET modelling suite, JET-ITC-Report 2008. [14] V. Parail, R. Albanese, R. Ambrosino, J-F. Artaud, K. Besseghir, M. Cavinato, et al., Self-consistent simulation of plasma scenarios for ITER using a combination of 1.5D transport codes and free boundary equilibrium codes, in: 24th IAEA Fusion Energy Conference, October 8–13, 2012, San Diego, USA, 2012, ITR/P 1-10. [15] P. Gohil, T.E. Evans, M.E. Fenstermacher, J.R. Ferron, T.H. Osborne, J.M. Park, et al., L–H transition studies on DIII-D to determine H-mode access for operational scenarios in ITER Nuclear Fusion 51 (August (10)) (2011). [16] M. Bécoulet, G. Huysmans, Y. Sarazin, X. Garbet, Ph. Ghendrih, F. Rimini, et al., Edge localized mode physics and operational aspects in tokamaks, Plasma Phys. Cont. Fusion 45 (2003) A93–A113. [17] RF-DA, Design and simulation of 15MA DT scenarios with DINA code: ITER design “Baseline 2010”, ITER private communication, September 2010. [18] EU-DA, Study of control of plasma current, position and shape, ITER private communication, December 2011. [19] D. Testa, M. Toussaint, R. Chavan, J. Guterl, J.B. Lister, J.-M. Moret, et al., The magnetic diagnostic set for ITER, IEEE Trans. Plasma Sci. 38 (March (3)) (2010). [20] G. Ambrosino, R. Albanese, M. Ariola, G. De Tommasi, R. Fresa, M. Mattei, et al., Plasma position and shape control for ITER scenarios, Final Report on EFDA Study Contract 07-1702/1579, 2007. [21] EU-DA ITER private communication. [22] J.A. Snipes, Physics of the conceptual design of the ITER Plasma Control System, this conference. [23] A. Winter, Design progress of the ITER Plasma Control System, this conference.
Please cite this article in press as: L. Zabeo, et al., Overview of magnetic control in ITER, Fusion Eng. Des. (2014), http://dx.doi.org/10.1016/j.fusengdes.2014.03.051
ARTICLE IN PRESS Fusion Engineering and Design xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Fusion Engineering and Design journal homepage: www.elsevier.com/locate/fusengdes
Overview of magnetic control in ITER L. Zabeo a,∗ , G. Ambrosino b , M. Cavinato c , Y. Gribov a , A. Kavin d , V. Lukash e , M. Mattei f , A. Pironti f , J.A. Snipes a , G. Vayakis a , A. Winter a a
ITER Organization, Route de Vinon sur Verdon, 13115 St. Paul Lez Durance, France CREATE/Universitá di Napoli Federico II, Dip. Ingegneria Elettrica e delle Tecnologie dell’informazione, Naples, Italy Fusion for Energy (F4E), Josep Pla 2, Torres Diagonal Litoral - B3, 08019 Barcelona, Spain d D.V. Efremov Scientific Research Institute, 196641 St. Petersburg, Russia e Kurchatov Institute, Moscow, Russia f CREATE/Seconda Universitá di Napoli, Dip. Ingegneria Industriale e dell’informazione, Naples, Italy b c
a r t i c l e
i n f o
Article history: Received 24 May 2013 Received in revised form 28 February 2014 Accepted 11 March 2014 Available online xxx Keyword: Plasma control
a b s t r a c t ITER is targeting Q = 10 with 500 MW of fusion power. To meet this target, the plasma needs to be controlled and shaped for a period of hundreds of seconds, avoiding contact with internal components, and acting against instabilities that could result in the loss of control of the plasma and in its disruptive termination. Axisymmetric magnetic control is a well-understood area being the basic control for any tokamak device. ITER adds more stringent constraints to the control primarily due to machine protection and engineering limits. The limits on the actuators by means of the maximum current and voltage at the coils and the few hundred ms time response of the vacuum vessel requires optimization of the control strategies and the validation of the capabilities of the machine in controlling the designed scenarios. Scenarios have been optimized with realistic control strategies able to guarantee robust control against plasma behavior and engineering limits due to recent changes in the ITER design. Technological issues such as performance changes associated with the optimization of the final design of the central solenoid, control of fast transitions like H to L mode to avoid plasma-wall contact, and optimization of the plasma ramp-down have been modeled to demonstrate the successful operability of ITER and compatibility with the latest refinements in the magnetic system design. Validation and optimization of the scenarios refining the operational space available for ITER and associated control strategies will be proposed. The present capabilities of magnetic control will be assessed and the remaining critical aspects that still need to be refined will be presented. The paper will also demonstrate the capabilities of the diagnostic system for magnetic control as a basic element for control. In fact, the noisy environment (affecting primarily vertical stability), the non-axisymmetric elements in the machine structure (affecting the accuracy of the identification of the plasma boundary), and the strong component of eddy current at the start-up (resulting in a poor S/N ratio for plasma reconstruction for Ip < 2 MA requiring a robust plasma control) make the ITER magnetic diagnostic system a demanding part of the magnetic control and investment protection systems. Finally the paper will illustrate the identified roles of magnetic control in the PCS (plasma control system) as formally defined in the recent first step of the design and development of the system. © 2014 ITER Organization. Published by Elsevier B.V. All rights reserved.
1. Introduction
∗ Corresponding author. Tel.: +33 4 42 17 65 24; fax: +33 4 42 17 65 00. E-mail addresses: [email protected] (L. Zabeo), [email protected] (G. Ambrosino), [email protected] (M. Cavinato), [email protected] (Y. Gribov), [email protected] (A. Kavin), lukash@nfi.kiae.ru (V. Lukash), [email protected] (M. Mattei), [email protected] (A. Pironti), [email protected] (J.A. Snipes), [email protected] (G. Vayakis), [email protected] (A. Winter).
Axisymmetric magnetic control for a tokamak device is the basic control for maintaining the plasma properly shaped (performance) and positioned in the vacuum vessel (avoid close proximity with the internal components at high plasma performance). In the case of ITER [1], magnetic control will also play an important role in machine protection because of the high plasma energy and electromagnetic forces.
http://dx.doi.org/10.1016/j.fusengdes.2014.03.051 0920-3796/© 2014 ITER Organization. Published by Elsevier B.V. All rights reserved.
Please cite this article in press as: L. Zabeo, et al., Overview of magnetic control in ITER, Fusion Eng. Des. (2014), http://dx.doi.org/10.1016/j.fusengdes.2014.03.051
G Model FUSION-7396; No. of Pages 5
ARTICLE IN PRESS
2
L. Zabeo et al. / Fusion Engineering and Design xxx (2014) xxx–xxx
Fig. 1. Layout of the magnetic actuators in ITER. Poloidal field (PF) coils, central solenoid (CS) coils and vertical stabilization systems (VS1 and VS3).
It is a common approach allocating three main actions to the magnetic control areas such as driving the plasma current, control the shape and the position of the plasma [2]. Plasma initiation is also usually included in the magnetic control area by means of requiring the control of the magnetic actuators. The paper will illustrate the present understanding of the magnetic control in ITER with attention also to the uncertainties and open questions still under investigation. In the paper, space will be given to both the plasma scenarios development and the controller design. 2. Plasma scenario optimization A wide set of analysis have been performed so far with the primary purpose to validate the capabilities of the ITER machine in achieving the desired plasma performance while still remaining inside the technical limits of the device. Those limitations are on the performance of the actuators (e.g. maximum current and voltage at the coils) and the machine structural integrity (e.g. maximum thermal load on first wall or electromagnetic forces at the coils). Scenarios have been simulated in all the phases from the plasma breakdown, ramp-up, flat-top to ramp-down showing that the present design of ITER is able to pursue its goals. Naturally, in the process of the development of the scenarios, modifications have been made and simulations repeated each time a change in the performance of the actuators or an additional restriction coming from one of the plant system of the machine was identified. Alternatives and adjustments to the scenarios have been tested in order to verify the largest set of possible conditions that ITER would encounter. Input for the design includes also specific plasma events and conditions that could compromise the performance of ITER and in some cases jeopardize the controllability of the plasma with the risk of damaging first wall components. In the following, the main activities in the respect of the scenario design will be reported with a particular emphasis on the magnetic control aspects (Fig. 1). 2.1. Plasma initiation Plasma initiation in ITER defines the interval of the start of discharge of the central solenoid (CS) followed by the gas breakdown and impurity burn-through till plasma current of about 0.5 MA, when the control of the plasma position is activated. Breakdown
occurs under specific conditions [3] such as the applied voltage history, the quality of the field null region and the vertical field evolution. Additionally, a radio frequency source from an EC system is envisaged for ITER [4]. The magnetic system needs to prove the ability to realize a high quality field null region (<∼2 mT) simultaneous with the maximum toroidal electric field (>∼0.3 V/m), followed by the correct evolution of the magnetic field to provide stable radial and vertical plasma equilibrium for the rising plasma current. Usually performed with pre-programmed actions, optimization in the magnetic distribution will be probably required by means of a feedback control. Optimization is also desired in order to maximize the flux at the breakdown leaving more flux for sustaining the long in ITER. The analyses have been carried out by means of simulations with different codes. Transport and equilibrium codes have been used together with the purpose to simulate the full conditions for the plasma formation in a realistic way. A first set of simulations makes use of the TRANSMAK [5] code for approaching the conditions for plasma breakdown and the current ramp-up till about 0.5 MA [6]. In parallel a different set of simulations has also been proposed [7]. That optimizes the set of voltages to apply to the coils prior the breakdown. Then the plasma equilibrium was calculated by using the CREATE-NL [8]. Several representative scenarios of plasma initiation have been studied such as inboard and outboard with partly (∼60 Wb) or fully (∼118 Wb) charged CS and ‘first plasma’ scenario. Some important results can be summarized as follows: • The operating space for Ohmic breakdown is extremely limited. Electron cyclotron assistance is required to guarantee a reliable breakdown and impurity burn-through. • There is a rather large region of low field (<3 mT in the area with minor radius ∼1.6 m) in which one or more field null points can be found at the breakdown time. Non-axisymmetric eddy currents, ferromagnetic elements closely located to the plasma (ferromagnetic inserts and test blanket modules) and measurement errors in the coil currents could reduce precision in the control of magnetic configuration with the possibility of undesired breakdown. • The flux loss at breakdown is in the range of 8 Wb for all the cases with an external voltage at the center of the plasma formation region between 12 V and 14 V. An Ip current ramp rate of about 1 MA/s can be achieved up to Ip = 0.5 MA for all the cases. • PF3–PF5 voltages are fully saturated in most of the cases. CS voltages are used at the maximum of their capabilities leaving a reduced margin of control for the plasma rise. • Forces and fields in coils remain within the imposed limits although for fully charged CS cases some of them are very close to limits (vertical forces on the coils PF1 and PF6, max field on the CS conductors). • The present limitation of the grid is on the order of 500 MW, where the analyses show a variation of the required peak power during the breakdown between 400 MW and 800 MW. Optimization is still possible to remain inside the limit. 2.2. From plasma rump-up to termination The baseline plasma scenarios [1] have been designed and simulated. Two major sets of analysis have been performed in order to provide a representative operational space for ITER within the specifications of the machine. In here we will focus on the axisymmetric magnetic control aspects without disputing the physics results, assuming that the physics target has been achieved. We will report the main results on the 15MA–DT scenario, main goal for ITER, although many others have been simulated.
Please cite this article in press as: L. Zabeo, et al., Overview of magnetic control in ITER, Fusion Eng. Des. (2014), http://dx.doi.org/10.1016/j.fusengdes.2014.03.051
G Model FUSION-7396; No. of Pages 5
ARTICLE IN PRESS L. Zabeo et al. / Fusion Engineering and Design xxx (2014) xxx–xxx
2.2.1. DINA simulation of plasma scenarios Simulations of the whole scenarios have been performed using the DINA code [9] from the start of the discharge till the end of the plasma. Some results are reported in [10], [11]. In DINA, the simulations start with the resistances of the switching network units (SNU) and waveforms of the feed-forward voltages in all CS and PF coils as taken from the corresponding TRANSMAK simulation up to the first available equilibrium (around 0.1 MA). When the plasma current increases to about 0.4 MA the feed-forward voltages were corrected by feedback voltages produced by the system controlling the plasma vertical and radial positions. During the simulation the profile of plasma current is calculated in a self-consistent way. The simulations assumed an early X-point formation (at 3.5 MA) during the plasma current ramp-up, the plasma current rampdown in a divertor configuration, and vertical stabilization with external coils (vertical stabilization system VS1, varying differential current in the coils PF2 to PF5) and/or in-vessel (system VS3, varying the current in the in-vessel coils connected in series for generating of radial magnetic field). The simulations take into account the dynamic characteristics of the power supplies and the limits on the coil voltages and currents. The current limits were taken into account in the control algorithm for limiting also the value of the magnetic field on the coil PF6 during the current ramp-up at low li . One of the engineering constraints imposed on the PF system is the limit on vertical electromagnetic forces acting on the coils and in particular on the CS modules. The most critical appears to be the vertical downward force acting on the CSL3 module. This vertical force depends on the current in this module. One proposed solution to cope with the limit is a separated feedback loop on the current in CS3L applied via a feed-forward current although in this case CS3L does not participate in the feedback control of the plasma current, position and shape. The majority of the DINA simulations were performed with the plasma vertical stabilization by VS1 system only by using a new more robust controller with higher control margin, allowing plasma vertical stabilization with a RMS value of low frequency noise in the vertical speed diagnostic signal up to 0.6 m/s (1 kHz bandwidth). That choice generated a significant oscillation and an increase in the AC losses [12] in the PF, coils requiring additional studies and controller optimization. Fewer scenarios were simulated using the in-vessel coils with a higher noise up to 1.2 m/s on the vertical speed and with two different controllers (VS3 only or combination of VS3 and VS1). One of the main results indicates that the use of only VS3 could end up in an excessive temperature of the conductors while the combined set would perform inside the limits. It should be noted, that different designs of the plasma current and shape controllers are required depending on the version of the plasma vertical stabilization system. Based on the power supply and coil capabilities, these analyses indicate that the fastest plasma current ramp-up to 15 MA can be performed in 50 s where the fastest ramp-down can be performed in about 65 s. The longest ohmic plasma current ramp-down lasts about 300 s (with auxiliary heating even up to 1400 s obtained for an extended burning plasma scenario simulation [11]). 2.2.2. CREATE-NL and JINTRAC simulation of plasma scenarios Simulation of ITER plasma scenarios from plasma current rampup till plasma termination was also performed with a suite comprising the 1.5D code JINTRAC [13] and the free-boundary equilibrium evolution code CREATE-NL. Some results are reported in [14]. Results on a set of the most representative scenarios have demonstrated the capabilities of the ITER PF and CS systems to meet the target confirming the previous analyses. The results are not exactly the same because of the different control solutions and
3
scenario development choices but the order of magnitude is the same. The 15MA-DT inductive scenario has been designed and simulated adopting different solutions. Standard ramp-up of about 80 s followed by a ∼500 s of flat-top and with the longest ramp-down of about 200 s is achievable with the ITER PF/CS. There is a low voltage margin in the PF system with cases of saturation during breakdown, initial ramp-up and L–H/H–L transitions [15]. A beneficial H–L approach would be delaying the event until lower plasma current (around 10 MA) during the ramp-down solution. During the flat-top, there is a tendency of an increase on the vertical forces in the PF5 and CS modules In addition, for long pulses, the CS currents are fully consumed leaving a small control margin at the final stage of the discharge, if needed. Space for optimization is nevertheless available on both scenario definition and control strategy. As for the previous simulations, the use of VS3 is strongly suggested not only to guarantee a higher controllable vertical displacement (up to ∼16 cm) but also to improve control during H–L transitions and ramp-down. Additional analyses are in progress with the purpose of providing higher performance vertical controllers using VS1, VS3 or a combination of the two systems. It is important to mention that the number of scenarios investigated is much wider than the single 15 MA scenario presented here. Although the overall conditions for plasma magnetic control may be less demanding, they are extremely important for quantifying the operational space of ITER. For example an additional scenario of interest is the 7.5 MA half field with a flat-top duration of about 1000 s. This case shows that the ITER magnetic system is able to guarantee the necessary conditions. The ramp-up phase is fundamental for allowing the plasma to reach the needed conditions, such as li , ˇp or density while avoiding conditions that could lead to instabilities. The ramp-down as well is a critical phase, not only for the need to guarantee a proper control of the plasma to avoid proximity with the first wall, but also to provide a set of optional termination scenarios based on the overall conditions, including failure modes. Those two phases, of course, depend on the plasma scenario at the flat-top requiring dedicated development. 2.3. Fast events ITER needs to be able to respond properly to fast plasma events such as ELMs [16], minor vertical displacements and confinement transitions (H–L and L–H). The slow response time of the vessel and the limits in the actuators make the control complex. Moreover, those events must be controlled in ITER because of the significant impact on in-vessel components (i.e. disruptions and high thermal load on the wall). Dedicated control solutions need to be identified and implemented in the control system. Discrete ELMs at 10 MA and minor disruptions at 15 MA during the flat-top seem to be controllable with the present ITER design and with a minor optimization of the controllers. With the in-vessel coils (VS3), a vertical displacement of up to ∼16 cm appears achievable, although only a few centimeter displacements is permitted with the external coils (VS1) alone. Instead, an H to L transition at 15 MA is difficult to control, causing voltage saturation. For avoiding plasma contact with the wall in the case of unexpected H to L mode transition, minimum value of the innermost plasma-wall gap needs to be increased depending on how close to the engineering limit is the current of CS1 module [17]. Feed-forward associated with feedback control could also help in the transition as demonstrated in [18]. However the feasibility of the technique depends on the capability of predicting the H–L transition at least a few seconds in advance, which still needs to be demonstrated.
Please cite this article in press as: L. Zabeo, et al., Overview of magnetic control in ITER, Fusion Eng. Des. (2014), http://dx.doi.org/10.1016/j.fusengdes.2014.03.051
G Model FUSION-7396; No. of Pages 5
ARTICLE IN PRESS
4
L. Zabeo et al. / Fusion Engineering and Design xxx (2014) xxx–xxx
Fig. 2. Reference control schema for plasma current, position and shape.
3. Controllers, diagnostics and actuators The results presented clearly indicate the importance of the design of optimized controllers able to perform inside the stringent ITER constraints. The design needs to carefully take into account the machine characteristics (i.e. double layer thick vessel with a ∼300 ms response time), actuator performance (i.e. voltage and current limits and coils forces), and the diagnostics details (i.e. noise level). 3.1. Controllers The standard control schema (Fig. 2 ) used in many of the previous simulations makes use of two feedback loops operating on different timescales. The fast loop performs the plasma vertical stabilization and the slow loop the plasma current, position and shape controls. More complex solutions have been proposed for both the loops making use of different optimization techniques aimed at improving robustness and performance to the control. The two available systems for the vertical stabilization, VS1 and VS3, provide different performance if used alone or in combination. Performance needs to be estimated in all cases to define the operational space under all possible conditions. The internal coils will operate in a very critical environment with a risk of long term failure. As already mentioned, the plasma needs to be maintained far from the vessel and prompt action is required when abrupt movements occur in case of rapid transient (drop or step) in the plasma. Developments are planned with the goal of providing ITER with a set of controllers to be used in simulations and later during operation. Evolution in the control area will, nevertheless, be required during the operational phase when the real plasma behavior and environment conditions will be available.
been used for the scenario design. A low frequency noise in the signal is expected for ITER. Studies performed in JET and ASDEX-U [20] extrapolated for ITER, indicate that the expected plasma noise can be assumed band-limited at 1 kHz white noise with a standard deviation between 0.2 and 0.6 m/s. Those values could limit the performance of VS1 where a higher noise level is still acceptable if VS3 is used. In order to investigate more realistically the vertical control issue, an estimator for the vertical velocity of the plasma has been designed starting from the magnetic measurements. The performance of the vertical control system has been tested also making use of this estimator proving positive results [20]. Plasma shape: the position and shape of the plasma is detected by means of a set of gaps (distance plasma-wall in pre-defined positions) that can vary in number from one controller to another. An extended set of studies reports generally positive responses on the achievable performance of the magnetic system with an accuracy within the ITER requirements [21]. Unfortunately, especially for long pulses the drift in the integrators for the measurements introduces a significant error in the estimation. Those cases could be critical for control although the low plasma current and energy during those phases make the problem less dangerous for the machine integrity. The effect of the non-axisymmetric components of the magnetic field due to distributed non-axisymmetric sources could also be a problem reducing the accuracy of the measurements. A final assessment on the capability of detecting those corrupting components with sufficient accuracy so to correct the parameters for control has not been given yet although preliminary results have been provided. The high level of eddy currents together with noise on the measurements could pose some limitations in the accuracy in the plasma reconstruction during the first phase of the discharge for Ip < 2 MA. An estimation of the eddy current distribution is probably needed although the decay time should be short in comparison to the control of the shape. A final evaluation still needs to be provided.
3.3. Actuators The power supply system in ITER (Fig. 3 ) has been already established within the constraint posed, in particular, by the structural limits of the superconducting coils. Those limitations concern not only the maximum voltage and current but also the maximum allowable electromagnetic forces as due to the cross coupling between coils. All the scenarios have been designed trying to keep those forces far from the limits although some cases show a small margin. Not being space for new a design with more powerful power supply or coil systems, it the control system mandate to optimize their usage.
4. PCS roles 3.2. Diagnostics The primary diagnostic dedicated to the axisymmetric magnetic control is the set of magnetic measurements from which the plasma current, plasma shape and position can be calculated. A large number of magnetic measurements will be installed in ITER [19]. The high coverage of the machine and the redundancy in the detectors seem to be sufficient for ITER purposes. Nonetheless achievable accuracy in the plasma geometry and the level of noise, partially still unknown, could make the control of the plasma difficult. Plasma current: the requirements seem to be achievable with the present set of diagnostics and with a good redundancy. Vertical velocity: simple models for the vertical velocity for the vertical stabilization with the injection of an artificial noise have
Axisymmetric magnetic control is part of a wide set of controls that will be installed in the plasma control system in ITER [22], [23]. Recently a first step in the design of the PCS – plasma control system has concluded. The requirements and functionalities for the system have been identified for each of the control areas. The requirements for the diagnostics and the actuators have been also partially discussed. Axisymmetric magnetic control has been extensively discussed as well although it is probably the most advanced in development. The functional breakdown given to axisymmetric control includes also support functionalities and additional control loops as support for the magnetic control. Plasma boundary reconstruction, magnetic field distribution identification together with current
Please cite this article in press as: L. Zabeo, et al., Overview of magnetic control in ITER, Fusion Eng. Des. (2014), http://dx.doi.org/10.1016/j.fusengdes.2014.03.051
G Model FUSION-7396; No. of Pages 5
ARTICLE IN PRESS L. Zabeo et al. / Fusion Engineering and Design xxx (2014) xxx–xxx
5
Fig. 3. Power supply system design for PF, CS, VS1 and CC (correction coils).
loop control and saturation avoidance were some of the supporting elements widely discussed during the review. The next milestone for the PCS is to provide a first possible solution for the implementation of the system together with a possible implementation for the algorithms and control system architecture as it is required for the first plasma and early operation phase. In the axisymmetric magnetic control area the important targets would be the basic control of the actuators with the possibility to optimize the field distribution for the null point maybe with a feedback loop as well. 5. Future work and conclusions The paper has presented the actual understanding of the axisymmetric magnetic control in ITER. Although in some conditions ITER will probably work with small margin of control, the general considerations are positive confirming the achievable ITER targets with the actual design. Studies and analysis need still to be performed so to clarify and confirm critical situations that could increase complexity in the control of the plasma. Acknowledgments The authors would like to thank all the people from the IO and Domestic Agencies for their support and for reports and analyses essential for the success of the ITER project. The views and opinions expressed herein do not necessarily reflect those of the ITER Organization. References [1] Progress in ITER physics basis. Nucl. Fusion, 2007, 47 (Special Issue). [2] M. Ariola, Pironti A., Magnetic Control of Tokamak Plasmas, Springer-Verlag, London, 2008. [3] Y. Gribov, D. Humphreys, K. Kajiwara, E.A. Lazarus, J.B. Lister, T. Ozeki, et al., Plasma operation and control, Nucl. Fusion 47 (2007), Chapter 8. [4] B. Lloyd, P.G. Carolan, C.D. Warrick, ECRH-assisted start-up in ITER, Plasma Phys. Control. Fusion 38 (1996) 1627.
[5] V.A. Belyakov, Plasma initiation stage analysis in tokamaks with TRANSMAK code Plasma Device and Operations 11 (3) (2003). [6] RA-DA, Study of plasma initiation using TRANSMAK code: ITER design “Baseline 2010”, ITER private communication, December 2010. [7] EU-DA, Task on the study of plasma start-up, ITER private communication, December 2011. [8] R. Albanese, M. Mattei, G. Calabrò, F. Villone, ISEM, 2003, pp. 404–405. [9] R.R. Khayrutdinov, V.E. Lukash, Studies of plasma equilibrium and transport in a tokamak fusion device with the inverse-variable technique, J. Comp. Phys. 109 (1993) 193–201. [10] V.E. Lukash, A.A. Kavin, Y.V. Gribov, R.R. Khayrutdinov, A.B. Mineev, Simulation of ITER plasma scenarios starting from initial discharge of central solenoid, in: 38th EPS Conference on Plasma Physics, 27 June–1 July 2011, Strasbourg, France, 2011, P2.109. [11] V.E. Lukash, A.A. Kavin, Y.V. Gribov, R.R. Khayrutdinov, A.B. Mineev, Study of burn duration in ITER with reduced internal stress in the bottom module of the central solenoid, in: 39th Conference & 16th International Congress on Plasma Physics, 2–6 July 2012, Stockholm, Sweden, 2012, P5.068. [12] L. Bottura, P. Bruzzone, J.B. Lister, C. Marinucci, A. Portone, Computation of AC losses in the ITER magnets during fast field transients, IEEE Trans. Appl. Superc. 17 (2) (2007) 2438–2441. [13] S. Weisen, JINTRAC-JET modelling suite, JET-ITC-Report 2008. [14] V. Parail, R. Albanese, R. Ambrosino, J-F. Artaud, K. Besseghir, M. Cavinato, et al., Self-consistent simulation of plasma scenarios for ITER using a combination of 1.5D transport codes and free boundary equilibrium codes, in: 24th IAEA Fusion Energy Conference, October 8–13, 2012, San Diego, USA, 2012, ITR/P 1-10. [15] P. Gohil, T.E. Evans, M.E. Fenstermacher, J.R. Ferron, T.H. Osborne, J.M. Park, et al., L–H transition studies on DIII-D to determine H-mode access for operational scenarios in ITER Nuclear Fusion 51 (August (10)) (2011). [16] M. Bécoulet, G. Huysmans, Y. Sarazin, X. Garbet, Ph. Ghendrih, F. Rimini, et al., Edge localized mode physics and operational aspects in tokamaks, Plasma Phys. Cont. Fusion 45 (2003) A93–A113. [17] RF-DA, Design and simulation of 15MA DT scenarios with DINA code: ITER design “Baseline 2010”, ITER private communication, September 2010. [18] EU-DA, Study of control of plasma current, position and shape, ITER private communication, December 2011. [19] D. Testa, M. Toussaint, R. Chavan, J. Guterl, J.B. Lister, J.-M. Moret, et al., The magnetic diagnostic set for ITER, IEEE Trans. Plasma Sci. 38 (March (3)) (2010). [20] G. Ambrosino, R. Albanese, M. Ariola, G. De Tommasi, R. Fresa, M. Mattei, et al., Plasma position and shape control for ITER scenarios, Final Report on EFDA Study Contract 07-1702/1579, 2007. [21] EU-DA ITER private communication. [22] J.A. Snipes, Physics of the conceptual design of the ITER Plasma Control System, this conference. [23] A. Winter, Design progress of the ITER Plasma Control System, this conference.
Please cite this article in press as: L. Zabeo, et al., Overview of magnetic control in ITER, Fusion Eng. Des. (2014), http://dx.doi.org/10.1016/j.fusengdes.2014.03.051