ct white paper-fesac-v4

FESAC TEC White Paper on CT System, R. Raman, May 28 (2017)

1

Momentum Injection and Precise Core Fueling for Reactor Grade Plasmas R. Raman

University of Washington, Seattle, WA, USA Email: [email protected]

1. Technology to be assessed

The technology to be addressed is the need for momentum injection and variable depth core fueling for reactor grade fusion plasmas based on the tokamak and ST concepts, and deep fueling for reactor grade Stellarator plasmas. High-performance ST and tokamak plasmas greatly benefit from plasma rotation and rotation shear to increase energy confinement and sustain high beta. This has been possible due to the injection of substantial momentum from tangentially injected neutral beam systems that also contribute to important core fueling in such plasmas. In larger devices such as ITER or DEMO, higher beam injection energies are required to penetrate to the plasma core, and this reduces the momentum input per unit power. As a result, ITER is projected to have low toroidal rotation relative to present devices. With this not-to-be-ignored issue in toroidal momentum input, the stability of ITER and DEMO plasmas at acceptable levels of performance needs serious re-investigation. Compact Toroid (CT) fueling [Figure 1] has the potential to fill this very serious gap in momentum injection, and in addition provide a source of deep controlled fueling for density profile control in burning reactor grade plasmas. 2. Application of the technology

A Compact Toroid (CT) is a self-contained plasmoid with embedded magnetic fields. The structure is very robust and it can be accelerated to the high velocities needed for fusion reactor fuelling. The CT injection concept was first proposed by Perkins et al., [1] and Parks [2]. CT

acceleration was first demonstrated on the RACE facility at LLNL, where accelerated CT velocities of up to 2000 km/s were achieved at high acceleration efficiencies [3]. Further experiments on the ITER scale MARAUDER device at a US Air Force laboratory demonstrated acceleration of mg sized CTs to velocities of over 300 km/s [4]. These are the parameters required for a reactor CT fueller. The accelerator design was further improved by the Canadian Fusion Fuels Technology Project

(CFFTP) [Figure 2] in collaboration with the University of Saskatchewan, LLNL and the University of California-Davis, for the purpose of injecting these plasmoids into high temperature tokamak plasma [5]. This resulted in the first successful tokamak injection experiments being conducted on the TdeV tokamak [6, 7]. Subsequent experiments on the STOR-M and JFT-2M tokamaks showed that CT injection can also trigger advanced confinement modes [8, 9].

Fig. 1. Pictorial representation of CT fueling


2

A reactor CT injector [10,11] would typically inject 2.2 mg toroids of DT plasma or pure tritium plasmas at a fuelling rate of up to 20 Hz. This 5 MW injector will impart the same momentum as a 69 MW, 500 keV neutral beam injector, while supplying 14 times more core fueling.

These would be injected at a nominal velocity of 300 km/s, but have capability to vary the velocity (200-500 km/s) in order to vary the fuel and momentum deposition location. The injector would be positioned with some tangency with respect to the radial direction to be able to inject toroidal momentum.

Two to three injectors positioned at different toroidal locations, and with different tangency could be used to control rotation shear.

3. Critical variable(s)

Localized fuelling: It is necessary to show that by altering the CT injector parameters, that the CT could be used to deposit fuel at an arbitrary location within the tokamak. Momentum injection: A tangential CT installation, on a large tokamak or ST, to demonstrate the momentum injection capability of the CT.

Repetition rate operation: As part of future activities, CT injection from a 10-20 Hz injector into a large tokamak plasma at higher values of the toroidal field is required. The present off-the-shelf hardware capability for high current switches and capacitor banks is such that a 20 Hz system can be built without the need for further research development in pulsed power technology.

NSTX-U is a particularly good choice for the near-term experiments as it has a large plasma cross-section and low toroidal field, and it is an ST. As a consequence NSTX-U has a very large toroidal field gradient. The toroidal field on the inboard side is an order of magnitude higher than on the outboard side. Since the CT penetration criterion depends on the toroidal field, it means that on NSTX-U the CT stopping location can be more precisely defined. This makes NSTX-U or any large spherical tokamak an ideal candidate in which to study the CT penetration scaling laws, and to develop the CT injection concept. 4. Design variables

The design variable are the mass of the CT, the density of the CT, velocity of the CT and the repetition rate frequency of the injector. The mass is controlled by the amount of gas that is injected to form the CT, as well as by the programming of the pulsed power system used to form the CT.

Fig 2. Top- Shown are the different regions of a CT injector. Bottom – The CT injector in storage at PPPL. The device is 3 m long and powered by two capacitor bank power supplies (not shown)


3

The density of the CT is controlled by the magnitude of the poloidal field used to confine the CT and by the electron temperature in the CT. The voltage of the accelerating capacitor bank power system controls the velocity of the CT. The repetition rate is governed by proper design of the pulsed power system, as well as by the design of the vacuum pumping system so that the residual gasses could be sufficiently pumped out during the time between discharges. 5. Risks and uncertainties

CT injection is more involved than a frozen pellet injection system, but it has enormous potential, and something like this may be necessary for economical operation of a reactor system. As noted in the numerous publications listed in this white paper, many aspects of the technology have been separately studied, but experimental verification of the operation of a high frequency CT injector and large tokamak tests using such an injector are essential next steps. Design of the tokamak interface is particularly important, as the CT needs to travel through high magnetic fields, and the front end of the injector itself needs to be shielded from external magnetic fields as described in References [11,12]. The development effort is similar to the level of effort invested in the development of high power neutral beam systems. CT systems also have a simpler fuel system without the need for tritium cryogenics [13,14]. 6. Maturity The TdeV results show that CTs can be sufficiently clean for the purpose of tokamak fueling. As shown in Ref. [6], during the fueling of a 1.4 T single null divertor discharge the tokamak plasma was not adversely perturbed. While some fuel was deposited deep inside the separatrix, there was no localized fuelling and a large fraction of the fuel was deposited near the edge. This is an inherent difficulty with small tokamaks because the CT axial dimensions are comparable to the tokamak minor radius. These issues can be avoided by selecting a larger cross-section target plasma. A TdeV CT injector sized CT can penetrate toroidal fields of about 1 T. TdeV CT injection experiments were conducted at 1.4 T. Since the CT axial length was about the same as the TdeV minor radius, localized fuelling was not possible in past experiments. NSTX-U is a 0.4-1 T machine with a minor cross-section much larger that the CT plasmoid length. NSTX-U experiments would allow for a localized fuelling demonstration. The steep toroidal field gradient in NSTX-U makes it an excellent test bed for establishing the penetration

Fig. 3: Layout of the CT injector on NSTX-U for variable angle CT injection to study momentum injection and density profile control. The CT hardware is in storage at PPPL.


4

scaling laws. Figure 3 shows the proposed layout of a CT injector for such a momentum injection test on the NSTX-U device. The needed experimental results are:

Localized fuelling: It is necessary to show that by altering the CT injector parameters, that the CT could be used to deposit fuel at an arbitrary location within the tokamak. NSTX-U has a large cross-section plasma at a nominal toroidal field of 0.4-1 T. Thus there is adequate overcapacity in the present injector design to demonstrate this capability in NSTX-U.

Momentum injection: As shown in Figure 3, a tangential CT installation is possible in NSTX-U to demonstrate the momentum injection capability of the CT.

Repetition rate operation: CT injection from a 10-20 Hz injector into a large tokamak plasma at higher values of the toroidal field is required. The present off-the-shelf hardware capability for high current switches and capacitor banks is such that a 20 Hz system can be built without the need for further research development in pulsed power technology.

NSTX-U is a particularly good choice for the near-term experiments as it has a large plasma cross-section and low toroidal field. NSTX-U is a spherical tokamak. As a consequence NSTX-U has a very large toroidal field gradient. The toroidal field on the inboard side is an order of magnitude higher than on the outboard side. Since the CT penetration criterion depends on the toroidal field, it means that on NSTX-U the CT stopping location is more precisely defined. This makes NSTX-U or any large spherical tokamak an ideal candidate in which to study the CT penetration scaling laws.

7. Technology development for fusion applications Present capabilities are between the TRL2 and TRL3 levels. The steps outlined below will bring it up to the TRL 6 level readiness in four to six years. Supporting modeling and theory support is also needed. The development cost for a reactor-sized system is significant, and similar to the effort invested in the development of NBI systems for ITER. ThisworkissupportedbyU.S.DepartmentofEnergycontractnumbersDE-SC0006757,DE-FG02-99ER54519andDE-AC02-09CH11466. References: [1] L.J. Perkins, S.K. Ho, J.H. Hammer, Nuclear Fusion 28 (1988) 1365 [2] P.B. Parks, Phys. Rev. Lett. 61 (1988) 1364 [3] C.W. Hartman, J.H. Hammer, Phys. Rev. Lett. 66 (1991) 165 [4] J.H. Degnan, et al., Phys. Fluids B 5 (1993) 2938 [5] R. Raman, et al., Fusion Technology 24 (1993) 239 [6] R. Raman, et al., Nuclear Fusion 37, No. 7 (1997) 967 [7] R. Raman, et al., Phys. Rev. Lett., 73, 3101 (1994) [8] C. Xiao, A. Hirose, S. Sen, Phys. Plasmas 11 (2004) 4041 [9] T. Ogawa, et al., Nucl. Fusion 39 (1999) 1911 [10] R. Raman, P. Gierszewski, Fus. Eng. Des. 39–40 (1998) 977 [11] R. Raman, Fusion Engineering and Design, 83 (2008) 1386 [12] R. Raman and K. Itami, Journal of Plasma and Fus. Res., 76, No. 10 (2000) 1079 [13] R. Raman, Fusion Science and Technology, 50, (2006) 84 [14] R. Raman, Fusion Science and Technology, 54 (2008) 71

ct white paper-fesac-v4

Documents