FRC Based ITER Fueler Design

by

Alan L. Hoffman

Professor of Aeronautics & Astronautics

February, 1996

This project is to design a central fueler for an ITER scale tokamak based on accelerated compact toroids of the FRC variety. Central fueling of tokamaks has many advantages related to both minimizing tritium inventory and maximizing energy lifetimes, and may be essential for optimum ITER operation.

The design points for ignited operation in ITER call for operating either at 1.18 times the "Greenwald density limit", (ngw)*, with normal confinement (H93 = 1) and the full power output of 1500 MW, or at 1.0 times the Greenwald limit, with enhanced confinement (H93 = 1.3) and power reduced to 1200 MW.¹ Additional design points have been described for operation in a driven mode with lower confinement factors, but they also require densities at or above ngw. The density must be high in order to obtain sufficient power density for ignition and ultimately for viable reactor economics. In addition, the divertor scenarios chosen by ITER require high scrape-off layer (SOL) and divertor plasma densities in order to exhaust a large fraction of the alpha particle power by radiation and charge exchange. Density profiles obtained with the conventional fueling technique of gas puffing are predicted to be very flat in ITER, and this is well documented, for example, in JET. However, recent theoretical modelling² suggests that the confinement quality of a reactor-grade tokamak discharge could be significantly enhanced if the density profiles could be peaked. If this theory is correct, density peaking would improve the ITER ignition margin substantially.

Achieving high line-averaged density with high confinement in ITER's chosen ELMy H-mode operating scenario is proving to be a very challenging task experimentally. In steady state ELMy H-modes produced in JET by using neutral beams, but without any additional fueling by gas puffing or pellets, the steady state density attained is typically 50 - 60% of ngw, and the confinement quality is good.³ When the density is raised by gas puffing, confinement remains acceptable up to about 0.85 ngw. Further gas puffing increases the density to nearly the Greenwald value, but confinement deteriorates strongly and the discharges usually make a back transition into the L mode. Similar results have been reported from ASDEX Upgrade.⁴ In recent JET experiments in which ELMy H-modes were produced by RF heating rather than neutral injection beams, so that no central fueling was present, it was found difficult to obtain densities much above 0.6 ngw⁵, suggesting that a certain amount of central fueling is required even to approach the Greenwald value.

Concerns about ITER ignition would be significantly alleviated by the peaked density profiles that could be created by central fueling. Solid pellets, the currently envisioned fueling method, cannot be accelerated to the velocities necessary to penetrate far into an ITER scale plasma.⁶ It may be possible to inject these pellets on the inboard side and let them form a plasma cloud which would then be accelerated toward the plasma center by the higher magnetic pressures at smaller R. Higher fueling efficiencies using inboard pellet injection have been demonstrated in experiments on ASDEX⁷, but this technique is still experimental and may not work in larger tokamaks where the distance the cloud would have to travel before dispersal along field lines is much longer. In fact, if long acceleration distances had occurred on ASDEX, the plasma cloud would have been accelerated out the far side.

A more certain method of central fueling with 'plasma clouds' would be to inject small diamagnetic plasmoids from the outboard side, with sufficient velocity to penetrate to the tokamak center. This is calculated to require kinetic energy densities, equal to or greater than the toroidal field pressure.

Such a fueling scheme was proposed over 15 years ago using compact toroids⁸, of which there are two types, spheromaks which contain both toroidal and poloidal fields, and tend to be oblate, and field reversed configurations (FRC), which have negligible toroidal fields and tend to be prolate. Spheromak acceleration experiments were initiated over 10 years ago, and the acceleration and injection of spheromaks into small tokamaks has already been demonstrated⁹.

An experimental study called TRAP (Tokamak Refueling by Accelerated Plasmoids) was also initiated four years ago at the University of Washington to demonstrate acceleration of FRCs and their ability to penetrate simulated tokamak fields. FRCs offer three significant advantages over spheromaks with reference to fueling.

1. FRCs are high beta and thus can be much higher density for a given confinement field.¹⁰
2. FRCs are formed and accelerated inductively with no electrode contact, and have been demonstrated to have a minuscule impurity content.¹¹
3. FRCs are highly elongated and will thus require smaller vacuum penetrations for transport to the tokamak plasma.

The experimental TRAP demonstration utilized the already existing LSX facility¹² which, while originally designed for forming and studying large, stationary FRCs, was sufficiently flexible for performing the TRAP experiments. In fact, the modified LSX, called LSX/mod, both formed and accelerated high density FRCs with the correct mass to fuel either ITER or JET, and with kinetic energy densities nearly sufficient to penetrate JET level toroidal fields, limited only by the high inductance LSX power supplies which could not supply energy to the coaxial acceleration coils at a fast enough rate to match the high FRC velocities during for the latter stages of acceleration.¹³

The intent of this grant application is to design an FRC based fueler which could initially be used on JET, and then have additional acceleration stages added to be used on ITER. If we utilize the already demonstrated 2x10²⁰ ion (0.7 mg) LSX/mod FRCs, for which the formation technology is well developed, a 3 - 6 Hz repetition rate would be required to reach the Greenwald density limit on JET, and a 1.5 - 40 Hz (using 2 injectors) repetition rate would be required to satisfy the projected fusion replacement or total particle replacement rate projected for ITER. The accelerator parameters required for fueling JET have already been examined, and calculations will be conducted to optimize them plus the additional stages needed for ITER fueling. The principal engineering requirement will be to make the device repetitively pulsed, which will involve the help of industrial contractors. Strong coordination with JET and ITER design groups will also be a major emphasis to insure the compatibility of the fueler with the tokamak access and plasma physics constraints. In order to do this, the program will employ scientists knowledgeable about both tokamak edge physics / divertor design and FRC production.

The output of this effort will be a consistent engineering design (everything except detailed drawings) of FRC based fuelers suitable for operation on JET and ITER. A small scale demonstration of the repetitive operation of one scaled coil module will also be made.

* The "Greenwald density limit" is an empirical limit which has been found to describe the highest line-averaged density attainable in edge-fueled diverted tokamak discharges. It is given by
n(10²⁰ m^-3) @ I_p(MA)/a²(m).

[1] ITER Preprints, Sixteenth IAEA Fusion Energy Conference, Montreal, Canada, 7-11 October 1996.

[2] Dorland & Kotschenreuther, "First principles model".

[3] The JET team (presented by D. Stork), Montreal IAEA.

[4] J. Neuhauser et. al., Plasma Physics and Controlled Fusion 37, Supp. 11A, A37 (1995).

[5] George Vlases, private communication, Jan. 1997.

[6] S. Milora, J. of Fusion Energy 1, 15 (1981).

[7] P.T. Lang et. al., "Pellet Injection into ASDEX Upgrade Plasmas with Improved Scenario from the Magnetic High-Field Side", Max-Planck Institute fur Plasmaphysic Report, IPP 1/304 (October 1996).

[8] J.H. Hammer and C.W. Hartman, "Mirror Theory Monthly", LLNL (1981).

[9] R. Raman et. al., "Experimental demonstration of non-disruptive, central fueling of a tokamak by compact toroid injection", Phys. Rev. Lettr., 73, 3101 (1994).

[10] M. Tuszewski, "Field Reversed Configurations", Nucl. Fusion 28, 2033 (1988).

[11] J.T. Slough & A.L. Hoffman, "Experimental study of the formation of FRCs employing high-order multipole fields", Phys. Fluids B 2, 797 (1990).

[12] A.L Hoffman et. al., "The large s FRC Experiment", Fusion Technology 23, 185 (1993).

[13] J.T. Slough & A.L. Hoffman, "Acceleration of a FRC for central fueling of ITER", Sixteenth IAEA Fusion Energy Conference, Montreal, October 1996.

[14] A.L. Hoffman et. al., "Formation of FRCs using scaleable, low-voltage technology", Fusion Technology 9, 48 (1986).

[15] R.D. Milroy & J.U. Brackbill, "Numerical studies of a FRC Plasma", Phys. Fluids 25, 775 (1982).