Journal of Fusion Energy, Vol. 7, Nos. 2/3, 1988

The IGNITEX Fusion Project R o d o l f o Carrera I

reaction was discovered, a non-self-sustained fission reaction was produced in a laboratory, and in one more year a self-sustained reaction was achieved at the University of Chicago (1942). However, after almost 40 yr of fusion research, a self-sustained fusion reaction has not yet been produced in a laboratory experiment. This fact, of course, indicates the much greater difficulty of the fusion experiment. Because of the extreme difficulty involved in the production of a self-sustained fusion reaction, it is necessary to propose such an experiment with maximum ignition margins, maximum simplicity, and minimum financial risk. As a result of the considerations just mentioned, the IGNITEX concept proposes an experiment utilizing the most reliable regime of fusion plasma operation and unconventional fusion technology that can satisfy the physics, engineering, and cost requirements imposed on the experiment. Emphasis on experimental simplicity will maximize the chances for success. The basis for the IGNITEX concept is Bruno Coppi's idea for a compact ignition experiment and recent technology advances in high-current systems. The purpose of the project is to build a machine in which thermonuclear ignited plasmas can be produced and controlled so that the physics of these plasmas can be studied. Understanding of the basic physical phenomena of ignited plasmas is essential in the analysis of the potential applications of fusion as an inexhaustible source of energy. The IGNITEX project has been until now a very modest project. Highly motivated people fiom the Center for Fusion Engineering, Center for Electromechanics, Institute for Fusion Studies, and Fusion Research Center at the University of Texas have been involved in the project over the last year. The funding for this work has been provided by the Texas

In my presentation I talk about the recently proposed fusion ignition experiment IGNITEX. I emphasize the basic ideas of this novel concept rather than the specific details of the physics and engineering aspects of the experiment. The original idea of the IGNITEX experiment was proposed by Marshall N. Rosenbluth (Director of the Institute for Fusion Studies at the University of Texas at Austin), William F. Weldon (Director of the Center for Electromechanics at UT), and Herbert H. Woodson (Director of the Center for Fusion Engineering at UT) (Texas Atomic Energy Research Foundation Report No. 38, April, 1984). This concept is a good example of the importance of maintaining an adequate balance between the basic scientific progress in fusion physics and the new technologies that are becoming available in order to make fusion work. I will first say a few words about the IGNITEX project and its organization. Then, I will describe the machine proposed for the experiment: a single-turncoil tokamak. Next, I explain the physics basis for the production and control of ignited plasmas, and give some fundamentals on the high-field tokamak technology proposed for the experiment. I also comment on the cost estimates for the construction of IGNITEX. I end this brief presentation discussing some of the advantages of the IGNITEX approach and giving some conclusions of the preliminary work completed on this concept. The objective of the IGNITEX project is to produce and control ignited plasmas for scientific study in the simplest and least expensive way possible. Being able to study this not-yet-produced regime of plasma operation is essential to fusion research. As you know, only two years after the fission nuclear XCenter for Fusion Engineering, the University of Texas, Austin, Texas 78712.

143 0164-0313/88/0900-0143506.00/09

PlenumPublishing,Corporation

144

Atomic Energy Research Foundation, whose members are the 10 investor-owned utilities that operate in Texas and by the Bureau of Engineering Research at UT. A proposal has been recently presented to the Office of Fusion Energy of the U.S. Department of Energy for the continuation of the project (M. N. Rosenbluth, W. F. Weldon, and H. H. Woodson,

Basic Design Report for the Fusion Ignition Experiment (IGNITEX), March, 1987). The proposal has attracted strong national and international interest, and we are starting some collaborations with a number of groups inside and outside the United States. The tokamak is the fusion device with the most immediate prospects for a fusion ignition experiment. The ohmic regime of operation in tokamaks is the one with the most extensive data base and is, certainly, the simplest and best known. Furthermore, record energy confinement times have been obtained in this regime of operation. Thus, an ohmically heated tokamak machine was a natural selection for the IGNITEX concept. A high magnetic field was thought to be needed for, among other plasma physics consequences, the plasma to support a very highinduced electric current. The high-current-carrying capability of the plasma is further increased in a compact device. Also, a compact device was selected to reduce the cost of construction of the machine. In order to have a high ignition margin as well as other plasma characteristics that I discuss later, a 20-tesla (T) magnetic field on the plasma axis was deemed to be necessary. This value for the toroidal magnetic field is almost twice as large as has been produced to date utilizing conventional tokamak technology. However, when a single-turn-coil toroidal field (TF) system is considered, such a magnetic field is feasible. This type of magnet has a very high strength and is relatively inexpensive to build. It requires a power supply that can provide very high currents at very low voltages. Homopolar generators are especially well-suited for operation of a singleturn-coil system. These generators are quite simple and inexpensive. In summary, the IGNITEX device is a compact, high-field, single-turn-coil tokamak capable of reaching and controlling fusion ignition with ohmic heating alone. The plasma in the IGNITEX experiment will be confined in a compact device (major radius 1.5 m, minor radius 0.5 m) with a very high toroidal magnetic field (20 T). Very high currents (in excess of 12 MA) will be induced in the plasma. The approach to ignition conditions will be at very low /3 values

Carrera

(0.6%) to maintain the discharge far from stability limits. Good plasma equilibrium configurations can be obtained (including two X-point equilibria) with about a 3-cm shift in the magnetic surfaces at the time of ignition. The magnetic flux requirement is lower than in conventional tokamaks. Predictions based on neoclassical orbit theory indicate a high a-particle containment in IGNITEX discharges. The neutron wall load can be passively maintained at acceptable values at the time of ignition (3 M W / m 2) and during the ignited phase (below 8 MW/m2). The IGNITEX experiment offers the possibility of passively controlling the thermal instability at ignition so that an ignited phase of the discharge can be maintained without exceeding disruptive limits. Moderate values of elongation (1.6) are sufficient in IGNITEX to obtain a high-ignition margin prediction, which in the case of the Kaye-Goldston scaling is 340% (this is assuming that all the power produced in the plasma, i.e., fusion plus ohmic, degrades confinement as auxiliary heating power does). In order for the IGNITEX experiment to reach ignition conditions with ample margin, neither H mode of operation nor auxiliary heating are needed. Table I summarizes the main quantities describing the IGNITEX experiment (at the time of ignition). The magnet system proposed for the IGNITEX machine is a single-turn coil for the TF system and a poloidal field (PF) system internal to the TF coil. Figure 1 gives a conceptual layout of the IGNITEX machine.

Table I. IGNITEX Experiment Minor radius Major radius Plasma elongation Toroidal field on axis Safety factor at the plasma edge Plasma current Average plasma density Average plasma temperature Average toroidal fl Average energy confinement time Confinement product Ignition margin Fusion power Neutron wall load Neutron production rate a containment factor TonozE product

47.0 cm 150.0 cm 1.6 20.2 T 2.72 12.0 MA 3.6X10 TM cm -3 6.5 keV 0.6% 0.54 s 3.9 • 1014 s//cm3 3.4 149.0 MW 3.0 lvl~V/ m2 5.3 • 1019 n / s 3.2 4.7 X 1015 keV/cm3/sec

The IGNITEX Fusion Project

145

A single-turn coil provides very high filling factors and then a high-strength magnet. Low current density in the TF inner leg is possible. The TF coil will have no windings in it and will be made of copper-alloy plates. It will be operated at very low voltages so that insulation requirements are minimized. These characteristics make the TF magnet system very simple to construct. The PF system is made of single-turn coils, which facilitate assembly and maintenance. The PF system coupling to the plasma is very high because of its proximity to the plasma column. This characteristic facilitates plasma shaping and control, and it decreases power supply requirements. Also it allows us to locate a central compression bar in place of the conventional ohmic transformer. The use of a compression bar, together with radial and axial preloading, make the stresses in the magnet tolerable. Active axial preloading is supplied by a hydraulic press. Passive radial preloading is provided by thermally fitted steel rings located in the top and bottom of the magnet structure. These two rings will hold the single-turn-coil plates together and will prevent radial gaps in the structure. The

experiment pulse is lengthened (10 s) and resistive losses are decreased by precooling the machine at liquid nitrogen temperature (77 K) before operation. Homopolar generators are the appropriate power supplies for this type of tokamak device. These generators are specifically designed for pulsed operation: they will store energy during a relatively long period of time and then will provide a large pulse of power in a short period of time. These generators produce pulses of direct current, and therefore complex and expensive rectifier systems are not required. Iron-core technology and high-speed composite flywheels will be employed in the construction of the IGNITEX generators. The machines needed for the TF system in IGNITEX are quite compact (3.75 m diameter by 2.25 m length). Extensive experience in the operation of these generators shows their reliability. The single-turn-coil tokamak configuration proposed for the IGNITEX experiment permits higherthan-usual fields to be produced in the plasma. These high fields, in turn, make the following predictions possible: ohmic ignition, operation far from marginal stability, control of thermal instability at ignition.

3'

Plasma b o r e

Cryostat

TF toll preload and support structure

I

~

~-.--.~

Hydraulic press assembly~---'---Stainless steel back up plate Insulatlng~ TF compression block

O- f

--]

--I~

I~._...,.

Fig. 1. IGNITEX conceptual layout.

~

Steel radial pretoad ring

.--.------Steel

compression bar

-~'------ Cryostat

146

An unconventional single-turn-coil tokamak requires an unconventional power supply because the magnet will be an intrinsically low impedance system. Therefore, a power supply that can handle very high currents (150 MA) at very low voltages (10 V) is needed. Homopolar generators are well-suited for operation of this experiment. In fact, the IGNITEX approach seems to be a feasible, simple, and low-cost approach to high-field tokamak confinement and then to fusion ignition. Traditionally, ohmic heating has not been considered sufficient by itself to heat a plasma to ignition conditions. The reason is that as the electron temperature increases, the plasma resistivity decreases and so does the efficiency of joule heating on the plasma. In fact, when a conventional high-magnetic-field tokamak (about 10 T magnetic field) is considered, this statement is correct; that is, the regime of ignited plasmas is not accessible by ohmic heating alone in a conventional tokamak with the constraints imposed here. A gap exists between the low-temperature regime of operation dominated by ohmic heating and the high-temperature regime of plasma operation dominated by a heating. A plasma located in that gap will cool down so that ignition conditions are not directly reachable. Some sort of auxiliary heating is needed to cross the gap. The situation changes when a single-turn-coil tokamak with much higher fields is considered. The regimes of low- and high-temperature converge and no energy-loss gap exists in between. Therefore, a suitable path to ignition from the ohmic-dominated regime to the a-dominated regime exists. Once ignition conditions are attained, the plasma temperature will increase rapidly without external action because of the fusion energy being generated. The a particles will heat primarily the plasma electrons and then the electrons will heat the ions by coulomb collisions. This thermal excursion would easily make the plasma exceed the disruptive limits. Avoiding a plasma disruption is critical to an ignition experiment because of the consequences on the device itself and because production of stable ignited plasmas is the objective of the experiment. In the IGNITEX experiment the thermal instability can be damped by the plasma itself. The reason for this is the following. Ignition conditions are approached in IGNITEX at low-plasma temperature (6.5 keV). Before ignition, bremsstrahlung is the dominant radiation emission mechanism. However, when ignition is achieved, because of the low plasma

Carrera

j3 and the fast electron temperature increase, cyclotron emission becomes the dominant radiation plasma cooling mechanism. This cooling mechanism basically saturates the thermal excursion in the plasma. Therefore, the IGNITEX machine is predicted to produce ignited stable plasmas. Plasma discharges have been simulated with the assumption that the currents and the fields in the experiment will be ramped in 3 s and that a flat top of 5 s will follow, with 2 s for shutting down the discharge. Ignition conditions can be achieved in less than 1 s once a steady current has been established in the plasma. Well before ohmic heating decreases significantly (due to the improved conductivity of hot plasmas), a heating takes over as the dominant plasma-heating mechanism. Power losses to the first wall are dominated by neutron radiation. The peak value of the neutron wall load is 8 M w / m 2 (assuming uniform deposition) during the ignited phase of the discharge. The plasma conditions can be maintained far from disruptive thresholds all along the discharge. Specifically, during the ignited phase the plasma beta value is well below the Troyon limit, and the discharge evolves within the Hugill diagram of tokamak operation all along the pulse. In summary, the IGNITEX concept utilizes the best-known and best-performing regime of plasma operation. It has an ample margin for ignition without consideration of H mode of operation and auxiliary heating. It maintains the plasma conditions far from stability limits to maximize the probability for success in the production of ignited plasmas. In addition, a passive means of thermal runaway control is provided. The TF magnet system in IGNITEX will be toroidally shaped, single-turn coil divided, for fabrication purposes, into 12 pie-shaped sectors. Each sector will be driven by an independent homopolar generator. A set of 12 generators will work in parallel around the tokamak. A thin resistive metal shim between sectors is included in the design to prevent the coil current from deviating from the intended poloidal path. An electromechanical analysis for the design of the TF coil has been carried out under the assumption that the current will be ramped up in 3 s and that a 150-MA flat top is maintained for the rest of the pulse. The internal PF system will be formed by five pairs of single-turn coils independently powered by five homopolar generators. In order to reduce stress

The IGNITEX Fusion Project and heating, swinging of the current will be utilized. The generators will provide a maximum combined current output of 22 MA at the time of plasma breakdown. During the subsequent circuit oscillation, the current in the PF coils will pass through zero and reach a negative value of -15.7 MA. As presently designed, the PF system can induce up to 13.8 MA of plasma current. The material to be employed in the fabrication of the IGNITEX magnet system is dispersion-strengthened copper. The specific material considered in the design has been Glid-Cop AL-15, which contains 0.15% aluminum oxide. This copper-based material has a yield strength of 84 ksi (1000 lb/in2; 1 ksi = 6.9 MPa) when 96% cold work is applied to it. In the calculations, it has been pessimistically assumed that the material wilt retain 54% of the room-temperature copper conductivity with an increase to 246% at liquid nitrogen temperature. The fatigue limit of this material is rather high (30 ksi). The axial preload to be applied to the inner leg of the TF magnet system will be 68 ksi. The TF coil axial preload and support structure will include a hydraulic press assembly with a stainless steel backup plate that will preload the magnet through an insulating compression block that penetrates the cryostat. Steel rings will preload the TF coil radially to 34 ksi. Stresses will be further reduced by elongating the plasma bore and introducing a steel compression bar (diameter 0.8 m) in the center of the toroidal structure. Nonlinear calculations of the thermomechanical stresses indicate that a maximum stress of 78 ksi is produced at the inner side of the plasma bore. In the compression bar, the maximum stress is 115 ksi. Cryogenic precooling of the system prevents excessive heating for the pulse lengths of interest. A maximum temperature of 145~ is reached along the discharge. The average current density in the inner leg of the TF system is fairly low considering the magnitude of the field being generated (57 MA/m2). This low value is possible because of the high tilling factors in a single-turn-coil system. Preliminary fatigue considerations indicate that more than 1000 pulses at 20 T should be possible in this machine and more than 100,000 pulses at 15 T. The PF system includes 10 coils located inside the TF coil bore. The top and bottom coils are elongation coils that permit two X-point plasma equilibrium configurations to be formed. At plasma

147 initiation the PF system will be subjected to a maximum stress of 69 ksi. This is the instant of maximum stress because at that time the current is maximum at the PF coils and the plasma current still has not been built up. However, this stress is relieved promptly when the current is swung on the PF circuit. At the end of the pulse, the maximum temperature in the PF system will be 87~ The innermost coil in the PF system will reach the highest temperatures and stresses. The PF electromechanical calculations span over 13 s because the first 3 s before plasma breakdown are employed to ramp up the current in the PF coils. The magnetic flux consumed along the discharge (inductive plus resistive) is 27 V/s. This is much less than the 40 V / s that would be consumed in a comparable conventional tokamak. The presently proposed PF system can produce 31 V/s, which will suffice to induce up to 13.8 MA of plasma current. As a result of the analysis of field and current distribution in the IGNITEX magnet system, the energy requirements for the proposed pulses have been estimated. The total energy requirements are 12.0 GJ and 2.1 GJ in the TF and PF systems, respectively. These results account for magnetic energy stored and resistive dissipation in the magnet, terminals, and power supply circuit. Twelve homopolar generators (HPGs) will supply the power required to the TF system. Each generator will be rated at 12.5 MA and 1 GJ. During the fiat top of the discharge, the voltage of operation will be 10 V. Because insulation requirements are minimum, very high filling factors are possible. Compact, low-cost generators with iron-core rotors and composite flywheels are proposed. The operation of the TF generators will proceed as follows. First the excitation will be increased to maximum level (1.7 T) to ramp up the coil current in 3 s. During the flat top, the excitation will be decreased so that the current output is maintained steady at 12.5 MA; some excitation increase will be required as plasma ignition progresses to compensate for rotor deceleration. In the last 2 s of the pulse, both excitation and current output are ramped down. Critical design variables of the required HPGs are the current density in the brushes and the flywheel maximum tip speed. For the IGNITEX experiment, 1.25 k A / c m 2 in the brushes will be required. Current densities up to 8 k A / c m 2 have been tested already. About 500 full-load pulses will be possible in the IGNITEX generators without brushes being re-

148 placed. The flywheel will be spun to a maximum tip speed of 500 m/s. Composite flywheel technology has allowed tests over 1400 m / s to date. Five independently operated HPGs will serve as power supply to the PF system. The total combined energy of the set will be 2.6 GJ, delivering a maximum current of 22 MA. The open circuit voltage will be 60 V. The PF generators will be very similar to the TF generators (this feature will reduce construction costs), but their operation will be rather different. The rotors of the PF generators will be brought up to full speed before the current is discharged into the PF coils. As the current is built up in the coils, the rotors will decelerate to zero speed. Since the PF circuit is designed to be underdamped, the rotor will be accelerated again in the opposite direction by the magnetic energy stored in the PF system as the current in the plasma is ramped up. Maximum speed in the rotor will be reached when the current in the PF system has decreased to zero. A new oscillation of the circuit will maintain the current variation in the PF circuit so that the inductive and resistive flux components are produced. The maximum current attained in the second oscillation is 15.7 MA (due to resistive losses). The circuit is opened before a third oscillation begins. At this time, both the plasma and the PF coil current are at zero. A preliminary cost estimate of $110M for the basic construction of the IGNITEX experiment has been obtained. The magnet and power-supply systems will cost $50.3M, including instrumentation and control for the power supplies. The magnet and power-supply technologies utilized in the IGNITEX design are much less expensive than their counterparts in conventional tokamaks. In addition, the IGNITEX design emphasizes simplicity: neither auxiliary heating nor divertor systems are considered; no winding requirements exist; and low-voltage, directcurrent operation is proposed. Other components included in the cost breakdown are: project costs ($20M), support structure ($7.5M), facilities ($8M), vacuum vessel ($2.5M), first wall ($1.5M); shielding ($1.0M), diagnostics ($2.0M), instrumentation and control ($3.0M), cryogenics ($2.5M), fueling ($0.5M), vacuum pumping ($3.5M), disposal ($4.3M), and remote maintenance ($3.0M) (which includes a robotic arm for first-wall tile replacement). An important consideration in the design and cost estimate for IGNITEX has been the basic idea that the sole objective of the experiment be to produce ignited plasmas for scientific study.

Carrera The IGNITEX concept has significant advantages over previously proposed conventional designs: 9 Ignition is predicted without need for auxiliary heating. Thus, a source of design complexity and experimental uncertainty can be eliminated. 9 Ignition is predicted with ample margin in the standard L mode of operation. Problems associated with divertor design and operation can then be eliminated as well. 9 Ignition can be reached far from stability limits. Approach to ignition at very low fl values, together with stabilization by the thick conducting wall surrounding the plasma column, anticipate rather stable plasmas in the IGNITEX device. 9

a containment is predicted to be very high.

9 In the IGNITEX design the plasma thermal runaway associated with plasma ignition is stabilized by the plasma ignition itself so that disruptive limits are not exceeded. Therefore, an important problem in conventional designs is addressed. 9 The current density in the TF magnet system is lower than that in conventional designs. Hence, stress in the TF coil is comparable to the stress in conventional magnets that produce lower magnetic field levels. 9 The IGNITEX internal PF system minimizes power and energy requirements. 9 The problems associated with magnet insulation are minimized in the IGNITEX design because of the low-voltage operation. The magnet design is simpler than usual because, for example, no need exists for windings and turn-to-turn transitions. 9 Finally, the IGNITEX experiment has a relatively low cost because of its simplicity and unconventional technology. I would like to conclude my presentation making the following remarks: IGNITEX makes use of the best-known fusion physics and unconventional fusion technology. IGNITEX offers a simple approach to the scientific study of ignited plasmas. IGNITE)( has a high probability of achieving fusion ignition. IGNITEX has a relatively low cost. IGNI-

The IGNITEX Fusion Project TEX is an innovative idea that can make a contribution to the fusion program in the short term. ACKNOWLEDGMENTS Contributors to the work presented here have been M. Driga, J. H. Gully, K. T. Hsieh, E. Montalvo,

149 C. Ordonez, M. N. Rosenbluth, W. A. Walls, W. F. Weldon, and H. H. Woodson. Helpful conversations with many members of the UT fusion community are gratefully acknowledged. This work has been supported by the Texas Atomic Energy Research Foundation and the UT Bureau of Engineering Research.