4
Magnetic Confinement Fusion

INTRODUCTION

Plasma science has played a major role in magnetic fusion research from its inception and, in many ways, the quest for controlled fusion has been crucial in the development of modern plasma science. In a fusion reactor, a mixture of deuterium and tritium is ionized and the resulting plasma, which is confined by a magnetic "bottle," is heated to temperatures of the order of a few hundred million degrees centigrade. As illustrated in Figure 3.1, the deuterium and tritium nuclei would fuse upon colliding, thereby forming helium nuclei and very energetic (~14-MeV) neutrons. These neutrons may be captured in a thermalizing blanket and the energy used for electric power generation. The needs of magnetic fusion research required a far better understanding of collective interactions in plasmas than existed in the 1950s and 1960s. After the initial series of experiments, of particular concern was the gross magnetohydrodynamic stability of magnetic confinement configurations, the anomalous transport of energy and particles, and the heating and fueling of confined plasmas to reactor-relevant temperatures and densities. Some of the fundamental properties of collective interactions can be probed in relatively simple plasma configurations, the kind of basic experimental plasma research discussed in Chapter 8. However, many collective phenomena can be observed only in hot and dense plasmas in complex magnetic field geometries. The investigation of such phenomena required the development of new diagnostics to probe the properties of hot and dense plasmas, giving birth to experimental plasma research in fusion grade plasmas. Progress in all of these research areas will be required for ultimate success in



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 71
Plasma Science: From Fundamental Research to Technological Applications 4 Magnetic Confinement Fusion INTRODUCTION Plasma science has played a major role in magnetic fusion research from its inception and, in many ways, the quest for controlled fusion has been crucial in the development of modern plasma science. In a fusion reactor, a mixture of deuterium and tritium is ionized and the resulting plasma, which is confined by a magnetic "bottle," is heated to temperatures of the order of a few hundred million degrees centigrade. As illustrated in Figure 3.1, the deuterium and tritium nuclei would fuse upon colliding, thereby forming helium nuclei and very energetic (~14-MeV) neutrons. These neutrons may be captured in a thermalizing blanket and the energy used for electric power generation. The needs of magnetic fusion research required a far better understanding of collective interactions in plasmas than existed in the 1950s and 1960s. After the initial series of experiments, of particular concern was the gross magnetohydrodynamic stability of magnetic confinement configurations, the anomalous transport of energy and particles, and the heating and fueling of confined plasmas to reactor-relevant temperatures and densities. Some of the fundamental properties of collective interactions can be probed in relatively simple plasma configurations, the kind of basic experimental plasma research discussed in Chapter 8. However, many collective phenomena can be observed only in hot and dense plasmas in complex magnetic field geometries. The investigation of such phenomena required the development of new diagnostics to probe the properties of hot and dense plasmas, giving birth to experimental plasma research in fusion grade plasmas. Progress in all of these research areas will be required for ultimate success in

OCR for page 71
Plasma Science: From Fundamental Research to Technological Applications controlled magnetic fusion. In the following sections, the panel concentrates on the so-called tokamak confinement concept. A tokamak is a toroidal plasma chamber in which confinement is produced by an axial magnetic field and a toroidal plasma current, usually driven inductively by a transformer. The panel also considers important non-tokamak confinement geometries, however. MAGNETOHYDRODYNAMICS AND STABILITY Introduction and Background To achieve the densities and temperatures required for a successful thermonuclear reactor, a plasma must be contained by magnetic forces (such a confinement geometry is sometimes called a ''magnetic bottle") for a sufficiently long time to produce net thermonuclear power. In the attempts to achieve this confinement, stability has emerged as one of the most important problems. A plasma confined by a magnetic field is not in thermodynamic equilibrium and therefore is potentially able to break out of the confinement system by a large variety of instabilities. Past Achievements During the past 10 years, great progress has been made in understanding the equilibrium and macroscopic stability properties of the tokamak plasma. The majority of tokamaks today routinely produce equilibria that are much more complicated than those of the original circular-cross-section "doughnut" concept. When a tokamak plasma is deformed from its axisymmetric equilibrium state, macroscopic MHD instabilities usually set in. The most virulent of these are the "ideal" instabilities, which tap the free energy associated with the pressure gradient or the plasma current. The unstable modes grow rapidly and can result in sudden loss of the stored plasma energy. Theoretical predictions of the stability boundaries for MHD modes have been corroborated by experiments. A significant achievement is the validation of the dependence of the beta limit (β is the ratio of plasma kinetic pressure to magnetic pressure) on plasma current, minor radius, and magnetic field. Stable operation regimes have been developed based on a good understanding of the dependence of MHD stability on global plasma parameters. In recent years, MHD research has begun to focus on the next level of understanding, namely, the impact of internal profiles on stability. The so-called second stable regime, for example, is a consequence of the localized high pressure creating a favorable "magnetic well" that stabilizes pressure-driven instabilities. Experiments have already shown that the beta limit can be doubled by optimizing plasma profiles. Furthermore, recent tokamak results have indicated a possible correlation between stability and confinement, offering

OCR for page 71
Plasma Science: From Fundamental Research to Technological Applications the promise of an operational regime of enhanced stability and improved confinement. A second class of important but less virulent MHD instabilities is the resistive instabilities. They allow the magnetic field lines to open up and reconnect and may be accountable for the degradation of confinement. A third class of MHD instabilities is that driven by energetic particles. Experiments have observed instabilities in beam-heated plasmas that resulted in the ejection of energetic beam particles. Other experiments have used energetic beam ions to simulate the alpha particles in igniting plasma. Of particular concern are the effects associated with self-generated plasma currents ("bootstrap" currents) owing to high plasma pressures at high temperatures and finite density gradients. High bootstrap current fractions (>50%) have been self-consistently calculated and found to be stable at reasonable values of βp in major fusion devices. Nevertheless, as the pressure is increased, the plasma may become unstable, and this is observed in some of today's experiments. Once we develop a better understanding of these processes, there are plans to improve the stability at high pressures and high bootstrap current fractions in future tokamak experiments. Future Prospects In recent experiments, so-called toroidal Alfvén eigenmodes (TAEs) were driven unstable with neutral beams and high-energy particles driven by radio-frequency power. These TAE instabilities are important since they may be driven unstable by the alpha particles produced in the deuterium-tritium (D-T) fusion reaction. Theoretical calculations of fast particle destabilization thresholds for TAE modes are in reasonable agreement with the experimental results. Upcoming tritium experiments at the Tokamak Fusion Test Reactor (TFTR) in the United States and the Joint European Torus (JET) in Britain will test these models for the first time in tokamaks that have significant densities of alpha particles. Despite the stability problems as the limits of plasma pressure and currents are approached, significant progress has been achieved by building larger and "smarter" machines. Tokamaks' confinement and stability continue to improve, and it is important to continue to improve the tokamak concept for eventual use as an economical power-producing reactor. The DOE is proposing to build the Tokamak Physics Experiment (TPX), a new "steady-state" national tokamak research facility at the Princeton Plasma Physics Laboratory. This device is now being designed by a national team of physicists and engineers. The plan is to start operation at the beginning of the next decade. One of the TPX's key objectives will be to push the stability limits by controlling the toroidal current profile with current-drive methods (see below).

OCR for page 71
Plasma Science: From Fundamental Research to Technological Applications TOKAMAK TRANSPORT Introduction and Background Since plasmas of sufficiently high density and temperature must be formed to accomplish thermonuclear ignition, understanding the transport of energy and particles is the key to the design of a fusion reactor. Virtually all tokamak experiments worldwide have demonstrated a common consistent level of energy and particle confinement in the so-called L-mode (L standing for "low confinement") regime of operation. In addition to L-mode, a variety of regimes have been observed and studied that have improved confinement, typically in response to changes in the plasma boundary conditions. The most ubiquitous of these is the H-mode regime (H standing for "high confinement"), which differs from L-mode primarily by having a transport barrier at the plasma edge. In the H-mode, the energy confinement is typically a factor of two better than in the L-Mode. More recently, even better confinement has been achieved in the so-called VH-mode ("very-high'') regime, where confinement times up to four times longer than those of L-mode plasmas have been achieved. Characteristics of the H-mode and VH-mode are illustrated in Figure 4.1. Past Achievements During the past decade, tokamak plasma parameters comparable with those estimated to be required for fusion reactors have been achieved (Figure 4.2), including maximum central ion temperatures of Ti ~ 37 keV (4 × 108 °C), confinement times of τ ~ 1 s, and central densities n ~ 3 × 1020 m-3. These parameters were not all achieved simultaneously, but simultaneous measurement has been achieved of a fusion triple product nτTi ~ 1.1 × 1021 keV s m-3. Current experiments operate with the same dimensionless parameter values as those expected in reactors, with the exception that reactors will require two to three times greater confinement times and electron temperatures than those obtained in today's machines. Tokamak experiments in the 1980s demonstrated that the transport of thermal plasma energy and momentum across the confining magnetic fields is ~100 times faster than predicted by theory. This theory considers only the effects of Coulomb collisions and essentially is an extension of gas-dynamic models to magnetized plasma transport. These observations have shifted the emphasis in tokamak cross-field transport theory from collisional models to ones that include the effects of plasma turbulence. Dimensionless scaling experiments have begun measuring the scaling of tokamak cross-field thermal transport with respect to the theoretically important dimensionless parameters, in a manner analogous to windtunnel experiments. It has been found that the scaling is unlike that predicted by most theories. In

OCR for page 71
Plasma Science: From Fundamental Research to Technological Applications particular, the experiments imply that the size of the turbulent eddies is not always controlled by the electron or ion gyroradius size, but presumably is set by macroscopic scale lengths. New diagnostic techniques used in recent measurements of the turbulent fluctuations within tokamak plasmas have found that the spectrum and implied transport are dominated by moderately long-wavelength modes. Measured changes in the plasma transport are well correlated with changes in the fluctuation amplitude, implicating them as the cause of the transport. In contrast, the measured transport parallel to the magnetic field is in good agreement with the predictions of "neoclassical" theory. One of the major scientific achievements of tokamak transport and turbulence studies has been the development of a model explaining the formation of the transport barrier in H-mode. This model is based on stabilization of turbulence by sheared E × B flow (here E is the radial electric field observed in the vicinity of the tokamak edge, and B is the toroidal magnetic field). The measured levels of E × B flow shear are well above those theoretically required for such stabilization. Furthermore, the increased E × B flow shear is correlated with the reduction of density fluctuations, cross-field energy, and particle transport both spatially and temporally. These results provide some of the best evidence to date of the close connection between fluctuations and transport. Similar E × B shear stabilization effects may also take place in the core of tokamak plasmas (e.g., the confinement improvement from H-mode to VH-mode). The concept of electric field flow shear stabilization of turbulence may be one of the most fundamental contributions of tokamak physics to general fluid dynamics. Although flow shear stabilization can take place in ordinary fluids, a sheared velocity field is usually a source of free energy; hence it usually drives instabilities rather than stabilizing them. Only in a plasma can magnetic shear prevent instabilities driven by velocity shear (e.g., Kelvin-Helmholtz instabilities) so that flow shear can then affect the other instabilities. Another recent result is the importance of the plasma current density profile in controlling confinement. Experiments modifying the current profile transiently, by inductively driving a skin current, changing the plasma shape, or using external current drive, have shown that, with peaked current profiles, the cross-field transport can be substantially reduced. Although this effect is not understood theoretically, measurements of the local ion thermal transport indicate that it may be reduced by increasing magnetic shear. Future Prospects The eventual goal of the magnetic fusion program is the realization of a commercial reactor to generate electricity. Present limitations on confinement and total pressure in the plasma force reactor designers in the direction of multigigawatt units. Understanding and reducing turbulent transport in tokamaks

OCR for page 71
Plasma Science: From Fundamental Research to Technological Applications FIGURE 4.1 The loss of thermal energy from magnetically confined plasmas is dominated by turbulent transport. The high-confinement regime (H-mode) in tokamaks is characterized by reduced turbulence at the plasma edge, which inhibits this outward transport of energy. The existence of this regime is believed to be caused by a shear in the drift velocity, associated with a radial electric field. Recently, an even higher confinement regime (the VH-mode) has been discovered, in which the region of high shear and low turbulence extends deeper into the plasma. Shown in this figure are, from top to bottom, the electric field, the velocity shear, and the thermal diffusivity as functions of the normalized minor radius, ρ, for the H- and VH-modes. The inset at the top of the figure shows the time evolution of the normalized density fluctuations, as measured by microwave scattering. (Reprinted, by permission, from K.H. Burrell, T.H. Osborne, R.J. Groebner, and C.L. Rettig, Proceedings of the 20th European Physical Society Conference on Controlled Fusion and Plasma Physics, European Physical Society, Geneva, 1993, vol. 17C, part 1, pp. 1–6. Copyright © 1993 by the European Physical Society.)

OCR for page 71
Plasma Science: From Fundamental Research to Technological Applications could have a major payoff in reduced-size commercial reactors in the next century. A detailed theoretical understanding of tokamak turbulent transport from first principles continues to elude us. However, continued development and exploitation of novel diagnostics and experiments, coupled with the new nonlinear numerical simulations, should allow identification of the dominant turbulence drive and damping mechanisms during the next decade. Theoretical understanding of the turbulent transport mechanism should allow the development of new techniques for controlling transport to improve tokamak reactors. Present experiments have shown that control of the plasma current and electric field (flow shear) profiles can have significant effects on turbulence and confinement. Similarly, optimizing the current profile, pressure profile, and plasma shape are important for increasing the beta limit. However, in many cases present experiments in current profile control have been done with techniques that are inherently transient. They demonstrate the principle, but they are not necessarily sufficient to prove that such profiles can be maintained in steady-state or in long-pulse reactors. The challenge for future experiments will be to demonstrate active control of the plasma current, electric field, and pressure profiles by techniques that are economical and applicable to long-pulse or, preferably, steady-state devices. EDGE AND DIVERTOR PHYSICS Introduction and Background Divertors are magnetically separated regions at the boundaries of magnetic confinement fusion devices. Their originally envisioned functions (Spitzer, 1957) were to exhaust heating power and helium "ash" from fusion reactions and to protect the reacting core plasma from impurities. A separatrix is a surface that separates the flux region of the core plasma from the "burial chamber," the region remote from the core plasma where magnetic field lines intercept material surfaces. The plasma that diffuses across the separatrix from the hot core is "scraped off" on those material surfaces, thus giving rise to the name scrapeoff layer. The edge plasma is located just inside the magnetic separatrix and the

OCR for page 71
Plasma Science: From Fundamental Research to Technological Applications FIGURE 4.2 The approach to power plant conditions of magnetic confinement fusion experiments: The performance of various plasma devices is shown as a function of the "fusion parameter," niTiτE, and the central ion temperature, Ti. The boundaries are indicated for Q = 1 in a deuterium-tritium plasma (i.e., "scientific break-even," where fusion power out equals power in) and for ignition.

OCR for page 71
Plasma Science: From Fundamental Research to Technological Applications scrapeoff layer is just outside. The properties of the edge and scrapeoff plasma regions have been shown to exert a profound influence on the confinement and transport properties of the main plasma. This arises through modifications to the boundary conditions on charged and neutral particle and energy flows. Recent Advances Research on divertor physics was invigorated in the early 1980s by the discovery of an enhanced core plasma confinement mode (H-mode) for auxiliary-heated plasmas. A doubling of energy confinement was soon verified in many subsequent divertor experiments and, more recently, in limiter-bounded plasmas, as well as in stellarators (an alternate form of toroidal confinement device). New regimes of enhanced confinement were also discovered in ohmically heated tokamaks and in auxiliary-heated tokamaks with special vessel wall treatments. The common link between all these improved regimes of energy confinement was control over hydrogen recycling in the edge plasma. To understand the transport of energy in the edge region, two-dimensional fluid modeling of the scrapeoff layer has yielded important predictions. A regime of high recycling divertor (HRD) was predicted in which the flux of particles onto the divertor plate greatly exceeded the flow of particles out across the separatrix. This greatly reduces the average energy of the ions hitting the divertor plate and, hence, the impurities generated there by sputtering. A flux enhancement factor of ~20 has been measured in several tokamaks, which confirms the model. In addition, the flux enhancement in the divertor acts as a flow against which impurities created at the divertor surfaces must fight to reach the plasma core region. Though not directly within the category of plasma edge physics, activities on wall conditioning deserve special recognition. New methods to coat and condition the walls of tokamaks (e.g., carbon, boron, silicon, and lithium coatings and helium discharge conditioning) have resulted in the greatest improvements in core plasma phenomena. At present, this activity is more of an art than a science and is not well understood. Future Research and Technical Opportunities The requirements of the International Thermonuclear Experimental Reactor (ITER) program have given new momentum to edge and divertor physics research. The ITER activity, supported by the European Community, Japan, Russia, and the United States, is now in the engineering design phase. ITER will be the first large-scale device built for the purpose of demonstrating controlled thermonuclear ignition and burn as well as the engineering feasibility of a tokamak-based reactor. However, analysis by the ITER team has shown the inability

OCR for page 71
Plasma Science: From Fundamental Research to Technological Applications of present conceptual designs of the divertor to handle the intense power loads with sufficient safety margin or lifetime. Attention has returned to certain old, though still-untested, ideas and recent variations on how to widen the thin power-carrying scrapeoff layer. Ergodization of field lines is under scrutiny. Another scheme relies on the injection of cold gas into the scrapeoff layer to reduce, by charge exchange, the power carried to the divertor plate by plasma ions. Intense radiation, mainly by impurities entrained in the plasma flow toward the divertor plate, may be able to drain the power out of the electron channel. At the extreme, a completely successful embodiment of these approaches would result in volumetric plasma recombination and the replacement of a solid divertor plate by a gas target (an idea that originated in 1974). The fluid codes presently used to model the scrapeoff layer do not represent a sufficiently accurate description of the physical processes. These must be improved by the direct incorporation of kinetic effects, drift motion, nonambipolar flows, and better atomic physics. With such improvements, these codes could better evaluate helium flows and impurity effects. Comparisons with helium exhaust experiments would prove enlightening. Methods to control helium flows by interactions with waves and MHD activity appear promising. Such activities are under consideration both for U.S. tokamaks and in collaborations on foreign tokamaks. PLASMA HEATING AND NON-INDUCTIVE CURRENT DRIVE Neutral-Beam Heating and Current Drive Introduction and Background Neutral-beam injection has been the principal method of heating the past several generations of tokamak plasmas and has also found utility in driving toroidal plasma currents. A neutral-beam injector consists of a high-current ion source, with multiaperture grids that electrostatically accelerate ions into a conductance-limiting duct, where a portion of the ions charge-exchange with gas to become fast neutrals. The residual ions are swept out of the beam by a deflection magnet, leaving the high-energy neutrals to pass through a duct across the tokamak's fringing magnetic fields into the plasma. Once inside the plasma, the beam neutrals are ionized, and these high-energy ions are captured by the magnetic field. As they circulate many times around the plasma, they collide with the plasma electrons and ions, transferring energy to them until the beam ions are thermalized. If the neutral beams are injected predominantly in one direction tangential to the plasma's major radius, they can drive net plasma current, reducing the need for inductive current drive after the plasma startup phase.

OCR for page 71
Plasma Science: From Fundamental Research to Technological Applications Past Achievements Neutral-beam injection has several features that have made it attractive for implementing experiments on numerous tokamaks over the past two decades. Since the beam is trapped by the plasma and transfers energy to it through two-body interactions, the physics involved is relatively straightforward and calculable, allowing the power deposition profile to be predicted accurately. The experimental flexibility has enabled neutral-beam injection to provide most of the heating for transport and confinement studies on tokamaks since the 1970s. Over this period, the injected power levels have been increased from the 100-kW range to 33 MW. Recently, neutral beams have produced ion temperatures up to 44 keV in some of the world's major tokamaks. Beams, including tritium, were used in some cases to produce more than 10 MW of fusion power. (See Figure 4.3.) Beams played an essential role in the discovery of peaked density profile enhanced confinement regimes (called Supershots) and the first demonstration of "bootstrap current," a gradient-driven self-current, on a tokamak. Neutral beams also are important for driving plasmas into H-modes and VH-modes. Future Prospects The neutral beams used on tokamaks over the past 20 years have all been based on positive ion sources (with an electron added in the neutralizer to form the neutral beam). The practical neutralization efficiency that can be achieved with the positive ions decreases very rapidly at deuterium beam energies of 120 keV and greater. Therefore, for future applications (MeV energies may be desirable in reactors), negative ion beams are more attractive. The achievable neutralization efficiency in an optimized-thickness gas cell is high for negative ion beams (58–60%) and is nearly independent of energy in the hundreds of keV to many MeV range. The roles for neutral beams in the future will include reliable plasma heating and central plasma current drive. Technical opportunities abound for improving the current density, brightness, and gas efficiency of negative ion sources, and for perfecting photodetachment neutralizers and plasma neutralizers that could permit still higher neutralization efficiencies. Radio-Frequency Heating and Current Drive Introduction and Background An alternative way to heat plasmas to high temperatures is by means of radio-frequency waves. Radio-frequency heating spans a very large range of frequencies, from a few megahertz (MHz or 106 Hz) to a few hundred gigahertz (100 GHz or 1011 Hz). (One hertz designates one cycle per second oscillation frequency.) The low-frequency end corresponds to the regime of Alfvén waves,

OCR for page 71
Plasma Science: From Fundamental Research to Technological Applications FIGURE 4.3 Photograph of the Tokamak Fusion Test Reactor (TFTR), located at the Princeton Plasma Physics Laboratory. The major radius of the donut-shaped plasma is 2.5 m. Typical plasma currents are 2.5 MA at toroidal magnetic fields of 52 kG. Powered by intense neutral beams and with a deuterium-tritium fuel mixture, TFTR has achieved record ion temperatures of 44 keV and fusion powers of 10.7 MW in second-long pulses. (Courtesy of Princeton Plasma Physics Laboratory.) while the high-frequency regime corresponds to electron cyclotron waves, which are resonant with electrons gyrating at their gyro (cyclotron) frequency or its harmonics. Other frequencies of interest include the ion-gyro frequency or its harmonics (30–200 MHz) and the ion plasma frequency (more accurately the so-called lower-hybrid frequency) at 1–4 GHz. The basic premise of rf heating is that an antenna installed in the vicinity of the vessel wall radiates electromagnetic waves that deliver rf power from a transmitter to the high-temperature plasma core where the power is absorbed by wave-particle resonances. For example, Alfvén waves and lower-hybrid waves may transfer their energy and momentum to electron motion parallel to the magnetic field by the process of "Landau damping" (named after the famous Russian theoretical physicist who first predicted this kind of resonant wave-particle interaction nearly five decades ago). The accelerated resonant particles eventually dissipate their energy in the background plasma by collisions, thereby heating the bulk plasma particles. If, in addition, the waves travel in a preferred toroidal direction (which can be

OCR for page 71
Plasma Science: From Fundamental Research to Technological Applications arranged by proper phasing of the antenna elements), net momentum is transferred to charged particles (usually electrons), thereby generating a net toroidal plasma current. Since rf generators can be operated continuously (cw, or continuous wave operation) at the megawatt power level, a steady-state tokamak operation may become feasible. In spite of its complexity, the physics of waves in plasmas presents one of the best-understood and most scientifically established disciplines in plasma science. Many aspects of this subject have been verified in fundamental "basic physics" experiments, while some aspects are still under study in the complex field geometry of toroidal configurations. However, the nonlinear aspects of wave propagation still are not well understood. Past Achievements Over the past decade, the scientific discipline of radio-frequency heating and current drive in plasmas has seen rapid growth both in the United States and abroad. As in the case of neutral-beam heating, most rf experiments on fusion plasmas in the 1970s consisted of sources amounting to only a few hundred kilowatts, increasing to the ~1-MW level in the early 1980s and culminating recently at the 22-MW injected power level in the ion-cyclotron (ICRF) frequency range. Impressive heating results have been obtained recently in large tokamak devices, where central electron temperatures up to 13 keV have been achieved. In agreement with theoretical projections, directly accelerated "minority ion" species with up to MeV energies have been observed. Detailed energy analysis of these energetic ions shows excellent agreement with "quasi-linear" wave-particle interaction theories and large simulation codes, one of the triumphs of modern plasma theory. Perhaps even more striking results have been obtained in the area of noninductive current drive by rf waves. In this case, the waves not only heat the plasma (i,e., transfer wave energy to particles) but also transfer net momentum in the toroidal direction. In the past decade, current drive by rf waves has been verified in nearly all frequency regimes. Recently, in Japan, currents at the 3.5-MA level have been driven by multimegawatt lower-hybrid waves. Since these currents are often driven by electrons with energies of 100s of keV, it has been possible to study the transport of these energetic electrons by x-ray imaging techniques. Another scientific spinoff of these experiments is a better understanding of stochastic acceleration of charged particles in electromagnetic wave fields. This may have important application to astrophysical and space plasmas. Future Prospects Radio-frequency heating and current drive will likely be used in all future toroidal plasma devices. While ICRF power is eminently suitable for bulk

OCR for page 71
Plasma Science: From Fundamental Research to Technological Applications plasma heating, even under reactor-like conditions, electron cyclotron resonance heating may be used either for bulk plasma heating or for special-purpose localized heating (temperature profile control). In principle, localized heating offers the possibility of controlling the pressure profile and thereby improving MHD stability. Lower hybrid waves has been most successful at driving toroidal plasma currents (lower-hybrid current drive, or LHCD). In future experiments, LHCD will be used mainly for driving off-axis plasma currents (current profile control), while the central currents may be driven with neutral beams or with fast magnetosonic waves in the ICRF regime. For purposes of disruption control, highly localized edge current drive with electron cyclotron waves is contemplated. In the future, rf wave theory will concentrate on the nonlinear regime. In many cases this leads to the study of strongly nonlinear regimes in plasmas, including turbulence, chaos, and stochastic particle acceleration. An understanding of these phenomena will have a large impact on our understanding of similar phenomena in astrophysical and space plasmas, including solar physics. Radio-frequency heating is a strong technology driver. Higher-power radio-frequency sources are under development in nearly all frequency regimes. In the ICRF regime, new tetrodes have been developed by industry with cw power levels up to 3 MW; future directions include the possible development of 5-MW tube capability. In the lower-hybrid regimes, cw tubes (klystrons) up to 0.5 MW have been developed at 2.45–3.7 GHz, and for future applications, 1 MW tubes are good prospect for development. Finally, in the electron-cyclotron resonance heating (ECRH) regime, ~1-MW pulsed tubes (gyrotrons) have been developed at frequencies in the 100-GHz range, and future development work promises cw tubes at the 1-MW level at frequencies up to 150 GHz. DIAGNOSTIC DEVELOPMENT Introduction and Background The need to measure detailed plasma parameters in fusion-grade plasma environments has led to many creative applications of plasma science. In turn, numerous advances in plasma science have been inspired directly by the need to understand the plasma properties with ever-increasing precision. Important examples from the last decade of research are indicated below. In addition, a summary of future directions is presented. Past Achievements Density and electron temperature profiles are routinely measured by laser Thomson scattering and laser interferometry. Electron temperature profiles in hot plasmas are also measured by electron cyclotron emission (ECE). Ion tem-

OCR for page 71
Plasma Science: From Fundamental Research to Technological Applications peratures are often measured by spectroscopic techniques. Among more novel diagnostics, one of the most important and, therefore, most intensely investigated areas is that of incoherent fluctuations and their relationship to energy and particle transport. The fluctuating quantities of interest include density, temperature, and plasma potential. To relate these to transport, it is necessary to measure frequency and wavenumber spectra, along with relative phase, as functions of spatial location. Many techniques have been developed to attack these problems. Beam emission spectroscopy (BES) allows for the measurement of electron density fluctuations by looking spectroscopically at line radiation that results from plasma excitation of high-energy neutral-beam atoms injected into the plasma. BES has led to the discovery of large-scale structures that are now the subject of intense theoretical investigations. Other techniques that have been invented to measure density fluctuations include reflectometry (backscattering from the critical layer), laser scattering, and phase-contrast imaging. Utilizing heavy-ion beam probes of very high energy (~1 MeV), measurements of both density and potential fluctuations have been carried out. Probes have long been used in low-temperature plasmas to measure both density and potential fluctuations, and these are used routinely in the scrape-off region of tokamak plasmas. For the first time, fast scanning probes have allowed access to hotter regions of plasma, inside the last closed flux surface. The ability to measure density and potential fluctuations simultaneously has allowed the first direct measurements of fluctuation-induced energy and particle cross-field transport. The desire to know the detailed structure of the magnetic field in the hot confinement region of tokamaks has spawned several creative new diagnostic techniques. These include the application of Faraday rotation, Zeeman polarimetry using neutral beams and pellets, and the imaging of the ion clouds that result from pellet ablation. The approach that probably has the most potential for highly precise internal field measurements with good spatial resolution involves an application of the motional Stark effect (MSE), also using an energetic neutral beam. This has led to detailed q profile (safety factor) measurements and new insights into MHD phenomena. The ability to measure detailed profiles of plasma parameters also has matured significantly over the last 10 years. By combining measurements from multiple arrays of soft x-ray sensitive diode detectors with new tomographic inversion algorithms, a wealth of new physics information on the structure and evolutions of fusion plasmas has been gleaned. Charge exchange recombination spectroscopy (CXRS), whereby excited states of hydrogen-like ions of low-Z impurities, such as carbon and oxygen, are populated by charge transfer from atomic hydrogen beam atoms, has enabled detailed local measurements of ion temperature profiles. This is crucial to our attempts to understand the mechanisms responsible for cross-field energy transport. Perhaps even more important, this approach has allowed for the measurement of plasma rotation and, particularly near the edge of plasma, has provided important clues to the rela-

OCR for page 71
Plasma Science: From Fundamental Research to Technological Applications tionship between shear in the radial electric field and the transition from the relatively poor L-mode confinement to the significantly improved H-mode confinement regime. Future Prospects There is a clear synergism between the need for improved diagnostic capabilities and advances in plasma science that either result from meeting that need or are the means to drive the improvement. Of the many areas that will continue to demand attention in the future, the most important may be the need to develop diagnostics for measuring properties of burning fusion plasmas. Most state-of-the-art diagnostic techniques have to be reexamined when the harsh neutron and radiation environment of a power-producing plasma is considered. With burning plasma, we are faced with the additional task of diagnosing the confinement and slowing down the alpha particles that must ultimately provide the power to sustain an ignited fusion plasma. Although much has already been accomplished in these areas, innovation in the next decade must proceed at a pace at least as rapid as that of the last 10 years. In addition, measurements of other fusion products, as well as stability and confinement of reacting plasmas in the presence of copious amounts of alpha particles, must proceed. NON-TOKAMAK CONCEPTS Introduction and Background Fusion plasma physics has historically evolved through the exploration of a variety of magnetic configurations. The tokamak is the most highly developed concept, and the most of the discussion presented above concentrated on this configuration. However, the need for innovative and diverse ideas is as vital as ever in view of the projected multidecade development that lies ahead for fusion. The main non-tokamak concepts presently under investigation are the stellarator, the reversed-field pinch (RFP), and the compact torus. Each has potential advantages over the tokamak and is a unique source for new plasma physics information. Stellarators have the potential for a steady-state reactor without the need for the inductively driven current. Reversed-field pinches, with a relatively weak magnetic field, offer the potential for a compact, high-power-density reactor with normal (non-superconducting) coils. Compact tori offer a reactor geometry in which the plasma torus does not link external conductors. The magnetic mirror approach to fusion is not discussed here since research on this concept ended in the mid 1980s. The stellarator and reversed-field pinch share the same magnetic topology with the tokamak: toroidally nested, closed magnetic surfaces produced by helical magnetic fields. The relative strengths and origin of the toroidal and poloidal

OCR for page 71
Plasma Science: From Fundamental Research to Technological Applications components of the magnetic field distinguish the different approaches. In the stellarator, the magnetic field is produced by current in external windings; in the tokamak, by plasma current for the poloidal field and an external current for the toroidal field; and in the RFP, predominantly by plasma current. Recent Advances Microwave and radio-frequency waves are now used to create ''true" stellarator plasmas without a net internal current, thus avoiding the problem of "disruptions" that destroy plasma confinement. The absence of a net internal current has allowed detailed demonstrations of the nature of the pressure-driven "bootstrap" current (which must be maximized in tokamaks and minimized in stellarators for optimum performance) through its dependence on magnetic field curvature, as well as its control (e.g., reversing its direction). The plasmas are quiescent with no global instabilities. Operation in the "second stability" regime (in which plasma stability increases with increasing pressure) has been obtained and a connection made with energy confinement. External control has allowed a wide range of magnetic configurations. The similarities in plasma confinement between stellarators with very different magnetic configurations, and between stellarators and tokamaks, suggest that similar mechanisms may be responsible for global transport. The confinement scaling is similar in some tokamaks and stellarators, although stellarators exhibit a more favorable density dependence. The edge fluctuation levels, the corresponding particle transport, and the properties of the edge plasma are similar in both devices. Particle confinement is controlled by the edge properties. It has been increased by using electrically biased limiters and decreased by using magnetic error fields. The improved energy confinement behavior seen in tokamaks is now also seen in stellarators. Initial experiments with biased plates that intercept field lines exiting the plasma (divertors) are encouraging for eventual steady-state particle control in stellarators. The RFP has evolved significantly during the past decade, both in the understanding of the physics and in the plasma parameters achieved. A fascinating property of the RFP is that it spontaneously generates a portion of its confining magnetic field. This constitutes a laboratory demonstration of the "dynamo" effect, analogous to the astrophysical dynamo responsible for magnetic field generation in stars. A thorough first-principles understanding of the dominant magnetic fluctuations in the RFP has been established through three-dimensional, nonlinear, magnetofluid computations. This theory agrees with experimental observation of fluctuation spectra and nonlinear three-wave coupling. It also offers a detailed explanation of the dynamo mechanism. The equilibrium magnetic field also is understood as a minimum energy state arising from plasma relaxation. These concepts carry over to tokamak phenomena, such as relaxation oscillations and disruptions. Recent attention has turned to the transport result-

OCR for page 71
Plasma Science: From Fundamental Research to Technological Applications ing from the fluctuations. Techniques have been developed to measure directly the energy and particle transport driven by magnetic fluctuations. As a result of the intrinsically low magnetic field in the RFP, all devices operate at high, reactor-level values of plasma beta. However, the quality of energy confinement, which is typically an order of magnitude worse than that in comparable-size tokamaks, requires improvement and is a topic of continuing research. Another line of research is experiments with the so-called compact tori, with low aspect ratios. Stable equilibria have been produced in the laboratory, despite the absence of a strong toroidal magnetic: field thought to be necessary for stability. This concept has the great advantage of a small unit size, which would significantly lower the cost of an eventual reactor based on this concept. Future Prospects Present experiments can develop much of the physics basis needed for improving stellarators, including tests of the contrasting optimization principles for the two main types of stellarators. The largest efforts involve complementary experiments in the United States, Japan, and Germany. Increased plasma heating power will permit fusion-reactor-relevant properties (higher pressure and improved confinement) in long-pulse (30-second) operation, optimization of the stellarator configuration and operational techniques for future large stellarators, and development of steady-state power and particle handling. Large superconducting-coil stellarators now under design and construction will allow true steady-state disruption-free plasmas without the need for externally driven currents or internally driven "bootstrap" currents. The higher heating powers available in these large stellarators will allow studies at higher plasma parameters (pressure, temperature, confinement time) needed for stellarator reactor development. The evolution of the RFP as a fusion concept requires improvement in energy confinement. From the present experimental database, it is anticipated that transport will be reduced with plasma current. The highest RFP plasma currents operated to date are about 0.6 MA, for durations of less than 0.1 s. Currently, experiments are starting up in Italy that will produce 2-MA plasmas for 0.25 s. This will permit observation of the scaling of confinement on critical parameters, such as the Lundquist number (a measure of the electrical conductivity), which is particularly important for the MHD phenomena prevalent in the RFP. In addition, the evolving understanding of RFP fluctuations and transport is beginning to provide a scientific basis to develop methods to enhance energy confinement. Experiments are beginning in this area. A major question in compact torus research is whether stable plasmas exist in which the ion radius of gyration about the magnetic field is small. Early experiments possibly were stabilized by the presence of ions with large gyroradii. Compact tori experiments with smaller gyroradii are just beginning.

OCR for page 71
Plasma Science: From Fundamental Research to Technological Applications CONCLUSIONS The contribution of magnetic fusion research to the field of plasma science has been very significant. Besides being a driver for the development of modern plasma physics, fusion also has benefited greatly from the many advances in basic plasma science. Perhaps the most important area of future research is to learn how to "control" high-temperature plasma in modern confinement devices, which will require learning more about transport and devising effective means of controlling it. This also implies finding stable equilibria at the upper end of the high-beta regimes achieved to date and going beyond present beta limits, especially at high values of βp (i.e., reduced plasma current in tokamaks). We must learn how to control radial plasma profiles, including those of temperature, density, and current density. At high currents, we must learn how to control disruptions, especially through current profile control. Control is clearly essential for achieving a more attractive fusion reactor based on the tokamak concept. In addition, pursuing other confinement concepts is important, particularly if attempts at control of the tokamak plasma fail or become too complex and expensive. It is also conceivable that a more effective confinement concept than the tokamak could emerge, especially if a steady-state reactor is desired because of technological considerations. However, in the past, funding limitations have often prevented a thorough development of alternate confinement concepts, with the possible exception of the stellerator. In all confinement concepts, the issue of power and particle exhaust (helium removal) must find a solution in plasma science. This problem is just beginning to be addressed by the scientific community, and its solution will require a thorough theoretical analysis, often involving large codes, and experimental research in the area of "plasma edge" physics. To succeed, this study must include a combination of plasma science, atomic physics, and materials science. Finally, as the next generation of tokamaks enters the thermonuclear regime with burning D-T fuel, the generation of copious amounts of 3.5-MeV alpha particles will open the door to the study of alpha-particle-related plasma phenomena, including stability and transport. New diagnostics may have to be developed to study the interior of the burning plasma environment. Unfortunately, in the past, many opportunities for fundamental scientific exploration were missed, in some instances because of funding constraints and in others because of changing priorities within the fusion program. Perhaps the biggest problem in funding more scientific investigations in magnetic fusion is that the level of funding of this fusion program has decreased, in real dollars, during the past decade. Thus, painful choices have often had to be made between upgrading larger facilities to operate in high-performance regimes and increasing the scope of scientific investigations in intermediate-scale devices. Given the mission-oriented mandate of DOE's Office of Fusion Energy (OFE), further research and development will continue to shift toward issues

OCR for page 71
Plasma Science: From Fundamental Research to Technological Applications related to burning plasmas, nuclear technology, reactor relevant materials, and so on. At the same time, other funding agencies, such as the National Science Foundation (NSF) and other offices in DOE, have not funded scientific investigations in high-temperature plasmas. If this trend continues, a serious void in the continued growth of high-temperature plasma science will result, despite its scientific merits. The international fusion community is now engaged in the design of a major fusion facility, the International Thermonuclear Experimental Reactor. This facility will investigate the behavior of burning plasmas under conditions of intense self-heating by alpha particles, and it will demonstrate integration of the nuclear technologies required for a fusion reactor. In addition, the U.S. program has proposed construction of a national facility, the Tokamak Physics Experiment, to investigate modes of continuous operation under advanced performance conditions. The TPX, which is illustrated schematically in Figure 4.4, is planned to begin operation by the end of this decade, when it would become the "work-horse" for research in high-temperature plasma science in the United States. RECOMMENDATIONS For the continued development of plasma physics as a scientific discipline it is essential that there be a continued experimental capability to investigate high-temperature plasma phenomena. It is clear that a commitment to increased high-temperature plasma research and training of scientific manpower should be made now. With appropriate funding, the number of graduate students working toward a PhD in plasma-related fields is sufficient to meet such a commitment. The mainstay of this kind of research will remain the DOE Office of Fusion Energy (OFE). However, increased support for energy-relevant basic plasma science by the Office of Basic Energy Sciences (BES) at DOE, in cooperation with the OFE, which is recommended in Part I, would greatly benefit all energy-relevant plasma science and technology. This program could help fund specific experiments on large machines, as well as the operation of small and medium-sized experiments. Funding at the level of several hundred thousand dollars per year per investigator would be of considerable value to university and industry efforts, even for participation in a large experiment. Initial investment in equipment at the level of a few hundred thousand dollars would also be of additional value. Diagnostic-type experiments could be carried out "piggyback" style at existing facilities. Many plasma physics problems are best addressed in small- and medium-scale devices. Such devices can be used to test innovative confinement concepts, and the panel sees a need for two to three devices in the United States. In addition, somewhat larger-scale facilities would be desirable to continue basic research in high-temperature (a few keV) plasmas. Such devices might in some

OCR for page 71
Plasma Science: From Fundamental Research to Technological Applications FIGURE 4.4 Schematic diagram of the advanced superconducting Tokamak Physics Experiment (TPX), proposed as a national facility to develop the scientific basis for a compact and continuously operating tokamak fusion reactor. Operating at a toroidal magnetic field of 40 kG and plasma currents of 2 MA, the TPX would investigate modes of enhanced plasma confinement and non-inductive current drive for plasmas lasting longer than 1000 s. (Courtesy of Princeton Plasma Physics Laboratory.) cases function as user facilities, supported by a consortium of institutions and funding agencies. The panel envisions at least two such devices operating as user facilities. These facilities may be converted from currently operating and/or mothballed devices, most likely tokamaks. Appropriate nonohmic heating and current drive capability should be available on such a device. Finally, it is also important to maintain a strong parallel program in theory and modeling, for it is the interaction between experiment and theory that facilitates the greatest progress in plasma science. In any case, it is clear that a commitment to continued support of research in high-temperature plasma science should be made now.