An experimental fusion reactor that may help quench the world's thirst for energy will soon be taking shape. The International Thermonuclear Experimental Reactor, or ITER, is part of a program that intends to demonstrate the feasibility of fusion energy for peaceful purposes.
The program involves China, the EU, Japan, Korea, Russia, Switzerland, and the U.S. Construction will begin in 2007 at a site in Cadarache, France, with the $5 billion plant expected to be up and running in 2016.
ITER's goal is not electricity generation, per se, but rather to serve as an experimental test bed for optimizing fusion processes and verifying the hardware and systems that will be needed in tomorrow's electricity-producing fusion-power plants. Up to now, research has only addressed the scientific feasibility of fusion, with little attention to how it can be turned into a practical power source. The design is based around a hydrogen plasma torus operating at over 100 million °C, and it should produce 500 MW of fusion power.
A range of experimental operating modes will address issues such as plasma confinement and stability and getting rid of impurities. Specific goals include demonstrating power amplification generating up to 10 times more power than it consumes and steady-state "burning" of plasma. The reactor will also test components such as heat shields and superconducting magnets needed to make fusion energy generation practical. If these efforts prove successful, the next step will be building a prototype plant to demonstrate reliable electricity generation using fusion power.
Experts fear the combination of a growing population and higher living standards will considerably increase world electricity demand in coming years. At the same time, demands to reduce fossil fuel use for environmental and political reasons leads to the need for other energy options.
Fusion holds promise as an essentially unlimited energy source with manageable environmental impact. The physics and engineering of a fusion-power station are not completely understood. But, according to ITER officials, the basic principles have been worked out and no technical roadblocks have been identified that will stop its development into a viable source of electricity. But the unanswered questions facing fusion are how to optimize the process and make it economically viable.
For energy production, experts generally agree the most suitable reaction involves the two heavy isotopes of hydrogen, deuterium and tritium. A sufficiently high rate of fusion requires temperatures hot enough to separate electrons from atomic nuclei, changing the fuel to plasma hotter than the center of the sun. Deuterium-tritium fusion creates helium, neutrons, and energy. For continuous operation, helium must be constantly removed and new fuel fed to the plasma. Surrounding structures absorb the neutrons and transfer heat to turbogenerators to make electricity.
The process is inherently safe, say ITER designers. No chain reaction is involved and cutting off fuel rapidly extinguishes the plasma. Likewise, closing the exhaust poisonsthe plasma with impurities and brings fusion to a halt. Large heat-transfer surfaces and watercooled heat sinks in ITER maintain low temperatures and prevent components from melting. And leak-tight containment barriers, needed for the process to work, confine any contaminants.
Radioactivity is a concern, though not nearly at the levels found in fission reactors. Tritium has a half-life of 12.5 years (meaning only 0.4% of the radioactivity remains after 100 years). Components in the reaction chamber will become radioactive, but most will return to safe levels 50 to 100 years after shutdown.
The heart of ITER is a tokamak, a toroidal device that uses electric currents and magnetic fields to confine and heat the 800 m 3 gaseous plasma. The tokamak also ensures that any plasma particles that strike the surroundings are at low energy and hit heat-tolerant components. Otherwise, the plasma could cool, generate impurities, and cause damage.
A superconducting-magnet system consists of 18 D-shape toroidal-field (TF) coils, six poloidal-field (PF) coils, and a central solenoid (CS) coil. (Poloidal field lines loop the short way around a torus, whereas toroidal field lines loop the long way.) In addition, saddle-shape correction coils outside the TF magnets compensate for field errors due to manufacturing inaccuracies and misalignments during assembly. They also help control plasma instabilities.
The coils are massive, and will store more energy at higher fields than any ever built. The CS coil, for instance, weighs 840 t and stands about 12 m high and 4 m in diameter. Each TF coil weighs 290 t and measures 14 m high 9 m wide. These coils use superconducting Nb3Sn in a cable-in-conduit configuration. For TF coils, some 1,100 wires about 0.7 mm in diameter twist together inside a 4-cm diameter metal tube 820-m long. Cryogenic circulation pumps send supercritical helium through the tube, around the wires, and down a central gap to cool the coils.
Nb3 Sn is brittle, so manufacturing is time consuming and expensive. Initially, separate Nb and Sn wires (in a copper matrix) are wound into shape, then joined during a 200-hr heat treatment at 650°C. Next, it is electrically insulated with a wrap of glass fiber and Kapton polyimide, structurally reinforced, filled with liquid epoxy resin, and cured. Finally, it's further reinforced in a steel housing.
PF coils are in a field region that permits less-expensive NbTi strands. The PF coils, however, link with many systems, making replacement difficult. So each carries redundant turns so "incipient" short circuits are detected and isolated before causing damage. As a further precaution, the coils can be rewound in place or replaced, although at a price in machine downtime.
Reactor operation will verify coil performance and reliability, as well as the integrity of joints, conductors, and insulation and overall manufacturing and assembly processes.
Fusion takes place in a welded, double-walled stainless-steel reactor vessel with internal shield plates and ferromagnetic inserts to reduce magneticfield ripple. It maintains an ultrahigh vacuum required for the plasma to form. The vessel holds a range of components, including blanket modules that absorb heat and provide shielding; a divertor that removes helium and other impurities; and access port plugs for heating, fueling, monitoring, and remote maintenance.
The vessel and internal components absorb and reduce neutron energy to levels tolerable for the magnets and surrounding equipment. It also confines coolant leaks so that radioactive materials cannot spread outside the plant. Separate water circuits cool the blanket, divertor, and vacuum vessel.
A thermally shielded cryostat houses the reactor vessel and superconducting magnets, and maintains ultralow temperatures for superconductivity. The cryostat is a reinforced single-wall cylinder 24-m high and 28-m in diameter (volume ≈14,000 m3) which also serves as a second confinement barrier.
Cryostat vacuum minimizes convective heat transfer, and thermal shields stainless-steel panels cooled by 80°K helium gas between the cold magnets and warm vessel minimize radiative heat transfer.
A bioshield surrounds the cryostat and reactor vessel. This concrete structure reduces radiation to safe levels and lets personnel access equipment soon after the tokamak stops operating.
Components near the reacting plasma eventually become radioactive. If replacement becomes necessary, parts must be removed by a remote handling system, placed in casks, and transported to hot cells for repair or disposal. A rail-mounted vehicle in the plasma chamber will service blanket modules, and a cantilevered transporter will maintain the divertor and port plugs. It is estimated that a single blanket module will take 25 days to replace, and the complete blanket (approximately 440 modules) 9 months.
A combination of radio-frequency and neutralbeam systems heats the plasma and drives internal currents necessary for long burn pulses. The plasma absorbs energy at specific frequencies from RF waves, which are transmitted by antennas in the port plugs. Neutral beams are high-energy deuterium atoms injected into the plasma.
Experiments with combinations of neutral beam and RF systems, with a total heating power exceeding 110 MW, will determine the best mix.
Steady-state operation requires active control of the plasma current in which the "bootstrap" effect generates a large percentage of current needed for fusion. Bootstrap current results from a complex interplay between particles in a tokamak. It is produced by the plasma itself, not an external source. Thus, bootstrap current can take the place of inductive current drive, reducing the input power needed for steady-state burning.
Plasma exhaust deuterium, tritium, helium from fusion reactions, and impurities leaves through vacuum pumping ports and is separated into its constituents. Deuterium and tritium are reinjected to the plasma, either as frozen pellets from the inboard plasma side for good penetration into the plasma core, or by gas injection at the top of the plasma. Impurities are removed as waste and helium recycled if sufficiently pure.
ITER's elongated plasma would be vertically unstable without active feedback control. For instance, one control loop adjusts the up-down asymmetry of outer PF coil voltages to minimize movement in the plasma current center. Magnetic diagnostic loops on the inner vacuum-vessel surface keep track of plasma currents. Another loop linked to the main PF coil voltages helps limit changes in six measured gaps between the plasma and wall.
More than 40 diagnostic systems will monitor the plasma. The seven main types are magnetic, neutron, optical/infrared, bolometric (heat radiation), spectroscopic, microwave, and plasma-facing-component. They are also divided into three categories: basic machine protection and control, advanced plasma control, and evaluation and physics studies.
A central supervisory control system (SCS) and local subcontrol systems handle control, data acquisition, and communications during normal operation. The SCS also monitors and ensures plant subsystems operate within proper limits.
ITER will occupy a site of about 100 acres and construction is expected to take seven years. It will operate for an estimated 21 years, followed by six years of decommissioning to remove radioactive materials.
Materials Under Fire
Materials close to the plasma in a fusion-power reactor are bombarded by high-energy neutrons yet must produce as little radioactive waste as possible. Thus, low-activation materials those with a limited propensity to become radioactive under neutron bombardment are a must if fusion is to become a viable energy source. ITER will typically produce damage of 3 dpa (displacements per atom) in the austenitic stainless steel of the first containment wall. With judicious use of low Nb and Co grades of steel, most radioactive waste in ITER (except part of the vessel and its internal components) will be cleared for unrestricted reuse a century after decommissioning.
For commercial power reactors, damage to the first material walls if made of stainless steel would be approximately 300 to 500 dpa over a 30-year life. Even if the walls could be changed every few years, this amount of damage is beyond the capability of austenitic steels, which significantly swell above damage levels of 30 dpa.
Materials that last longer or experience less damage will thus be needed. Some promising candidates, low-activation ferritic steels and SiC composites, withstand more than 150 dpa without swelling.
Despite the relatively low damage rate, ITER will be the first facility where materials face a true fusion-neutron spectrum. How components stand up is of great interest. For instance, diagnostic systems must maintain optical, electrical, and structural properties under high radiation doses. Plasma-facing materials must ensure plasma purity and efficiently remove heat. Joints to underlying heat sinks and structures, and the cooling system, will also be scrutinized.
The magnet support structure, on account of its large size, must be assembled using welds. This introduces weak points liable to fatigue. And how the superconductors behave in operation is a key consideration.
By far the most widely used structural material in ITER is austenitic stainless steel, which has largely been qualified for nuclear use through fission and fast-breeder development programs. Beryllium, tungsten, and carbon-fiber composite are used in the first wall facing the plasma. These materials join to copper-alloy heat sinks and, in turn, to stainless-steel supports.
Operation will include development and testing of lithium-based, tritium-breeding blankets essential for tritium self-sufficiency in future fusion-power reactors. These blankets may be connected to a turbogenerator and generate electricity from fusion for the first time.