Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
OTHER APPLICATIONS OF FUSION FISSILE FUEL PRODUCTION It should be evident from the preceding discussions that the parameter Q, the ratio of nuclear power yield to power input, is of considerable importance to any fusion reactor design. Insofar as a fusion reactor may be viewed as an energy-multiplying device, practical considerations warrant as high a Q value as possible. For some time it has been recog- nized that one can achieve very high overall Q even though the fusion reactor, per se, operates at Q no greater than unity. The energy multiplication, then, results from conversion processes external to the fusion reactor, and is dependent largely on what applications one may foresee for the l4.l MeV neutrons, which carry some 80 percent of the energy released in the D-T fusion reaction. In a suitably designed blanket the fast l4.l MeV neutrons can be made to multiply and produce a larger number of slower neutrons through n-2n reactions and fast fission. As a result, it is possible to design blan- kets containing lithium and natural (or depleted) uranium, with or with- out thorium, so that for every fast neutron entering from the fusion reactor core, one obtains one (plus a slight excess) tritium atom to replace the one lost by fusion, a number of fissile atoms of either plu- tonium or 233U or both, plus an amount of heat. A blanket designed to optimize fissile fuel production is usually termed a fusion-fission symbiont, while one designed to optimize power production is termed a fusion-fission hybrid. Notice that up to five fissile atoms can be obtained in this cascade process for each fusion-produced neutron, and recalling that with fuel recycling each fissile atom can yield 300 MeV in an external fission burner (thermal reactor), we readily perceive that the total energy release by the hybrid or symbiont mode can be two orders of magnitude greater than by the pure fusion mode, for each D-T event. Thus, for commercial introduction of early generation fusion, approaches with intrinsic Q near unity (for example, two-component Tokamaks, mirror devices, or pellet fusion based on simple targets) enjoy better prospects as hybrids or symbionts than they do as pure fusion reactors. Moreover, we might argue, as some U.S.S.R. spokesmen have, that the inherent ad- vantages of hybrids or symbionts tend to rule out any practical interest in pure fusion development at this time. The opposite view, however, 29
30 can also be readily stated: the engineering complexities and environ- mental problems that would result from adding fissioning or fissile- fuel-producing blankets to a fusion reactor are so formidable that the pursuit of fission-fusion concepts would constitute a counter productive diversion to the fusion program. It should be noted that fusion-fission systems based on the plasma design goals of the next generation Tokamaks appear to yield attractive fission fuel breeders. The symbiont concept is clearly an alternative to the fission breeder, with possible advantages in terms of controlling accidental increases in reactivity, and with additional possible advantages under circumstances where fissile fuel production for an existing reactor complex is more important than the expansion of generating capacity.* In this context, it is also necessary to consider the "electrical breeding" or spallation concept. In the spallation process, high-energy accelerators are used to produce energetic beams of protons or deuterons, which are subsequently fired at liquid metal primary targets (lithium, bismuth, or lead being possible candidates), or appropriately cooled solid targets of uranium or thorium, to produce large numbers of neutrons. In effect, the neu- trons "boil off" the target nuclei and continue to multiply in a suitably designed blanket or secondary target. The cascade process is somewhat similar to what has been described above, and it is possible to conclude that one fissile atom would be produced for each l0 to 20 MeV of beam energy in a suitably designed blanket. The concept dates back to the late l940's and early l950's, when E. O. Lawrence initiated the materials testing accelerator (MTA) project in the U.S. There has been a revival of interest in spallation, particularly among Canadian workers at Atomic Energy of Canada, Ltd. (AECL). Most current concepts envision ion beams of about 0.5 to l GeV in energy with currents ranging from 0.25 to 0.5 amperes. Advances in RF power supplies have contributed to the renewed interest in building such large (approximately l kilometer in length) ion beam accelerators. Some rather preliminary estimates indicate that both symbiont and spallation approaches might lead to fissile fuel production at a cost of $50 to $l00 per gram of 233U or 239Pu. As these estimates are in the potentially attractive range for future considerations, a more detailed discussion of this matter is included in a separate appendix. One needs to bear in mind the distinction, however, in that the "electrical breeder" using conventional accelerators has a considerably more advanced tech- nological base than the fusion-fission symbiont, where the latter is based on as-yet-untested fusion reactor concepts. *Since a fusion-fission symbiont, in principle, could fuel four to five LWR's, we can envision a converter reactor economy fueled by fusion- fission devices, ad infinitum. Whether this is on interest depends ul- timately on the relative economics of such a hybrid complex versus a fission breeder economy.
3l CHEMICAL FUEL PRODUCTION Some amount of attention has also been given to using the l4.l MeV fusion neutrons to produce chemical fuels by radiolysis. For example, production of CO from CO2 by neutrons, alpha particles, secondary gamma rays, and other secondary nuclear reaction products may be considered. The CO may then be used to generate hydrogen by the familiar water- shift reaction. Measured efficiencies for radiolytic production of CO from CO2 by gamma radiation is l5 percent and by alpha radiation is 30 percent. There is no data available for radiolysis efficiencies by l4-MeV neutrons, though conservative guesses place it midway between the above two figures. Approximately, one estimates that a one meter thickness of CO2 blanket at 200 atm (gas phase) would be sufficient to attenuate the l4.l-MeV neutrons. Direct radiolysis of water by ultraviolet radiation to product hydro- gen may also be considered, and more complex schemes using recyclable hydrogen halides have also been proposed. Advanced fusion fuel cycles, such as p - 11B, in which much of the energy is emitted as hard x-rays, offer unique possibilities for chemical fuel production. Given our present state of understanding, it is difficult to draw any conclusions about the merits of configuring fusion reactors such that the bulk of their net energy output is realized through the production of chemical fuels. The proposition certainly deserves more careful analysis. NUCLEAR WASTE DISPOSAL In principle, any neutron-rich source might be used to transmute accumu- lated radioactive waste products to nuclei that either are stable or decay more rapidly to stable ones. Thus, transmutation provides a pos- sible means of bringing the radioactivity levels of nuclear wastes down to values comparable to natural radioactive material. Were this to be economically feasible, transmutation would offer an alternative to disposal in geological formations. It has been proposed that some fusion reactors could be dedicated to burning fission wastes. Preliminary studies indicate that for the transmutation of the actinides and most long-lived fission products, neutron wall loadings on the order of 5 to l0 MW per square meter would be required. Such values of wall loadings are about a factor of two higher than those proposed in current pure-fusion reactor designs; however, limitations on wall loading are yet to be determined experi- mentally and values of 5 to l0 MW per square meter might be feasible. Transmutation of 90Sr and 137Cs, on the other hand, would require on the order of 200 to 500 MW per square meter of neutron wall loading, and it is difficult to see how this can be sustained in currently conceived fusion reactors; consequently, new concepts, such as the Linus experi- ment at the Naval Research Laboratory, would have to be explored to meet this requirement.
32 SUMMARY In conclusion, we do not believe that any of the alternative approaches to pure fusion have been examined in sufficient detail to warrant a marked change in present plans, which have as their primary goal the develop- ment of pure-fusion reactors.
