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.
Appendix C Nonreactor, Nonrepository Disposal of Excess Weapons Plutonium: Technical Issues INTRODUCTION Chapter 6 describes a variety of options for dealing with long-term dispo- sition of weapons plutonium. The options involving nuclear reactors, accelera- tor-driven subcritical assemblies, and reactor wastes are described in more detail in the report of the Panel on Reactor-Related Options for the Disposition of Excess Weapons Plutonium.i This appendix offers additional detail on the disposal options not covered in that report, including: · disposal in deep boreholes; · sub-seabed disposal; · ocean dilution; · space disposal; and · underground nuclear explosions. Consistent with the discussion in Chapter 6, options should: ~ Management and Disposition of Excess Weapons Plutonium: Report of the Panel on Reactor- Related Options (Washington, D.C.: National Academy Press, 1994). 245
246 APPENDIX C 1. minimize the time during which the plutonium is stored in forms readily usable in weapons; 2. preserve material safeguards and security during the disposition process, seeking to maintain the same high standards of security and accounting ap- plied to stored nuclear weapons; 3. result in a form from which the plutonium would be as difficult to recover for weapons use as the larger and growing quantity of plutonium in com- mercial spent fuel; and 4. meet high standards of protection for public and worker health and for the environment. This criterion must include not only expected situations but possible failures (particularly during the disposal process), and not only safety but the extent to which that safety is demonstrable to the public. In addition, this appendix examines the costs of the various options, issues such as public and institutional acceptance, and possible conflicts with existing agreements and policies. Several of the disposal options that might be pursued for disposition of ex cess weapons plutonium have also been considered for disposal of spent fuel or high-level waste (HLW), for which the international consensus today favors burial in mined geologic repositories, rather than any of the disposal options outlined in this appendix. There are a number of important differences, how ever, between weapons-grade plutonium and spent fuel or HLW. These include: · Heat. The heat output of weapons plutonium is roughly 3 watts per kilo- gram, or 30 watts for a package containing 10 kilograms of plutonium. A comparable disposal package of 10-year old spent fuel or HLW typically gives off 1,000-2,000 watts of heat. · Radioactivity. The gamma radiation from a typical package of weapons plu- tonium at 1 meter would amount to only thousandths of a rem (roentgen- equivalent-man) per hour, while for spent fuel assemblies or vitrified logs of HLW, the equivalent figure is thousands of rems per hour (for the first few decades after these products are produced). · Toxicity. Weapons plutonium, spent fuel, and HLW are all highly toxic, primarily because of their radioactivity. The alpha radiation from plutonium is particularly damaging if the small particles are inhaled and lodge in the lungs. Environment and health risks from all of these materials must be carefully considered over very long times in evaluating disposal options. Mass and Volume. The nominal stock of excess weapons plutonium is 50 tons each for the United States and Russia, which could in principle be stored in a single large warehouse. The global stock of spent fuel by the year 2000 will amount to more than 150,000 tons (containing over 1,400 tons of plutonium), occupying a vastly larger volume. Thus some options that might be prohibitively costly for spent fuel might not be so for excess weapons plutonium. .
APPENDIX C 247 · Perceived Value and Security Risk. HLW has virtually no value to anyone; hence, it is extremely unlikely that anyone would seek to recover the mate- rial, except perhaps to monitor the disposal mechanism or to correct a per- ceived failure of that mechanism. By contrast, weapons plutonium might have significant economic value as a fuel decades from now, and even sev- eral kilograms of weapons plutonium could be extremely valuable to a prolif- erator for use in nuclear weapons. For a particular disposal approach to meet the "spent fuel standard" outlined in Chapter 6, it must be as difficult (measured in likely cost, time, and availability of the needed technologies) to retrieve the plutonium for use in weapons as it would be to separate a similar amount of plutonium in spent fuel for the same purpose. DISPOSAL IN DEEP BOREHOLES Description Disposal in deep boreholes has been considered in several countries for spent fuel or HLW (generally as a backup to the currently preferred approach of disposal in mined geologic repositories nearer the surface), and this is a possi- ble approach for plutonium disposal as well. Studies in Denmark, the United States, and Sweden have examined the borehole approach in some detail.2 The approach appears technically feasible, though a substantial period of additional development would be required to answer outstanding questions and provide information for licensing. For example, wastes might be emplaced in the lower 2,000 meters of a hole drilled to a depth of 4,000 meters, with a diameter of 1 meter. Rather than simply placing plutonium pits in canisters in such a hole, some processing be- fore emplacement would be required, to eliminate void space and to prevent the possible development of conditions in which the plutonium could sustain a nu- clear chain reaction, producing heat and fission products in the hole so-called "criticality."3 Nevertheless, even if the plutonium itself were only 10 percent of the weight of the final product (and an even smaller fraction of its volume), it 2 For a recent summary discussion, see J. Swahn, The Long-Term Nuclear Explosives Predicament: The Final Disposal of Militarily Usable Fissile Material in Nuclear Waste from Nuclear Power and the Elimination of Nuclear Weapons, ISBN 91-7032-6894 (Goteborg, 1992). A Swedish summary report is particularly useful: Svensk Karnbranslehantering AB (Swedish Nuclear Fuel and Waste Management Co.), Storage of Nuclear Waste in Very Deep Boreholes: Feasibility Study and Assessment of Economic Potential, Technical Report 89-39 (in English), December 1989. Earlier work is reported, for example, in Woodward-Clyde Consultants, Very Deep Hole Systems Engineering Studies, 0NWI-226 (San Francisco: Woodward-Clyde Consultants, April 1981). The baseline concept in the latter is a 20,000-foot (6-kilometer) borehole, in contrast to an even more challenging initial proposal of 10-kilometer depth. 3 For example, the plutonium might be vitrified in a borosilicate glass before emplacement, as described in Chapter 6 for placement in a mined repository; in this case, vitrification could be without HLW, since the difficulty of access to the deep borehole would provide the primary barrier to retrieval of the plutonium.
248 APPENDIX C would be possible in principle to place the nominal 50-metric-ton stock of ex- cess weapons plutonium in a single borehole. Boreholes have been drilled to depths between 1,500 and 6,000 meters in the United States, Germany, the former Soviet Union, Italy, and Sweden. Thus the feasibility of drilling holes to appropriate depths can be considered demon- strated, although emplacing large quantities of material at depth would involve some additional engineering challenges. A deep borehole for disposal of either HEW or weapons plutonium would have to provide a substantial volume at depth, in contrast with the usual deep hole for exploration or for the production of oil or gas. For example, a detailed Swedish study on HEW disposal in deep holes fo- cuses on a "preferred option" that would involve a hole 80 centimeters in di- ameter at depth, into which would be placed canisters with a length of 4.4 meters and a diameter of 50 centimeters, centered in the hole.4 Each canister would be separated from the next by sealing plugs of compressed bentonite clay. After the "deployment zone" was filled, an additional long length of ben- tonite clay would be used to seal the hole. For the depth range of 250-500 meters, asphalt would be emplaced in the hole, and for the top 250 meters, a high-density concrete would be used to provide a cap. The hole would be drilled with normal drilling equipment, with the upper portions having a wider diameter than the actual deployment zone. The upper portion of the hole in the Swedish concept would have a casing with an internal diameter of 100 centimeters, while the lower portion would have a "liner" throughout the deployment zone with an internal diameter of 60 centimeters (2 feet) to allow the deployment of the 50-centimeter-diameter canisters. The up- per casing might be pulled after deployment. As the Swedish report points out, the United States took a similar approach in the drilling of a deep hole on Amchitka Island, Alaska, in 1969. The casing emplaced there was 1,860 meters long, with a diameter of 137.5 centimeters, a weight of 1,820 tons, and a wall thickness of 6.4 centimeters. Such a hole could be prepared in less than a year, and filled and sealed in another year or so. The "liner" for the hole envisioned in the Swedish study used to keep the borehole open for deployment of the canisters would be quite different from anything used in normal drilling practice today, in that a substantial fraction of it would be holes open to the surrounding rock, so that the sealing clay could readily extrude through the liner to make good contact with the rock wall of the hole. The clay would provide support for the hole wall and would help seal any cracks or fissures. Crystalline rock at great depth is under substantial stress. Because of the uneven grain of the rock and the slow strain of the earth's crust, the horizontal stress is not uniform the rock is being pulled in some directions more than others. Typically, the maximum horizontal stress may be 1.5 times the vertical 4 See Svensk Karnbranslehantering AB, op. cit.
APPENDIX C 249 stress ("lithostatic pressure"), whereas the horizontal stress at right angles to the maximum may be 0.5 times the vertical stress. Because of the uneven stress on the rock at depth, when a hole is drilled into it the rock tends to be pulled in such a way that the hole becomes elliptical and usually develops "ears"-extensions of the hole at the ends of the ellipse. To minimize this "spelling" problem, the driller must choose an optimum den- sity for the drilling fluid. Drilling with water would mean that the pressure in the rock outside the borehole would be much higher than the support pressure of the water on the wall of the borehole (at the depths of interest), which could lead to the borehole collapsing inward. Using very dense mud can lead to the opposite problem, which can cause the surrounding rock to crack. Even the optimum mud density can do no more than to provide a pressure at depth that is the same in all directions; some cracking and elliptical growth of the cross section of the hole would still result. This must be taken into account, both as it affects deployment of the HEW or plutonium in the hole and as it affects the effectiveness of containment in the hole after the canister disintegrates in time. Environmental Impact Would the borehole reliably prevent the plutonium from being released into the environment at harmful concentrations? Although boreholes have not received anything like the technical scrutiny that has been applied to mined repositories, the great depth of the hole and the very low permeability of crys- talline rock (granite) suggest that the risks of radioactive releases from such holes might be even lower than those from mined repositories. The small area of disturbed material, the long path to the surface, and the possibility of plug- ging many hundreds of meters of hole with diffusion and convection barriers may make this concept effective. Furthermore, the relatively small area exposed means that the materials will be exposed to only a small water flow, and poorly soluble materials such as plutonium will dissolve quite slowly. The main ques- tions are how the plutonium might be conveyed to the surface once emplaced and the potential for accident during emplacement. Crystalline rock at depth has very low porosity and hydraulic conductivity (the ability of water to move through the rock). This means that movement of the plutonium through uncracked rock would be extremely slow, even if water that might be in the borehole ultimately contacted and became saturated with the plutonium.S But to keep the plutonium isolated for many millennia, the deployment hole must also avoid faults in the rock mass. The influence of hori- zontal, angled, and vertical faulting has been modeled, and it is clear that holes must not be located near vertical faults that might allow radionuclides to mi- grate toward the surface. If the borehole were connected at depth to a large, near-vertical fault and a similar connection were available near the top, density s Data are provided in Svensk Karnbranslehantering AB, op. cit.
250 APPENDIX C differences between the fluid in the borehole and that in the fault (for instance, due to fission product decay heat in HEW or spent fuel but not in weapons plutonium) would drive fluid circulation, leading to far more flow than would be available from circulation confined to the pore fluid of the borehole itself. Thus, it will be necessary to characterize candidate regions for deep bore- holes (using normal seismic techniques), and to make measurements from one or more pilot boreholes, in order to avoid emplacing containers in regions of major faulting. The possibility of major faulting over many millennia is one important area of uncertainty that requires further study. For similar reasons, it is important to choose drilling methods that will minimize cracking of the sur- rounding rock. Another important issue is avoiding transport up through the hole itself. This is the purpose of the 2,000 meters of clay, asphalt, and concrete envi- sioned to seal the hole. Assuming that parts of the hole are likely to be satu- rated with groundwater, it is important to ensure that there are no ready means for convection in this water to transport radionuclides upward. For example, dissolved gas, heat, or differences in salinity could in principle reduce the den- sity of the water in the part of the hole where the waste was emplaced (compared to the water above it), causing the lower water to rise slowly through the hole plug. Hence, it is important to choose materials for the waste package that do not generate more gas in the borehole (due to corrosion) than can be dissolved in the water (determined by the solubility limit at depth).6 The increased salinity of water at great depth may essentially eliminate upward convection~uite beyond the limits on convection posed by engineer- ing means such as bentonite seals, avoidance of major faults, and the like. In general, when drilling deep holes, water is encountered whose salinity, in the words of the Swedish report, "increases dramatically with depth in most areas and in some cases approaches saturation." Because saline water is denser than fresh water, it could not rise connectively into an overlying region of fresh wa- ter, even if the saline water were heated (as it would be in the case of disposal of HLW). As the Swedish report concludes: "Clearly, a repository in a saline environment with fresh water above is highly desirable. If the water is highly saline, it appears that no radionuclides at all will be transported to the surface by convection." Highly saline water may be found in drilling from islands or the seabed, or near the margins of the sea. For example, the Swedish study reports that "boreholes on the island of Gotland show a salinity content between four to eight percent, which is much higher than . . . the seawater today." In cases in 6 For the disposal of spent fuel, the Swedish concept (ibid.) would probably involve casting the fuel rods themselves in copper. This would reduce the otherwise large void volume within canisters containing spent fuel assemblies, which (if undisplaced) would reduce the effectiveness of the clay seal after the canister corroded. Copper is suitable for the chemical environment common in granite in Sweden (a reducing, rather than an oxidizing, environment), though it would not be suitable, for example, for waste packages for the U.S. Yucca Mountain geologic repository.
APPENDIX C 251 which the depth to saline water is greater than the depth of the nearby sea, there would generally be no convection upward of the dense saline water to- ward the sea although upward transport could still occur (in principle) in the case of a pressurized aquifer, as in a large-scale Artesian well. If the material to be disposed of generated substantial quantities of heat (as is the case with HEW and spent fuel), the decrease in density resulting from the warming of the surrounding water could lead to upward convection in the ab- sence of such saline water, though it would still not rise through a major salin- ity gradient. This, however, is not relevant for the disposal of plutonium. It is important to note that sorption of plutonium to the small particles in the clay or bentonite used to seal the borehole would provide another major barrier to its emergence. The effect is dramatic: the rate of diffusion through a material that is sorbing the plutonium is reduced by a factor known as the "distribution coefficient," Kit, which for bentonite might be factor of roughly 100,000.7 In reality, however, the rate at which plutonium reaches the surface will be affected by the less effective and less readily analyzed sorption of plu- tonium on the particles in the small faults in the granite. But some preliminary estimates of the flow of water and plutonium through a partially sealed well can be made. Assume, for example, an upflow in the hole of 1 cubic meter of water per year. Solubility of plutonium in analyses of repositories is often given as 10-3 grams per cubic meter; so if the water were fully saturated with plutonium, 10-3 grams would come to the surface each year. As described in the discussion of ocean dilution below, new U.S. regulations will enter into effect on January 1, 1994, limiting the allowable plutonium concentration in water to which mem- bers of the public might be exposed to 2 x 10-8 curies per cubic meter. To meet that standard, each gram of plutonium would have to be diluted in 4 x 106 cubic meters of water. This standard would be satisfied by diluting the plutonium- saturated effluent assumed above with about 4,000 cubic meters of rainwater per year, which falls, on average, over an area of about 0.008 square kilometers (about 2 acres). In reality, two factors would reduce the amount of plutonium transported to the surface still further: first, it will take water itself some 1,000 years or more to move to the surface in the well at the assumed flow rate. Second, and more important, most of the plutonium would be sorbed to the bentonite and other materials in the hole; the ultimate plutonium transport rate of 10-3 grams per year would be achieved only after the 2,000-cubic-meter bentonite column was loaded with plutonium. If plutonium is sorbed to the material used to plug the well with a distribution coefficient of 1Os (the ratio between mass sorbed per cubic meter and plutonium concentration in the pore water), then even at the - below. ' For more on sorption and the distribution coefficient, see the discussion of sub-seabed disposal,
252 APPENDIX C assumed solubility of 10-3 grams per cubic meter, the bentonite column would be fully loaded with plutonium at 1O: x 10-3= 100 grams per cubic meter. Since there are 2,000 meters of bentonite and the plutonium is assumed to be flowing upward at only a thousandth of a gram per year (the solubility limit in the as- sumed water flow), it would take some two hundred million years for the ben- tonite to be loaded-10,000 half-lives of plutonium-239 (Pu-239~. More complex calculations are required to assess the degree to which plu- tonium might emerge through small faults in the rock, which would not have nearly the absorptive capacity of the bentonite-filled borehole. Recall, however, that the above calculations do not include the effects of salinity: if high salinity nt rlenth can he Guaranteed. even extensive faulting would not bring plutonium ~ A- ~ 7 ~ ~_, to the surface. In short, it appears that if the borehole site is chosen appropriately, and the material emplaced correctly, only volcanism or meteor impact would bring the material to the surface in significant quantities. The risks from either of these types of events are lower for boreholes than they would be for mined reposito- ries, the closest comparison, because of the much greater depth of the boreholes. The borehole option, however, would have to be analyzed for various acci- dent possibilities during emplacement, in order to define facilities and proce- dures to reduce their likelihood and to provide means to proceed in case of accident. Borehole collapse during drilling would require redrilling, but collapse after emplacement of canisters begins is a more complicated problem that would need to be addressed during a development program for this option. A set of open questions concerning the borehole option is described in Chapter 6. Cost Swedish estimates place the cost of deep-hole disposal of spent fuel in the range of $100 million per hole, although a Russian group advertises that it will drill a set of holes for much less. There is clearly less processing necessary to transform weapons plutonium to a suitable waste form and to handle the resulting canisters than is the case for HEW or spent fuel, because of the intense gamma radioactivity of these lat- ter products. In any case, it appears certain that in the United States, the costs of devel- opment of the borehole option, and particularly of gaining the needed licensing and approvals (if they were eventually obtained), would substantially exceed the costs of the actual emplacement.
APPENDIX C 253 Retrievability The ability to monitor and retrieve the canisters, once emplaced, would be desirable from the point of view of ensuring that the system was working as expected. But retrievability is not a virtue if the goal of the disposal method is to create major barriers to reuse of the plutonium in weapons. At various times, deep-borehole disposal of canisters of high-level waste has been described as irretrievable, when that was considered desirable, or retrievable, when that was regarded as a virtue. As the Swedish report on this concept put it: It was initially thought that the VDH [very deep hole] concept would not allow the canisters to be retrieved once they had been deployed. Further considera tion of this aspect of the concept indicates this not to be the case. There is no reason why the plugged section tof the original hole] cannot be drilled or washed out with high pressure fluids. Once the canisters have been reached they could be fished out using overshot tools, a standard oilfield practice. This procedure assumes that the canisters are still intact.8 As this quotation suggests, the simplest retrieval approach would involve redrilling the hole, which would be conventional, even easy, for the section filled with bentonite clay. In this way, one could reach the string of canisters and fish them out one by one, assuming they remained intact. The only major differences from conventional drilling would be the requirement to follow the pilot hole and the details of access to the canister. If the operation were to be conducted at a time when the canisters had ruptured or dissolved, a more com- plex approach requiring greater environment, safety, and health precautions would be needed, but the material would remain retrievable, at somewhat greater cost. Clearly, however, it would not be possible for anyone to retrieve the pluto- nium without the permission of the host country, as long as political control in the host country remained intact. Moreover, because such drilling activities would be highly visible, the host country could not retrieve the plutonium with- out detection. Of course, what powers if any will control a particular bore- hole site after centuries or millennia have passed cannot be known. To make retrieval more difficult, one might make the hole harder to redrill by embedding extremely hard material in the mud and concrete with which the hole is filled. One might make it more difficult to find the precise location of the hole by choosing a site in which the hard rock began at a depth of hundreds of meters or more from the surface, and by filling the zone above the sealed hole in the rock, and the region between there and the surface, with rubble. Still, if the location of the hole were known, it could eventually be found. If the goal of retrieval were only to acquire a few tons of plutonium, and reactors and reprocessing facilities were available, it might turn out to be easier to make new plutonium or to separate reactor-grade plutonium from spent fuel; ~ Svensk Karnbranslehantering AB, op. cit.
254 APPENDIX C but since the hole would only have to be redrilled once, retrieval from the deep borehole would probably be a cheaper route by which to acquire a large quan- tity of plutonium. Policy Issues While disposal in very deep holes appears technically feasible, and appears to offer the potential for superior isolation of plutonium from the biosphere, it has received far less critical study than has disposal of spent fuel and HEW in mined repositories. Thus, a substantial additional research and development effort would have to be focused on the deep-hole approach if this were to be a leading contender for plutonium disposal. The Swedish study suggests a future work plan that includes: · continued review of data from past deep boreholes in crystalline rock; · drilling-related research; · research on plugging and sealing; · modeling of water convection in and around the hole; · pilot studies to determine the depth to saline water, using electromagnetic methods; and · drilling of a 3-kilometer borehole, to test the geological assumptions. The deep-borehole option is not yet ready for "development." In the ab- sence of a crash program, it would take more than a decade to formulate a plan, carry out research on drilling and emplacement, and use existing holes to evaluate the effectiveness of sealing techniques. A critical issue, at least in the United States, would be the likely difficulty of gaining the needed licenses and approvals for a deep-borehole disposal ap- proach. As noted in Chapter 1, decades of effort and billions of dollars of ex- penditure have been devoted to developing the mined repository approach in the United States, and it is not expected that such a repository will be approved and opened for at least another two decades. In the case of the borehole ap- proach, the relevant data would in some respects be more difficult to acquire. In the course of drilling the hole itself (and the smaller-diameter pilot hole), a great deal of data on the properties of the rock being drilled through and the geology of the site could be acquired. But to assess the homogeneity of the site would probably require drilling a number of additional holes to comparable depths nearby. (Means would have to be provided to ensure that these addi- tional holes themselves did not provide a potential means of transport of radi- onuclides to the surface.) Even so, the degree of detail available on the geology of the area at 4,000-meter depth is unlikely to match that attainable for a mined repository at 500 meters. Finally, developing a technical licensing approach that did not rely on the monitoring and retrievability possible with a mined re- pository concept would be difficult and time-consuming.
APPENDIX C 255 These difficulties might be somewhat less in the different regulatory envi- ronment in Russia, but that cannot be predicted. Deep-borehole disposal might appeal to some Russian officials because it would permit eventual retrieval of the plutonium when its use as reactor fuel became cost-effective, while barring theft in the intenm. SUB-SEABED DISPOSAL Description As described in Chapter 6, disposal in the sub-seabed has long been the leading alternative to mined geologic repositories for disposal of spent fuel and high-level wastes. A detailed 1988 study by the Nuclear Energy Agency (NEA) of the Organization for Economic Cooperation and Development (OECD) con- cluded that "sub-seabed burial appears to be a technically feasible method of disposal of high-level radioactive wastes or spent fuel."9 The sub-seabed option faces major problems of public and international acceptability as well as major legal restnctions, however. Moreover, a substantial period of further develop- ment would be required before it could be implemented. The U.S. program was canceled in 1986, and there is now no country in the world actively pursuing research and development on sub-seabed disposal. The idea of sub-seabed disposal is to put the material in metallic canisters that would be placed in the "abyssal clay formation" several kilometers beneath the ocean surface. The canisters would be placed perhaps 30 meters below the surface of this deep ocean mud, which core samples demonstrate has been un- disturbed in some areas for millions of years. This could be done by the use of free-falling "penetrators" dropped from ships, which would fall through the ocean and embed themselves in the mud; by a long drill stem from a ship; or by towering an emplacement package by cable from a ship (see Figure 6-4~. An alternate concept would be to drill through these sediments into the bedrock below and place the canisters in holes drilled there. This in essence combines the deep-borehole and sub-seabed concepts. This approach would be 9 See Nuclear Energy Agency, Organization for Economic Cooperation and Development, Feasibility of Disposal of High-Level Radioactive Waste in the Seabed (1988), 8 volumes. The previous NEA/OECD study' Seabed Disposal of High-Level Radioactive Waste (1984) is also helpful. See also C.D. Hollister et al., "Subseabed Disposal of Nuclear Wastes," Science, 213, September 1981, pp. 1321-1326; U.S. Congress, Office of Technology Assessment, Subseabed Disposal of High-Level Radioactive Waste (Washington, D.C.: Government Printing Office, May 1986); and The Subseabed Disposal Project: Briefing Book 1985, JK Associates, 1985. This sub-seabed option in mid-ocean areas should not be confused with the idea that wastes should be placed in the "subduction zone," where one tectonic plate is slipping beneath another and the wastes would therefore be carried deep beneath the earth's crust. The problem with this approach is that even "fast" seafloor motions proceed at a rate of the order of 1 centimeter per year, meaning that in all of historic time (some 5,000 years) the material would only have moved 50 meters. Furthermore, the subduction zones are geologically active and unpredictable prone to volcanoes, among other phenomena. For these reasons this report does not consider the subduction-zone option any further.
256 APPENDIX C substantially more costly, but might make the plutonium nearly impossible to find and recover once emplaced, which would address the retrievability prob- lem discussed below. Because of the higher cost, however, and because prevent- ing retrieval is not as serious an issue in the case of HEW or spent fuel, this concept of drilling through the abyssal mud to the underlying bedrock has not previously been examined in detail. The remainder of this discussion focuses on the more conventional approach of emplacing the canisters in the mud. If, for example, 10 kilograms of plutonium were placed in each canister, the nominal 100-ton excess weapons plutonium stockpile (counting both U.S. and Russian plutonium) would require 10,000 canisters. If each of these weighed one ton, this would involve extremely careful sea transport and em- placement of about 10,000 tons of materials. Strict safeguards would need to be enforced over the operation of the ship in view of the high price that might be paid to steal even a small fraction of the cargo. Environmental Impact As with deep boreholes, the first question to consider is reliable isolation of plutonium from the environment. The mud of the abyssal plains has several advantages. It is located at depths of more than 4 kilometers, far from the edges of tectonic plates or other regions of geologic activity, and far from shorelines or other areas of human activity; as mentioned, it has been stable in some areas for millions of years. In what follows, one should note that such abyssal plains exist within the 200-mile exclusive economic zone (EEZ) of U.S. shores, a re- gion that has different legal status from the broad ocean areas. The barrier to release of the plutonium provided by the canister itself would be short-lived compared to the plutonium.~° Hence the long-term barrier to dispersal is the mud itself. The first issue is whether the hole created by the emplacement of the canis- ter would quickly recluse. If not, this hole would provide a potential route for radionuclides to reach the surface. Modeling suggests that reclosure would oc- cur very rapidly (within a fraction of a second) as a result of the large negative dynamic pressure (suction) associated with the high impact speed of the pene- trator, but field experiments would be required to assess this prediction. This is a "make or break" question, and thus should be tested before any significant expenditures on other aspects of the option. If the canisters were emplaced appropriately with the mud reclosed above them, the issue becomes: How would the plutonium move through the mud? The extremely fine particles of this abyssal mud greatly retard motion of water through them. Convective transport of plutonium in these sediments can be a The best cladding material for the canisters yet proposed, an alloy designated as Ticode 12, has a corrosion rate of 6 millimeters per 1,000 years, compared to plutonium's half-life of 24,000 years. See Hollister et al., op. cit.
APPENDIX C 257 treated in a fashion similar to that sketched for the deep-borehole option, by considering first the various driving forces for convection (density differences arising, for example, from heating or gas evolution, but also the very slow flow of pore water through the material) and then the transport of dissolved ions, taking into account the sorption of plutonium on sediment particles. Although it appears that convective transport of weapons plutonium could be reduced to extremely low levels by correct emplacement, convection is still likely to domi- nate diffusive transport over very long times since transport distances increase linearly with time for convection, and only as the square root of time for diffusion. As in the case of the granite and bentonite described above, plutonium would be expected to bind to the mud particles, further delaying any possible movement. First, consider the movement without such adsorption. The rate at which a material will move by diffusion if it is not bound to particles in the medium depends on the diffusion coefficient D of the material in the medium. The mean square displacement of the material, <x2>, is proportional to D and the time elapsed: <x2> = Dt. The diffusion coefficient of dissolved plutonium may be estimated as some 2 x 10-6 cm2/sec some ten times lower than that of water itself. This would correspond to a root mean square displacement of about 25 meters (almost enough to reach the surface) in 100,000 years, at which time (4 half-lives of radioactive decay) roughly one-sixteenth of the original plutonium would remain. This is the expected diffusion in still water- that is, water immobilized in fine sand that does not sorb the plutonium. Although some radionuclides (such as technetium) do not sorb on the mud particles and therefore would move through the mud at the rate described above, plutonium tends to bind strongly to the mud particles and therefore its motion through the mud would be drastically slowed. As noted in the discus- sion of deep boreholes, such sorption is often described by using a parameter known as the "distribution coefficient," K~.~2 The rate of diffusion through Bleach mud particle, however, has a limited number of surface sites capable of sorbing a particular class of charged atom, or ion, such as plutonium. Thus, the volume of mud immediately around the canister may become saturated with plutonium, at which point the remaining plutonium not yet bound to the mud will diffuse through it at the rates described above, as though no sorption were taking place. The mud's capability to bind plutonium might be saturated at a concentration of something like 1 kilogram of plutonium per cubic meter of mud, even if the plutonium concentration in the pore water is well below the solubility limit. At that concentration, the 10 kilograms of plutonium that might be held in a single canister would be sorbed in 10 cubic meters of mud' so the saturation zone would extend less than 1.5 meters from the canister itself. Hence this discussion focuses only on the simpler unsaturated problem. i2 The units of Ky are milliliters per gram of sorbant each gram of sorbant adds as much "hidden volume" as would Ky milliliters of actual solvent. For y grams of sorbant per milliliter of gross volume, and for y x Ky >> 1, the mass of plutonium per milliliter of gross volume would be larger than that in the solution without sorbant by a factor y x Kit; this applies to the linear range, for which Ky is nearly constant. Regarding the Ky concept, the 1984 NEA/OECD report (op. cit.) cautions that the various nuclides will be changing chemical form, making colloids, etc., all of which may behave differently.
258 APPENDIX C mud with sorption taking place can be modeled as if it were diffusion through still water with a different diffusion coefficient, D', defined as (D x c)/~ x Kd).~3 In other words, the material will diffuse (1 + yK4/~) times more slowly than it would in the absence of sorbant. If Kit = 1,000 ml/g, y = 2 g/ml, and ~ = 0.20, the 100,000 years it would take for plutonium that was not sorbed to diffuse 25 meters would be increased to 1 billion years; if Kit were as large as 10,000 ml/g, the diffusion time would be 10 billion years- greater than the age of the earth. In general, a strongly sorbed material with a high distribution coefficient, like plutonium, would take yKy/e times longer to penetrate significantly through a diffusion barrier than it would if it were not sorbed.~4 The sorption sites in the mud might also be saturated by some material other than plutonium that binds to the particles at the same sites, reducing the effective Kit for plutonium and hence the delay in its transport. Studies done to date suggest that the chemical character of the sites does not greatly affect the adsorption of plutonium, but one would want to make the canisters from a ma- terial that would not compete with plutonium for the same adsorption sites. A variety of mechanisms that might speed the transport of the plutonium to the surface have been examined. In the case of spent fuel or HEW, an important However, the NEA and the JK Associates reports (op. cit.) both indicate that plutonium in particular will definitely be bound by the mud and is well studied. i3 The formula for flux in an unsaturated region where there is linear behavior is F = -D dCldx, where C is the concentration of the diffusant (plutonium, in this case) in the pore water and D is the diffusion coefficient in the pore water. (At this point the analysis will ignore a relatively unimportant correction for "tortuosity.") If the volume fraction available for pore water is £, in a steady-state gradient the mass flow due to diffusion is multiplied by the same factor £; however, the amount of diffusant species present per unit volume of mud is C x (£ + y x Kit). The equation for the time rate of change of diffusant concentration in the pore water is then y x K4(dC/dt) = -£ x dF/dx. This recognizes that a unit volume of sediment at diffusant concentration C in the pore water contains about y x Kit x C mass of diffusant (insignificantly different from (£ + y x Kit) x C for a heavily sorbed species). Combining the equations, we have the relation between time and space behavior of the concentration of diffusant in the pore water dCldt = D'd2C/dx2, in which D' is defined as £D/(£ + yogi). The 1984 OECD/NEA study (op. cit., p. 171) indicates a figure for the plutonium diffusion coefficient in "ocean sediments" of 1 x 10-~°, and about 30 percent larger for "bentonite backfill." This is entirely compatible with the figures used here, considering both the diffusion coefficient in water and the distribution coefficient between the pore water and the sediment. On the other hand, the 1988 NEA summary (op. cit., Vol. 1, p. 19) has a significantly lower diffusion time in the absence of sorption than calculated above, indicating that radionuclides that were not sorbed in the mud would reach the surface "on the order of 104 years after their release from waste packages emplaced at a depth of 50 +/- 20 meters in the sediment because of the time required for diffusion through the sedimentary pore water." The calculation in the text would have reached a similar conclusion had it used a diffusion coefficient equal to that of water, rather than the lower diffusion coefficient of plutonium. ]4 The flux across the barrier once the plutonium has begun to penetrate it significantly, however, will not be reduced by sorption. Under the same assumptions, the steady-state flux across a barrier for instance, between water saturated with plutonium at concentration CO on one side of a sediment layer of width W. and the sea on the other side is simply F = £DCo /W, unaffected by the sorption on the sediment.
APPENDIX C 259 issue is the possible effect of the high temperature from the decay heat of the waste. This would make the surrounding pore water less dense, and hence buoyant. Studies to date suggest that "the thermally-induced movement of the pore waters, the waste package, and the sediments is small,"~5 but a major field test to assess this conclusion, the In-Situ Heat Experiment, was canceled in the late 1980s. In the case of excess weapons plutonium, this is not an issue be- cause the heat output from the weapons plutonium would be only about 30 watts per canister (assuming 10 kilograms of plutonium could be placed in each canister), compared to some 1.5 kilowatts considered for the HEW case. Hence the weapons plutonium would raise the temperature of the mud at 1 meter by only 2-3°C, and the induced pore water flow would be less by the ratio of the heat produced. One issue that had not been fully addressed when the seabed studies were terminated is the existence of faults in some of the sediments examined, above and below the planned emplacement depths, which could create avenues for more rapid transport to the surface. For plutonium disposal, where only a small area is needed, one could simply avoid the known areas of faulting.~7 While one could imagine that living organisms might burrow through the mud and bring plutonium to the surface, tests indicate that such organisms exist only in approximately the top meter of sediment though small organisms such as bac- teria might well be brought to lower depths by the act of emplacing the canis ters itself. Even if the plutonium did reach the surface of the mud before it had nearly all decayed, it would be released in very deep water that does not mix rapidly with water at higher levels, and where concentrations of marine organisms are lower than they are near the surface-though seabed photography of these re- gions does show numerous benthic animals such as sea cucumbers. Preliminary modeling suggests that any doses to humans would be delayed many thousands of years and would be small compared to normal background radiation. (The dose to the most exposed population if much of the plutonium escaped into the sea is estimated below, in discussing the ocean dilution option.) Another issue is possible criticality of the plutonium. A sustained chain re- action would greatly increase the heat in the emplacement, potentially creating a "bubble" of plutonium and fission products that would rise through the mud more rapidly than described above. To limit this risk, some processing of the plutonium would probably be necessary before emplacement. The canister could be filled with solid material, such as a plutonium alloy, or a cast matrix con- taining 3 atoms of normal boron per atom of Pu-239 (which is enough to keep Is NEA/OECD, 1988, op. cit., Vol. I, p. 42. is If, however, the canisters were given the ability to pump a plutonium-bearing solution into the mud, as described below, this solution might be designed to have a different density than the surrounding water in the mud, causing it to either rise or sink very slowly. ]7 NEMOECD, 1988, op. cit., Vol. 1, p. 20.
260 APPENDIX C the system subcntical). Boron, however, might leach away preferentially after the canister corroded, leaving the plutonium behind in a package moderated by seawater, which could lead to criticality; gadolinium, with its lower solubility, might prove to be a better choice. In addition to adding neutron-absorbing ma- tenals, the amount of plutonium placed in each canister might be limited so as to mitigate such criticality concerns. The design would have to prevent chain reactions not only for single canisters successfully emplaced, but in accident scenarios as well, in which multiple adjacent canisters might be immersed in water as a result of accidental flooding of the ship's hold, for instance. That, however, would be an easier problem than handling the less well-specified configuration that might evolve after thousands of years under the seabed. For both the deep-borehole and sub-seabed options, the question of long-term cnti- cality after differential leaching needs to be addressed because the issue is quite different from those addressed in the studies of these options for the disposition of HLW. As noted above, because of the extremely slow predicted movement of ra- dionuclides through the mud, some studies have concluded that plausible hu- man doses of radioactivity from spent fuel or HLW correctly emplaced in the sub-seabed would be many thousands of times lower than doses from natural background radiation. Doses to maximally exposed individuals from a worst- case accident, however, such as a ship bearing spent fuel or HLW that sank near shore with none of the canisters recovered, could be several times normal background. The overall risk of a transportation accident would generally be greater for options like sub-seabed disposal that add ocean transport to the land transport required for all the options. The potential impact of an accident dur- ing shipment or emplacement requires further study. Not surprisingly, the environment, safety, and health (ES&H) aspects of sub-seabed disposal are dependent on the details of the process. If this approach were to be pursued, either for spent fuel and HLW or for weapons plutonium, a phased and interactive program of analysis and testing would be appropriate, focusing first on the potential "showstoppers," with the goal of eliminating the option as quickly and cheaply as possible, or of discovering approaches that survive analysis and field test. ~8 See NEA/OECD, op. cit., 1984, 1988; M.F. Kaplan and R.D. Klett, "Biological and Physical Oceanographic Sensitivity Analysis for Subseabed Disposal of High-Level Waste," SAND83-7107, Sandia National Laboratory, November 1984; and M.F. Kaplan, R.D. Klett, C.M. Koplik, and D.A. Ensminger, "Radiological Protection Options for Subseabed High-Level Waste Disposal," SAND84- 0548, Sandia National Laboratory, March 1985, cited in U.S. Congress, Office of Technology Assessment, op. cit. In this report's analysis of the ocean dilution option for disposal of weapons plutonium, the dose to the most highly exposed subpopulation identified in International Atomic Energy Agency (IAEA) studies is estimated in detail, for 100 metric tons of weapons plutonium injected into mid-ocean, without the barrier provided by the sub-seabed mud.
APPENDIX C 261 Cost The committee has not carried out an analysis of costs for disposal of plu- tonium in the sub-seabed. Past analyses for the case of spent fuel and HEW, although preliminary, suggested that the costs of sub-seabed disposal would be competitive with those of mined geologic disposal. But in the case discussed here, the cost of the repository must be considered "sunk," and only the net additional cost of adding this plutonium to it would have to be considered, whereas the entire cost of developing, demonstrating, gaining approval for, and implementing the sub-seabed option would have to be counted against the plu- tonium disposition mission. It appears likely that the cost of the actual emplacement would be in the range of a few hundred million dollars. The de- velopment and demonstration program necessary to meet licensing requirements might well cost billions of dollars (as in the case of land repositories). Retrievability As with deep boreholes, retrievability would be desirable from the point of view of ensuring that the disposal concept was working as expected, but unde- sirable if the goal was to create substantial barriers to recovery and use of this material in weapons. The state emplacing the plutonium-bearing canisters could, and presuma- bly would, maintain a detailed record of where the canisters were emplaced.~9 Given the distance from the ocean surface to the mud, retrieval could be han- dled by a major ocean-drilling ship; alternatively, it would be possible to develop a deep-water derrick that could be lowered from an ordinary ship, which could be positioned and repositioned on the seabed to retrieve one pene- trator at a time. The deep-ocean drilling ships needed for the first approach have an operating cost of a few hundred thousand dollars a day, and less than a day would probably be required to retrieve each canister, corresponding to a recovery cost of perhaps $10,000 to $50,000 per kilogram of plutonium (not counting the cost of locating it in the first place). The seabed derrick option would probably be cheaper, after development costs were paid. Similarly, other states with sophisticated deep-ocean technology that were aware of the approximate location of the disposal site would be able to find the canisters and recover them. The mud is homogeneous, with the consistency of "melted Godiva chocolate,"20 and is quite transparent to sonar. Indeed, it is conventionally surveyed acoustically by sonar of various types, including very high resolution, towed near-bottom vehicles, and such sonar could locate the ~9 Indeed, in some concepts the penetrators would be equipped with transponders for this purpose, though the life of these transponders would be extremely short compared to that of the plutonium. 20 C.D. Hollister, Woods Hole Oceanographic Institution, personal communication, November 3, 1992.
262 APPENDIX C canisters quite precisely years or centuries after emplacement. The canisters, if intact, could then be retrieved as just described. For less-developed countries and subnational groups, recovering plutonium by such means might be more costly and time-consuming than recovering plutonium from spent fuel, thus meeting the spent fuel standard. To limit such recovery possibilities, the area in which the plutonium-bearing canisters had been deposited could be monitored for an indefinite time, if that were agreed on. So long as the site was monitored, essentially nothing could move along the bottom without being detected. In that case, it is necessary to assess the cost of preventing retrieval, who will pay that cost, what kind of forces would have to be involved, and what country or group of countries would retain the responsibility. It is by no means assured that 1,000 years from now, for example, the United States will continue to exist or con- tinue to maintain an interest in safeguarding plutonium disposed of in this way. Alternatively, to avoid this requirement for surveillance and possible inter- vention, one might consider options to reduce the observability of the canisters (such as through "stealth" canisters designed to be difficult to detect by sonar) or to eliminate the canisters sooner rather than later. Such concepts have not been examined in the case of HEW disposal because preventing retrieval is not an issue for disposal of HEW, and because the canister provides one of the sig- nificant barriers to release of the radioactivity (since the lifetime of the canister is long compared to the period required for most of the radioactivity to decay). In the case of plutonium, preventing retrieval is an issue, and no material has been proposed that one can say with confidence will survive in ocean mud for the tens of thousands of years required for most of the plutonium to decay- which means that in limiting the total long-term hazard (as opposed to the early hazard), the canister is essentially irrelevant. A variety of mechanisms for eliminating the detectable canisters can be envisioned. These range from disin- tegrating the canister soon after emplacement by chemical dissolution or small explosive to doing without the canister in the first place, by injecting a plutonium-bearing solution into the mud with a small drill, to a depth compa- rable to that at which the canister would have been emplaced (with care taken to ensure that the hole created by drilling was appropriately reclosed).2~ A country with precise knowledge of the location of the disposal sites would still be able to "mine" the mud to recover the plutonium, but it seems unlikely that this would be competitive with recovering plutonium from spent fuel. None of these methods has been studied in any detail, but none seems to present funda- mental technical obstacles. Close study of such methods and their ES&H impli- cations would be essential to any further consideration of the sub-seabed option for plutonium disposition. 2] In concepts involving a plutonium solution injected into the mud, designers could in principle ensure that the plutonium-bearing "water" were somewhat denser than the water in the mud, meaning that it would gradually sink deeper into the mud with the passage of time, rather than rising toward the surface.
APPENDIX C 263 Another solution to the retrievability problem, as noted above, would be to emplace the canisters in bedrock below the abyssal sediments, rather than in the sediments themselves. The canisters would then be in equally solid mate- rial, and would not be visible to sonar. Once a hundred meters or more of mud had reclosed over the hole in the rock, the hole would be extremely difficult if not impossible to find. Policy Issues As with the deep-borehole option, sub-seabed disposal would require sig- nificant additional development if it were to be a contender for plutonium dis- posal. At the time the U.S. program was canceled in the mid-1980s, it was estimated that another 25 years would be required before a sub-seabed disposal approach could be operational.22 Development and emplacement costs are likely to be lower for the sub-seabed approach than for deep boreholes, but li- censing and policy questions are much more problematic, whether a site in the broad ocean is contemplated or one in the 200-nautical-mile exclusive eco- nomic zone. Sub-seabed disposal would face intense political opposition from many quarters, as well as a complex web of national and international legal hurdles and regulation. The U.S. Marine Protection, Research, and Sanctuaries Act, also known as the Ocean Dumping Act, forbids all dumping of high-level ra- dioactive waste at sea and has been interpreted as including sub-seabed dis- posal.23 Thus this legislation would have to be amended before the United States could implement the sub-seabed approach. The London Dumping Convention of 197224 incorporates a similar prohi- bition on "ocean dumping" of high-level radioactive waste. The convention was negotiated before the sub-seabed disposal concept began to be widely discussed in the scientific community and thus does not mention the idea specifically.25 The parties to the convention have never agreed on whether the agreement prohibits emplacement of wastes beneath the ocean floor, but a majority of the parties have expressed that view in the past. The parties have agreed that if the technical feasibility of the concept is demonstrated and one or more countries wished to pursue such a disposal approach, the convention would be the appro- priate forum in which to consider the matter. 22 JK Associates, op. cit. 23 ibid. 24 Formally known as the Convention on the Prevention of Ocean Pollution by Dumping of Wastes and Other Matter. 25 However, He L\EA, in a first draft of definitions for implementing He treaty prepared in 1973, mentioned the future possibility of burying wastes beneath the seafloor, and suggested that this should be considered as a variant of land burial, and therefore excluded from the convention. See NEA/OECD, 1984, op. cit.
264 APPENDIX C On November 12,1993, the United States and 36 other nations voted to ex- tend the convention to ban dumping of low-level radioactive waste as well, to impose a "permanent, legally binding ban on the dumping of all types of radio- active waste at sea." Belgium, Britain, and Russia abstained, and 29 of the 71 signatories to the 1972 London Dumping Convention were absent from the meeting. Proposals for an explicit prohibition on sub-seabed disposal are re- portedly slated to be discussed in 1994 or 1995. At a minimum, it is clear that any proposal for sub-seabed disposal would require the acquiescence of a majority of the parties to the convention and that such approval might be very difficult to attain. Further, the Law of the Sea Treaty, if it enters into force, would create an international authority that would regulate activities on the seabed, which would presumably assert authority over sub-seabed disposal. In addition to this legal framework, any proposal for disposal in or below the oceans is likely to provoke intense public and political opposition, both within the United States and internationally. In short, gaining approval from a majority of the parties to the London Dumping Convention for sub-seabed disposal of plutonium, and overcoming the political, legal, and regulatory hurdles (including providing experimental data that do not yet exist) would be difficult, uncertain of success, time-consum- ing, and expensive. Furthermore, it appears that the sub-seabed option for disposal of weapons plutonium is different in fundamental ways from the sub- seabed option for disposal of high-level waste. DILUTION OF PLUTONIUM IN THE OPEN OCEAN Description Another possibility is to dilute the plutonium in the open ocean, striving to achieve maximum mixing with ocean water so as to reduce the concentration of plutonium in any one location. By this method, one could relatively cheaply and quickly make the plutonium so dilute as to be completely irrecoverable. If only the nominal 100 tons of excess weapons plutonium were disposed of in this way (rather than the much larger global stock of reactor plutonium, dis- cussed below) and if it were possible to dilute it uniformly in the entire ocean which has a volume of roughly 1.4 x 109 cubic kilometers the resulting con- centration would be 7 x 10-~4 grams per liter. In an average liter of seawater, one atom of plutonium would decay to uranium every three hours (10-4 disinte- grations per second). The natural concentration of uranium-238 (U-238) in seawater, 3 x 10~ grams per liter, creates a background disintegration rate 300 times greater than 100 tons of plutonium would contribute. This calculation is misleading, however, because it is not remotely possible to mix the plutonium evenly throughout the ocean, biological processes will reconcentrate the plutonium, and one cannot compare the toxicity of plutonium
APPENDIX C 265 and uranium by looking only at the disintegration rate. The great difficulty of adequately ensuring that there would not be dangerous reconcentration of plu- tonium, the high probability of intense public and international opposition, and legal hurdles even higher than those facing the sub-seabed option probably make this option infeasible. Environmental Impact Existing regulations, national and international, provide a framework in which to consider the environmental impact of diluting 100 tons of excess weapons plutonium in the ocean. In the United States, new regulations enter into force on January 1, 1994, limiting the allowable plutonium concentration in water to which members of the public might be exposed to 2 x 10-8 curies per cubic meter.26 These limits were set with the goal of ensuring that a person drinking 2 liters a day of this water would receive a radiation dose of no more than 50 millirem per year. To remain below this new limit, each gram of weapons plutonium would have to be diluted in 4 million cubic meters of ocean water. If one set a criterion that the concentration of plutonium even when first released from the dilution ships should not exceed this legal limit, then the ships would have to dilute the nominal 100 tons of plutonium into 400,000 cu- bic kilometers of water. To do this, one might use ships or submarines with long drag lines containing thousands of nozzles, dispersing plutonium solution at depths of several kilometers beneath the surface of the ocean. A single ship or submarine proceeding at a rate of 10 kilometers per hour would cover 240 kilometers per day. If it towed a drag line equipped with a sheetlike array of perhaps 3,000 nozzles, spanning a depth range of 2 kilometers (at a mean depth of perhaps 4 kilometers), with the nozzles extending some 250 meters to each side of the line, the cross-sectional area into which the plutonium would be dispersed would be some 1,000,000 square meters, or 1 square kilometer.27 26 These are new Nuclear Regulatory Commission limits (see Inside NRC, May 4, 1992, p. 3). For plutonium, these limits are 250 times more stringent than previous standards, which can be found in the Code of Federal Regulations, Vol. 10 (Energy), Part 20 (Standards for Protection Against Radiation), rev. 1 (Washington, D.C.: U. S. Government Printing Office, January 1992), pp. 321-457. The previous limits were based on an allowable dose 10 times higher, and a belief that plutonium ingestion was less effective in generating radioactive doses than new studies suggest. 27 While the nozzles would be spaced a meter or two apart, the concentration in the column of water to which they were adding plutonium would quickly become uniform. If the eddy diffusion constant De in the deep ocean is roughly a typical 1 cm2/s, then without any induced turbulence, even with nozzles spaced at intervals L of 18 meters, the affected column would be rendered uniform in plutonium concentration in about (Lid ~I)2/De= 1 day. (See S.M. Flatte, ea., Sound Propagation Through a Fluctuating Ocean (Cambridge, U.K.: Cambridge University Press, 1979), p. 7. Even in unusually still ocean depths, it would seem to be a simple matter to arrange the towline structure to provide enough small-scale turbulence to mix the injected fluid very quickly within the 1-square- kilometer trail. To leave random motions of 1 cm/s over this trail from a ship moving at 10 km/hr corresponds to an increased towline power of 200 kilowatts.
266 APPENDIX C Thus, such a ship could disperse plutonium into 240 cubic kilometers each day, and some 4.6 ship-years would be needed to reach the required 400,000 cubic kilometers into which to disperse 100 tons of plutonium. International regulations would also apply to this case. As noted above, the London Dumping Convention forbids all disposal of high-level radioactive waste into the sea, and a majority of the parties to the convention have recently voted to bar low-level dumping as well. Any proposal to dispose of tens of tons of weapons plutonium in the ocean would surely be seen as directly contrary to the intent of the convention, even if the material could be diluted enough that it would meet current standards for disposal of low-level waste. Nevertheless, a discussion of what would be required to meet those standards is of interest. Those standards, set by the IAEA, are slightly less stringent in their objec- tives (aiming to limit doses to the most exposed individuals to 100 rather than 50 millirem), but they include two factors not considered in the U.S. drinking water regulations. First, the IAEA limits are based on the assumption that the radiation releases in question should not be approved unless it would be accept- able for them to continue for 1,000 years. For the case of a one-time disposal of weapons plutonium, the assumption of continuing releases would not be correct (though it would be if the continuing global production of reactor plutonium were added, as discussed below); but this continuing release assumption never- theless forms the basis for existing regulations. (The underlying principle of this approach is that our generation should not claim a greater right to or need for the capacity of the ocean to absorb radiation than later generations will have available to them.) Second, the IAEA regulations take into account the fact that some forms of sea life concentrate plutonium in their edible tissues. Molluscs accumulate plutonium in their edible tissues at concentrations 3,000 times higher than those in the surrounding water; edible seaweed, 2,000 times; crus- taceans, 300 times; and fish, 40 times.28 The IAEA estimates that some coastal populations consume 600 grams of seafood per person per day, consisting of 300 grams of fish and 100 grams each of crustaceans, molluscs, and seaweed. Applying the IAEA approach (continuing releases assumed, bioconcentra- tion in species consumed by humans talcen into account) to the U.S. regulatory standard would lead to much more stringent concentration limits. Consider first the bioconcentration issue. The U.S. drinking water standard is based on an exposed individual consuming 2,000 grams of water a day; if, instead, that in- dividual consumes seafood at the rate and with the concentrations estimated by the IAEA, the concentration must be reduced by a factor of 270 to maintain the same radiation dose. Thus, to limit doses to exposed populations to 50 millirem per year (as the U.S. law requires), the volume of ocean into which 100 tons of weapons plutonium would have to be diluted would not be 400,000 cubic 28 Intemational Atomic Energy Agency, Definition and Recommendations for the Convention on the Prevention of Marine Pollution by Dumping of Wastes and Other Matter, Safety Series No. 78, 1986 Edition (Vienna: IAEA, 1986).
APPENDIX C 267 kilometers, as estimated above, but more than 100 million cubic kilometers. This is more than three times the volume of the mixed surface layer of the oceans, and nearly 8 percent of the volume of the entire ocean. It is extremely unlikely that the plutonium could be successfully mixed into such a large volume, with no local "hot spots" where the concentration would be significantly higher, at any reasonable level of effort. For example, with ships such as those envisioned above, more than 1,000 ship-years would be required, assuming that one could somehow guarantee that one ship was not adding plutonium to the same volumes of water that other ships had. Moreover, this figure assumes that the weapons plutonium is allowed to consume the en- tire legal limits for radioactivity of this volume of water, with no other sources of radioactivity permitted for as long as that plutonium, with its half-life of 24,000 years, remains. If one adds to this the problem of limiting the equilib- rium concentration, on the IAEA's assumption of releases that continue for 1,000 years, the problem would grow even more difficult. All this assumes that the dilution would proceed without incident. If, for example, one of the ships collided with another ship and the plutonium was released, or one of the lines used collided with something else, resulting in greater plutonium releases, a local plutonium "hot spot" would be created. All of the above calculations assume that the weapons plutonium alone would be allowed to pose the maximum allowable risk. Clearly this would make it impossible to dispose of other radioactive elements in the oceans without ex- ceeding the regulations on maximum permissible doses. Moreover, in Chapter 6, the committee argues that if options for eliminating weapons plutonium nearly completely from international human access involve substantial addi- tional risks, costs, or delays compared to options that make it as inaccessible as plutonium in spent fuel as the ocean dilution option would-these additional problems should not be borne unless global stocks of civil plutonium are to be treated in a similar way. The excess global stock of civil plutonium is drasti- cally larger than the excess weapons plutonium stock (some 800 tons compared to 100 tons), more toxic (roughly seven-fold, as a result of the presence of more of the more radioactive isotopes), and growing (by approximately 70 tons per year). To keep the dose to an exposed population consuming 600 grams a day of seafood below the legal limit would require diluting the reactor plutonium that already exists in a volume more than three times as large as the entire vol ume of the oceans. Cost The implementation cost of diluting 100 tons of weapons plutonium in the oceans in this way would be minimal, if the requirement is simply to meet the 1994 U.S. standards for fresh drinking water (without considering bioconcen- tration or continuing releases). The cost would amount to several tens of mil- lions of dollars for ship modifications and for operations for only a few ship
268 APPENDIX C years. But meeting the new standards while taking into account the bioconcentration that would increase doses by a further factor of 270 would require larger nozzle arrays and additional ship-years, thus raising costs. These implementation costs would in any case be greatly exceeded by the likely costs of the studies and licensing efforts that would ultimately be involved if this ap- proach were to be seriously pursued. Retrievability In this case, the plutonium would be so dilute that there is no possibility of intentional retrieval. This is one of the few absolute statements possible in the complex subject of plutonium disposition. Policy Issues As noted above, public and international opposition to any such proposal- based on the environmental objections already outlined would predictably be so overwhelming as to effectively rule out ocean dilution as a viable option. Moreover, even more than the sub-seabed concept, ocean dilution would face a number of national and international legal hurdles. The U.S. ocean dumping law would presumably be interpreted as forbidding this option, as with the seabed option. As noted, the majority of the parties to the London Dumping Convention, including the United States, have voted to bar dumping of low-level radioactive waste as well as high-level waste. Such a ban would make all the preceding calculations concerning IAEA regulations on low-level waste disposal academic, unambiguously forbidding an ocean dilution ap- proach. SPACE DISPOSAL Description Disposal by launching into deep space has been studied extensively for dis- posal of high-level waste and could also be considered for plutonium. In this option, a number of rockets would be used to launch the plutonium onto a path unlikely to encounter the earth. The plutonium would have to be placed in packages designed to limit any possible releases in the event of a ma- jor rocket failure (such as explosion on the pad or during ascent or reentry from a failed orbit). The safety risks, cost, delay, likely intense public and interna- tional opposition, and other aspects of this approach do not seem to put it high on the list of options. Even if space disposal were economically competitive, the design, demonstration, and operation of the systems required for high-reliabil- ity launch and high-confidence handling of inevitable accidents are daunting.
APPENDIX C 269 Environmental Impact The main risks in this case would result from potential launch accidents, reentry from failed orbits, and if launch had been successful, possible long-term risks of collision of the payload with meteors in space. For example, a 1980 study examined the risk if a payload carrying HLW in a "cermet" (metallic ceramic) waste form were to reenter the atmosphere from a failed orbit.29 The study predicted that with the package design envisioned, 11 percent of a S-metnc-ton waste package would burn up during reentry. The study estimated that the result, depending on circumstances, would range from "a few cancer deaths to as many as 100 or so" (for HLW rather than plutonium). Space disposal of plutonium would require designs that would reliably pre- vent criticality accidents (which are not a problem in disposal of HLW). On the other hand, however, the minimal gamma radiation from plutonium makes the design and conduct of the missions easier than for HLW. There is little doubt that large (multiton) or small (10-kilogram) payloads of weapons plutonium can be designed that would reliably survive plausible accidents, including launch explosions or fires, reentry into the atmosphere, and high-speed impact on the ground although demonstrating such safety to regulators and the pub- lic would be problematic. Yet, unlike HLW, the plutonium payload, if it re- tumed to earth intact, would be a matter of great concern because it could be used to fabricate nuclear weapons. Thus, the inevitable risk of launch accidents is a fundamental problem for the space disposal approach. Launching the plutonium into low-earth orbit (which requires a velocity of about 8 kilometers per second) would not be sufficient, because material in low- earth orbit falls back to earth on a time scale shorter than the decay time of plu- tonium. Therefore one would have to launch the material into an orbit around the sun unlikely to encounter the earth (which requires at least 11 kilometers per second (km/s)), to a path that will escape the solar system (16.8 km/s), or into the sun itself (more than 18 km/x).30 Because the rocket launch mass re- quired grows exponentially with the required velocity, options requiring high velocity would greatly increase the cost of the project. 29 For an exhaustive analysis of the issues involved, see Analysis of Nuclear Disposal in Space, Vol. I, Executive Summary, and Vol. II, Technical Report, Phase 3, Battelle Columbus Laboratories, March31, 1980. 30 The velocity requirement from the earth's surface for rocket propulsion into the sun is often quoted as 32 km/s sufficient to overcome the earth's potential well (measured by the escape velocity of 11 km/s), while retaining a velocity of 30 km/s to cancel the earth's orbital speed. (The kinetic energy is proportional to the square of the velocities: 32-= 11-+ 302. ) However, it is clear that in principle, 16 km/s would suffice to reach the sun if the rocket almost escaped the solar system and then used a very small delta-V to cancel its tangential velocity so that it then falls into the sun. Specifically, a rocket burn giving 16 km/s near the earth's surface will carry the rocket around the sun to 18.25 astronomical units, at which time a retro-fire of 2.26 km/s will allow the payload to drop radially into the sun. The total delta-V in this example is thus 18.26 km/s, rather than the 32 normally considered not much more than the 16.8 km/s required to escape the solar system.
270 APPENDIX C If the payloads were successfully launched, the main long-term hazard would be dispersal of the plutonium after collision with a meteor-a risk that applies only to the options involving continuing orbit around the sun, rather than escape from the solar system or disposal in the sun itself. The 1980 study estimated that the probability of a collision that would release even 0.2 percent of the payload in small particles would be about 4 parts per billion per year, while complete fragmentation (meaning collision with a larger meteor) would be 100 times less probable.31 Thus, unless the payload were extremely unlucky, the plutonium would have completely decayed to uranium by the time a colli- sion with a meteor might occur. If the payload were in a circular orbit at 0.85 astronomical units (AU) (85 percent as far from the sun as the earth) the study estimated that 0.12 percent of the fine particles resulting from such a collision would fall to earth, taking an average of 100,000 years after the collision to do so. In contrast, if the payload were placed beyond earth's orbit, at 1.19 AU, 6.7 percent of the small particles produced in a collision will intercept the earth, taking an average of only 50,000 years to do so.32 The concept of launching the material directly into the sun would eliminate any risk that the plutonium would re-encounter the earth, but it would require a higher velocity (more than 18 kilometers per second) than any of the other options. Cost The cost of this approach would depend on the mass of the material that had to be launched, the velocity that the material had to reach, and what new systems had to be developed. The mass that needs to be launched would be much more than the mass of the plutonium itself. First, one must consider that the plutonium should be in a form that will reliably be noncritical even if, for example, an accident results in it being immersed in ocean water. Thus, the plutonium must be combined with some neutron-absorbing material. Detailed calculations show that PuO2 x 3B is subcritical in any quantity, density, or configuration (using boron of normal isotopic composition); it is 81 percent plutonium by weight.33 The reentry ve- hicle (RV) needed to ensure integrity in the event of an accidental explosion or reentry from orbit would roughly double the mass of the composite. Thus, the overall RV mass would total 2.5 times the mass of the contained plutonium. 31 Analysis of Nuclear Disposal in Space, op. cit. 32 This major difference arises from the Poynting-Robertson drag on the small fragments, which encounter more solar radiation on their leading face Man on the trailing face, thereby experiencing a relativistic drag that forces them to spiral in toward the sun. 33 J. L. Richter, Los Alamos National Laboratory, personal communication, April 1993.
APPENDIX C 271 The concept usually discussed for space disposal is to begin by launching the material into a circular low-earth orbit (LEO) at roughly 300 kilometers altitude (which requires a rocket velocity gain of some 8 km/s), and then to use an additional burn to move the material onto the desired deep-space path, whether to another orbit around the sun, to escape from the solar system, or to go into the sun. Such staging from LEO is the least-energy and probably the least-cost approach. An appropriately timed additional burn of some 3.4 km/s in LEO would place the payload at a radius of 0.85 AU from the sun six months later, and a further burn of 1.16 km/s will place the payload into a cir- cular orbit of radius 0.85 AU, inclined 1° to the plane of the ecliptic. If one assumes a solid propellant of equivalent specific impulse (Isp) of 200 seconds (corresponding to an exhaust velocity of 2 km/s, instead of detailed computa- tion using a real Isp of 270 seconds, which would then be reduced by reasonable mass fractions for structure, tanks, and engines), the combined velocity gain of these two last burns (4.51 km/s) would require a mass ratio of 9.5:1 between the rocket mass in LEO and the final inert payload in circular solar orbit. The initial velocity gain to LEO of 8 km/s at an Isp of 200 seconds corre- sponds to a mass ratio of 55 from launch to LEO, which combined with the 9.5:1 means an overall mass ratio of 524:1. Thus, launching 10 kilograms of plutonium would mean an overall RV mass of 25 kilograms, a mass in LEO of 238 kilograms, and a launch mass of 13 metric tons. Today, large payloads cost roughly $10,000/kg to launch to LEO. At this cost, launching the 238-kilogram payload to LEO would cost some $2.4 mil- lion, or about $240 per gram of plutonium. The launch costs for disposing of 100 tons of excess weapons plutonium in this way not including any other costs, such as development and licensing-would come to $24 billion. Thus, without any consideration of extra costs for development, licensing, rescue of the payload, or tracking and prompt retrieval of an aborted launch or reentry, one finds a cost that is truly out of sight in comparison even with building dedicated power reactors to consume the plutonium. Can launch costs be reduced? Probably so. A variety of new launch con- cepts have been proposed in recent years; at one time the goal of development in the Strategic Defense Initiative (SDI) program was to reduce launch costs 100-fold compared to today's prices. For this mission, it may be that smaller rockets would turn out to be cheaper than large heavy-lift vehicles, because of the great miniaturization that is now possible, and the economies of scale in procuring many units of a single design, together\with the much reduced cost of development of a small rocket compared with a large one of comparable tech- nology. If 10 kilograms of plutonium were launched on each rocket, some 10,000 launches would be required to dispose of 100 metric tons of military plutonium. How low would launch costs have to go to make space launch competitive with reactor or vitrification options that would cost $2.5 billion or less? If one imagines 10,000 launches, then $1 billion allocated for sensors and electronics
272 APPENDIX C would require that this element of each launch not exceed $100,000. The re- maining $1.5 billion would be consumed at an average cost for engine, struc- ture, and fuel of $10 per pound. With many thousands of ballistic missiles built over the years, however, no such modest costs have been achieved, and the extreme reliability required for launch of plutonium payloads is likely to in- crease the cost beyond that associated with a system based on con- sumer-standard components. Moreover, developing and licensing the launcher and reentry vehicle, and developing, demonstrating, and building a highly reliable rescue system for payloads that might get stuck in LEO the only way to prevent their eventual fiery return to earth would be substantial additions to the total cost. DESTRUCTION WITH UNDERGROUND NUCLEAR EXPLOSIONS Description Shortly after the dissolution of the former Soviet Union, the Russian firm CHETEK, associated with the Russian nuclear weapons laboratory Arzamas-16 and the Ministry of Atomic Energy, proposed that plutonium could be disposed of by using underground nuclear explosions.34 A single 50-kiloton device could be surrounded by some 5,000 plutonium pits. The 50-kiloton blast would melt both the pits and 50,000 tons of the surrounding rock, into which the plutonium from the pits would be dissolved and distributed. Several such blasts would be required to dispose of the pits from the tens of thousands of nuclear weapons now slated to be dismantled. More recent proposals are to destroy perhaps 100 pits at a time by using much smaller nuclear explosions. This would require the same excavation per pit as the more aggressive proposal, about the same total nuclear explosive yield, but more underground detonations. Another possibility would be to use this approach to destroy intact nuclear weapons, saving the time and cost of disassembly. One-point safety calculations would be needed for this case because the weapons would be affected by a nu- clear blast wave combined with substantial numbers of neutrons from the origi- nal explosion. This option would also throw away the substantial value of the highly enriched uranium in the weapons. These concepts raise obvious environmental and policy issues; CHETEK no longer appears to be actively pursuing the idea. 34 See, for example, proposal by Y.A. Trutnev and A.K. Chernyschev (Arzamas-16), presented at the Fourth International Workshop on Nuclear Warhead Elimination and Nonproliferation, Washington, D.C., February 26-27, 1992.
APPENDIX C 273 Environmental Impact The safety of such an operation has not yet been sufficiently analyzed, but need not be, since the idea can be rejected on other grounds. This concept would amount, in effect, to placing 100 tons of plutonium into an underground repository. In this case, however, the waste form would be created explosively, with no ability to ensure that it was a suitable form for dis- posal, and it would be located at a site never intended as a geologic repository. Either factor alone would essentially rule out this option from competition with approaches in which the waste form is carefully engineered and placed in an engineered repository. For example, concerns over potential long-term critical- ity of the underground plutonium, after possible differential leaching of differ- ent constituents in the rock, would be far more difficult to address than in the case of the vitrification or spent fuel options, since there would be no opportu- nity to engineer the resulting waste form with this problem in mind. Further- more, a much larger surface area of plutonium-bearing material would ulti- mately be exposed to water than is the case in an engineered repository. The explosion would be situated either above the water table (which is at a depth of some 550 meters at the Nevada Test Site) or below it. If above, then rainwater may penetrate the debris and transport the plutonium into an underlying aquifer. If below the water table, flowing water may transport the debris. Disposing of plutonium from tens of thousands of weapons in this way would mean a very large increase in the amount of plutonium already at the U.S. and Russian test sites, from hundreds of past tests. Cost The size and depth of the hole in which the explosion was carried out would significantly affect the cost of this option. CHETEK has referred to placing the pits in a mined cavem, but it would probably be cheaper to use a relatively normal drilled hole of some 8-foot (2.4-meter) diameter.35 To avoid venting of the explosion, normal test practice calls for a depth of burial that increases with the one-third power of the explosive yield, according to the for- mula B = 125Yi/3 (with B in meters and Y in kilotons). This gives a depth of 125 meters for a yield of 1 kiloton, and 270 meters for 10 kilotons. For example, one might use a 1.3-kiloton explosive detonated at a conser- vative depth of 300 meters. Since such an explosive will create a cavity with a radius of about 15 meters, one could imagine placing racks of pits above and below the explosive, each holding ten bays, 5 feet apart, each loaded with five warheads or pits, for a total of 100 to be destroyed by this blast. Alternatively, one could pack the pits considerably tighter. Using the Russian-design pit stor- age containers-50 centimeters in both diameter and length and stacking 35 C.J. Anderson, W.G. Sutcliffe, et al., Livermore National Laboratory, personal communication, January 25, 1993.
274 APPENDIX C them with 50 percent volume efficiency (four containers per cubic meter), one could put as many as 1,200 pits into the cavity for a 1-kiloton explosive. Yet another approach would be to use a string of explosives, each with its 100- 1,000 target pits, installed in the same hole and detonated simultaneously. Since a 2.5-meter hole 300 meters deep in Nevada costs about $1 million, the cost would be $10,000 per pit destroyed just for the hole (at 100 pits per hole); for 20,000 pits, the cost would be $200 million. There would be additional costs for the nuclear weapon used and in overhead for the various preparations re- quired; the overall program cost for a single 100-pit destruction event might be $20-$30 million. Perhaps after the first several, the cost would drop to about $10 million per shot, for a per-pit cost of $100,000 and a total program cost for 20,000 pits of $2 billion. This does not include costs of development and . . licensing. Retrievability In addition to the environmental and other policy issues associated with this approach, the plutonium would remain readily retrievable for the host na- tion. If everything goes as claimed, in the aggressive approach involving 5,000 pits at a time being destroyed by 50-kiloton blasts, an average of one pit would be dissolved in 10 metric tons of melted and solidified rock. Similarly, for the concept involving 100 pits destroyed by a 1.3-kiloton explosive, there would be one pit in every 13 tons of rock. If 1,200 pits were packed around such a 1.3- kiloton explosive, there would be one pit in every ton of rock. The only fission products in this rock would come from the single nuclear explosion, and thus the rock would be much easier to handle than spent fuel, even if that fuel were many decades old. Since gold can be mined profitably at a level of 1 ounce per ton (at a market price of some $10 per gram), it is clear that any of these options would create a very rich plutonium mine. On the other hand, casual access to the plu- tonium would be precluded, and that would be of some value. Policy Issues Clearly public opposition to this disposal method would be intense. Moreo- ver, setting off nuclear explosions for this purpose would contravene the current moratorium on nuclear testing, and undermine the current U.S. and Russian policy of pursuing a comprehensive ban on nuclear explosions. BRIEF SUMMARY OF CONCLUSIONS Of these five options, all but deep boreholes and the sub-seabed approach may be rejected at this stage. Disposal of plutonium in containers in deep bore- holes merits further analysis and evaluation. Sub-seabed disposal should be
APPENDIX C 275 explored further, but only in a form in which the plutonium is injected into the surrounding mud, rather than remaining contained in a set of compact canisters that could be readily located. It should not be pursued over the objections of the international community. Further exploration of these two options will evidently need to emphasize different aspects of the process. Deep-borehole disposal has many technical questions relating to the drilling of the hole, characterization of the environ- ment, emplacement of the containers, and sealing of the borehole. The contin- ued isolation of plutonium from the biosphere can be investigated rather inde- pendently, including the potentially important benefit of highly saline water at depth. Policy issues play a minor role in deep-borehole disposal. The sub-seabed option, on the other hand, poses major policy questions. The technical aspects of injection of plutonium solution into the deep-ocean sediment are relatively simple technically, and a program for analysis and field test of the isolation provided by the sediments can be laid out that would have high confidence of resolving the necessary issues. The crucial aspects of sub- seabed disposal are the policy implications, and these questions can be explored on the assumption that the technical aspects can be resolved satisfactorily. The policy issues may differ somewhat depending on whether the plutonium would be emplaced within the exclusive economic zone with 200 nautical miles of the United States or Russia, or in international waters- though the two zones are strongly coupled by ocean circulation and both are covered by the London Dumping Convention.