The management of long-lived radioactive wastes is one of the most vexing and difficult challenges created by modern technology. Some radionuclides will persist for millions of years. Plutonium-239, present in substantial quantities, can be used to make nuclear weapons, making the reversal of any disposal attractive for future proliferators. Solutions to reduce the longevity of the wastes by transmutation, possible in theory, create intolerable proliferation risks and leave residual contamination and waste that would still require long-term management. Also see table of rejected high level waste management methods .

In other words, there are no ideal options for managing highly radioactive waste. The menu is a poor one and any "solution" will be from among options that each have some drawbacks. That is one reason why phasing out nuclear power and stopping nuclear weapons production, both of which should be done for other reasons as well, are important complements to the search for the least environmentally destructive waste management approaches. Difficult as it may be to accept, it is highly unlikely that there will be any future technological "silver bullet" that addresses all of the important technical, environmental and proliferation issues simultaneously, even if cost is left out of the picture. Above-ground storage for the indefinite future is also not an option (see article on short- and medium-term steps). Inaction is a recipe for even more problems.

Further, in the real world, the resources devoted to any one problem are necessarily limited. So far, massive amounts of money have been spent on unsuitable, politically-driven projects, notably the Yucca Mountain project in Nevada and the Waste Isolation Pilot Plant in New Mexico. As the recent placement of wastes in WIPP without a state hazardous waste permit shows, spending a lot of money on a hole in the ground creates political pressure to open such repositories, no matter how environmentally unwise that might be.

The placement of waste in WIPP proves nothing more than that the economic and political forces behind such placement are, at least for the moment, more powerful than those opposed to the opening of the repository. It does not change the fact of the pressurized brine reservoirs in the area, or of the resources located there that make the possibility of human intrusion a severe problem at the site. Ignoring these problems is an expensive and dangerous continuation of the nuclear establishment's "out of sight out of mind" approach to nuclear waste management. This is a poor way to approach the scientific and technological challenge of minimizing the potential and actual damage from the waste that has already been generated.

A sound waste management program needs to be structured so that sufficient resources can be expended on several options, which will enable reasonable comparisons to be made. Of course, sound comparisons will require sound science, which makes the institutional framework for the long-term research at least as important as the technical issues. (See accompanying article, Institutional Reform for Long-Term Waste Management)

This article outlines three broad approaches that may in some measure meet the goal of isolating waste from the human environment for the necessary period (hundreds of thousands or millions of years):

1. geologic disposal - disposal in a land-based deep repository within the Earth's crust
   
2. sub-seabed disposal - disposal in the ion-absorbing soft clay sediments beneath the sea floor
   
3. disposal under the Earth's crust.
   

Geologic Disposal

Geologic disposal has been the most studied approach to long-term nuclear waste storage. The basic concept is to dispose of the waste in a deep repository in containers surrounded by other engineered barriers such as special backfill materials. The only location being investigated in the United States for spent fuel and military high-level waste is the Yucca Mountain Site in Nevada, which consists of volcanic tuff. A five-mile long tunnel has been drilled into the mountain. The Waste Isolation Pilot Plant (WIPP) deep salt bed repository has been approved to receive TRU waste by the Environmental Protection Agency, but still lacks a permit for the non-radioactive hazardous wastes present in most containers.¹ Granite and clay sites are being investigated in countries such as Sweden and France.

There are three principal difficulties with geologic repositories:

1. It is likely that some radioactive wastes will leak from the canisters and other barriers built to contain them.
   
2. Prediction of the performance of the repository over very long time periods is very difficult.
   
3. It is essentially impossible to guarantee that there will not be inadvertent or deliberate human intrusion.
   

These problems can be addressed to varying degrees by a sound site selection process, by adequate research and development on engineered barriers, and by carefully consideration of the causes of human intrusion. Let us address the last question first.

One of the thorniest issues relating to human intrusion is whether and how to warn generations far into the future of the dangers of radioactive waste. Warning systems to keep people away have dubious utility, at best, and encourage unwarranted complacency, at worst.² Furthermore, techniques to warn future generations against inadvertent intrusion would draw attention to the disposal site and increase the danger of deliberate intrusion to get at plutonium or other materials in the waste.

The chance of deliberate intrusion can be minimized by designing the repository and engineered barriers so that it would be technically and economically far more difficult to recover spent fuel and bring it to the surface than to build a new nuclear reactor to produce plutonium. The chance of deliberate intrusion is also reduced if there are no permanent markers warning of the disposal site and its contents.

The most important safeguard against inadvertent intrusion is to select a site where it is highly unlikely that human beings will search for resources. Following this logic, the best guarantee against intrusion is to select a site where:

• the water resources at the repository location or in its vicinity are highly unlikely to be used, for instance because of poor quality, so that their contamination does not present a probable hazard to human beings;
  
• there are no known commercially-important resources at the site or in its vicinity;
  
• Essentially all elements and geologic minerals are more easily and abundantly available in the general geographic region than at the repository location or in its vicinity.
  

The Yucca Mountain site fails on the first and third counts. Water is generally scarce in the region, while groundwater is available and of high quality. Although the water under the repository site itself lies under a mountain, groundwater in the immediate vicinity of the site is more accessible to drilling, making intrusion a real possibility. Further, groundwater as near as 20 miles from the site, in Amargosa Valley, is currently used for irrigation. Yucca Mountain is also located in a mineral rich area. The mountain itself has not been exploited for mineral resources, but silver and gold mining are carried out within sight of it.³ The WIPP site fails the second criterion because there are petroleum and potash resources in the vicinity.

A recommendation for study by a 1983 panel on waste isolation of the National Research Council of the National Academy of Sciences (NAS-NRC) appears to meet these criteria⁴ (but fails on other grounds - see below). The suggested type of site would be in a granite layer containing brackish groundwater that lies under a sedimentary aquifer. Such sites are found in some locations near the eastern seaboard of the United States, where surface fresh water is relatively abundant. Since there would be a fresh water aquifer above the site, intrusion to get at brackish water would be highly unlikely. As regards other resources, granite is abundantly available close to the surface in eastern locations, so that drilling for any other known resources in deep granite is also unlikely.

However, human intrusion is only one of the concerns that must be addressed in a repository program. In addition, the repository (or any other disposal means) must meet environmental, health and technical criteria. Some of the essential ones are:

• The repository and engineered barriers should each be able to meet strict health-based performance criteria as separate systems in order to provide a minimal element of redundancy. This is essential since there will remain considerable uncertainties in estimating performance of either system over long periods of time.
  
• Repository performance, including that of the engineered barriers, should be characterizable to a degree to allow statements about compliance with strict health protection standards to be made with a high degree of confidence.
  
• The site should not have the potential to destroy or seriously disrupt unique ecological resources. For instance, putting unique species at risk would be unacceptable.
  

In addition to the many problems already mentioned, Yucca Mountain also fails on the first of these criteria because the geology is not expected to provide a meaningful barrier in the long term. The one specific location suggested by the NAS-NRC panel is unsatisfactory because it fails to meet the third criterion. It would be near at or near the Chesapeake Bay, one of the richest and most sensitive natural environments in the United States. The introduction of vast quantities of nuclear waste and the accompanying large-scale construction in the region would be highly disruptive of a unique ecological and economic resource.

Finding an appropriate repository site is a very difficult and complex process that must balance a wide range of considerations, as is illustrated by the preceding discussion. Thus, it is very premature at this time to select actual repository sites or even to engage in a site selection process. Much more basic research on various geologic settings is needed before sites can be scientifically screened. Further, repository types need to be considered in tandem with the development of engineered barriers.

IEER's recommendations for the US repository program are:

1. Convert WIPP and Yucca Mountain into world-class centers for research on geologic disposal, testing of materials for engineered barriers, etc., using only non-radioactive analogs. This would be contingent on consent by the states of New Mexico and Nevada, and in the case of Yucca Mountain, of the Western Shoshone people. WIPP and Yucca Mountain would be permanently off the table as potential repository sites because they are poor repository locations. Waste already placed in WIPP should be removed from it since it is a poor site and since the presence of radioactive wastes in it will limit and compromise research activities that would serve a sound long-term management program.
   
2. Expand and intensify research into the study of natural environments in which radioactive materials have been contained for millions of years and couple this work to an engineering program of developing analogs of these natural materials. The objective would be to design and manufacture engineered barriers around the spent fuel that would mimic these natural materials and environments.
   
3. Study various kinds of repository locations by doing theoretical research, computer modeling, and laboratory, geologic and other field work for ten to fifteen years without any attempt to rank or screen these locations as potential repository sites. Waste would be stored as safely as possible on site or as close the point of generation as possible during this time.
   

Sub-seabed disposal

Sub-seabed disposal has been studied to a much lesser extent than geologic repository disposal. It is important to distinguish sub-seabed disposal from sea dumping of radioactive waste. Sea dumping involves the disposal of waste into the water, where it is certain to become dispersed. By contrast, sub-seabed disposal would place the waste beneath the sea floor. If successfully accomplished, the waste would not disperse into the oceans.

There are two approaches to sub-seabed disposal as it has been considered so far:

• put the waste in holes drilled tens of meters deep into the ocean floor
  
• put the waste in canisters shaped like long projectiles that penetrate into the ocean floor. The penetration depth in soft clays may be several tens of meters.⁵
  

A site in the North Pacific Ocean with 100 million square kilometers of ocean floor covered with soft red clays up to 100 meters deep has often been mentioned as a possible site. ⁶

The main advantage of sub-seabed disposal relative to geologic disposal is that large radiation doses via the drinking water pathway are highly unlikely. Water used for drinking and irrigation is generally regarded as the most important radiation exposure pathway that would result from geologic disposal.⁷ However, radiation doses via the food pathway are possible. Based on current technology, deliberate human intrusion would be far more difficult than with geologic disposal. Given that rapid technological change is likely to continue, deliberate intrusion might be possible, though the lack of markers or any other surface manifestations should make this less likely than for land-based repositories. Inadvertent intrusion would appear to be far more unlikely under the ocean, especially in areas away from coastal areas and where there are no readily accessible seabed mineral resources.

Because less research has been done into sub-seabed disposal, less is known about the potential problems with this storage method. However, troubling questions have been raised. For instance, oceanographers Hessler and Jumars have noted that while the density of living matter in the deep sea is low, life there is very diverse. Several factors promote this diversity of life in the deep-sea environment, notably the fact that it is very stable:

"Such stability minimizes the likelihood of extinctions even for species maintaining extremely low population densities, and thereby allows the diversity of communities to build to high levels....

"While no one has yet measured the tolerances of abyssal [deep-sea] organisms, it is almost a certainty that they can adjust to only a small degree of environmental change....Thus any kind of human activity on the deep-sea floor - be it waste disposal, nodule mining, or anything else - is likely to have a far more deleterious effect than would a comparable disturbance in shallow water."⁸

In the long run, questions of isolation from the human environment in the case of sub-seabed disposal may be broadly similar to those facing geologic disposal. Transportation, waste emplacement, and licensing also pose significant challenges. Finally, the international convention against sea dumping of radioactive wastes may prohibit sub-seabed disposal.

Given the potential vulnerability of life in the deep-sea to human activity, sub-seabed disposal cannot be viewed as a "solution" to the waste disposal problem. But its relative problems may not be more severe than those with geologic repositories, though the specific issues are somewhat different. Hence, at the present time, sub-seabed disposal should be allocated significant research resources. These resources should not be used to add radioactive materials into the oceanic or sub-seabed environment. International collaborative sub-seabed disposal research could be a major component of the conversion of Cold War naval apparatus in the nuclear weapons states to peaceful purposes.⁹

One disadvantage of sub-seabed disposal is that it would involve disposal in the global commons. Countries that have made inadvisable decisions regarding nuclear power and weapons would be able to dispose of waste without taking the commensurate domestic liability for the problem. To make matters worse, countries that have not generated high-level radioactive wastes would share in potential adverse consequences. The use of sub-seabed disposal or any other international approach should be considered only in the context of the complete and irrevocable phase-out of nuclear power and of fissile materials and tritium production for weapons purposes.

Disposal outside the biosphere

There are two alternatives for disposal of nuclear wastes outside the biosphere: either above it, in space, or below, beneath the Earth's crust in the upper mantle.

The first is impractical and should be rejected due to the large volumes of waste involved, due to both cost and safety concerns. The remaining option is disposal beneath the biosphere. It is difficult to define an exact lower boundary to the biosphere, because there are interactions between the various layers of the Earth. For instance, volcanic eruptions bring up magma from outside the biosphere into it. The operational definition of "biosphere" for the purposes of nuclear waste disposal would itself have to be the subject of considerable research. Two somewhat different definitions may be satisfactory:

• deep regions of the Earth's crust where there is no water even in the pores of the rocks.
  
• stable portions of the upper mantle (which lies below the Earth's crust) that do not exchange material with the biosphere on time scales smaller than tens of millions of years.
  

Figure 2: Earth, cross section

Figure 3: Earth's layers

Figures 2 and 3 show the various layers of the Earth. The Earth's crust is roughly 5 to 10 kilometers thick beneath the oceans; by contrast, it is between 20 and 70 kilometers thick under continental areas.¹⁰ The boundary of the Earth's crust and the upper mantle - called the Mohorovicic discontinuity or Moho for short - is characterized by a sudden increase in density with depth. This enables the upper mantle to be identified as a distinct layer geologically (and hence also for disposal). In some areas, the rock in the upper mantle is in a molten or semi-molten state, but in most areas it is solid. Investigation of the layers of the Earth where boreholes cannot yet be drilled is carried out by indirect methods such as study of changes in the velocities of seismic waves at boundaries between layers.

Disposal in the uppermost region of the mantle would have some of the same characteristics as deep borehole disposal in the Earth's crust.¹¹ In the case of disposal in the upper mantle, the waste containers would be lowered into extremely deep boreholes extending below the Earth's crust. The boreholes would be drilled in a geologically stable area, that is, away from areas where tectonic plates converge (at the continental margins) or diverge (as, for instance, in the mid-Atlantic and East Pacific ridges).

Stable areas in the upper mantle may be able to keep the waste out of the biosphere for millions of years, though this hypothesis would have to be carefully investigated before this method is selected. Upper mantle disposal would also address the thorny questions of deliberate or inadvertent human intrusion better than the other two approaches.

The safety, technological and scientific questions surrounding this option are as immense as its theoretical promise and it is unclear whether or not they can be resolved. For instance, the technology for drilling into the upper mantle does not exist and is not now under development. It is highly unlikely to be developed in the near future. However, drilling very deep boreholes may become more feasible with new technology such as cutting rocks with lasers.¹² It may also be possible to dispose of the waste in the stable areas of the upper mantle beneath the ocean floor, where the Earth's crust is less thick than under continental areas.

There are a host of safety issues surrounding upper mantle disposal. For example, even if sufficiently deep boreholes can be drilled, would they be stable enough to allow disposal of the waste all the way into the upper mantle? How would mishaps in lowering the waste be handled? How would the various layers of groundwater be sealed at great depths in order to permit waste emplacement?

Finally, the science around the estimation of the performance of upper mantle disposal is not developed. For example, drilling holes into the upper mantle may provide a path for magmatic flow to the surface, bringing radioactivity with it. The likelihood of such an event at a specific site would need to be assessed in the licensing process. Further, the upper mantle is presently inaccessible to direct measurement and investigation, so that its properties must be inferred in various ways. While these indirect techniques allow for an understanding of general structure and composition, it is not at all clear that sufficiently detailed knowledge can be developed to use this disposal technique with confidence. In the absence of new investigation techniques, the process of actually licensing this disposal method would be open to question.

In weighing the factors mentioned above, we have concluded that the theoretical potential of upper mantle disposal to keep long-lived radioactive waste out of the biosphere is high enough that it deserves significant financial resources, even though it appears unlikely at present that this approach would bear fruit.

Conclusion

Selecting sites for land-based disposal of nuclear waste in geologic repositories is very premature. There has not yet been enough research to determine whether this approach is the best one. Moreover, even within the framework of geologic disposal, programs have been compromised by political expediency.

We have discussed three broad approaches to long-term waste storage that IEER believes should be studied in parallel: geologic disposal on land, sub-seabed disposal, and upper mantle disposal. The main aim of this research should be to yield sufficient data and analysis in one to two decades to enable a comparison between these options. If the first phase of the process reveals sufficient promise in sub-seabed disposal or upper mantle disposal, further work might be required before repositories can be ruled out, because problems with repositories are better known and those with the other two might take longer to emerge. At that time, one or two could be ruled out, if the data warrant, and further resources concentrated on the remaining approach(es). That would also be a more appropriate time to reconsider the question of whether and how a site selection process for disposal should be undertaken.