The Department of Energy’s Office of Environmental Management, this committee and other NRC committees,1 and many citizens who are concerned with DOE site cleanup realize that it is not possible to totally remove all of the legacy waste and environmental contamination from the EM sites. In its accelerated cleanup program, EM intends to leave much of its buried waste and other “in-place” waste and contamination at DOE sites. However, ensuring that left-in-place wastes in fact remain in place is a responsibility that continues for potentially very long times. In assessing step improvements in efficiency in waste characterization and treatment, the committee found that options to characterize and treat (stabilize) wastes in place offer significant opportunities for accelerating EM’s program. In this chapter, the committee provides advice on technical and scientific approaches for optimizing the trade-offs between short-term expediency and long-term liabilities.
COCOONING APPROACH TO MANAGING WASTE IN PLACE
In its fact finding, the committee noted that reactor “cocooning” at Hanford is an instructive conceptual approach to managing waste in place. Reactor cocooning, also known as Interim Safe Storage (ISS), involves demolishing and removing all of a reactor’s ancillary buildings and the reactor building itself, except for the thick shielding walls around the
defueled reactor core. A roof that is expected to last for about 75 years is built over the remaining structure, which is sealed, and the door is welded shut. A minimal but prudent amount of external monitoring is maintained, and the structure is reentered for inspection every five years. It is expected that the roof will have to be replaced one or more times before the residual radioactivity in the reactor core will have decayed sufficiently, and/or new technology developed, so that final disposition (removal, abandonment, a combination of these, or a currently unknown option) can be undertaken. Reactor cocooning has support from Hanford’s regulators and other stakeholders.
The committee believes that the cocooning concept can be adapted and applied to other left-in-place wastes and facilities. Conceptualizing left-in-place wastes as cocooned can facilitate EM’s accelerated cleanup in terms of both simplifying the technical work itself and gaining acceptance by stakeholders. According to the cocooning concept,
Needs for characterization and treatment are greatly reduced compared to the handling, characterization, and treatment necessary if the waste were removed—but characterization and treatment are not totally eliminated;
Immediate advantages versus long-term liabilities (monitoring, rebuilding, eventual disposition) are displayed clearly to all stakeholders, including EM, contractors, regulators, state governments, communities, and other interested parties;
The waste is stabilized in a scientifically responsible way that meets today’s regulatory requirements and uses well-established technologies, but does not foreclose future disposition options;
Monitoring and modeling are ongoing, providing verification that the waste is safely stabilized in the same sense as periodically inspecting cocooned reactors; and
Future remedial activities are adapted to new developments in science, technology, regulation, and the changing (usually decreasing) risk of the waste with time.
The concept of cocooning is to adaptively manage wastes that do not pose immediate health or environmental threats and to avoid actions that involve costly or inappropriate treatment activities and result in little gain. Radioactive decay generally serves to reduce the hazard over time. New technologies as well as new knowledge to inform better decision making for the wastes’ eventual disposition can emerge during the cocooning period. Simply leaving wastes in place without an adequate monitoring program and periodically revisiting characterization and treatment options cannot be considered cocooning.
For waste that EM considers leaving in place, the committee recommends that EM broaden the use of the cocooning concept as currently applied to the Hanford reactors. The cocooning approach provides stabilization and monitoring of wastes left in place, a clear understanding of current benefits and future liabilities for all stakeholders, and the possibility of adapting to changes that will inevitably arise in the future.
In its site visits, the committee noted instances where the time and money that were being spent on engineered trench caps appeared unwarranted, cap design appeared inappropriate and/or the projected performance of caps, and also engineered liners, appeared overly optimistic. The trench cap at the SWSA-6 site at Oak Ridge is one case in point.2 In 1989, a multi-million dollar high-density polyethylene (HDPE) cap was placed over SWSA-6, despite the fact that continued lateral groundwater flow through the buried waste, the primary cause of contaminant transport from this site (ORO, 2002), rendered the cap insignificant as a hydraulic barrier. Maintenance and repair of the cap ceased in 1996, reducing its reliability even as a protective barrier between the underlying waste, ecological receptors, and site workers.
Another example is the Environmental Remediation Disposal Facility (ERDF) at Hanford,3 where risk-assessment analyses assume that leachate from the facility will not reach groundwater for 1000 years (DOE, 2000a). These analyses presuppose that the ERDF’s engineered cap and liner will remain effective for close to 1000 years.4 There are no data to support such expectations of longevity. Furthermore, the ERDF cap includes a layer of compacted clay despite concerns that, in arid and semi-arid regions, such layers might desiccate and crack over time, reducing the cap’s effectiveness as a barrier to water infiltration (Suter et al., 1993). Continued monitoring, with repair and some rebuilding as necessary, would be more reasonable expectations.
Rather than installing unwarranted or inappropriate barriers, and/or over-rating barrier performance without adequate scientific basis or means, the cocooning concept would lead site managers to use more cost-effective and scientifically justifiable barrier designs, and help ensure that all stake-
By analogy to reactor cocooning at Hanford, the cocooning concept applied to managing left-in-place waste encourages the use of established, cost-effective technology while making long-term responsibilities clear to all stakeholders.
holders recognize their continuing responsibilities for maintaining barriers and managing the left-in-place wastes (see Sidebar 4.1).
The following sections outline the steps for managing waste in place according to the cocooning concept.
The purpose of characterization is not to characterize a site or waste area group5 as thoroughly as possible, but to gather sufficient information to make decisions. Characterization needs for managing left-in-place wastes include knowledge of the waste boundaries, including its footprint area and
vertical limits, the local subsurface hydrostratigraphy and the waste properties. The use of geophysical methods (described in Chapter 3), real-time field analyses and decision support, and geostatistical techniques for optimizing sampling locations can produce timely results and reduce overall costs (Ditmars, 2002; NRC, 2003c). One example of a successful application of geophysical methods is the Argonne National Laboratories’ Adaptive Sampling and Analysis Program (ASAP), which is based on noninvasive surveys and computer analyses to facilitate decision making. The ASAP led to an estimated saving of $40 million for characterization of radioactive soil contamination at the Fernald Site (Ditmars, 2002).
Detailed characterization of buried waste, such as the Glovebox Excavator Method (GEM) used for Pit 9 at the Idaho National Engineering and Environmental Laboratory (INEEL), is time consuming and difficult (see Sidebar 3.1). The committee does not consider such detailed characterization routinely necessary for sound decision making. The use of remote sensing, historic photos, and discharge from groundwater seeps at WAG-4,6 Oak Ridge, where waste burial records were lost due to fire, is an example of an approach to characterization that the committee felt to be reasonable (Huff et al., 1996). In the case of groundwater plumes, characterization approaches need to recognize that plumes are unlikely to be stationary, either spatially or temporally (Read et al., 2004).
Site characterization should be commensurate with the complexity of the site and/or waste in question. For example, hydrogeologic characterization of the Oak Ridge site might justify more effort than is needed at the Savannah River site, because waste migration in the fractured bedrock underling Oak Ridge will be more complex than waste migration in the Coastal Plain sediments underlying the SRS. Hanford’s Data Quality Objectives program provides advice on obtaining sufficient, but not unnecessary data, to define the risk at the site, demonstrate the need for remedial action, and support the rational for selecting a remedial action alternative.7 Similar advice is provided by the EPA Office of Federal Facilities Restoration and Reuse.8
BASIS FOR MANAGING WASTE IN PLACE
Only after waste and its surroundings are adequately characterized can scientifically based decisions about leaving it in place be made. Although the decision to leave waste in place is site specific, the committee provides the following broad guidelines and suggests that if one or more guidelines are met leave-in-place should be considered. These guidelines are largely compatible with the EPA (1997) “rules-of-thumb” for when it is appropriate to contain wastes, rather than retrieve them:
There is currently no clear disposition pathway for the waste if it were retrieved, e.g., contaminated soils at the Oak Ridge Corehole 8 site,9 and irradiated lithium-aluminum targets used for tritium production and now buried at Hanford and the Savannah River Site (SRS),
The waste poses little threat when left in place, but will expose workers to risk if it is moved, e.g., calcined waste from nuclear fuel reprocessing now stored in large silos at INEEL,
With current technologies, it is technically infeasible to remove all of the waste, e.g., the deep (> 60 m below ground surface) trichloroethylene plume in groundwater beneath WAG-1 at INEEL,10
The subsurface area that is affected is very large, e.g., large groundwater plumes at each of the sites, such as the tritium plume beneath the 1,500 acre low-level burial grounds at Hanford’s 200 East area,
The threat posed by the waste will diminish to acceptable levels within the next few decades, e.g., tritium groundwater plumes at SRS,
Removing the waste will cause greater risk or damage to ecosystems than the risk posed by the waste itself, e.g., contaminated sediments in the Clinch River system at Oak Ridge and Par Pond at SRS (Whicker et al., 2004).
The committee also recognizes that there are times when the risk posed by existing buried waste is unacceptable, for example due to its toxicity or location. The 618-10/11 caissons at Hanford, described in Chapter 2, are examples of waste that should not be left in place.
The committee considers buried waste or contaminated soil to be stabilized when the waste or soil does not release unacceptable levels of contaminated liquids or gases into the environment and is not accessible to ecological and biological receptors at risk through direct contact. In cases where existing buried waste or contaminated soil is not already stable, strategies that involve leachability reduction, hydraulic isolation and/or physical isolation can help achieve stabilization. In the case of contaminant plumes, stabilization means that the plume is either contained in a controlled area where is does not pose an unacceptable risk to human health or the environment or it is treated at discharge points to prevent the unacceptable release of contaminants. Stabilization can be achieved by physical, chemical or biological treatment (Wentz, 1995).
In order to minimize the flow of water through waste materials, and thus minimize leachate generation, the cumulative permeability through the wastes and associated barriers should be less than the surroundings. For permeability reduction in place, in situ grouting (Conner, 1990) or dynamic compaction (Massarsch, 1999) is adequate in many situations but may not eliminate all of the permeable pathways through the waste when used for large-scale applications. As an example, grouting could be considered for the “Corehole 8” contamination discussed during the committee’s Oak Ridge site visit. Although this contamination is in an area near buildings that are in active use, it does not pose a threat at the land surface. Excavating the contamination would expose the waste and put workers at risk. In situ grouting could reduce both cost and risk. Less established practices for reducing waste leachability, such as in situ vitrification, are suitable only under select situations, and are not likely to offer widespread advantages for EM’s accelerated cleanup.
Isolation: Caps and Barriers
The purpose of isolation is to separate buried waste from people and the environment and to control fluid (water and/or gas) migration into or from the waste. Isolation methods, such as caps and vertical walls, are usually intended to prevent fluid migration (Bedient et al., 1999) and can be used as the sole stabilization strategy or to augment leachability reduction techniques that reduce liquid migration through the waste but do not attain targeted end goals. However, caps are also a way to restore the appearance of the environment and encourage the return of local ecology, and consid-
eration of these factors should also be a part of a cap’s design (NRC, 2003b). Furthermore, appropriate cap designs in humid climates can be different than those in arid and semi-arid climates. For example, low-permeability covers are suitable in humid climates, while evapotranspiration covers are suitable in arid climates (Parker, 2004).
Although current knowledge indicates that properly designed surface caps might have design lives of hundreds of years, this assumption is predicated on continued monitoring and repair of the cap (see Sidebar 4.2). Furthermore, it is hard to anticipate site-specific environmental stresses and conditions, such as subsurface deformations or the health of local vegetation, which might adversely affect the integrity of a cap (Parker, 2004). The cocooning concept will enable an adaptive approach to using surface caps that confirms their service life through continued monitoring and/or extends it through repair or even rebuilding when necessary. Whenever possible, existing monitoring wells should be preserved when installing caps. The design of caps that are able to incorporate performance monitoring systems is also important.
Caps alone are insufficient when subsurface water can leach the buried waste (e.g., WAG 6 at Oak Ridge). Additional means of controlling water movement into and/or out of the waste will be needed. These means could include vertical barriers (e.g., soil-bentonite walls, sheet pilings), vertical (French) drains and/or permeable reactive barriers.
For plumes that cannot be isolated or contained in a suitable area using engineered or natural hydraulic, physical, chemical or biological methods, and need control before the discharge point, reactive barriers and in situ redox manipulation are an option. Several DOE sites are already making use of this technology for both organic and inorganic plumes, as well as mixed plumes (e.g., the In Situ Redox Manipulation (ISRM) permeable treatment zone for chromate contamination at Hanford). Permeable reactive barriers can both capture a plume and treat it in situ. Although these barriers require monitoring and periodic repair/replenishment, they are one of a few solutions for treating mixed waste and limiting plume discharge to surface sources (see Sidebar 4.3).
MONITORING AND MODELING: CONTROLLING FUTURE LIABILITIES
Continued monitoring to improve conceptual understanding of subsurface contaminant behavior and, hence, theoretical modeling efforts, are essential to maintaining the safety of human health and the environment at
legacy waste sites. Demonstrating, through observation, that the fundamental concepts of contaminant migration are understood scientifically can lead to greater public confidence that the sites are indeed safe. Many of EM’s technology needs are related to modeling the fate and transport of subsurface
HDPE is a material that is used as a hydraulic barrier in caps and liners (Figure 4.1). It undergoes UV degradation when exposed to daylight, as it is at SWSA-6. It is vulnerable to puncture and burrowing mammals and can be uplifted if gas emissions from the underlying waste are not vented. HDPE is only recommended as a cover system when no other liner is practical (Daniel and Koerner, 1993). Multilayer caps can be effective in controlling water and gas flow into and out of underlying waste, isolating waste from bio-intruders, and restoring the appearance of a site. However, there are many mechanisms that adversely affect the performance of multi-layer cover systems (Daniel and Gross, 1996; Parker, 2004). As a result, EPA’s draft guidance document on landfill coversa presently considers a design life of hundreds of years feasible for cover systems, but only provided that performance monitoring and long-term maintenance of the systems are sustained.
FIGURE 4.1 Trench caps made of high-density polyethylene (HDPE) are expected to last at least one hundred years. There are no field data to confirm this expectation, rather experience has shown punctures, tears, and deformations. Photo courtesy of Oak Ridge Operations Office.
Monitored natural attenuation (MNA) relies on natural processes (chemical and biological breakdown of contaminants, radioactive decay, dilution, and dispersion) to achieve remediation goals within a time frame that is reasonable compared to that offered by active methods (EPA, 1999a, 1999b). MNA requires very detailed site characterization, which can be more expensive and time consuming than required for active technologies.
Cocooning is distinct from MNA. Cocooning is not a final remediation, but rather waste stabilization that allows for future remediation options. Further, cocooning does not require extensive, in-depth site characterization. Finally, although cocooning of nonaqueous phase liquids (NAPL), such as oils, and radionuclide contamination might be feasible, MNA is not expected to remediate NAPL and has not been demonstrated for most radionuclide plumes other than tritium (NRC, 2000b, 2003b). The committee recognizes that natural attenuation processes could contribute to remediation goals for some left-in-place wastes. However, the number of cases in which MNA will be a viable alternative to cocooning is likely to be limited.
contamination (Colwell, 2004). The committee observed that in some instances, e.g., the East Tennessee Technology Park at Oak Ridge, privatization of parcels of land is planned while underlying groundwater remains contaminated. This is a case in which the cocooning approach and specific monitoring plans could help to prevent contaminants from becoming a health hazard.
Better use of existing capabilities and adaptive use of new technologies can reduce the financial and time burdens of monitoring efforts. Both stabilized wastes and contained plumes require monitoring. Although wells for monitoring should be preserved, hydraulic receptors, such as French drains, seeps, and rivers that integrate large areas, should also be used in monitoring schemes. A good example of the use of hydraulic receptors for monitoring is the use of seeps at WAG-4 Oak Ridge. These seeps were used to identify specific trenches that were contributing most to radionuclide release in order to focus remediation efforts on smaller areas (Huff et al., 1996).
The use of “key indicator species” to monitor the effectiveness of stabilization strategies for leave-in-place wastes is also valid. The monitoring of fall Chinook salmon in the Columbia River is a good example of this approach at Hanford. The committee recognizes that no single species is
likely to be indicative of ecological health at a site. Appropriate key indicator species should be identified in consultation with qualified experts, including those with local environmental knowledge.
For monitoring the conditions of contaminant source areas, such as stabilized tanks and trenches, noninvasive geophysical techniques should be pursued more aggressively, as discussed in Chapter 3. The committee believes that simple measurements, using durable low cost sensors that monitor, for example, changes in fluid and material resistivity, could provide useful, first-order information on how cocooning is performing. Long-term monitoring programs that routinely involve the collection and laboratory analysis of a large number of samples should be avoided. However, reasonable triggers for this activity might be unexpected monitoring results from in situ sensors, changes in site information and conditions, or improvements in conceptual understanding. The committee encourages the development of complex-wide networked information systems, such as the U.S. Geological Survey’s (USGS) Water Watch project,11 which could provide real time information on environmental conditions at sites that would be valuable to decision makers and other stakeholders.
Modeling is another important component of a viable leave-in-place strategy. In addition to data collection, monitoring involves interpretation of data. Modeling can provide a focus for this interpretation if handled in an appropriate manner. However, as noted in an earlier report (NRC, 1990a), modeling is only as good as the modeler’s assumptions and interpretations. Because our current ability to accurately predict subsurface flow and transport, especially in a complex setting like the vadose zone, is limited (Lenhard et al., 2004), the purpose of modeling should be to develop concepts of waste behavior that can be tested and updated. These new concepts can enhance the judgments and decision making of personnel responsible for ensuring that the waste remains safe.
Since a key feature of models is incorporation of long-term knowledge of a site, modeling is important for transition from EM to long-term management of left-in-place wastes. Models can help ensure that known geohydrological features and anticipated processes for waste and water movement at a site are not lost, and that the limitations of assumptions and model parameters are well understood. Changes in site personnel often result in loss of essential historic data (Versteeg et al., 2004). This should be anticipated and avoided.
Waterwatch provides real-time maps of water flow in streams and rivers throughout the United States. See http://water.usgs.gov/waterwatch/
The management of left-in-place wastes can benefit from changes in knowledge over time. As an example of how advances in technology have changed site remediation practices in just over a decade, the committee cites the fact that the number of projects at Superfund sites that used “innovative technologies” for contaminant remediation, e.g., air-sparging, bioremediation, dual-phase extraction, permeable reactive barriers, phytoremediation, chemical treatment, in-well air stripping, went from zero to more than 90 from 1985 to 1999 (EPA, 2002). Although the service life of cocooning will vary with the waste, its location, and the threat it poses, periodically reexamining its condition and newer stabilization and treatment options on the order of decades is appropriate.12 This length of time is understandable to stakeholders and, through adaptive management, provides opportunity for using future technology advances to ultimately reduce the costs and risks to society.
As noted at the beginning of this report, the EM cleanup is not intended to remediate all sites for unrestricted future uses (“greenfield”). After EM has completed its work, the Office of Legacy Management, created by DOE in 2004, will assume the long-term responsibility for closed sites—according to DOE’s current planning. For managing the very-long-term responsibilities that are inherent in radioactive waste management, the BRWM has long advocated an adaptive or flexible approach: “[T]ime to assess performance and a willingness to respond to problems as they are found, remediation if things do not turn out as planned, and revision of the design and regulations if they are found to impede progress toward the health goal...” (NRC, 1990b, p. viii). This theme has been echoed in more recent reports on management of high-level waste (NRC, 2001c, 2003d) and excess nuclear materials (NRC, 2003a). Adaptive management is also appropriate for chemical, biological, and hazardous waste, and for classified materials that are now designated as waste as noted in Chapter 2.
According to its task statement, the committee has sought near-term opportunities to support EM’s accelerated cleanup program. The next step, the responsible, long-term management of wastes and contamination that remain after EM cleanup, can best be undertaken through an adaptive approach.