Page 49
4 Research Recommendations
In this chapter the committee offers its views and recommendations on research opportunities for the Environmental Management Science Program (EMSP) to best affect important deactivation and decommissioning (D&D) problems not addressed effectively by existing technologies. Based on its discussion of future D&D challenges and their underlying causes in Chapter 2, the committee concluded that the most significant needs and opportunities lie in the characterization and decontamination steps of the D&D process, robotics and intelligent machines to enhance worker safety, and the scientific basis for determining objectives of a D&D project (facility end states). The committee has been selective in this identification to encourage the EMSP to concentrate its limited funding in a few specific areas where the committee believes research can make the most significant contributions to meeting future D&D challenges. Some technology areas, although clearly important, were excluded because in the committee's view the science and technology base already exists to address them on a relatively short time scale—less than five years.
Each recommendation is illustrated with a brief discussion of the current state of art, technology gaps, and research opportunities. Examples are included for illustration, but these should not be construed as the only opportunities that may be envisioned by the research community. Although the selection of examples was influenced to some degree by the backgrounds and expertise of the committee members, the research recommendations were arrived at by a consensus process that considered input to the committee, the needs of the end user, 1 the existence of critical knowledge gaps, the potential for future cost savings, and the possibility of achieving technology breakthroughs.
1End users are those who will use a given method or technology to accomplish a D&D task. They are usually contractor personnel at DOE sites.
Page 50
Characterization
Characterization of contaminated materials, such as concrete, stainless steel, and packaged wastes, is required at nearly every stage of D&D (see Table 2.2). Labor-intensive sample collection and measurement methods expose workers to radiation and other risks and contribute to the high costs of characterization, presently estimated at 15 to 25 percent of the total D&D budget (Hart, 2000). Initially an assessment must be made to determine the types and quantities of radionuclide and chemical contamination present to ensure that adequate precautions are taken to protect workers and the environment and to assess options for eventual cleanup and disposal. Progress during decontamination must be monitored, which requires repeated characterizations during the work. Finally, for disassembly or demolition the nature and extent of residual contamination must be assessed to determine the final classification and disposal pathway of the material in question. Characterization of very low levels of residual contamination is required if a facility is to be reused.
The varied nature of Department of Energy (DOE) facilities (e.g., reactors, reprocessing canyons, laboratories, infrastructure, support) has led to a wide range of contaminant types and site-specific characterization challenges (see Chapter 2). In each case, characterization generally requires a detector tailored to the contaminant being measured and its matrix, for example, concrete, metal, liquid, or air (Janata, 1989; Webster, 1999; DOE, 2000b). In many instances, reliance is placed on characterizing and mapping sites by physically removing samples (e.g., wipes, cores), sending these to an offsite lab, and conducting chemical analysis and physical characterization. When onsite measurements are obtained, use is typically made of handheld monitors, and the data are recorded manually. Often, measurements must be repeated several times at each step of the D&D process.
As an example, over 400,000 survey measurements were made in the course of decommissioning the Fort St. Vrain commercial power reactor. Over half (221,000) were required for the final survey, which required 22 months to complete. In addition, the allowable levels of residual contamination had to be reduced by about 25 percent below the regulatory guide to account for nuclides such as Fe-55 and tritium that could not be detected with available field instrumentation (Holmes, 2000).
The committee has three recommendations for research that could lead to new or improved methods of characterizing the contaminated construction materials that are found in most obsolete DOE facilities: characterization of surfaces, characterization beneath surfaces (depth profiling), and remote mapping of contamination.
Page 51
Characterization of Surfaces
The committee recommends basic research leading to ultra-sensitive devices for rapid characterization and certification of amounts of radionuclides and EPA-listed substances 2 on the surfaces of construction materials and equipment (e.g., pumps, motors).
Current Status
There are in excess of 180,000 tons of metal as well as significant quantities of machinery and equipment in surplus DOE buildings and temporary storage facilities comprising some 65 million square feet of floor area. Most surfaces are contaminated or potentially contaminated. As noted previously, these surfaces must be characterized repeatedly throughout the D&D process. The greatest opportunities for research are in characterizing surfaces bearing very low levels of radionuclide and chemical contamination. For example, before these materials can be disposed of, they must first be characterized and, if certified to be below a derived concentration guideline level (DCGL) for a specified radionuclide, may be free-released for recycling and reuse (MARSSIM, 2000). If not, they must be decontaminated and re-measured before release or, alternatively, sent to an appropriately licensed waste disposal site. Although most of the contamination is located on the surface, a significant quantity lies in microcracks and other surface defects. In addition, the presence of oils, moisture, rust, and dirt impedes the detection of low-energy alpha and beta emissions. Complicating characterization are certain radionuclides that will be essentially impossible to measure at the DCGLs in situ using current state-of-the-art instrumentation and techniques because of the types, energies, and abundance of the radiation (MARSSIM, 2000). Examples include very-low-energy pure beta emitters such as 3H and 63Ni and low-energy photon emitters such as 55Fe and 125I. In these instances, wipes or samples must be taken and analytical chemical methods used.
Opportunities
The difficulties and limitations in characterizing low levels of contamination on surfaces are important gaps in current technology. In addition faster methods that give better spatial resolution than current wipe
2EPA-listed substances are designated as hazardous by the Environmental Protection Agency (EPA). They are regulated by EPA under provisions of the Resource Conservation and Recovery Act of 1976 and its amendments.
Page 52
or core methods are needed. To address these needs the committee suggests that the EMSP give special attention to research proposed for surface characterization by analytical chemistry methods, nanosensors, and biosensors.
Analytical chemistry methods
Modern methods of chemical analysis that can detect small numbers of molecules are often capable of measuring far lower concentrations of radionuclides than direct radiation measurements. 3 There are opportunities for research to extend analytical chemistry methods to develop more sensitive methods to measure small amounts of long-lived radionuclides.
Laser ablation mass spectroscopy (LA/MS) is an example of a rapid characterization technique that is suitable for solids such as concrete and requires no sample preparation (Van Hecke and Karukstis, 1998; DOE, 1998b). An intense pulsed laser is used to vaporize surface material (ablation). An inert carrier gas (argon) transfers the ablated material to an inductively coupled plasma torch, where the sample plume is disassociated into ionized atomic species. A mass spectrometer subsequently identifies the species and determines its abundance in the sample. Continued ablation of the surface provides the possibility of obtaining a three-dimensional profile of the contamination, as discussed in the next recommendation.
Sensitivity and dynamic range are such that constituent concentrations of most elements in the periodic table can be measured from parts-per-billion to tens of percent with a single analysis. The sensitivity of this analytical chemistry approach is many orders of magnitude better than can be achieved by direct radioactivity measurements for most radionuclides. Typical potential minimum detectable levels are 1 × 10−9 pCi/g for 238U, 10−3 pCi/g for 239Pu, 1 pCi/g for 137Cs, and 10 pCi/g for 60Co (MARSSIM, 2000, p. H-35). The technique is applicable to organic and inorganic species. Variations of laser ablation spectroscopies are attractive as well. These include laser ablation inductively coupled plasma atomic emission spectrometry (LA-ICP-AES) and laser-induced breakdown spectroscopy (LIBS). Research to adapt LA/MS and related approaches for D&D applications that require ruggedness, portability, and high sensitivity would most likely involve basic principles of energy beam-material interactions, including energy coupling, mass removal by vaporization and ablation, particle generation, gas dynamics, solid vapor entrainment, and transport processes.
3See, for example, Chapter IX, “Radiation Measurement” of Webster, 1999, and Appendix H of MARSSIM, 2000.
Page 53
Research toward the development of simple, inexpensive new chemical-sensing materials is another opportunity for applying analytical chemistry methods to D&D needs. The development of simple multi-analyte chemical-sensing color test strips, currently funded by the EMSP in the D&D environmental category, is based on a self-assembled nanostructured material (Asher, 2000; Holtz and Asher, 1997) displaying photonic band-gap behavior (Joannopoulos et al., 1995). Such sensing materials might be fabricated into visual color test strips or luminescent smears to sensitively and selectively report on the concentration and identity of both chemical and radioactive contaminants, including lead, uranium, plutonium, strontium, cesium, and cobalt (Asher, 2000).
Nanosensors
Nanotechnology will have a significant impact on many aspects of science and technology, 4 and the committee believes that research in support of nanotechnology development deserves special attention by the EMSP. Laboratory-on-a-chip sensor research and micro-electro-mechanical systems (MEMS) are already being supported by the EMSP, but there are many knowledge gaps and opportunities for research in this new field (NSTC, 1999).
The ability to manufacture materials with switchable molecular functions will provide completely new approaches to miniaturized valves, pumps, and chemical separations and detection. Nano-composite materials also display unique mechanical, electrical, magnetic, and chemical properties. The quantum-confined behavior of thin films and nanocrystals of semiconductors has opened the possibility of designing composite materials having desirable electromagnetic properties. In principle, it is thus possible to quantum engineer improved solid-state radiation detectors. Semiconductor nanocrystals (quantum dots) in glass have recently been shown to have better thermoluminescence dosimetry (TLD) performance than current state-of-the-art TLDs (Justus et al., 1999a). It is also potentially feasible to manufacture nano-robots having a simple form of locomotion and an ability to sense and harvest targeted chemical species. These would be placed on the surface of metals to migrate into and out of microcracks as they sense and remove specific radionuclides. For example, development of a fabrication method for creating individually addressable and controllable polypyrrole-gold microactuators has been reported. This microbotic arm can
4In April 1998, Dr. Neal Lane, Assistant to the President for Science and Technology, commented, “If I were asked for an area of science and engineering that will most likely produce the breakthroughs of tomorrow, I would point to nanoscale science and engineering” (NSF, 1999, p. 1).
Page 54
pick up, lift, move, and place micrometer-size objects, which are about the size of a biological cell (Jager et al., 2000).
Biosensors
Biosensor technology is another relatively young field (Updike and Hicks, 1967). There has been significant growth in the development and application of living cells, bacteria, yeast, and mammalian cell cultures as sensors for chemicals (e.g., mercury, chromium, toluene, polychlorinated biphenyls (PCBs), trichloroethylene (TCE), benzene), environmental stresses (e.g., ultraviolet or oxidative damage), and advantageous and deleterious pharmacological agents (xenoestrogens). Many of these biosensors are products of genetic modifications of living organisms resulting in organic systems that produce electrochemical, bioluminescent, fluorescent, or chromogenic signals once the organisms have been exposed to a specific chemical or a broad chemical class or stress-inducing agents. Such surrogate signals of agent exposure may offer real-time online sensing capacity, while others may integrate exposure in a form of biological dosimetry. In industry, stress response genes have been linked in transcriptional fusions to create enzymatic, bioluminescent, and fluorescent biosensors and have been employed in a pattern recognition format for new chemical and environmental toxicological screening.
There has been tremendous growth in development and commercialization of a broad range of biosensor devices and applications (Kress-Rogers, 1997). Modern devices can range from fiber optic (Anderson et al., 2000) and microcantilever-linked immunoassays (Moulin et al., 2000; Thundat et al., 2000) to subcellular and cellular micro-electronic systems (Simpson et al., 1998). Analytes measurable by biosensors include a vast array of organic chemicals, biochemicals, inorganics, and metals and more recently ionizing radiation (Lee et al., 2000).
There are research opportunities to fundamentally develop and evaluate biosensor organisms to detect radionuclides and chemicals that are important for D&D applications. In addition, there are opportunities for system integration to interface the biosensor with the appropriate analytical chemistry or microelectronic platforms to provide robust detectors that can be used in D&D operations.
It might be possible for biosensors to be used to characterize occluded radioactive or chemical contamination. For example, a series of stress-inducible promoters responding to ionizing radiation could be fused with the gene for synthesis of green fluorescent protein (GFP). The biosensor organism containing the GFP fusion could be passed through decommissioned reactor lines and sampled online at an outlet to quan
Page 55
tify net GFP as an index of exposure. Such analytical capacity would take advantage of the microscopic proportions of the sensor organism to penetrate even the narrowest orifices and at the same time provide a direct measure of biological dose exposure. It may also be possible to create coupled MEMS, bio-microelectronic, or robotic sensor devices to provide both real-time and location-specific measurement of localized contaminants.
Research to integrate microelectronics and nanotechnology (Cunningham, 1998) with elements of gene array technology (Mecklenburg et al., 2000) and cellular engineering may lead to new sensor technology. Bionanosensors are recognized as a key research area in the National Science Foundation's first call for proposals related to the National Nanotechnology Initiative. 5 Such technology could create new capacity for continuous and remote monitoring in chemically and physically complex environmental and structural systems characteristic of DOE's site D&D needs.
Profiling Construction Materials
The committee recommends basic research leading to development of real-time and minimally invasive methods to characterize radionuclides and EPA-listed substances as a function of depth in construction materials, especially concrete.
Current Status
Concrete constitutes most of the volume and weight (estimated at over 27 million tons) of DOE's surplus facilities. Because of long-term exposure, the concrete is often contaminated to a depth of several millimeters beneath its surface (DOE, 2000b), and in some cases, such as for tritium, considerably deeper. Generally during D&D, concrete floors, walls, and ceilings are first characterized and mapped by physically removing samples and sending them to a laboratory for analysis. Once the degree of contamination is determined, an effort is made either to decontaminate the affected portions or to remove near-surface material as shown on the cover of this report. The development of minimally and non-invasive real-time in situ sensing technologies to characterize the concentration of contaminants, as a function of depth within concrete, could eliminate difficulties associated with core sample collection and subsequent analysis and greatly reduce characterization costs.
5http://www.nano.gov/press.htm.
Page 56
There are presently no real-time non-invasive means available to adequately determine the concentration depth profile of contaminants in concrete, hence there are good opportunities for research in this area. Gamma-ray spectroscopy has been used with limited success for selected isotopes but is not applicable for all radionuclides of relevance. X-ray fluorescence is limited to the measurement of surface contaminants.
Opportunities
There is need and opportunity for research to extend currently available or develop new surface analysis methods to provide a full three-dimensional picture of contaminants as a function of depth in construction materials. Research to extend laser ablation mass spectroscopy (LA-MS) along the lines described under characterization of surfaces is one example (Van Hecke and Karukstis, 1998). Research toward MEMS or bioorganisms that could tunnel into construction materials and provide contaminant profiles may offer significant potential. Contaminant profiling is an important need for D&D for which there appears to be little current applicable research.
Remote Mapping
The committee recommends basic research leading to the development of methods for remotely mapping radionuclides and EPA-listed substances.
Current Status
Remote-sensing systems can provide both economic and safety benefits by distancing the worker from hazardous work areas. The EM Office of Science and Technology (OST) and its deactivation and decommissioning focus area (DDFA) have recognized this advantage. Remote mapping of activity levels using gamma cameras is a recent innovation that has proven useful in D&D operations. These instruments display the relative strength and location of gamma radiation as a two-dimensional image superimposed on the corresponding visual image (DOE, 1998c, 1998d). However, there are still significant gaps in this technology because it is not able to survey areas with low levels of radiation (e.g., for release surveys). Nor are the current systems able to identify specific radioactive isotopes. Isotopes that primarily emit beta particles (e.g., tritium, technetium-99) or alpha particles (e.g., plutonium-239) are generally much harder to detect and quantify at low levels than are gamma-emitting isotopes (DOE, 1999a).
Page 57
Limitations of the current technology stem from the use of scintillator materials in the detector head that lack the ability to distinguish between gamma rays of different energies and, consequently, specific radionuclides. Currently available solid-state detectors, on the other hand, can distinguish energies but suffer other drawbacks, including poor sensitivity and the need for cryogenic cooling.
Opportunities
The committee believes that current technology for remote mapping can be significantly advanced by research in new detector materials, fiber optics, and fiber optic-based dosimeters.
Detector materials
There are research opportunities to discover or synthesize new detector materials to overcome current limitations (Squillante and Shah, 1995; Entine and Squillante, 1999). In general, a good material must be readily available in high-quality crystal form and be compatible with electronic device fabrication procedures. In addition, the material should strike an adequate balance between physical properties such as charge carrier mobility, carrier lifetime, resistivity, resistance to radiation damage, and stopping power, all of which contribute to detection sensitivity. Recent advances are promising. Room temperature detection has been reported using silicon (Si), cadmium telluride (CdTe), and cadmium zinc telluride (CZT). Other new materials such as GaAs, TlBr, and Pbl have attracted attention.
Device geometry may also be altered to improve performance. For example, relatively large output pulses are produced with an improved signal-to-noise ratio when a p-n junction in a silicon wafer is used in an avalanche-photodiode geometry. Silicon drift detectors with concentric-ring geometry have inherently low capacitance, also improving performance significantly. In principle, enhanced radiation absorption and sensitivity may be obtained through the use of quantum-confined nanostructured detection materials containing quantum wells or dots. These composites are made from existing semiconductor materials, like GaAs, that are already extensively used in electronics but have properties that differ dramatically from their bulk crystalline counterparts. Their characteristics, especially their electronic and magnetic properties, are quantum dominated. Such devices are already being quantum engineered for uses in optics, electronics, and magnetics. The committee believes that research opportunities exist in the development of new solid-state detector materials and geometries with significantly improved radiation-sensing properties, especially nanostructured materials with desirable quantum-engineered behavior.
Page 58
Fiber optic chemical sensors
Fiber optic sensors are another example of remote sensor technology in need of further research. Fiber optic chemical sensing has been explored (Lieberman, 1996) and probes to measure various chemical species in varying environments (e.g., air, water, soil, in vivo) have been developed for environmental sensing and medical applications. Most often a coating that has been formulated to selectively absorb the chemical species of interest is applied to the outside cladding of the fiber. As the coating selectively absorbs the species of interest, the light transmission through the fiber is reduced. Fiber optic chemical sensors are already employed at former nuclear weapons sites primarily for sub-surface soil characterization and water analysis. For example, fiber optic sensors have been used at the Hanford Site to monitor carbon tetrachloride in water at parts-per-million levels and at the Savannah River Site to measure trichloroethylene at parts-per-billion levels. 6 Further research is needed to develop coatings that are specific to other species of D&D interest.
Fiber optic radiation sensors
Remote sensing of radiation using optical fibers as the detector material and a conduit for transmitting accumulated dose information is a recent development (Henschel et al., 1994; Huston et al., 1996; Borgermans et al., 2000). Opportunities exist for both improved performance and such novel features as optical interrogation (Miller et al., 2000). Fiber optic radiation sensors would find considerable use in surveillance and monitoring, for example, in fiber optic pipe surveys, in cone penetrometers for assessing the migration of radioactive species in the soil around buildings, and in monitoring liquid tank leaks. Early versions of fiber optic sensors relied on damage and the resultant decrease in transmission (darkening) with accumulated dose that occurs in glasses exposed to intense radiation (Henschel et al., 1994; Evans et al., 1978). The composition of glasses can be modified using a variety of dopants to significantly alter the radiation sensitivity from fractions of a Gy to 105 Gy (Beuker and Haesing, 1994; Tomashuk et al., 1999). In practice, rapid fading of the radiation-induced darkening complicates the determination of the absorbed dose.
Fiber optic luminescent sensors have recently been developed based on optically transparent electron-trapping glass materials containing low concentrations of metal ions such as copper, cerium, or europium and clusters of nanocrystals such as ZnS or Cu2O (Justus et al., 1999a, 1999b). Radiation exposure results in trapping of electrons at defect
6http://www.srs.gov/general/srenviro/erd/technology/a06.html.
Page 59
centers within the glass matrix. Entrapment can persist for years at room temperature, so this approach has a low fade rate. The trapped electrons can be stimulated to recombine with the ionized centers by heating (thermal luminescence, TL) or by illumination with near-infrared light (optically stimulated luminescence, OSL). The recombination process results in a luminescence signal that is proportional to the absorbed radiation dose. Moreover, after TL or OSL, the fiber is refreshed and can be reused, a requirement for permanent installation in a remote monitoring system.
These materials have demonstrated greater sensitivity than TLD materials currently in use and are remarkably linear with accumulated dose (over seven orders of magnitude) while exhibiting low fade rate. The glass can be drawn into long lengths of fiber and fused directly to commercial optical fibers (Huston et al., 1996; Miller et al., 2000). A fiber sensor based on OSL has been fabricated that provides a sensitivity of 10−4 Gy per meter (Miller et al., 2000). The doped glass materials can also be used in a variety of non-fiber forms such as chips, rods, or powders for other radiation monitoring applications (e.g., personal badge dosimeters). Improvements in both materials and analysis methods are required to develop reliable, robust fiber optic dosimeters for DOE radiation monitoring activities.
Decontamination
Like characterization, decontamination of equipment and facilities is necessary at most stages of the D&D process (see Table 2.2). Initially, radiation and contamination levels may have to be reduced to allow workers access or to limit their exposure to radiation and other hazards. Decontamination may be required before dismantling or demolition work to prevent the spread of radioactive or toxic materials, which can have adverse offsite as well as onsite consequences. Decontamination procedures are intended to result in a small volume of the most hazardous waste, and much larger volumes of waste that has low or no hazard, thus reducing the cost and long-term risk of disposal. Some decontaminated equipment or facilities might be recycled or reused. The end state of any decontamination activity must be consistent with both site-specific and overall DOE cleanup objectives.
For a D&D project, considerations of cost, schedule, and worker safety will lead to an optimization of how much time and effort should be put into decontamination efforts versus simple waste conditioning and disposal. For example, at present, bulk waste materials that are decontaminated to USNRC Class C or below can be disposed of at rela-
Page 60
lively low cost in disposal facilities on DOE sites, whereas materials greater than Class C will go to a special repository. There are significant uncertainties in future waste disposal costs, however, which provide an incentive for research that may make decontamination faster and more effective.
For large-scale applications, almost all current decontamination methods are time consuming, involve risks to workers, produce significant volumes of secondary waste, and often leave residual contamination, especially actinide contamination. They usually require direct, hands-on work such as the concrete spalling work shown on the front cover of this report. Other available methods include wiping the surface with cleaners (e.g., detergents, acids, complexants), washing with high-pressure water, immersing objects in various cleaners, and electro-polishing (which removes a thin layer of the surface of metals). Two other methods in use are blasting with solid CO2 or sodium carbonate.
With current technologies D&D contractors usually choose to send large amounts of contaminated materials (e.g., concrete and steel) to licensed disposal facilities rather than attempt to decontaminate them for possible reuse. However, even for concrete, a relatively cheap raw material, recycling can be economical (Parker et al., 1998). Presentations to the committee indicated a need to improve current technologies for removal of radionuclides and EPA-listed organics and metals from equipment and building structures and metal, concrete, and wood debris. In many instances, paints, sealers, and varnishes create a laminate problem, with aged materials being harder to decontaminate than more recent deposition. Deep penetration of contaminants into porous structural material, such as concrete, also makes decontamination difficult.
Fundamental Interactions and Modeling
The committee recommends basic research toward fundamental understanding of the chemical and physical interactions of important contaminants with the primary materials of interest in D&D projects, including concrete, stainless steel, paints, and strippable coatings. Results should be used to develop first-principle models that describe the interactions and can be used to investigate improved approaches to decontamination.
Current Status
While there exists a good deal of chemical data on the contaminants themselves (Delany and Lundeen, 1990) and on their transport in
Page 61
the environment (van der Lee et al., 1997), there is little information of direct relevance to D&D problems. Both radioactive and toxic contaminants are known to exist in a variety of chemical forms (e.g., in different valence states, complexes, or as colloids) that exhibit very different behaviors.
Modeling of radionuclide and chemical contaminant behavior that is relevant to D&D problems is almost nonexistent. Conversely, there has been extensive modeling work in other DOE problem areas, such as subsurface contamination and high-level waste disposal (NRC, 2000a, 2001a). Available models are not adequate for developing improved decontamination processes. For example, surface oxides are known to sorb metal ions. The sorption has been described by a wide variety of models. Most of these models are based on measurements taken under a specified set of conditions (e.g., sorption isotherms) rather than on fundamental parameters. They are not generally applicable to the variety of conditions encountered in decontamination activities. In most studies neither the surface nor the metal ion is explicitly treated. Often, the role of the chemical form of the contaminants (speciation) is neglected.
Present decontamination approaches are usually based on experience or trial and error rather than quantitative prediction of how the contaminants are bound to construction materials and how chemical or physical methods can best remove them. Currently trial and error is often the only game in town because the history of how a facility was contaminated is unknown (see Chapter 2) and available methods to characterize the contamination are not adequate.
Actinide-contaminated materials are a major problem at many DOE sites. Contaminated materials include glove boxes, shielded cell liners, concrete, lead bricks, lead glass, and plastics. Radioactively contaminated lead, which is also chemically toxic, is a particular challenge. The DOE has a large inventory of contaminated lead due to its use as shielding material. The ability to efficiently remove actinides from the surface of construction materials will allow recycling or cheaper disposal of the material.
Opportunities
Scientific understanding of the interactions among contaminants and construction materials is fundamental to developing more effective D&D technologies. Such information includes how contaminants bind to steel and concrete surfaces; how they penetrate into these materials; their migration into pores, fissures, and welds; and time-dependent aging effects (Dzombak and Morel, 1990).
Decontamination studies should focus on a few fundamental parameters and interactions that can provide useful data for developing
Page 62
new decontamination methods. Investigations on radionuclides particular to the DOE, such as actinides, should be stressed. A variety of conditions (pH, temperature, ionic strength) should be examined. The interactions should be kinetically and thermodynamically described to facilitate applying the data to a variety of decontamination conditions and to ease the incorporation of data into first-principle models.
Modeling from first principles provides an opportunity to integrate the results of fundamental research in both chemical interactions and biological processes that are relevant to D&D problems. Such models can be the first step in bringing new knowledge to bear on improving decontamination approaches and processes. Properly integrating chemical and radionuclide speciation into D&D models is likely to be especially informative—adding new knowledge in general—since the most important species will likely be different from those in high-level waste or in subsurface contamination because of their different chemical environments. Beyond their use in decontamination the models can help provide a more general scientific basis for predicting behavior of contaminants in construction materials as a scientific underpinning of facility end states, which are described in the final recommendation in this chapter.
Biological Processes
The committee recommends basic research on biotechnological means to remove contaminants from surfaces and from within porous materials found in surplus DOE facilities.
Current status
The capacity of microbiological processes to destroy, transform, mobilize, and sequester toxins, pollutants, and contaminants is well established (Young and Cerniglia, 1995). Microbially produced surfactants and bioemulsifiers have been examined as replacements for organic chemicals in surface cleaning, stripping, and remediation. Microbial processes have been investigated for coal and tar liquefaction. A variety of microbial process activities are documented in the degradation, transformation, and immobilization of organic and inorganic chemicals and radionuclides. Bioleaching is a well-developed technology that has long been practiced for metal ore processing and has been developed for valuable metals recovery. Recent work at the Idaho National Engineering and Environmental Laboratory has employed bacteria for surface decontamination studies (Rogers et al., 1997). DOE has recognized the potential value of biotechnical remediation methods and supports research
Page 63
through its Natural and Accelerated Bioremediation Research (NABIR) and Subsurface Science programs as well as the EMSP (DOE, 1995).
Opportunities
The committee believes that the rapidly growing field of biotechnology affords many opportunities for research that is relevant to DOE's facility decontamination needs. In addition, a fundamental understanding of the biological processes would help to ensure that waste byproducts from the decontamination could be minimized, safely treated, and stabilized. Other research opportunities lie in determining effective ways to apply biological agents. For example, can high pressure be used to infuse bacteria or required nutrients into and through a contaminated concrete matrix? Biological treatment of in-place structural materials may differ from treating debris rubble. Biotechnical research areas that the committee recommends for emphasis include bioleaching, biosurfactants, and biocatalysts. A few examples of specific research opportunities are also given.
Bioleaching
Research to extend well-known technology in mineral ore leaching and metal recovery (Torma, 1988; Ehrlich and Brierley, 1990) can provide biochemical capacities for removal of metals and radionuclides from construction materials. Applications for contaminated concrete could be profound if volumetric reductions were achieved.
Bioleaching generally uses a Thiobacillus-driven production of H2SO4 and a Fe++ to Fe+++ couple to create a leaching solution. Numerous metals, including uranium, can be recovered commercially. Metal-rich ore differs significantly from concrete because the pH of the ore leaching solution is relatively easily maintained at pH <2. Concrete, being highly alkaline and buffered, poses difficulty for establishment of sulfuroxidation-driven bioleaching, but alkali-tolerant strains of Thiobacillus have demonstrated potential. Further exploitation of such organisms and nitrification-driven nitric acid concrete leaching is warranted.
Biosurfactants
There are opportunities for research that could lead to the use of microorganisms in aqueous or surfactant mixtures (with or without solvents or other surface-tension-lowering agents) for surface treatment and contaminant removal. Analogous technology has been developed for problems of soil bioremediation. In application for structural materials and debris it may be anticipated that previous problems (such as loss of solvents and surfactants due to mass loadings on soils or poor
Page 64
recovery and recirculation) can be avoided due to the different nature of the contamination matrix. In general, the area of biosurfactants and emulsifiers is under-investigated and could have broad implications not only for the EMSP but also for defense and industry (Sullivan, 1998; Banat et al., 2000).
Biocatalysis
Uranium can be microbially mobilized to the hexavalent form aerobically and can be reduced and precipitated (even from tridentate organic complexes) as tetravalent U by sulfate-reducing bacterial process technology. Other radionuclides and metals can likewise be precipitated as sulfides. Similarly, a variety of hazardous contaminants (e.g., PCBs, TCE, and perchloroethylene) can be independently or cooperatively degraded by joint aerobic and anaerobic bacterial processes with or without synergistic action of physical-chemical treatment, such as photocatalytic oxidation, or Fe++-driven reduction. Process technology has been created that couples such treatments with surfactant or solvent flushing and washing. While originally developed for hazardous waste soil remediation, technologically these processes may be well suited, perhaps better suited, for more defined applications in decontamination.
Examples
Research in the following areas, in the committee's opinion, is most likely to lead to relevant new knowledge and technologies for DOE facility decontamination problems.
-
selection, isolation, or biological engineering of acid, alkali, and solvent-optimized bacteria for focused application in decontamination;
-
cell-free enzyme development for decontamination, including gel-polymer immobilization matrices;
-
improved kinetics for aerobic and anaerobic biocatalytic decontamination;
-
improved resistance to chemical matrix toxicity; and
-
microbial (bacterial and fungal) production of extracellular chemicals and enzymes for application in surface cleaning and delamination.
Robotics and Intelligent Machines
DOE has recognized the potential for robotics and intelligent machines (RIM) to meet some of its greatest challenges, including
Page 65
-
lowering the cost of operations in the face of budgetary pressures,
-
meeting the rising regulatory standards and continuing to improve the health and safety of its workers, and
-
carrying out necessary operations that are too hazardous for humans to perform.
Because RIM crosscuts almost all of DOE's programs, DOE has laid out long-range plans for developing this technology in a Robotics and Intelligent Machines Roadmap (Sandia, 1998). Technology roadmaps are planning documents that DOE uses to call attention to future needs for development in technology, provide a structure for organizing technology forecasts and programs in order to avoid gaps or overlaps, and communicate needs and opportunities throughout the R&D community. For EM activities, the roadmap lays out ambitious goals for RIM, which include
-
increasing productivity by 300 percent,
-
reducing personnel exposure by 90 percent, and
-
reducing secondary waste by 75 percent.
The DDFA estimates that 30 percent of its needs include RIM requirements (Haley, 2000). Facility D&D presents workplace hazards that are unique among EM's challenges. Most D&D baseline technologies require that workers routinely enter areas with radiation and many other industrial safety hazards and perform hands-on work with powerful and heavy equipment, including cutting devices that can instantly penetrate protective clothing. This routine work includes sampling (for characterization), decontaminating, and eventually dismantling the barriers that were originally constructed to protect workers from radioactivity and toxic chemicals.
As discussed in Chapter 2, DOE facilities are massive, they are often crowded with complex, heavy equipment, and in many cases details of how the equipment was designed and operated have been lost. The reality of nuclear facility D&D is that the physical tasks are unstructured (not repetititve) and involve a wide variety of highly contaminated components (e.g., piping, valves, wiring, tanks). When performed directly by human workers this work represents a significant safety risk and a high cost in resources and time (see Appendix D). The first goal of remote systems technology is to remove the workers from harm's way, which dramatically improves safety. The second goal is to increase productivity and reduce costs and project schedules—all of which would make the D&D enterprise more manageable. Considering the present timeline for this D&D work, there is at least a decade available during which an accelerated science-based development program could be pursued to revolutionize the technology to meet the goals of the RIM roadmap.
Page 66
Intelligent and Adaptable Remote Systems
The committee recommends basic research toward creating intelligent remote systems that can adapt to a variety of tasks and be readily assembled from standardized modules, with special emphasis on actuators, universal operational software, and virtual presence.
Current Status
DOE's RIM roadmap lays out needs and plans to greatly advance current technologies, but the envisioned leaps in technology are not likely to occur without new knowledge. Even though industrial robot systems have reached a level of maturity providing high availability at low cost, they are not well suited to the complex and unpredictable demands of the D&D task environment. D&D has special needs primarily because it is unstructured, requiring continuous kinesthetic input and supervisory oversight, and also because the loads and motion ranges are very large and impact loading can be very severe.
Current technology available for D&D is best represented by the Chicago Pile 5 (CP-5) large-scale demonstration of a dual arm robotic system (see Table 3.2). This system used an older hydraulic control system with nominal software. Simple tasks were performed, maintenance and downtime were extensive, the rate that the work was accomplished was low, and resulting confidence in the technology was low (NRC, 1998a). The system did perform the overall dismantlement function, but it also raised the question of efficiency (cost and time). Most efforts to improve the technology have resulted in one-of-a-kind demonstrators (not based on a full architecture with components enhanced by a continuous science effort), which are hard to certify as to their real performance.
Opportunities
Research opportunities exist primarily in making RIM more human-like in their abilities to adapt to a variety of tasks, both physically and intellectually. In this regard, the committee believes that the best opportunities for research relevant to D&D tasks involve actuators, universal criteria-based software, and virtual presence.
Actuators
The actuator is the power (muscle) of remote systems and as such it is the key to performance, reliability, and cost. Except for better materials and improved control electronics, most actuator technology has not
Page 67
changed for several decades. Today's actuators typically use only one sensor (for position) so that virtually no real-time data (e.g., force and velocity) are available to make them “intelligent.” More complete sensory input coupled with decision-making software (see below) can produce intelligent actuators that are able to adapt to a variety of tasks. Achieving a relatively inexpensive modular design to allow plug-and-play deployment of these devices would be especially useful in D&D projects, because equipment that fails or becomes contaminated is usually discarded. Research to answer the question of granularity (What is the minimum number of required standard modules?) to enable the assembly on demand of the maximum number of remote systems would make the overall system substantially more cost effective in deployment and maintenance.
Universal criteria-based software
Criteria-based decision making is the essence of intelligence in robotic systems. Today's control of robotic devices is derived from techniques developed during World War II in which control is linear (based only on the difference between two measured parameters). A robot capable of mimicking human adaptability, however, would require a non-linear control system in terms of many highly coupled parameters corresponding to the physical features that accurately represent performance of the task. The criteria-based software could be universal in the same sense that operating systems on microcomputers are universal— one system supports many different applications (Miller and Lennox, 1991; Sturzenbecker, 1991; Stewart et al., 1992).
Virtual presence
In the initial planning and characterization phases of D&D work, workers often must enter an area of high radiation and contamination that is also congested with left-in-place equipment and materials for which removal inevitably involves physical stress (fatigue) and the potential for personal injury (Fournier et al., 1998). Virtual reality systems could allow workers to perform essential survey and decision-making functions from a remote location (Lloyd et al., 1999), thus enhancing their safety and productivity. Advances in the state of the art as, for example, in deep sea exploration could improve overall system performance by providing force feedback, remote vision, collision avoidance, and radiation-resistant sensor technology.
Summary
There are research opportunities to develop basic scientific understanding to support the new actuator, software, and virtual presence
Page 68
technologies that the committee believes are the keys to achieving DOE's objectives for RIM. To effectively organize this research the committee believes there is need and opportunity to establish a small number of technology demonstrations at the DOE laboratories to convincingly document measured performance of the present state of the art. This would be technology that exists or could be aggregated in the next two to three years (not one-of-a-kind systems or extensions of industrial robot technology). These demonstrations would not only show where investment in research is needed but would also provide a baseline to measure progress.
The RIM roadmap contains a list of applications and a tabulation of the desired technology (Sandia, 1998). Some of the most significant science opportunities to achieve these technology goals are listed below. A more complete discussion is given in Appendix D.
-
Sensors for site characterization and acquisition of performance data are essential to support decision making either through software or by visualization and human judgment.
-
A special need is the kinesthetic interface to human operators to enhance their motor skills and input commands to the remote system.
-
Mobile platforms that in themselves are modular and highly dexterous must be further developed to gain access to the work environment and to transport size-reduced facility components.
-
Quick-change end-effector tools based on a science of tools (design, modeling, and operation) are needed to perform the incontact physical tasks for D&D.
-
Dexterous high-load robot manipulators capable of tool management, size reduction, parts transport, and parts packaging under human supervision will be especially important for D&D tasks.
-
Intelligent and standardized actuator modules to build all D&D remote systems on demand from a minimum number of high-performance and low-cost modules just as we now build computers on demand.
-
A universal operating software for any intelligent machine used for D&D. Similar to the operating systems in today's micro-computers, this software would support operation of mobile platforms, gantries, small automation subsystems, and dexterous manipulators, all under the centralized supervision of operators in a remote position.
-
Electronics will be pervasive in a modern remote systems technology. Hence, their hardening against radiation, temperature, shock, and particulates is necessary.
Page 69
End States
The purpose of facility D&D is to leave a building in an agreed-upon condition, referred to as the end state. As discussed in Chapter 2, end states are established though input from many involved people and organizations through considerable negotiation. End states for facility D&D are usually a part of the goals for overall cleanup of a site. These goals can be general, such as to protect the Columbia River at Hanford, or more specific, such as to re-industrialize parts of the Oak Ridge Site. At Rocky Flats most buildings will be demolished and the debris removed from the site. After completing its site visits and in the continuing deliberations after its interim report, the committee agreed that there is need for a better scientific basis for evaluating the acceptability of alternative D&D end states.
Fate and Behavior of Contaminants in Construction Materials
Research should be directed toward understanding the fate and behavior of treated and untreated contaminated material by determining the fundamental chemical species of the contaminants in the host material and how the species behave. The effect of time and changing ambient conditions should be considered in these investigations. Further research should be directed at incorporating these results into risk assessments to evaluate and compare long-term safety that can be provided by different end-state options.
Current Status
Considerable effort has been directed at scientific understanding of the fate and behavior of radioactive wastes in the environment (NRC, 2000a) and, to a lesser extent, of hazardous wastes (NRC, 1999b). There is very little scientific understanding of the fate and behavior of radionuclides or EPA-listed chemicals in construction materials. Understanding how contaminants may move through or out of concrete and metal is important for providing scientific input toward establishing the end state. From a research perspective, developing this scientific knowledge is clearly linked to the committee's recommendations in the areas of characterization and decontamination.
Some of DOE's largest facilities slated for D&D have potential future uses. Evaluating the different uses requires an understanding of the long-term behavior of the contaminated material. The Hanford Site pro-
Page 70
vides four examples: (1) After the end of operations for the N reactor, some buildings were returned to service for spent fuel storage (McDaniel, 2000); (2) the B reactor is being preserved as a museum; (3) building entombment for low level waste disposal is under consideration for former chemical separation plants (MacFarlan, 2000); and (4) there are plans to re-industrialize parts of the site (the 300 area).
Opportunities
Research is needed on understanding the physical and chemical forms (speciation) of contaminants in building construction materials. Buildings may be in storage for decades before D&D and potentially in use for decades thereafter. Understanding the speciation and behavior of the contaminants, how the speciation evolves with time, and the impact of decontamination activities on the chemical speciation is critical for developing a scientific basis for determining end states.
Decontamination often uses chemical or biological processes that can impact the behavior and performance of construction materials as well as the contaminants themselves. The use of chemicals or bacteria for decontamination can dramatically affect the local environment by changing pH or inducing chemical reduction or oxidation reactions through respiratory activity. This activity, coupled with physical changes due to material cutting, melting, or polishing from decontamination efforts, can impact the behavior of the contamination. For example, acids dissolve concrete; less dramatic reductions in pH can also have profound effects, but details of these changes over time and how they may affect the eventual release of contaminants are not well understood. Even if there are no decontamination activities the host material will change with time.
The behavior of hazardous airborne species (Egge, 2000) presents another opportunity for research. While water is the primary mover of contaminants in the ground (subsurface contamination), airborne pathways may be especially important in establishing a scientific basis for facility end states. Performance assessment modeling using fundamental fate and transport data could be developed as an important decision-making tool for establishing facility end states.
