The development of the atomic bomb during the Second World War was a stunning scientific and engineering achievement. Created by order of President Franklin Roosevelt in 1942, the Manhattan Project had produced by 1945 enough enriched uranium and plutonium for the August attacks on Hiroshima and Nagasaki that ended the Second World War. The sources of that enriched uranium and plutonium were the Oak Ridge site in Tennessee (then called Clinton Laboratories) and the Hanford Engineering Works in Washington, respectively—massive industrial complexes built from scratch to spearhead the development effort. After the war, these sites were expanded and additional sites established to meet the nation's growing demands for nuclear materials for the Cold War arms race with the Soviet Union.

click for full image and caption

The effort to build the bomb and win the arms race exacted a steep environmental price. Substantial quantities of radioactive and chemical contaminants were released to the environment, and even today large quantities of radioactive and toxic waste remain in storage at several sites. While it was being produced, few provisions were made for "solving" the problem of waste—that is, putting the waste into a stable form and placing it out of reach of people and the environment—and the passage of time has since exacerbated the difficulties. With institutional memories fading and facilities aging, management of this legacy has become a more difficult, expensive and perilous undertaking.

The federal government, under the auspices of the U.S. Department of Energy (DOE), is making large expenditures of taxpayer funds to address the environmental legacy of the arms race. But after almost 13 years of effort and outlays of over $70 billion, the goal of "cleanup" is proving elusive. Through the work of independent expert committees appointed by the National Research Council, we have been following the DOE cleanup program for many years. Here, in a personal and unofficial assessment of what has been learned, we examine that program's efforts to come to terms with the environmental consequences of weapons production, including its technical and societal dimensions. Our examination suggests that it is time for the cleanup program to redefine success based on reducing and managing the human and environmental health risks that will extend far into the future.

Nuclear-Weapons Production

click for full image and caption

The U.S. nuclear weapons complex is massive in scale and highly dispersed: some 5,000 facilities located at 16 major sites and more than 100 smaller sites, ranging from mills to recover uranium from mined ore to facilities for weapons assembly, maintenance and testing. The largest facilities in the complex are those built for materials production and processing—Hanford, the Idaho National Engineering and Environmental Laboratory (INEEL), Oak Ridge and Savannah River in South Carolina. At these locations, nuclear materials (enriched uranium, plutonium and tritium) were produced for use in weapons, naval fuel and civilian nuclear applications. Their massive scale was dictated by the need for secrecy and safety as well as the physics of nuclear-materials production.

Manhattan Project scientists believed that an atomic weapon could be constructed using either of two radioactive isotopes, uranium-235 or plutonium-239. But obtaining sufficient quantities of these materials (tens of kilograms per bomb) proved a technically difficult and expensive challenge. Uranium-235, the principal isotope of a uranium weapon such as that used at Hiroshima, makes up about 0.7 percent by mass of natural uranium, most of the remainder being uranium-238. To be usable in a fission weapon, uranium-235 must be concentrated (enriched) to greater than 20 percent and preferably, for greatest efficiency, to over 90 percent. Uranium enriched to 20 percent or more uranium-235 is "highly enriched uranium" (HEU).

Uranium enrichment during the Manhattan Project exploited the small mass differences between uranium-235 and uranium-238 using electromagnetic separation and gaseous diffusion—the latter a process still used in the United States to enrich uranium for civilian power plants. Thousands of separation stages were required to obtain sufficient uranium-235 enrichment, and the facilities housing these processes were scaled to yield the required quantities.

Two plants were built during the Manhattan Project to obtain the roughly 50 kilograms of enriched uranium for the "Little Boy" weapon that was dropped on Hiroshima: the K-25 building at the Oak Ridge Gaseous Diffusion Plant and the Y-12 electromagnetic-separation plant at Oak Ridge. After the war, the Atomic Energy Commission (AEC) expanded the gaseous diffusion plant and built two additional enrichment facilities, the Paducah plant in Kentucky and the Portsmouth plant in Ohio, to meet growing civilian demands.

Plutonium-239, the principal isotope in "weapons-grade" plutonium, was produced by irradiating uranium-238 fuel rods with neutrons in large production reactors built along the Columbia River at the Hanford site. The capture of a neutron by uranium-238 produces uranium-239, which subsequently decays to neptunium-239 and then plutonium-239.

Once a plutonium-239 atom is produced, additional neutron captures can cause it to fission or produce heavier "reactor grade" plutonium (primarily plutonium-240 and -241), which is not as suitable for weapons. To minimize additional captures, the Hanford operators removed the fuel rods from the reactor after a short irradiation time, but this required large throughputs of uranium to obtain the needed quantities of plutonium. Three reactors and two chemical processing plants were built at Hanford to produce plutonium for the war effort. Following the war, the AEC built six additional reactors and three chemical processing plants at Hanford and five reactors and two chemical processing plants at Savannah River to meet the growing plutonium demands.

Tritium, an isotope of hydrogen and a key component of fusion weapons ("hydrogen bombs"), which were first developed by the United States in the early 1950s, was produced by the irradiation of lithium-6 targets in the production reactors at Savannah River. An isotope-separation facility was constructed at the Y-12 site at Oak Ridge to produce lithium-6 for this purpose.

During the roughly 45 years of nuclear-materials production in this country, about 103 metric tons of weapons-grade plutonium were obtained from the production reactors at Hanford and Savannah River; 994 metric tons of highly enriched uranium were obtained from the enrichment plants at Oak Ridge, Portsmouth and Paducah. Some of these materials have been declared to be surplus to U.S. defense needs (see table). Plans are now in place to turn most of this excess into fuel for use in nuclear power plants. The spent fuel will be disposed of in a geologic repository. Dealing with the by-products of nuclear-materials production is another matter.

Environmental Consequences

click for full image and caption

Although the production of nuclear materials generated huge quantities of waste, good records of radioactive and chemical waste production and environmental discharges generally were not kept until the 1970s. What is known of that early history today is based on reviews of written records supplemented by process knowledge and mass-balance calculations. We have been selective in our use of data in the following discussion, preferring more recent sources that contain documentation for the estimates. We also have rounded the data except where there is demonstrated support for greater precision.

The uranium-enrichment process produced low-level solid and liquid wastes and other process liquids, and up to 200 kilograms of depleted uranium (enriched in uranium-238) for every kilogram of HEU. There are on the order of half a million metric tons (metal equivalent) of depleted uranium in storage at several sites, and although this material is not classified as waste, most of it has no agreed-upon disposition pathway.

The separation of lithium-6 for tritium production used on the order of 10,000 metric tons of mercury, of which about 900 metric tons is unaccounted for. The DOE estimates that about 110 metric tons was discharged into East Fork Poplar Creek at Oak Ridge, and some of this contamination has migrated offsite and into the Clinch River–Watts Bar Reservoir system that is used for recreation and municipal water supply. The inorganic mercury compounds in this waste are not thought to be toxic, but they can pose a hazard to human beings if transformed to methylmercury by soil and water microorganisms.

Plutonium production also produced large quantities of waste: During the 50 years of operation of the Hanford site, for example, about 67 metric tons of plutonium were produced from almost 97,000 metric tons of irradiated uranium fuel. Chemical processing of that uranium to recover plutonium produced some 2 million cubic meters (500 million gallons) of highly radioactive, chemically toxic waste, and another 1.7 billion cubic meters (450 billion gallons) of process liquids. The DOE reports that about 76,000 cubic meters of solid waste contaminated with actinides (principally plutonium) and an additional 1.2 million cubic meters of other solid low-level wastes have been buried at the site.

Some of the waste generated by nuclear materials production was released directly to the environment. Volatile gases from chemical processing were vented directly into the atmosphere, sometimes without filtering, especially during the early years of production. Reactor cooling water contaminated with conditioning chemicals such as chromium and with radioactive isotopes produced in the reactor (neutron-activation products) were also discharged. Waste liquids were discharged into large surface ponds or into subsurface soil and groundwater through injection wells and other drainage structures. Radioactive and chemically contaminated solid waste was burned or dumped into shallow pits and trenches.

Indeed, there are thousands of "release sites" that are current or potential future sources of contaminant releases to the environment. As a result of such releases, soil and groundwater at many sites are extensively contaminated with industrial solvents, toxic chemicals, metals and radionuclides.

click for full image and caption

Large volumes of waste remain in storage at several sites and could become significant sources of future environmental contamination if not managed properly. At Hanford, for example, there are about 200,000 cubic meters of high-level waste in storage in 177 large underground tanks. These tanks have been in service between 16 and 58 years; under current plans, the last tank will not be closed until about 2046. The older, single-containment tanks were designed with service lives of 10 to 20 years, although no one really knew how long they would last. One tank began leaking just six years after it was put into service, and to date 67 of these tanks are suspected to have leaked up to 5,700 cubic meters, or 1.5 million gallons, and possibly more than a million curies of high-level waste into the subsurface. (Curies are a measure of radioactivity in a material; for comparison, a ton of uranium-238 has 0.3 curies.) Some of this contamination has reached groundwater.

At Savannah River, there exist some 130,000 cubic meters (34 million gallons) of high-level waste stored in 48 underground tanks. Nine of the tanks have leaked waste into their secondary containments, and a few tens of liters of waste leaked into the environment from one tank when the secondary containment overflowed. Efforts are now under way to immobilize the sludge fraction of this waste in a borosilicate glass matrix.

The Idaho site processed naval spent fuel and some research reactor fuel to recover enriched uranium, but here, unlike at Hanford and Savannah River, the high-level waste was immobilized as a powdered ceramic (calcine), about 4,000 cubic meters of which are being stored in stainless steel bin sets inside steel-reinforced concrete silos. These structures were designed to contain the waste for up to 500 years. Additionally, another 4,000 cubic meters of so-called "sodium-bearing waste" liquids await disposition in some of the site's 11 underground storage tanks.

Past practices for managing the large volumes of waste generated by nuclear materials production, when judged by today's standards, appear ill-informed at best, bordering on reckless at worst. It is important, however, to judge these practices against the prevailing environmental attitudes and practices during the Second World War and Cold War. The Manhattan Project was created during a national emergency at a time when the future of Europe and Asia hung in the military balance. National priority was given to weapons production at the expense of waste management. This sense of urgency, and a shroud of secrecy that hid production activities from public view, carried over into the Cold War, although increasing effort was given to minimizing environmental releases as time went on.

click for full image and caption

Wartime shortages of materials such as stainless steel created further difficulties. Carbon steel was employed to construct the waste tanks at Hanford and Savannah River, with the result that the high-level waste, which was highly acidic, had to be neutralized with alkaline chemicals such as sodium hydroxide to reduce tank corrosion. The addition of these chemicals to the waste increased volumes and produced solid precipitates. Later chemical processing and volume-reduction operations to reduce radioactive heat generation and conserve tank space further increased physical and chemical heterogeneity. Characterization of this waste and removal of the precipitates from the tanks will be difficult and expensive, especially at Hanford.

The waste-management decisions made during the Manhattan Project and ensuing Cold War created the environmental problems that the nation now confronts. These decisions continue to exact a steep price, both in the high annual costs of managing the stored waste and environmental contamination, and also in the loss of trust by citizens in their government as the consequences of waste-management practices carried out in secrecy for almost five decades have become public knowledge.

From Production to "Cleanup"

The decline of large-scale nuclear-weapons production began in the late 1970s and accelerated through the 1980s, coinciding with the thaw in Cold War relations that culminated in the Strategic Arms Reduction Treaty (START) and the breakup of the Soviet Union, both in 1991. At the same time, the reactor accidents at Three Mile Island in 1979 and Chernobyl in 1986 raised public concerns about the continuing operations of U.S. production reactors. In May 1986, Energy Secretary John Harrington asked the National Academy of Sciences and National Academy of Engineering to review the safety of the government's production and research reactors. He also commissioned a group of experts to review the operation of the N-Reactor at Hanford. Based on that review, he shut down the reactor in 1987, commenting that the United States had no need for it because the country was "awash in plutonium."

During this same period states also were beginning to assert their authority to regulate environmental releases at the sites, prompted by a 1984 federal court ruling that the Y-12 site at Oak Ridge was subject to state regulation under the Resource Conservation and Recovery Act. Complaints from Colorado led to the June 1989 Federal Bureau of Investigation raid and closure of the Rocky Flats site, a 1951-vintage weapons-component manufacturing facility near Denver, for violations of federal environmental laws. Five months later, Energy Secretary James Watkins announced the creation of the Office of Environmental Restoration and Waste Management (now the Office of Environmental Management) and declared a new mission for weapons sites: environmental cleanup. The era of large-scale nuclear-weapons production had ended.

The new cleanup program contrasted, in many respects, with the production operations. From the earliest days of the Manhattan Project, weapons production had been conducted with scientific and technical rigor and a strong focus on meeting production goals that were noticeably lacking in the early years of the cleanup effort. One of the first actions taken by the new program, before it had developed an adequate understanding of the environmental insults at its sites or its scientific and technical capabilities to address them, was to negotiate legally enforceable cleanup agreements with states and regulators. Many of these original agreements had to be renegotiated after the problems were more fully understood. At present, the cleanup program is operating under some 70 separate agreements that contain more than 7,000 schedule milestones, many of which are potentially enforceable via court action.

Although the cleanup program has been in operation for over a decade, it has, until recently, accomplished relatively little actual cleanup. To be sure, DOE has had some important recent successes, both in site remediation and waste disposal. Perhaps its most notable waste disposal success was the 1999 opening of the Waste Isolation Pilot Plant near Carlsbad, New Mexico. This deep geologic repository will eventually be used to dispose of up to about 175,000 cubic meters of defense-generated transuranic waste (mainly plutonium-contaminated debris, clothing and tools, and the like) from nuclear weapons sites. Additionally, DOE has recommended Yucca Mountain, Nevada as the site for a deep geologic repository for spent fuel and high-level waste and is now in the process of developing an application for a construction license, which it plans to submit to the Nuclear Regulatory Commission in 2004. If constructed, this repository will be used to dispose of the immobilized high-level waste and spent fuel from nuclear weapons sites along with commercial spent fuel.

The notable remediation successes include the stabilization and capping of mill tailings piles and the cleanup of some Manhattan-era sites, the latter of which is presently being carried out by the Army Corps of Engineers. Also, successful efforts are being mounted at many sites to characterize the nature and extent of environmental contamination, halt the spread of contaminated groundwater, and cap waste burial sites to retard water infiltration and contaminant leakage. Work also is proceeding to decontaminate and demolish buildings and clean up contaminated soil and groundwater at some of the smaller sites (such as Fernald and Mound) so that they can be declared closed around 2006.

click for full image and caption

Perhaps the most significant technical success in the remediation program to date has been the construction and successful operation of a $2.5 billion plant at Savannah River for immobilizing high-level waste, which went into production in 1996 and has to date produced more than 1,200 canisters of borosilicate waste glass. At Hanford, work also has begun to cocoon the nine production reactors, remediate contaminated soil and groundwater along the Columbia River, and stabilize corroding spent fuel from the N-Reactor that has been stored for over decade in two unlined water basins next to the river, one of which is leaking. This fuel is being dried, canned and placed into temporary storage away from the river. With the notable exception of the immobilization program at Savannah River, however, none of these remediation actions has been technically demanding. In fact, attempts to undertake the technically demanding tasks have failed, due largely to inadequate scientific and technical understanding. Three examples serve to illustrate this point.

In the early 1990s, the Idaho laboratory began a project to excavate and treat waste and contaminated soil from a 1-acre site known as "Pit 9," one of a series of pits and trenches used for disposal of low-level and transuranic radioactive waste. Pit 9 is thought to contain about 7,000 cubic meters of sludge and other solids contaminated with plutonium from Rocky Flats, and the remediation effort was designed to demonstrate retrieval and processing technologies that could be applied elsewhere on the site. The DOE awarded a $200 million contract for this work in late 1994, but the project fell behind schedule, and costs exceeded the contract price before any waste had been retrieved or processed. The contract has been canceled, and the contractor has alleged that inadequate characterization of the waste in the pit contributed to this failure. Excavation of waste from this pit may not take place until 2004, fully a decade after the initial contract was awarded.

Efforts are now under way at Savannah River to develop a chemical process to remove radionuclides, principally cesium, strontium and plutonium, from the nonsludge fraction of its high-level waste for immobilization in glass. Savannah River contractors spent 10 years and almost $500 million to develop an in-tank precipitation process for removing cesium, but when this process was placed in production in one of the underground storage tanks, large quantities of benzene, an explosive hazard, were generated. Subsequent investigations and experiments failed to positively identify the benzene-generation mechanism, although a catalytic reaction involving trace elements in the waste was thought to be responsible.

DOE–funded scientists are now developing a solvent-extraction process that has a high selectivity for cesium. This process looks promising, but the schedule for waste retrieval and processing has been set back several years. The delay would likely have been much longer if not for the foresight of the department's research and development organizations, which funded research on alternative separation processes before the problems became evident.

There have been several attempts at Hanford, starting in the early 1990s, to begin retrieving and immobilizing high-level waste from its tanks using approaches similar to those at Savannah River. Construction of a facility to immobilize about 10 percent by volume and 25 percent by radioactivity of the liquid high-level waste finally began this year with the start of construction of a waste treatment and immobilization facility. This phase-1 project is slated to last until 2018 and cost about $15 billion. Hanford has not yet determined how it will process the remaining waste, or how it will retrieve the solid or semi-solid wastes from its single-containment tanks to meet the 99 percent removal milestone required by its compliance agreement with the State of Washington. Retrieval of this waste without damaging the tanks and releasing contaminants to the environment may be difficult using currently available technologies.

click for full image and caption

To be fair, the DOE has tried several times in recent years to improve the effectiveness of the cleanup effort. In 1995, the assistant secretary for environmental management announced a "10-year plan" for reducing the high annual carrying costs of the sites by accelerating the closure of smaller sites. Several sites, including the Mound (Ohio) and Rocky Flats sites, are now slated to be closed by 2006. The current administration is developing a plan to "accelerate cleanup" by focusing on risk reduction and negotiating with site regulators to shorten cleanup schedules. The objective is to reduce the current $220 billion to $300 billion estimated life-cycle cost of the cleanup program by $100 billion and 40 years.

Although these goals strike us as sensible, the success of this effort will hinge on several factors. Will regulators be willing to modify their compliance agreements with the DOE? Will state and local authorities and the site administrators themselves allow reallocation of budgets so that high-risk projects can be funded on an accelerated schedule—or conversely, will Congress allocate additional funds for this purpose? Will the DOE and contractors exercise good judgment in developing and applying remediation plans and, especially, learn from past experiences at the sites to avoid repeats of some of the problems described previously? Can DOE-funded investigators come up with timely solutions when new problems are identified, as they did for the cesium-separation problem at Savannah River?

Coming to Terms with "Cleanup"

The term "cleanup" poorly describes the current activities at DOE sites: Only a small portion of the approximately $7 billion in annual funding is actually used for contaminant removal and waste processing. Most of the budget is spent on site surveillance and maintenance. The cleanup program refers to these surveillance and maintenance costs as "mortgage costs."

In our view, these high mortgage costs are slowing work on high-risk problems that, if not addressed in a timely fashion, could lead to nasty future surprises. The slow progress in remediating the high-level tank wastes at Hanford and Savannah River is of particular concern. Many of the tanks are now well beyond their design lives and contain chemically complex and highly toxic waste, much of it in a liquid state. Some of the tanks are now leaking, and the number of "leakers" is likely to increase as the tanks age. Accidents, acts of God and terrorism are also concerns as long as the liquids remain in the tanks. Under current schedules, it will be several decades before all of this waste is recovered and immobilized, and some of the hardest work (such as retrieval of the "bottoms," rich in transuranic elements such as plutonium, from the single-containment tanks at Hanford) is being deferred until the later stages of the remediation effort.

The use of the term "cleanup" also suggests that the primary objective of the program is to remove waste and environmental contamination and return the sites to other productive uses. In fact, although some sites or parts of sites can be cleaned up and released for other uses, sometimes with few or no restrictions, the DOE has acknowledged that this will prove to be the exception rather than the rule and that parts of more than 100 sites are expected to be unacceptable for unrestricted release after cleanup. At many sites, and especially the large ones, contaminants are too widely dispersed in the environment to be recovered with current technologies. The stored wastes that exist at these sites can (and should) be processed to reduce volumes and stabilize the hazardous constituents, but after processing much of this waste will be reburied at the site. The hazards will be reduced or relocated, but not eliminated.

This fact is well recognized within the program, which defines cleanup as the completion of those actions necessary to meet agreed-upon standards and objectives, and not necessarily the removal of all waste and contamination. The expectations of regulators and local communities for achieving contaminant reduction and waste removal have been moderated since the cleanup agreements were signed as the technical difficulties and high costs of progress have become apparent. We sense, however, that expectations may still be higher than warranted in view of the difficult problems ahead, especially for the remediation of burial pits and trenches (such as Pit 9) and the retrieval and processing of high-level waste from the underground tanks, especially at the Hanford Site. Past success in site cleanup may not be a good harbinger of future prospects, because most of the difficult and costly problems have yet to be tackled.

Reducing and Managing Risk

Since its creation in 1989, the cleanup program has focused on developing and executing negotiated milestones to achieve specific cleanup tasks or levels of contaminant reduction, while at the same time (according to some critics) maintaining high levels of employment at sites that no longer have a national defense mission. It is becoming clear that many of these "activity-based" milestones may not be achievable with current technologies. Furthermore, the milestones are not designed around goals of protecting human and environmental health. If achieving such protection is the ultimate goal of the cleanup program, we believe that it may make more sense to organize major programmatic milestones around agreed-to levels of risk reduction without specifying in advance the specific remedial actions to be taken to achieve those reductions.

The judicious use of "risk-based" milestones could have several benefits. Such milestones could, for example, provide a better measure of progress and encourage the investment of funds where the greatest risk reductions could be achieved. They also could encourage greater creativity in the selection of "end states" for cleanup and the remedial actions to achieve them, creativity that is lacking in current activity-based milestone approaches.

Of course, the use of a risk-based approach requires that risk estimates be developed for site hazards. The cleanup program has had difficulty developing a risk-based analysis—the sites are complex, not all of the contaminants (groundwater plumes, for instance) have been located and characterized, nor is all of the waste adequately characterized. The cleanup program does not even use the "risk" concept consistently: Sometimes risk is defined based on effects on the health of off-site populations only, not including on-site workers, and other times risk is defined programmatically, that is, whether a particular action can be completed on time and within budget.

Long-Term Stewardship

As the title of this article suggests, a great deal of the environmental legacy of nuclear-weapons production may end up being managed, not eliminated. Many sites, or portions of sites, will not be remediated to levels deemed adequate for unrestricted access, and either the federal government or a state or local government will become landlords of last resort, with attendant responsibilities for protecting public and environmental health. In some cases this protection will come in the form of long-term surveillance to guard against human access or further environmental releases, and in other cases active measures such as groundwater treatment will be required. Some of these responsibilities may last indefinitely.

These long-term responsibilities have received little consideration by the cleanup program until recently, and even now "long-term stewardship" of contaminated sites is viewed as a separate activity from cleanup. Yet there is a very real trade-off between cleanup and stewardship—that is, protection against a hazard can be achieved either by eliminating it outright (through cleanup), managing it until it ceases to be hazardous (long-term stewardship) or a combination of both approaches. Over the short term, hazard management is usually less difficult and expensive than hazard elimination, but the long-term costs are not clear, and the effectiveness of long-term stewardship depends to a great extent on the continuing willingness and ability of future generations and institutions to manage the hazard, factors over which the current program has little or no control.

Given this trade-off between cleanup and stewardship, we suggest that both choices need to be put on the table when deciding on end states and remedial actions to achieve them, fully recognizing that a reliance on stewardship places a heavier burden on future generations. The use of risk-based cleanup approaches described earlier would help make these choices explicit.

There may be good reasons for relying on stewardship in some instances, especially if cleanup is not technically feasible or cost effective. Indeed, society routinely makes this choice for managing other kinds of waste, including chemically hazardous waste, although there is little or no evidence to demonstrate its effectiveness over multiple generations, and much evidence to the contrary.

Under current regulatory regimes, decisions to rely on long-term stewardship must be revisited periodically, and further actions to reduce hazards made if necessary and feasible. We believe that the ultimate success of long-term stewardship as a solution to the waste problem at DOE sites will hinge on advances in science, especially those elements of the social sciences that bear on the effective design and operation of durable institutions. There is reason to be hopeful given the rapid advances in the five decades since the Manhattan Project; yet continuing investments in building scientific and institutional capacities are essential to ensure the continued protection of people and the natural environment around these sites. The cleanup program is planned to last for several decades even under the most optimistic scenarios—consequently, wise research and development investments made today will likely pay great future dividends.

Acknowledgments

The authors are grateful for the assistance of Allen Croff (Oak Ridge National Laboratory), Kai Lee (Williams College) and Chris Whipple (ENVIRON International, Inc.), who provided information for this article and review comments on an earlier draft; Thomas Wood (Idaho Engineering and Environmental Laboratory), who provided photographs; Bruce Napier (Pacific Northwest National Laboratory), who provided information on environmental releases at Hanford; and Roy Gephart (PNNL), who provided information, comments and photographs.