water supplies, greatly reduce risk by eliminating exposure to contaminants. Other interim actions include source removal, capping, and pumping-and-treating groundwater.13

At the end of September 1994, 1,387 interim actions at 1,173 sites had been completed, and another 573 interim actions were underway at 551 sites. The number of interim actions completed and underway at any given time are key indicators of DOD cleanup progress, because interim actions involving waste removal and treatment may satisfy final cleanup requirements. For example, the Air Force took immediate action to stabilize a threat posed to a national wildlife refuge by a 10,000 cubic yard ash pile containing lead on Johnston Atoll in the Pacific. EPA Region 9 wanted timely action to control the contamination. The interim action involved spraying the ash pile with a liquid polymer. In two days, this action stopped contaminants from leaching to a nearby lagoon containing endangered species.'

DOD has not systematically collected data on actual risk reduction attributable to its program. DOD is currently evaluating the relative risk of its sites in order to categorize the threat of existing site conditions as either high, medium, or low. DOD expects to tie future funding levels to the reduction of relative risk of its sites. In addition, performance measures are being established for environmental work at closing bases. This will include categorizing the relative risk of sites on closing installations and measuring the availability of property environmentally suitable for transfer.

National Aeronautics and Space Administration.

NASA faces industrial cleanup challenges similar to those of DOD at a relatively small number of facilities. NASA is responsible for cleaning up approximately 730 potentially contaminated sites at ten major centers and seven smaller ones, and believes that this list is complete. At 155 of these sites, no further action is required or active remediation has been completed. Another 350 are currently undergoing site evaluation and preliminary assessment, and 75 are undergoing active remediation. NASA plans to investigate the remaining 150 sites over the next two years.

NASA produced these hazardous wastes through its research, operations, and maintenance activities. To facilitate meeting Federal, State, and local environmental requirements, NASA uses a decentralized management approach, with policy guidance, priority setting and oversight from a central Environmental Management Office. The central office has delegated responsibility for environmental compliance to its ten center directors.

"Department of Defense. Defense Environmental Restoration Annual Report to Congress for Fiscal Year 1994, Washington. D.C., p. 9.

"Ibid. pp. 14-15.

Contaminants at these sites are primarily fuels, solvents and industrial waste attributable to such problems as leaking underground storage tanks, exposed asbestos and mercury spills. The current NASA estimate for completing cleanup is $1.5-2 billion, in constant 1994 dollars. NASA plans to complete its cleanup program within the next 25 years. with annual funding requirements of $45-55 million to meet its environmental requirements. NASA has obligated over $136 million since 1988 for environmental restoration, compliance, and pollution prevention.

3. Land Manager Oversight of Hazardous Cleanups.

DOI and USDA face cleaning up sites where Federal operations caused

environmental contamination. In addition, as managers of Federal lands, these agencies may be forced to pay the costs of cleanup for sites contaminated by other parties where it is not possible to recover costs from the responsible parties.

Department of the Interior.

The DOI is responsible for a large number of potentially contaminated sites on the more than 440 million acres of Federal land that it manages. Much of this remediation challenge is the result of actions by users of Federal lands, including private parties, Indian tribes, other Federal agencies, and DOI. DOI estimates preliminarily that it may have as many as 26,000 sites requiring some cleanup. At many of these sites, responsibility would be shared with the parties who undertook the activities resulting in contamination. The most common site types are abandoned mines, oil and gas production sites, underground storage tanks, and landfills. To the best of DOI's current knowledge, the contaminants at these sites are primarily sedimentation in surface waters, acid mine drainage, chemically-reactive mining wastes, and industrial and household chemical wastes.

The current DOI estimate for completing cleanup is $3.9 billion to $8.2 billion in constant 1994 dollars. DOI does not yet have an estimate for the number of years to complete its cleanup program. DOI is funding cleanup activities at eight sites in FY 1995 through its Central Hazmat Fund (CHF), a centralized no-year fund for remedial investigations/feasibility studies and cleanups at hazardous waste sites for which DOI may be liable. DOI may fund activities at these and additional sites during FY 1996 using the CHF, contingent upon the final funding levels. Additional site cleanup activities are funded through the appropriations of the DOI bureaus with responsibility for the sites.

Department of Agriculture.

USDA estimates that there may be 3,500 potential CERCLA or RCRA sites on lands

it manages. Ninety-six cleanup projects were completed at such sites in FY 1994, while 272 sites were investigated and another 70 were identified as needing investigation. Of its estimated 25.000 abandoned and inactive mining sites, USDA estimates that only about 10

percent are likely to require environmental cleanup under CERCLA. Many of these sites occur on land under USDA's management, but, as with DOI, the contamination may not have been caused by USDA's operations. The most common site types are abandoned mines and landfills. Hazardous wastes from mining, chemical wastes, and sedimentation in surface waters are the primary contaminants at these sites. Not all agencies within the Department have completed their final inventories of potentially contaminated sites. Once these are completed in 1996, the estimate will be revised. The present USDA estimate for completing cleanup is $2.5 billion, in constant 1994 dollars. Some part of the total is anticipated to be paid by private sector responsible parties, particularly in case of abandoned mines at or adjacent to national forests.

Overall, its program is at an early stage of development, with basic site inventories remaining to be completed by the responsible agencies within the Department. USDA plans to complete its site cleanup program within the next 40 years and natural resource restoration within the next 50 years.

4. Cooperative Efforts Among Agencies.

In addition to single-agency efforts, some agencies have worked together to clean up sites. For instance, the Fish and Wildlife Service (FWS) is working with DOD on cleaning up the Crab Orchard National Wildlife Refuge in Illinois and the Rocky Mountain Arsenal in Colorado, and the National Park Service (NPS) and DOD are working together at the Presidio in San Francisco. In addition, DOD has teamed up with DOI bureaus, such as the Bureau of Land Management, the FWS, the NPS, and the USDA's Forest Service to deal with unexploded military ordnance and industrial explosives on lands managed by DOI and the Forest Service.

B. Building Knowledge About Relative Risk at Sites.

It may seem a simple expectation and task for agencies to complete risk assessments of all sites, but this task has proven a sizeable and expensive, yet essential, undertaking. DOE and DOD are the two Federal agencies that have made the greatest progress the effort to assess and evaluate risks. The other agencies are still in earlier stages of locating, assessing, characterizing, and evaluating the relative risks of their sites.

There nonetheless has been increasing progress in developing our understanding of risk at Federal facility sites and improving the risk-assessment protocols by which remediation decisions are made. The collective level of knowledge and experience in the use of risk assessment and analysis continues to build, and the science of risk assessment continues to evolve and be refined. It has been, and continues to be, very difficult to make detailed and accurate comparisons of relative risk across agencies, both because available information has been so limited and because the nature of risks varies so greatly from one agency to another. For example, EPA human health risk assessments for Superfund focus on cancer risk, while OSHA assessments focus on worker safety and health. While the quality of available information is often inadequate to make quantitative assessments of risk and comparisons of relative risk across sites and agencies, it is possible to use qualitative approaches incorporating such data and knowledge as do exist, to inform our understanding of those risks. Of course, assessments of the nature and magnitude of risks must be combined with additional information about technical resources. social, economic, and political values, and other factors to evaluate cleanup options and priorities in making sound risk management decisions."

Extrapolating risk information across DOE sites is a difficult task. Even trying to establish priorities within one site, given the different kinds of environmental settings, contaminations of concern and ongoing activities, can prove quite challenging. For example, just one site might contain the following competing priorities: remediating an Operable Unit under Superfund, siting of a hazardous waste site under RCRA, or stabilizing nuclear materials under

"National Academy of Sciences, Building Consensus through Risk Assessment and Management of the Department of Energy's Environmental Remediation Program (Washington, D.C.: National Academy of Sciences, Jan. 1994), pp. 2-3.