cleaning up the department of energy's nuclear weapons complex

CONGRESS OF THE UNITED STATESCONGRESSIONAL BUDGET OFFICE

ST UD-Y

CLEANING UPTHE DEPARTMENT OF ENERGY'S

NUCLEAR WEAPONS COMPLEX

The Congress of the United StatesCongressional Budget Office

NOTES

Unless otherwise indicated, all years referred to in this report are fiscal years.

Numbers in the text and tables may not add to totals because of rounding.

Cover photo shows drums of low-level radioactive waste stored on an asphaltpad in a trench at the Hanford nuclear facility in Richland, Washington. (Photocourtesy of the Department of Energy.)

Preface

T he Department of Energy faces daunting challenges as it attempts to correct theenvironmental problems that exist throughout its nuclear complex. As DOE'sbudget for its Environmental Restoration and Waste Management program has

grown, so has Congressional concern about how DOE intends to carry out its cleanupplan. This study, which was requested by the Chairman of the Department of EnergyDefense Nuclear Facilities Panel of the House Committee on Armed Services (nowknown as the Military Application of Nuclear Energy Panel), examines the key issuesthat bear on the potential costs of DOE's cleanup program. In keeping with the Con-gressional Budget Office's (CBO's) mandate to provide objective analysis, this studymakes no recommendations.

Elizabeth Pinkston of CBO's Natural Resources and Commerce Division andFrances M. Lussier of the National Security Division wrote the study under the generalsupervision of Robert Hale, Jan Paul Acton, Neil Singer, and Roger Hitchner. ElizabethPinkston was the primary author of Chapters 1, 2, and 3 and Appendixes A and B.Frances Lussier was the primary author of Chapter 4 and Appendixes C and D.

The authors would like to thank Perry Beider, Wayne Glass, Elizabeth Chambers,Douglas Elmendorf, and Kim Cawley of CBO for their valuable assistance. The studyalso benefited from the contributions of many people outside CBO. Paul Portney,Jonathan Gledhill, Mark Holt, Peter Johnson, and John Morrall made helpful commentson the draft. Though space constraints do not permit naming all of the DOE staff andcontractors who responded to numerous requests for information, the authors wouldespecially like to thank Eli Bronstein, Robert Muller, and Beverly Heffernan, who pro-vided many updates to budgetary data. (The assistance of external participants impliesno responsibility for the final product, which rests solely with CBO.)

Sherry Snyder edited the report, and Christian Spoor provided editorial assistance.Judith Cromwell, Gwen Coleman, and Donna Wood typed numerous drafts. KathrynQuattrone, with the assistance of Regina Washington and Martina Wojak-Piotrow, pre-pared the study for publication.

Robert D. ReischauerDirector

May 1994

Contents

SUMMARY ix

ONE OVERVIEW OF DOE'S ENVIRONMENTALPROBLEMS 1

Institutional Factors and ConstraintsAffecting Cleanup Policies 3

Factors Affecting CleanupDecisions 6

TWO DOE'S CLEANUP PROGRAM 7

Program Components 8Outlook for Spending 9Accomplishments of the Program 10Initiatives for Managing the

Environmental Cleanup Program 12

THREE STRATEGIC POLICY ISSUES FOR DOE'SCLEANUP PROGRAM 17

Background 17Understanding Risks 78Weighing Benefits and Costs 24Deciding When to Clean Up 27Setting Priorities 32

FOUR ISSUES RELATED TO DOE'S CLEANUP PROGRAM 35

Investing to Develop BetterTechnologies 35

Delaying Cleanup Until NewTechnologies Are Available 46

Spending for Administrationand Support 50

Maintaining DOE's Surplus Facilities 56

vi CLEANING UP DOE'S NUCLEAR WEAPONS COMPLEX May 1994

APPENDIXES

A Profile of Major Facilities and RemedialAction Programs 63

B Estimating the Costs of Cleaning Up DOE Sites 69

C Description of Specific IntegratedDemonstrations and Estimatesof Savings 73

D The Acquisition Process for a MajorWeapon System 77

CONTENTS VII

TABLES

1.

2.

3.

4.

5.

6.

7.

A-l.

A-2.

Appropriations for Environmental Restorationand Waste Management, 1989-1994

Spending Targets for Environmental Restorationand Waste Management, 1995-2000

Implicit Costs per Life Saved of SelectedRegulations

Costs and Cancer Fatalities UnderThree Alternatives for DecommissioningReactors at Hanford

Individual Initiatives in DOE's Research,Development, Demonstration, Testing, andEvaluation Program, by Key Problem Area

Summary of Funding for the Office ofTechnology Development, 1992-1995

Schedule for Records of Decision inSelected Technology Areas

Environmental Cleanup Budgets for DOE'sFacilities and Remedial Action Programs,1994 and 1995

Funding by Facility, 1990-1995

10

26

30

37

38

43

64

65

FIGURES

1.

2.

3.

4.

Department of Energy's Weapons Complex

Department of Energy's 1994 Budget forEnvironmental Management, by Program

Funding for Technology DevelopmentWithin the Office of EnvironmentalRestoration and Waste Management

Cost of Treating, Storing, and Disposingof Low-Level and Transuranic Waste, Withand Without Minimization of Waste

36

41

viii CLEANING UP DOE'S NUCLEAR WEAPONS COMPLEX May 1994

5. Portion of Environmental ManagementFunds Devoted to Administrative andSupport Activities 51

6. Comparison of DOE's EnvironmentalRestoration Costs with Those ofOther Sectors 53

BOXES

1. Risk Assessment and Risk Management 20

2. EPA's Analysis of Hazardous Waste Cleanup 28

Summary

A half century ago, the United States initiateda massive effort, cloaked in secrecy, thatproduced the most powerful nuclear arsenal

in the world. Fifty years later, much of the produc-tion capacity is no longer needed, but a legacyremains in the form of vast quantities of nuclear andother types of hazardous waste. As the demands forproduction of nuclear weapons decrease, the Depart-ment of Energy (DOE), which now manages boththe production of nuclear weapons and the cleanupeffort, will increasingly turn its attention to theproblem of cleaning up its complex.

DOE's Cleanup Program

The nuclear complex of the Department of Energyconsists of 15 major facilities or installations spreadover 12 states. Today that complex holds in storageover 100 million gallons of highly radioactivewaste, 66 million gallons of waste contaminatedwith plutonium, and even larger volumes of wastewith lower levels of radioactivity. In addition,radioactive and other hazardous substances havecontaminated soil and groundwater at DOE's instal-lations. Although some of DOE's environmentalproblems involve conventional contaminants that arecommon to many cleanup tasks, the vast majority ofits pollutants contain some level of radioactivity andso pose challenges unique to DOE. The departmenthas committed itself to meeting all applicable legalrequirements by the year 2019.

In 1989, DOE created the Office of Environ-mental Restoration and Waste Management (EM),

which has primary responsibility for cleanup activi-ties. Since its inception, the office has experiencedrapid budget increases. Its budget has risen from$1.6 billion in 1989 to more than $6 billion in1994, exceeding the funding for the production andmaintenance of nuclear weapons for that year.Funding devoted to the cleanup program is pro-jected to continue to increase, rising to more than$7 billion by 2000, based on the Administration'sout-year targets.

How much the cleanup program will ultimatelycost taxpayers is unknown. In 1988, DOE esti-mated that the cost would be between $66 billionand $110 billion, but estimates keep rising. In1993, DOE officials suggested the cost could rangefrom $400 billion to $1 trillion. But no one canmake an estimate with any degree of confidenceuntil the Congress and regulators clarify the ultimategoals of the program, which include reducing healthand safety risks to humans and mitigating damageto the environment. The goals may also includerestoring sites to make them available for otheruses-industrial, commercial, residential, or recre-ational. Setting goals and priorities will help deter-mine the specific steps DOE must take to achievethem, which will permit more accurate cost esti-mates. In turn, such estimates can help set prioritiesamong the various cleanup options.

Factors Contributing to the Growthof the Cleanup Program

Several factors have contributed to the rapid growthof the EM program. As the need to produce nuclear

x CLEANING UP THE DEPARTMENT OF ENERGY'S NUCLEAR WEAPONS COMPLEX May 1994

weapons has decreased, responsibility for someDOE facilities and programs for managing associ-ated waste have been transferred from the produc-tion program to the EM program. Another factor isthe increasing number of regulations that governhow DOE operates. Until the 1980s, DOE main-tained that the Atomic Energy Act, which it admin-istered, took precedence over other environmentaland safety laws and regulations. Now, however, thedepartment concedes that it must accommodatemany legal constraints established outside DOE. Itmust comply with national laws governing environ-mental impact statements and cleanup plans, and itis also subject to state environmental laws and re-lated requirements. Finally, many specific require-ments are set forth in agreements between DOE, theEnvironmental Protection Agency, and state regula-tory authorities. Many of these agreements containschedules and timetables for the start and finish ofcleanup activities at DOE's various sites.

Budget Constraints

While DOE's cleanup responsibilities have in-creased, the size of the budget function for nationalsecurity, which includes the bulk of DOE's cleanupbudget, continues to decrease in real terms in theaftermath of the Cold War. Although DOE hasreceived most of the cleanup funds it has requestedfrom the Congress, that situation could change.Indeed, in 1994 the Congress appropriated approxi-mately $300 million less than DOE requested for itscleanup program. To accommodate budgets thatmay no longer rise as rapidly as in the past, thedepartment may have to revise the priorities for itsEM program and seek more efficient means ofcarrying it out.

Establishing Goals andPriorities for the CleanupProgram

Policymakers need to establish realistic goals andobjectives for the cleanup effort. Returning allDOE sites to a pristine state by 2019 is clearly notrealistic, given the presence of such contaminants as

long-lived radioactive materials and substances thatpersist in groundwater.

There is general agreement that DOE shouldpromptly eliminate imminent hazards to the public,and DOE is moving to do so. For the remainder ofthe program, however, the Congress and DOE needto decide what to do and when to do it. They mustgrapple with the question of whether DOE shouldattempt to minimize all risks to human health andthe environment, regardless of cost, or whethersome amount of risk is acceptable. Moreover, theymust recognize that eliminating some risks willinevitably increase others. For instance, transport-ing hazardous wastes for disposal entails the risk ofa catastrophic accident. Policymakers also need todetermine the ultimate use of each site and howquickly to restore it to an alternative use.

Understanding Risks

To make informed decisions, DOE will need betterinformation about the risks posed by the hazardoussubstances within its complex. Polls suggest thatthe public perceives hazardous waste sites to beextremely dangerous. The negative publicity associ-ated with Love Canal and Times Beach, for exam-ple, help explain that concern. But two recent re-ports published by the Environmental ProtectionAgency (EPA) suggest that hazardous waste sitesrank considerably lower in risk than many otherenvironmental problems. In one report, expertswithin EPA attempted to assess the relative risks ofa number of environmental problems, such as haz-ardous waste sites, air pollutants, discharges of con-taminants into drinking water, and exposure ofworkers to chemicals. These technical and policyexperts ranked the problems in terms of four catego-ries of risk: cancer risks, noncancer health risks,ecological damage, and risk to economic welfare.In none of the four categories did hazardous wastesites such as those within the DOE complex rankamong the worst problems. The second report, byEPA's Science Advisory Board, identified someimportant shortcomings in the first report but didnot dispute its relative rankings of health risks.

These studies raise questions about how muchof the limited public funds available for environ-

SUMMARY XI

mental cleanup should be devoted to hazardouswaste sites such as DOE's. Both of the EPA stud-ies emphasize, however, that their results are notconclusive, because information about risks is lim-ited. Better information is clearly needed so thatpolicymakers can make informed decisions abouthow to allocate resources for cleanup activities. Forexample, it would be helpful to have more defini-tive studies regarding the long-term effects of sub-stances found at DOE installations on the health ofthe public and the environment. DOE also needs tohave a better idea of which hazardous substancesare found at its installations and whether or not theyare migrating off-site. To this end, DOE has ini-tiated an effort to evaluate the information availableabout risks to human health and the environmentand to fill the gaps in knowledge in order to de-velop a comprehensive strategy for reducing risks.The National Academy of Sciences is assisting withthis effort.

Only after such information is available canDOE show that it is making the appropriate choicesbased on scientific evidence and so continue to winboth Congressional and public support for its clean-up program. This problem is not unique to DOE;the Environmental Protection Agency and the De-partment of Defense need similar information, andany efforts to gather that information should there-fore be collaborative.

Estimating Costs

To establish priorities among cleanup projects bycomparing benefits and costs, DOE must have reli-able estimates of the cost of those efforts. Theaccuracy of such estimates depends, however, onunderstanding the difficulty of cleanup efforts, andcleaning up some DOE sites involves highly com-plex problems with which DOE has little previousexperience.

The department's ability to estimate costsshould improve as it gains more information con-cerning the extent of the cleanup problem it faces.As DOE performs more assessments of the contami-nants at its installations, it will have a better idea ofthe scope and seriousness of the cleanup task. Also,as DOE gains more experience with initiating and

completing specific tasks associated with the clean-up, it will have more cost data on which to basesubsequent cost analyses. Finally, DOE has under-taken some initiatives to improve its ability to esti-mate costs. They include a benchmarking initiativeto identify causes of cost growth in DOE cleanupprojects, and a performance tracking system de-signed to monitor the cost of projects over time.

Weighing Benefits and Costs

With more information about costs and risks, DOEcan decide how to allocate funds among the variouscleanup tasks. For each site requiring cleanup,DOE can employ benefit-cost analysis to help deter-mine whether remediation is needed immediately,can be delayed, or can be avoided. The analysisshould look at several factors: the cost; the benefitin terms of reduced risk to workers, the public, andthe environment; and alternative uses of land andfacilities that are not currently available to the pub-lic. The department could thus identify and proceedwith those tasks that, for each dollar spent, providethe greatest benefit in terms of these criteria.

The Hanford Example. In late 1992, DOE re-leased an analysis of the alternatives for disposingof eight surplus reactors at the Hanford site inWashington State. That analysis is subject to manysubstantial uncertainties about costs and risks, but itillustrates the effects of choices related to the clean-up program.

The department considered several options:maintaining the reactors in place indefinitely andmonitoring them to ensure safety; removing themimmediately for disposal at another Hanford loca-tion; and removing them to the disposal locationafter 75 years. It also considered whether to movethe reactor blocks intact or to dismantle them first.

DOE anticipated that the cost of removing thereactors would be about the same regardless ofwhether it was done now or many years later. Inthe time leading up to removal, DOE would incurmaintenance costs, but they would be relativelysmall. When future costs are discounted, however,the savings from delaying removal of the reactorsare substantial.

xii CLEANING UP THE DEPARTMENT OF ENERGY'S NUCLEAR WEAPONS COMPLEX May 1994

How delay would affect risks is somewhat morecomplicated. Removing the reactors immediatelywould entail additional exposure of workers tohigher levels of radioactive material; delaying re-moval would lower that risk by allowing some ofthe radioactive material to decay. Leaving the reac-tors in place indefinitely, however, could expose thepopulation to the potential risks of radioactive con-tamination. But even if the buildings fell into disre-pair or were abandoned, DOE estimated that just 20additional cancer deaths would result over a periodof 10,000 years. To put that figure in perspective,an estimated 400,000 people will die from cancer in1994 in the United States. Weighing all these fac-tors, DOE opted for one-piece removal of the reac-tors after a period of 75 years.

Similar Choices Elsewhere. Benefit-cost analysesare not available for most DOE cleanup sites. Inmany cases, DOE does not yet know the nature andextent of contamination at its facilities, nor does ithave sufficient information to make reliable projec-tions of risks or costs. Nonetheless, the type ofanalysis performed for the Hanford reactors is auseful tool for establishing priorities among thedepartment's cleanup tasks.

Involving Citizens in Setting Priorities

Weighing the benefits and costs of options forcleaning up the DOE nuclear complex requires theinvolvement of citizens affected by the cleanup,including taxpayers, workers at the facilities, neigh-bors whose environment is affected, and concernedmembers of the public at large. They can helpevaluate the benefits of various cleanup options byindicating how much they value them. Their prefer-ences about risks and land use are key factors inmaking trade-offs. For instance, some communitiesmay consider it more important to clean up a facil-ity to a standard acceptable for industrial use-andmake it available for manufacturing jobs—than toremove every trace of contamination. Others mayplace greater value on restoring the environment toits pristine condition. Incorporating such prefer-ences into a benefit-cost analysis can guide deci-sions about setting priorities and determining thelevel of cleanup to be done.

DOE has stepped up efforts to increase publicinvolvement in establishing cleanup policies. It isparticipating in an endeavor known as the Keystoneprocess that attempts to improve communicationbetween citizens and the federal agencies respon-sible for cleaning up hazardous waste sites. Amongother things, this effort provides a forum for dis-cussing priorities under constrained budgets.

Improving the Efficiency ofDOE's Cleanup Program

While DOE gathers information and performs theanalyses necessary to set priorities, it must alsocontinue to manage its ongoing cleanup efforts.Several policy changes might make those effortsmore efficient.

Place Greater Emphasis onTechnology Development

Current methods of cleaning up contaminants,which usually involve digging up soil or pumpingout water, are time consuming and costly, particu-larly for pollutants that have found their way intothe groundwater. To develop techniques for charac-terizing and cleaning up sites more quickly andcheaply, DOE established the Technology Develop-ment program. In its budget request for fiscal year1995, DOE has allocated 7 percent-$426 million-to that effort.

It may be appropriate, however, to invest evenmore funds in efforts to develop cheaper means ofcleanup. DOE should add funds only for promisingtechnology programs-those for which the presentvalue of the cost savings and other benefits associ-ated with the new technology exceed the cost to de-velop it. DOE and other organizations have identi-fied some programs that may meet this test. Candi-dates include research into methods to clean upheavy metals and techniques for removing from thesoil or groundwater those dense organic compoundsthat are not soluble in water.

SUMMARY Xlll

New technologies may be able to make thecleanup effort cheaper in the long run. DOE claimsthat using new rather than current technologiescould reduce the costs of some cleanup activities byas much as 80 percent. During the entire cleanupprocess, savings for the whole complex could ap-proach $100 billion from a total cost that could beas high as $1 trillion. These estimates are, ofcourse, subject to great uncertainty, but their sizesuggests that new technologies could have majoreffects on costs.

How much additional funding for technologydevelopment might be appropriate? A definitiveanswer would require a detailed analysis of candi-date projects, which is beyond the scope of thisstudy. But DOE and the Congress have both estab-lished 10 percent of the total cleanup budget as thegoal for funding technology development. To meetthat goal at the budget levels proposed by the Ad-ministration, DOE would have to add $200 millionto its funding for technology development in 1995and increasing amounts each year, with $250 mil-lion more needed in 2000. Since many of DOE'scleanup problems are similar to problems confrontedby the Environmental Protection Agency and theDepartment of Defense, some of the additionalfunds might be directed to agencies other than DOE.

Regardless of which agency controls the money,the personnel to carry out added development ef-forts could come from a variety of places, includingDOE's national laboratories and other researchorganizations whose defense funding is declining.Any added funds provided under this option, how-ever, should be spent on research. Of the fundsrequested for technology development for 1995,about 40 percent are earmarked for programs notdirectly involved in research and development. Ifdevelopment funds are increased, the Congresscould direct that all added funds be spent only onpromising projects for technology developmentrather than on increases in administrative or supportcosts.

Before adding funds for technology develop-ment, the Congress may also want to direct thatDOE implement a new management system to im-prove tracking of the costs and schedules of cleanupprojects. The new system might be patterned after

one now in place in the Department of Defense totrack major weapons programs. For each majorweapon system, the Defense Department's systemestablishes four milestones, each with its own costsand deadlines. Periodic reports compare progressagainst those goals. DOE could establish a similarsystem for major cleanup projects. Its milestonesmight start with the designation of a new cleanupsite and extend until remedial work begins.

Delay Technically DifficultCharacterizations and Remediations

While DOE develops new technologies to performcleanup tasks more cheaply, the department coulddelay projects that are costly and time consuming toaccomplish with current technology. Delaying theseprojects would save money in the short term and, ifmore efficient technology became available, couldalso reduce long-term costs.

Through a detailed analysis of the Environmen-tal Restoration program's five-year plans, the Con-gressional Budget Office (CBO) determined thatDOE could spend as much as 30 percent of itsbudget, on average, for environmental restorationactivities over the next six years on projects thatmay be difficult to accomplish with current technol-ogy. These projects include characterizing verylarge sites or buried waste, cleaning up contami-nated groundwater, remediating soil contaminatedwith radioactive substances or heavy metals, anddecontaminating and decommissioning surplusbuildings.

Delaying these technically difficult projects untilmore efficient technologies are available couldreduce costs substantially, but CBO cannot examineeach of these projects in detail to determine whichcan be performed at reasonable cost with today'stechnology. To illustrate the budgetary effect ofdelaying some projects, CBO examined the impactof reducing funding for all of the difficult projectsby 50 percent over the next six years. The resultingsavings would increase from $270 million in 1995to $300 million in 2000. The savings could be re-turned to the Treasury, or they could be used tooffset the added cost of increased investment intechnology development.

xiv CLEANING UP THE DEPARTMENT OF ENERGY'S NUCLEAR WEAPONS COMPLEX May 1994

DOE is proceeding with some of these difficulttasks, even in the absence of techniques to accom-plish them efficiently, because it is bound by manyagreements concerning the conduct of its cleanupprogram. Most of these agreements contain sched-ules and deadlines that determine when DOE mustbegin cleaning up certain sites. Delaying someprojects, as envisioned in this option, would resultin the breach of some of these agreements. Butmany agreements were signed in the early stages ofthe cleanup, before DOE knew the extent and typeof contamination it faced. The Environmental Pro-tection Agency, the states, and other parties to theagreements therefore might be amenable to renegoti-ating them if they felt that DOE was making agood-faith effort to find better ways to addressmutual problems. Indeed, a recent renegotiation ofthe agreement governing cleanup of Hanford de-layed treatment of highly radioactive waste stored intanks by 10 years. DOE conceded that it could notmeet the deadlines established in the original 1989agreement. Through renegotiation with the EPAand state regulators, DOE established a new set ofpriorities and deadlines acceptable to all parties tothe agreement.

Reduce Funding for Administration,Support, and Management

DOE may also need to improve the efficiency of itscleanup efforts. Since 1989, the department's Of-fice of Environmental Restoration and Waste Man-agement has received about $23 billion from theCongress to spend on the cleanup program. DOEhas been severely criticized because of the smallamount of visible cleanup that has been accom-plished. Some of the slow start reflects the diffi-culty associated with beginning a new operation andthe need to devote substantial sums to assess thenature and extent of cleanup problems. But DOEmay also be devoting too much of its budget toadministration and support, thus limiting the fundsavailable for actual cleanup work.

Several reviews of DOE's costs for cleanupactivities have concluded that EM devotes a largeproportion of its funds to administrative and supportfunctions. Three reviews, one by the Army Corpsof Engineers and two performed under contract to

DOE, have found that contractors charged the EMprogram project overhead rates of 20 percent to 28percent. These rates were higher than those chargedto private industry or other government agencies—insome cases by as much as a factor of 4 and 2, re-spectively. CBO and the Corps also found thatroughly 20 percent more of EM's funds were beingused to provide programwide support (such as pro-gram direction) and installationwide activities (suchas security and utilities). All told, at least 40 per-cent of EM's funds are devoted to administrativeand support activities, a level that many reviewershave considered excessive.

One of the analyses requested by DOE recom-mended a 7 percent reduction in the EM budget toreduce spending for project overhead, and the Corpsrecommended a 25 percent overall cut. If adminis-trative costs were cut by 25 percent, the total EMbudget would be reduced by 10 percent, a reductionthat is toward the lower end of the range recom-mended by the two reviews. A reduction of thismagnitude would yield savings of $630 million in1995. Annual savings would increase to $710 mil-lion in 2000, based on the Administration's out-yeartargets for EM spending. The funds freed up byreducing funds devoted to administration and sup-port could be used either to reduce the deficit or toaccelerate cleanup activities.

Various studies and the Assistant Secretary forEM, Thomas Grumbly, have suggested ways toachieve such savings. One suggestion is to increasethe number of DOE personnel in order to providebetter oversight of DOE's large number of contrac-tors. In the EM program alone, more than 49,000contractor personnel are engaged in cleanup, over-seen by fewer than 1,800 DOE personnel. Grumblyhas proposed adding 400 DOE personnel to monitorcontractors and perform some functions that arecontracted out but would be more appropriatelyperformed by government staff. He predicted thatincreased oversight would save $360 million in1995.

Another means of reducing administrativespending, particularly for contractor overhead, isthrough contract reform. Such reform would in-volve changes in DOE's contracts with the firmsthat manage its installations. Particularly with re-

SUMMARY xv

spect to contracts for environmental work, Crumblyrecommends limiting the time period covered by thecontracts, issuing separate environmental contractsat those installations where production is ongoing,and making individual contracts for some functionssuch as security or road maintenance. Crumbly isalso actively pursuing this approach and predictssavings of 10 percent to 20 percent in EM spendingon contracts-which represents the vast majority ofEM spending-over the next four years as a result ofincreased oversight and contract reform combined.

Safely Maintain DOE's SurplusFacilities

The end of the Cold War and the restrictions onstrategic arms that have resulted from internationaltreaties have made much of DOE's nuclear weaponscomplex unnecessary. DOE could declare as manyas 7,000 of its facilities surplus in the next 30 years,leading to their eventual decontamination and de-commissioning. In the meantime, the EM programis responsible for the security and maintenance ofan increasing number of surplus facilities.

In a recent report, the General Accounting Of-fice concluded that the EM program faces problemsconcerning maintenance, safety, and costs for thedisposition of its surplus facilities. Since maintain-ing inactive facilities is not a high priority amongEM's tasks, DOE's inactive facilities are deteriorat-ing physically. Repair projects for surplus buildingsare often deferred in favor of higher-priority workelsewhere. As a consequence, conditions at suchbuildings have violated regulations established byDOE and the Occupational Safety and Health Ad-ministration and have resulted in accidents amongworkers.

Furthermore, DOE has engaged in some prac-tices that can increase the cost and dangers associ-ated with cleaning up inactive facilities. Incomplete

or substandard work performed during the shutdownprocess can lead to unanticipated problems or acci-dents during subsequent decontamination and de-commissioning. Inadequate shutdown procedurescan also affect the cost of cleanup projects. Forexample, because equipment was not cleaned whenthe plutonium fuel facility at Savannah River wasput on standby in 1983, it is now so badly deterio-rated that it can no longer be used to remove theplutonium that remains in the facility. As a result,DOE has estimated that an additional $115 millionwill be needed to decontaminate and decommissionthe facility. Had the facility initially been cleanedadequately, subsequent higher cleanup costs couldhave been avoided.

In general, the cost to maintain surplus facilitiesawaiting cleanup is substantial and could growbecause of problems of the sort just noted. Increas-ing near-term funding designed to attain safe shut-down status at surplus facilities, thereby reducingannual security and maintenance costs, could pro-duce long-term savings in the DOE budget.

Conclusion

DOE's cleanup program must address a problemcreated and, for the most part, ignored over the past50 years. It must do so during a period in whichfunding for all federal programs, including envi-ronmental cleanup, is becoming increasingly tight.Better understanding of risks and costs, broughttogether by benefit-cost analyses, would be the bestmeans of establishing priorities for allocating scarcecleanup funds. DOE may also be able to improvethe efficiency of its cleanup efforts by policies suchas investing more heavily in technology develop-ment, delaying technically difficult projects, andcutting overhead costs. New management systemsmay also be necessary to permit DOE and the Con-gress to track the performance of cleanup projects.

Chapter One

Overview of DOE'sEnvironmental Problems

T he Department of Energy (DOE) presidesover a vast complex of facilities with mas-sive environmental problems. These prob-

lems have accumulated over the past 50 years as aby-product of producing nuclear weapons. Fromthe inception of the Manhattan Project in 1942through the 1980s, DOE and its predecessor agen-cies focused on developing, producing, and testingnuclear weapons. Protecting environmental safetyand health took a back seat to that primary mission.In recent years, however, concerns about environ-mental contamination have mounted. At the sametime, the end of the Cold War has diminished theperceived urgency of maintaining ambitious produc-tion goals. Together these factors have led to ashift in emphasis from production to environmentalcleanup.

The nuclear weapons complex consists of 15major facilities and a dozen or so smaller facilitiesat which production, research, and testing haveoccurred over the past five decades. In addition,DOE is responsible for environmental cleanup atthousands of sites formerly used in the weaponsprogram and sites where uranium was processed.Altogether DOE must contend with more than 100million gallons of highly radioactive waste, 66million gallons of waste contaminated with pluto-nium, and larger volumes of low-level radioactivewaste. It also must deal with huge volumes of othertoxic materials, including heavy metals, chemicalsused as solvents, acids, and other materials that aredifficult and costly to clean up.1

Six of DOE's 15 major facilities-Hanford,Savannah River, Oak Ridge, Fernald, Idaho NationalEngineering Laboratory, and Rocky Flats-accountfor more than 60 percent of the budget of DOE'senvironmental cleanup program. Hanford alone isresponsible for nearly a quarter of the budget. Themajor facilities are shown on the map in Figure 1and are profiled in Appendix A.

Although all of the weapons facilities presentcleanup challenges, they differ in ways that mayaffect goals, methods, and timetables of cleanupefforts. Some facilities are near population centerswhere releases of toxic materials might pose immi-nent hazards to public health; other facilities areremote and present less near-term risk. Yet thelatter may still impose long-term effects on humanhealth and the environment. Soil and drainageconditions vary, leading to differences in how toxicreleases might migrate into the groundwater andultimately to supplies of drinking water. All ofthese factors affect both the costs and benefits ofcleanup.

The potential cost to the federal government ofcleaning up the nuclear weapons production facili-ties is large, and the cost estimates keep rising. Noone knows what the ultimate cost will be, in partbecause the dimensions of the problem are still notclear. One estimate widely cited is $160 billionover the next 30 years—assuming technologicalbreakthroughs to reduce the costs of dealing withradioactive wastes. Recently, however, the head ofthe cleanup program has referred to that effort as"our trillion-dollar program."2

A thorough description of the nuclear facilities and their environ-mental problems is contained in Office of Technology Assess-ment, Complex Cleanup (February 1991).

2. Thomas Crumbly, Assistant Secretary of Energy for Environmen-tal Restoration and Waste Management, as quoted in InsideEnergy/with Federal Lands (newsletter published by McGraw-Hill,Inc., New York, July 19, 1993), p. 10.

2 CLEANING UP THE DEPARTMENT OF ENERGY'S NUCLEAR WEAPONS COMPLEX May 1994

People do not even agree on what is meant by"cleanup": some argue that the term should referonly to environmental restoration of sites that are nolonger in use; others would have it include suchactivities as managing wastes currently being gener-ated and developing technologies to promote envi-ronmental goals. This study adopts the broader def-inition, which encompasses all the activities ofDOE's Office of Environmental Restoration andWaste Management—the office charged with theprodigious task of planning and managing environ-mental policies for the weapons complex and forsome nondefense nuclear facilities.

The Office of Environmental Restoration andWaste Management spent $13.8 billion from 1990through 1993. Most of this money has been usedfor managing waste streams associated with ongoing

operations. At inactive waste sites designated forcleanup, most of the work to date has been to char-acterize the contaminants and assess what remedialwork is needed.

The relatively small amount of visible resultshas led to complaints that DOE is wasting money.DOE may have promised too much early in theprogram, before realizing the extent and complexityof its contamination problems. The department alsomay have been overly optimistic about technologicalbreakthroughs to aid in assessing and cleaning upwaste sites and overly ambitious in agreeing to meetcertain goals and schedules. A reassessment ofprogram objectives and priorities could result in aredirection of resources to achieve greater environ-mental benefits.

Figure 1.Department of Energy's Weapons Complex

Idaho National Engineering Laboratory

Hanford

Nevada Test Site

Lawrence LivermoreNational Laboratory

Rocky Flats Plant

Los Alamos National Laboratory

Oak Ridge

Savannah River

Pinedas Plant

Sandia National Laboratory

Waste Isolation Pilot Plant

SOURCE: Congressional Budget Office.

CHAPTER ONE OVERVIEW OF DOE'S ENVIRONMENTAL PROBLEMS 3

The cost of DOE's cleanup program largely de-pends on the ultimate goals. One goal is reducinghealth and safety risks to humans and damage to theenvironment. Another is restoring land and makingit available for other uses, which could range fromindustrial or commercial to residential or recrea-tional. The choice of the ultimate goal will affectthe type and extent of cleanup at individual sites, aswell as the cost and schedule of the entire cleanupprogram.

The Department of Energy has expressed acommitment to cleaning up the nuclear facilities bythe year 2019.3 That entails bringing operatingfacilities into compliance with all applicable lawsand regulations and cleaning up the contaminationat inactive sites. With current technologies, how-ever, the cost of such a complete cleanup would beextremely high. It may be appropriate to invest indeveloping better technologies, and wait until theyare available, before undertaking some cleanup ef-forts. Even if improved technology can reducecosts, which is not guaranteed, the cost of a com-plete cleanup program may be judged unacceptablewhen competing demands for resources are consid-ered. A more limited program of remediation couldfree up resources to provide greater benefits else-where.

Institutional Factors andConstraints AffectingCleanup Policies

In addition to budgetary limitations, the Departmentof Energy faces a number of constraints that affectits cleanup policies. Three factors are particularlyinfluential: nuclear weapons policies, legal require-ments, and public opinion.

Department of Energy, Environmental Restoration and WasteManagement Five-Year Plan, Fiscal Years 1994-1998, vol. 1(January 1993), p. 1-9.

Nuclear Weapons Policies

Questions about the role of nuclear arms in nationalsecurity policy in the aftermath of the Cold Warlead to uncertainty about the future mission of someof the nuclear weapons facilities, which in turncomplicates the task of setting policies for environ-mental restoration. Reduced reliance on nuclearweapons has prompted DOE to terminate productionat some facilities. Fernald, where uranium was pro-cessed, was the first facility to be turned over en-tirely to environmental restoration. Production hasalso ceased at Hanford and Rocky Flats. At somefacilities, production has been halted temporarily butmay resume, depending on plans for the strategicarsenal.

Some facilities are likely to remain in produc-tion for some time, although the size of the produc-tion area may shrink if output is reduced. If DOEdecides to consolidate production at a few facilitiesand close the rest, however, then production couldoccur on a relatively large scale at those sites. DOEis developing a programmatic environmental impactstatement that addresses these issues and plans toissue a draft of the statement early in 1995.Whether a facility is actively producing weaponsmakes a difference in carrying out environmentalrestoration. Cleanup is more complicated at activeproduction sites for a variety of reasons, includingthe need to keep production and cleanup activitiesfrom interfering with each other and the need forgreater security in production areas.

Legal Constraints

Current cleanup plans call for meeting all federaland state environmental laws, regulations, and re-quirements by 2019. DOE can make some changesin cleanup plans unilaterally, but most changeswould require approval by the Congress, the Envi-ronmental Protection Agency (EPA), or state regula-tors. The department must comply with a host offederal environmental laws and regulations and mustadhere to terms of agreements negotiated withstates. It is also constrained by authorization andappropriation legislation.


The major federal environmental laws governingcleanup of DOE's nuclear facilities are the Compre-hensive Environmental Response, Compensation,and Liability Act (CERCLA, the law that set upSuperfund) and the Resource Conservation andRecovery Act (RCRA). These laws establish re-quirements and procedures for dealing with hazard-ous wastes. CERCLA focuses on cleaning up inac-tive sites, and RCRA imposes "cradle-to-grave"requirements for tracking and properly dealing withhazardous materials throughout all stages of produc-tion. DOE must also comply with the NationalEnvironmental Policy Act, which sets forth require-ments for analyzing the likely impact of any activitythat would significantly affect the environment, aswell as the Clean Air Act, the Clean Water Act, theAtomic Energy Act, and numerous other environ-mental statutes.

DOE is also subject to state environmental lawsand related requirements. The Federal FacilityCompliance Act of 1992, for example, clarified thestates' authority over disposal of solid or hazardouswaste by federal facilities. Many specific require-ments, including interim milestones for accomplish-ing certain objectives, are set forth in agreementsbetween DOE, the Environmental ProtectionAgency, and state regulatory authorities. Theseagreements are commonly referred to as tripartyagreements.

The department considers complying with tri-party agreements to be one of its highest priorities.Analysis of data made available by DOE indicatesthat about 54 percent of the 1994 budget for envi-ronmental restoration and waste management is foractivities required by such agreements. The per-centage driven by these agreements differs substan-tially among facilities, ranging from 3 percent at theNevada Test Site to 89 percent at Savannah River.At five facilities-Savannah River, Rocky Flats,Hanford, Oak Ridge, and Fernald-more than 70percent of funding is attributed to meeting inter-agency agreements.4

4. These findings should be treated with caution because the data setfrom which they were derived has not been verified. Some obvi-ous errors throw into doubt the reliability of the data. They ap-pear to be the best data available, however. DOE is attempting togather better data on the role of legal requirements. For purposesof this analysis, Oak Ridge includes the K-25 Site, the Y-12 Plant,Oak Ridge National Laboratory, and administrative spending atOak Ridge; it excludes off-site facilities managed by the OakRidge field office.

The Congress has become increasingly con-cerned about the amount of funding required forDOE to meet its legal obligations. It has directedthe department to review compliance agreementsand to submit a report to the Committees on Appro-priations by June 30, 1995. The report is to evalu-ate the risks to the public health and safety posedby the conditions at weapons complex facilities thatrequirements in the compliance agreements address.5

Triparty agreements typically have been basedon the best information available at the time theywere signed-information about types of contami-nants, potential risks, expected cleanup capabilitiesand resources, and so on. In many cases, however,this information was quite limited. Parties to theagreements could only make educated guesses aboutpotential hazards and risks, and their estimates ofabilities to handle difficult cleanup problems wereoptimistic. In some instances, the estimates of risksmay have been pessimistic, and further investigationmay reduce concern about adverse effects on publichealth.

Moreover, in the time since some agreementswere signed, the mission of the facility has changed.At Rocky Flats, for instance, the agreement wasnegotiated when the plant was still producing pluto-nium components of weapons; now that the facilityhas been turned over to environmental restoration, adifferent set of cleanup options is available. Somelocal citizens have suggested a reordering of priori-ties-for example, placing more emphasis on restor-ing buildings to make them usable by other employ-ers than on cleaning up all the outside areas.

As time has passed and DOE has gained agreater appreciation of the overwhelming size, com-plexity, and potential costs of the cleanup problem,some facets of the triparty agreements have come tobe viewed as unrealistic and even counterproductive.For example, some agreements require removingand treating certain contaminants by fixed dates, butassumptions about the timely availability of treat-

5. U.S. House of Representatives, Making Appropriations for Energyand Water Development for the Fiscal Year Ending September 30,1994, and for Other Purposes, Conference Report 103-305, toaccompany H.R. 2445 (October 22, 1993), p. 94.

CHAPTER ONE OVERVIEW OF DOE'S ENVIRONMENTAL PROBLEMS 5

ment technologies and disposal facilities haveproved wrong. As a result, DOE would like to re-negotiate those milestones and requirements.

In January 1994, DOE reached agreement withEPA and the Washington State Department of Ecol-ogy to revise the triparty agreement at Hanford.The regulators allowed DOE to slip some of itsmilestones in return for a commitment to act moreaggressively on the most pressing safety problem-the storage tanks containing a mixture of radioactiveand hazardous wastes. The new agreement reflectednot only revised views about the relative seriousnessof different waste problems but also a realizationthat new technologies for treating wastes were moreelusive than originally anticipated.

Because each facility is unique, the Hanfordrenegotiation may or may not serve as a usefulmodel for others. Still, as additional informationbecomes available, the parties to other agreementsmay find that they too could gain by revising theagreements. For example, new information maylead the parties to conclude that some problems theyinitially thought to be most serious are less gravethan others, and vice versa. In addition, the initialagreements relied on assessments of when remedialtechnologies would be available. But treatmentmethods have not always emerged as anticipated.Reordering priorities to deal first with problems forwhich treatment technologies are available, safe, andeffective-and postponing work on problems forwhich no treatment exists—may benefit all parties.The troublesome issue of what to do about a seriousproblem for which no remedy is known remains.

Renegotiating an agreement also provides anopportunity to explore the question of the eventualuse of the land. Resolution of this issue will have abearing on the appropriate remedy as well as thespeed of restoration. If the local community iseager for DOE to make a facility available for in-dustrial use, it may be willing to accept somewhatlower cleanup standards than if the land was desig-nated for residential use. Cleanup might be expe-dited at facilities that could be put to commercialuse. Some of DOE's facilities are so large that avariety of future land uses may be appropriate fordifferent sites at those facilities.

DOE faces formidable obstacles to getting EPAand the states to renegotiate triparty agreements. Inthe typical case in which DOE has made commit-ments it cannot keep, it is hard pressed to findsomething to offer in return for forbearance. More-over, the department suffers from a serious credibil-ity problem. One legacy of the secrecy in whichDOE conducted business over several decades was aculture of not being held accountable to outsidersfor meeting health, safety, and environmental com-mitments. Although current management is tryingto change this, shaking a reputation formed overmany years is not easy.

Public Opinion

The triparty agreements responded in part to grow-ing public concerns about what hazards might lurkat DOE facilities. DOE's nuclear weapons programhad been cloaked in secrecy since its inceptionduring World War II, leaving neighboring commu-nities in the dark about potential environmentalrisks. The high level of secrecy shielded DOE fromthe public scrutiny that might have forced it to keepenvironmental problems from getting out of control.The department itself has acknowledged that "thesecretive, unresponsive nature of DOE during theweapons production years of the Cold War [has]undermined the public trust and created long-termsuspicion of DOE operations."6 Moreover, its credi-bility suffered as more information about environ-mental problems became available to the public.

Two factors converged in the late 1980s toincrease public interest and involvement in envi-ronmental remediation. DOE began to lift the veilof secrecy at the nuclear facilities, and a few well-publicized events-spills, releases, exposures ofworkers—alarmed citizens (especially those in neigh-boring areas, but also many environmental activists)about risks to health and safety. DOE recognizesthat its credibility suffers when it withholds vitalinformation about its environmental and safetyproblems, and it has stepped up efforts to informthe public.

Department of Energy, "Perspective on DOE's EnvironmentalRestoration Program," Section III of briefing materials for Bench-marking for Cost Improvement Initiative (June 1993), p. 4.


In addition to environmental concerns, manycommunities fear loss of jobs as neighboring nu-clear facilities are shut down. In the short run, em-ployment may actually increase at plants as theyshift from production to environmental cleanup.Over the longer term, however, unless facilities canbe cleaned up enough to be safe for other uses, thelocal economic base may be severely weakened.Neighbors of several facilities want DOE to speedup the decontamination of buildings to make themavailable for employers who can use the skills ofdisplaced workers.

Factors AffectingCleanup Decisions

What should DOE do about remedying the environ-mental problems it has sown? The answer is not atall clear, and the possible responses are numerous.At a minimum, DOE could focus on ensuring thatno significant amounts of contaminants migrate intodrinking water or otherwise pose imminent threatsto human health. At the other extreme, DOE couldattempt to restore all sites to their original pristinecondition. In between lie such options as cleaningsites up to a level acceptable for industrial or com-mercial use.

Underlying the laws requiring DOE to clean upits facilities is the premise that the nuclear weaponsfacilities pose risks to public health and the environ-ment and that these risks should be eliminated asrapidly as possible. As suggested in Chapter 3,however, the DOE sites probably encompass a widerange of risks. Some sites may pose substantial,imminent threats; others may be relatively benign,at least compared with other environmental risks.The state of knowledge about potential risks islimited, and until environmental and health assess-ments of DOE's nuclear weapons facilities are com-

pleted, one can only make informed guesses aboutthe extent of those risks. More information isneeded about the nature and extent of contamina-tion, whether and how contaminants are migratinginto the air or groundwater, and how they affect thehealth of people exposed to them.

A better understanding of the risks would permitinformed debate among the public and policymakersabout how much risk is acceptable. Eliminating allrisks is not only very costly but also virtually im-possible, since reducing one risk often entails in-creasing another. Identifying the trade-offs betweenrisks and costs can help policymakers set appropri-ate goals and priorities for the cleanup program.

Debate on acceptable risk should consider notonly ultimate cleanup levels but also how soonthose levels should be reached. DOE could adhereas closely as possible to the schedules set forth inagreements with EPA and the states. Alternatively,it could seek permission to delay cleanup at siteswhere existing technologies are ineffective andcostly, while moving forward with "easier" cleanupsand with the development of technologies to solvethe harder problems. If its budget becomes tighter,DOE, in consultation with regulators, stakeholders(people who work at or live near nuclear weaponsfacilities, taxpayers, and others with an interest incleanup), and the Congress, will have to decidewhich activities to defer.

Short of a fundamental reexamination of thegoals and scheduling of the cleanup program, DOEcould take several measures to make the existingprogram more effective. These measures, discussedin Chapter 4, include stepping up efforts to developnew technologies that would make cleanup safer,cheaper, and more effective, and cutting administra-tive and overhead costs. Attention to these mea-sures and to the fundamental objectives of the pro-gram could bring about significant improvements inbenefits per dollar spent.

Chapter Two

DOE's Cleanup Program

T he United States has produced nuclearweapons for the past 50 years, and for mostof that time environmental activities have

been subsumed under production activities. But asthe extent of environmental problems becameclearer and as public concerns mounted, Departmentof Energy officials decided in 1989 to form a sepa-rate office with primary responsibility for environ-mental cleanup. The Office of Environmental Res-toration and Waste Management (EM) assumedenvironmental responsibilities that previously hadbeen handled within the Offices of Defense Pro-grams, Nuclear Energy, and Energy Research.

The environmental cleanup program is the larg-est and fastest-growing part of DOE's budget, hav-ing risen from $1.6 billion in 1989 to nearly $5.9billion in 1994 (see Table I).1 Over this period, theCongress has appropriated more than $23 billion forenvironmental restoration and waste management.DOE runs the largest environmental cleanup pro-gram in the federal government, surpassing fundingfor the Environmental Protection Agency's Super-fund program and the cleanup of installations run bythe Department of Defense.

Some of the growth of EM's budget is not newfunding but rather reflects a shift in responsibilitiesto EM from other DOE offices, primarily the Officeof Defense Programs, which runs the nuclear weap-

The budget data throughout this chapter exclude the new UraniumEnrichment Decontamination and Decommissioning program,whose 1994 appropriation is $286.3 million, in order to facilitatecomparisons with spending in earlier years. This program is sepa-rate from the cleanup at the weapons complex. The total alsoexcludes the use of balances from prior years.

ons production program. Before EM was formed,that office managed all activities at the weaponsfacilities, including production, security, and envi-ronmental and safety activities. It continues to han-dle operations and maintenance responsibilities aslong as environmental management activities are arelatively small part of a facility's work, but as themission of a facility shifts to environmental restora-

Tablel.Appropriations for EnvironmentalRestoration and Waste Management,1989-1994 (In millions of dollars)

Year Appropriation

1989

1990

1991

1992

1993

1994

1,580a

2,274

3,600

4,308

5,520

5,888b

SOURCE: Congressional Budget Office using data from theDepartment of Energy, Office of Environmental Res-toration and Waste Management.

a. Budget authority.

b. Excludes funding for the Uranium Enrichment Decontamina-tion and Decommissioning program.


tion and waste management, EM picks up thesemanagerial and administrative functions—and a cor-responding budget. This shift makes it difficult totrack the growth of funding for DOE's cleanup ef-forts on a consistent basis.2

The $5.9 billion budget for 1994 includes $707million, or 12 percent of the total, for cleaning upDOE's facilities involving civilian uses of nuclearenergy. Over the years, the defense component hasbeen by far the larger share, accounting for between85 percent and 90 percent of the budget each yearsince EM's inception. Although this study is pri-marily concerned with the cleanup of defense facili-ties, the budgetary data included in the followingdescription of the program reflect the nondefensepart as well.

Figure 2.Department of Energy's 1994Budget for EnvironmentalManagement, by Program (In percent)

EnvironmentalRestoration, 30.3

FacilityTransition, 11.1

Other*, 1.8

TechnologyDevelopment, 6.6

WasteManagement, 50.2

Program Components

The Office of Environmental Restoration and WasteManagement is divided into several functional pro-grams. They include Environmental Restoration,Waste Management, Technology Development, Fa-cility Transition and Management, TransportationManagement, and Program Direction.

The program most closely associated with envi-ronmental cleanup is Environmental Restoration,whose 1994 budget is $1.8 billion, accounting for30.3 percent of the EM total (see Figure 2).3 Itsmission of cleaning up inactive facilities and sitesemploys two main types of activities: remedial ac-

2. Some people disagree about whether spending on the DOE clean-up program should be categorized as environmental spending or asnational security spending. These categories are not mutuallyexclusive, however, and DOE cleanup belongs in both. The totallife-cycle cost of nuclear weapons includes the cost of disposingproperly of the products and by-products. In this respect, DOEcleanup is attributable to spending on national security. But thecurrent objectives and the demands on current resources are envi-ronmental.

3. This is the amount specified for environmental restoration in U.S.House of Representatives, Making Appropriations for Energy andWater Development for the Fiscal Year Ending September 30,1994, and for Other Purposes, Conference Report 103-305, toaccompany H.R. 2445 (October 22, 1993). The legislation calledfor a general reduction of $280 million but did not specify fromwhich EM programs it should be taken.

SOURCE: Congressional Budget Office.

a. Includes Transportation Management and Program Direction.

tions and decontamination and decommissioning(D&D). Remedial actions generally deal with con-taminated soil and groundwater, and D&D appliesto buildings, tanks, and structures. DOE and itsregulators have defined about 3,700 sites subject toremedial action and about 500 contaminated struc-tures that require D&D. (Each major weapons pro-duction facility has numerous contaminated sites. Asite is essentially a discrete, well-defined unit atwhich cleanup activity can be self-contained.) Inaddition, DOE is responsible for cleanup at morethan 5,000 properties in the Uranium Mill TailingsRemedial Action Project and the Formerly UtilizedSites Remedial Action Program.4 EnvironmentalRestoration is the EM program that will be affectedthe most by decisions about the ultimate goals andobjectives of DOE's environmental cleanup mission.

Although restoring the environment is perhapsthe ultimate goal of EM, it is not its largest pro-gram. That position belongs to the Waste Manage-ment program, whose budget of $3.0 billion ac-

Department of Energy, Environmental Restoration and WasteManagement Five-Year Plan, Fiscal Years 1993-1997 (August1991), pp. 210-211.

CHAPTER TWO DOE'S CLEANUP PROGRAM 9

counts for 50.2 percent of EM's total in 1994.Waste Management is responsible for treating, stor-ing, and disposing of wastes from ongoing produc-tion as well as from environmental restoration activ-ities. It also constructs new treatment, storage, anddisposal facilities and performs cleanup actions re-quired to bring DOE facilities into compliance withall applicable laws and regulations as soon as pos-sible. These "corrective activities" account for only$26.5 million, less than 1 percent of the 1994 bud-get for waste management.

Since the inception of the EM program, about60 percent of the budget has been for waste man-agement and 30 percent for environmental restora-tion, with the remainder split among the other pro-grams described below. The lines between the twomajor activities sometimes blur because some wastemanagement activities are attributable to environ-mental restoration. For example, when wastes aregenerated as a consequence of environmental resto-ration, they generally are turned over to the WasteManagement program for treatment, storage, anddisposal.

Technology Development, with a budget of$397.5 million, accounts for 6.6 percent of EM'sbudget in 1994. The program is charged with theimportant mission of developing technologies forsafer, more effective, and less costly cleanup (seeChapter 4).

The Facility Transition and Management pro-gram was established as a separate function in 1993with the purpose of planning and implementing thetransfer of surplus facilities from defense productionto the cleanup program. Its budget rose sharply,from $17.9 million (0.3 percent of the EM budget)in 1993 to $671.8 million (11.1 percent) in 1994.The growth in this program stems primarily fromthe reclassification of some activities that wouldhave come under Waste Management but are nowassigned to Facility Transition and Management.

The remaining 1.8 percent of EM's budget isfor Program Direction and Transportation Manage-ment. These functions are handled primarily out ofDOE headquarters.

Outlook for Spending

As DOE began its environmental cleanup efforts, itattempted to estimate total costs through completionof the program. In 1988, the department estimatedit would cost $66 billion to $110 billion to clean upthe entire complex.5 In 1992, the General Account-ing Office reported an estimate of $160 billion andsaid costs could be much higher unless technologi-cal breakthroughs were made.6 At a June 1993workshop, a DOE official acknowledged that anestimate of $400 billion was not unrealistic.7 Thisstatement was followed in July by the AssistantSecretary's reference (cited in Chapter 1) to a $1trillion program.

Perhaps because of the uncertainties about ulti-mate goals, DOE has stopped publishing estimatesof total cleanup costs. Most experts now recognizethat uncertainties about requirements, tasks, andtechnologies make any overall cost estimate unreli-able. Thus, nobody knows the potential federalliabilities.

Estimating spending over the next few years issomewhat more tractable because it will be drivenprimarily by budget constraints. DOE has not pub-lished detailed information about its plans but hasreleased budget targets by program. The departmentassumes that budgets for the EM program will growabout 2 percent annually, from the requested levelof $6.0 billion in 1995 to $6.8 billion in 2000 (ex-cluding the Uranium Enrichment Decontaminationand Decommissioning program). Spending over thesix-year period from 1995 through 2000 would total$38.8 billion (see Table 2), and program shareswould remain about the same as in 1994 (see Figure2).

5. Department of Energy, Environmental Safety and Health Reportfor the DOE Defense Complex (July 1, 1988), p. 35.

6. General Accounting Office, Cleanup Technology: Better Manage-ment for DOE's Technology Development Program, GAO/RCED-92-145 (April 1992), p. 1. These estimates were not discounted.

7. Department of Energy's Benchmarking for Cost ImprovementKickoff Meeting, Washington, D.C., June 22-23, 1993.


Table 2.Spending Targets for EnvironmentalRestoration and Waste Management,1995-2000 (In millions of dollars)

Year

1995

1996

1997

1998

1999

2000

Total

TargetSpending Level

5,979 a

6,315

6,439

6,565

6,694

6.826

38,817

SOURCE: Congressional Budget Office using data from theDepartment of Energy, Office of Environmental Res-toration and Waste Management.

NOTE: Excludes funding for the Uranium Enrichment Decontami-nation and Decommissioning program.

a. Budget request.

DOE headquarters accounts for $5.8 billion(14.9 percent) of target funding. Some of thatmoney will be distributed to the field offices asDOE determines which ones have the most pressingneeds.

Accomplishments of theProgram

What has the Office of Environmental Restorationand Waste Management accomplished since itscreation five years ago? Some critics complain thatthe results are scant. For example, in its report onthe 1994 appropriation bill for energy and waterdevelopment, the House Committee on Appropria-tions expressed concern about the increasing costsof the program and the apparent lack of significantprogress in cleaning up contaminated sites. Thecommittee warned DOE not to expect unlimitedfunding for the EM program, especially if DOE isunable to show concrete results from the investmentto date.8 Defenders of the cleanup program respondthat it has had to overcome resistance to changefrom a culture in which environmental, health, andsafety factors were subordinated to weapons produc-tion to one in which they are the central mission.

Targets for 1995 through 2000 are also availableby field office. Data for field offices that manageenvironmental cleanup at more than one facility donot show how much each facility would receive.Still, the data provide a general picture of the fieldoffices' spending plans.

The Richland field office, which managesHanford, would receive $9.6 billion for the 1995-2000 period. That amount represents 24.7 percentof the total, substantially more than any other fieldoffice. Richland manages only the Hanford facility,so all of its funding would go toward cleaning upHanford. Savannah River would receive $4.4 bil-lion, and Albuquerque, Rocky Flats, Oak Ridge, andIdaho would each get about $4 billion over the six-year period.

Assessing the Cleanup Program

The EM program has faced the growing pains com-mon to a new and rapidly growing organization. Asthe budget has spiraled upward, the structure of theorganization has changed, with new program officesbeing formed. Decentralized operations at the facil-ities that produce weapons and reliance on contrac-tors to run the facilities-legacies of the weaponsproduction program-have hampered attempts byheadquarters to establish and coordinate effectivecleanup policies. Although decentralized decision-making can offer many advantages, differences instandard operating procedures among the facilitieshave made it difficult for headquarters to obtain cost

House Committee on Appropriations, Energy and Water Develop-ment Appropriations Bill, 1994, Report 103-135, to accompanyH.R. 2445 (June 17, 1993), p. 111.


and budgeting information on a consistent basis.That makes it exceedingly difficult to decide how toallocate resources to best advantage throughout thecomplex.

Some of the accomplishments of the EM pro-gram are, by their very nature, hard to recognize.The bulk of EM spending has been on waste man-agement activities, many of which are not visibleunless something goes wrong. An accidental releaseof hazardous substances is noticed, but the success-ful control of wastes on a daily basis is not.

Likewise, successes in the Technology Develop-ment program may receive little public attention.Innovations that cut costs or enhance the safety ofworkers are critical to the success of cleanup effortsbut do not directly result in a cleaner environment.Even where innovations yield more thorough remed-iation, the results may not be evident for manyyears to come. The inability to show immediatesuccess may lead DOE to underinvest in technologydevelopment, a problem that is discussed further inChapter 4.

Environmental Restoration Activities

The program that most people associate with clean-up is environmental restoration. Environmental res-toration is a popular activity because it promisesresults that people can see and enjoy. Many peoplefear the potential consequences for their health andthe environment from toxic wastes, especially radio-active materials. They also recognize that DOE'snuclear facilities will be off-limits to other uses un-til they are cleaned up. Accomplishments in theEnvironmental Restoration program, therefore, arewhat many people look for in assessing the successof DOE's overall cleanup efforts.

The legal requirements governing environmentalcleanup may explain in part why accomplishmentsseem elusive. Much of DOE's Environmental Res-toration program comes under the ComprehensiveEnvironmental Response, Compensation, and Liabil-ity Act (Superfund), although federal and state regu-latory agencies have given primary jurisdiction tothe Resource Conservation and Recovery Act atsome DOE sites. Both statutes impose specific pro-

cedural and substantive requirements on polluters;the triparty agreements among DOE, EPA, and stateregulators generally include the completion of theserequirements as milestones.

Before remedial actions or decontamination anddecommissioning can begin at a site, DOE muststudy it to see what contaminants are present andwhat kind of treatment or disposal is appropriate.At sites on Superfund's National Priorities List, thisstage is known as the remedial investigation/feasibility study. These studies have accounted forthe bulk of spending in the ER program and willprobably continue to consume a substantial amountof resources for the foreseeable future.9 Data sup-porting DOE's budget request for 1994 indicate thatabout 53 percent of the funding for environmentalrestoration would be spent for studies, about 35percent for cleanup, and the rest to support a varietyof management functions, including oversight ofcompliance, program direction, and landlord respon-sibilities.10 Inadequate records of past manufactur-ing and disposal processes increase the cost of stud-ies and the length of time they take.

After DOE completes the remedial investiga-tion/feasibility study, the Environmental ProtectionAgency prepares a record of decision, which setsforth what DOE must do to clean up a site. Only afew DOE sites have reached the record-of-decisionstage. That has two implications for the departmentas it tries to respond to criticisms of the cleanupprogram. First, it means that DOE has relativelyfew tangible results, such as decontaminated build-ings or restored sites, to show for the money spenton environmental restoration. Second, it means thatDOE does not have a clear idea about the magni-tude of the problem nor the extent of remediationthat will be required at individual sites.

9. DOE's August 1991 five-year plan was the last to identify assess-ment and cleanup separately. At that time, assessment spendingwas projected to be about equal to cleanup spending in 1993 andonly slightly less through 1997.

10. The data did not contain an explicit categorization. Categorieswere inferred from the description of the activity; many includedthe term "assessment" or "cleanup" in the title. For a small num-ber of activities, accounting for less than 1 percent of the budget,the category could not be inferred from the information available.


The often-heard criticism that the environmentalcleanup program is being studied to death is notunique to DOE. The Environmental ProtectionAgency has come under similar attack and haslaunched an effort to accelerate cleanup at siteswhere the appropriate remedy seems clear.11 DOEhas begun to explore interim remedies at such sites,but for many of them it has too little informationabout the scope of the problem to determine thebest approach for remediation. Achieving the rightbalance—conducting the assessment that is neededbut no more—is more easily said than done.

Perhaps the brightest spot in DOE's cleanuprecord is the Uranium Mill Tailings Remedial Ac-tion Project (UMTRAP), although the growth of itscosts has exceeded estimates.12 This project was au-thorized by the Uranium Mill Tailings RadiationControl Act of 1978, well before the EM office wasestablished. The act directed DOE to stabilize andcontrol uranium mill tailings at 24 designated sitesand about 5,000 properties near those sites in 10states and on two Indian tribal lands.13

UMTRAP has completed surface remediation at20 sites and plans to complete remedial action onall surface sites by the end of 1998.14 It still facesthe challenging task of cleaning up the groundwater.But DOE is confident that it can meet regulatoryrequirements for groundwater in a timely, cost-effective manner. The success of this programcannot serve as a model for DOE's other cleanupprograms, however, because UMTRAP's sites donot have the vexing problems of the larger, morecomplicated weapons facilities.

11. For example, at wood-treating sites on the Superfund list, EPA hasproposed dispensing with detailed assessments and instead pro-ceeding with the usual remedy-burning soil to break down or-ganic contaminants.

12. Uranium mill tailings are the sandy wastes that result from pro-cessing ore to extract uranium.

13. Department of Energy, Environmental Restoration and WasteManagement Five-Year Plan, Fiscal Years 1994-1998, vol. 2,Installation Summaries (January 1993), p. 11-68.

14. Ibid., p. 11-71.

Initiatives for Managingthe EnvironmentalCleanup Program

The Department of Energy has instituted severalefforts to help manage the environmental restorationand waste management activities. It has published aseries of five-year plans that provide blueprints forthe near term, established a program for controllingcosts and improving cost estimates, and stepped upefforts to track progress.

Five-Year Plans

Soon after the Office of Environmental Restorationand Waste Management was established in 1989,DOE published the first of a series of five-yearplans that set forth objectives, timetables, and bud-gets. In addition to serving as management tools,the plans provide a perspective on the growth andevolution of the program, especially the rise and fallof projections for budgetary growth.

The first plan envisioned spending $19.2 billionover the five years it covered, 1991 through 1995.One year later, the 1992-1996 plan called for spend-ing $31.6 billion.

The 1993-1997 plan, published in August 1991,contained two scenarios for budgetary growth. Thefirst, called the preliminary unvalidated case, envi-sioned spending of $40.7 billion over the five-yearperiod. This was the amount DOE estimated itwould need to comply with all legal requirements.The second scenario, called the validated targetlevel, assumed growth of about 10 percent a year--an amount DOE thought possible within overallconstraints on the federal budget. Spending in thatcase would total $28.8 billion over the five years.

The 1994-1998 plan projected budgetary growthof between 5 percent and 10 percent a year, yieldingtotal spending of about $35.5 billion over five


years.15 It is not clear whether DOE will publish aplan for the 1995-1999 period; it may instead skipto the 1996-2000 period since it has already madepreliminary projections for 2000.16 As noted earlier,these projections assume growth of about 2 percentannually and spending of about $32.8 billion overfive years.

Although overall budget constraints are set fromthe top down, the specifics of the five-year plansdepend on estimates made from the bottom up.DOE divides cleanup tasks into basic units calledactivity data sheets (ADSs). Each sheet shows thecost estimates, regulatory requirements, milestones,and other useful information for each year in thefive-year plan. ADSs should provide an effectiveway to track growth in costs, schedule slippages,and other developments over time and, indeed, theyare the building blocks of DOE's Progress TrackingSystem (discussed below).

The usefulness of the ADSs has been limited,however, because DOE has continually changed thescope of work contained in individual sheets. Forexample, for each of the principal sites needingenvironmental restoration at Rocky Flats, the 1993-1997 plan contained two ADSs—one for assessmentand one for cleanup~but the 1994-1998 plan col-lapsed them into one ADS per site, covering bothactivities. That change is easy to track, but otherchanges involved folding several ADSs into one orsplitting one ADS into two or more, thereby makingit virtually impossible to track progress. Overall,roughly 2,000 ADSs in the 1993-1997 plan wereconsolidated into 850 in the 1994-1998 plan.

Some realignment is inevitable in a relativelynew program. Undoubtedly some of the consolida-tions have resulted in more logical groupings oftasks. But continual change thwarts efforts to iden-tify the factors that cause costs to grow and sched-

ules to slip, and to analyze how to facilitate prog-ress in cleaning up the environment.

Improving Cost Control andCost Estimates

Rapidly escalating costs are a perennial problem forlarge-scale programs, and DOE's cleanup effort isno exception. Two independent reviews by analystsoutside the EM office-DOE's Independent CostEstimating (ICE) team and an interagency group ledby the Office of Management and Budget (OMB)and the Army Corps of Engineers—underscored theproblems of estimating and controlling costs.17

Team members visited DOE's field offices andreviewed a large sample of activity data sheets. TheOMB/Corps analysts found that DOE had overesti-mated direct program costs for about 12 percent ofADSs and that the estimates contained relativelyhigh overhead and contingency costs. For a sampleof ADSs, the OMB/Corps group found that DOEhad estimated overhead costs to be 139 percenthigher than for comparable Corps projects; the ICEteam found an estimated $350 million of seeminglyexcessive overhead costs and $450 million in exces-sive contingency costs.18 These findings are dis-cussed in greater detail in Chapter 4.

In response to the independent cost reviews,DOE has launched a "Benchmarking for Cost Im-provement" initiative.19 This project will gather andexamine detailed information about costs in anattempt to make cost estimates more reliable and tokeep costs under firmer control. If successful, this

15. This plan, published on January 19, 1993, was somewhat sketchybecause it was completed during the Presidential transition, whenpolicy changes appeared likely.

16. Alternatively, the department may abandon five-year plans infavor of other managerial tools and reporting mechanisms, such asthe progress report required by the National Defense AuthorizationAct for Fiscal Year 1994.

17. Interagency Review Group, Interagency Review of the Departmentof Energy Environmental Restoration and Waste ManagementProgram, Final Report (April 29, 1992); Army Corps of Engi-neers, Supplemental Report on Cost Estimates (Report to the As-sociate Director for Natural Resources, Energy and Science, Officeof Management and Budget, April 29, 1992); and Gilbert/Commonwealth, Inc., Independent Cost Estimate for the Environ-mental Restoration and Waste Management Five Year Plan, FiscalYears 1993-1997 (Reading, Pa.: Gilbert/Commonwealth, Inc.,November 1991).

18. Interagency Review Group, Interagency Review, p. iii.

19. Department of Energy, Office of Environmental Restoration andWaste Management, Benchmarking for Cost Improvement: FinalReport (September 1993).


effort will provide useful information about whichfactors drive costs and are responsible for cost in-creases at the site or project level.

Perhaps as important as the numerical findingsof the independent reviews is the more general con-clusion that DOE needs to improve its proceduresfor estimating costs. Whether improvements in costestimating by themselves can hold the line on futureincreases in program costs is debatable, but im-provements can certainly help to illuminate thetrade-offs involved in developing alternative pro-gram plans.

Experience at DOE and Superfund sites canserve as a guide for estimating costs of many com-mon remedial activities.20 The problems at someDOE sites are so difficult, however, that there islittle experience to draw on.

At many sites known to be contaminated, theexact nature, extent, and even types of contaminantsare unknown. In many cases, experts must conductextensive and costly studies before they can deter-mine the kinds of measures needed to correct theproblem. The hazardous and radioactive nature ofthe contaminants may require special methods andtechnologies with which DOE has insufficient expe-rience to estimate costs with confidence. The clean-up may require special training and equipment, andcontractors may insist on being indemnified againstresponsibility for future cleanup costs.21 Moreover,the legal requirements concerning the extent andtype of cleanup are subject to change. In short, costestimates made before a site has been analyzed andcleanup requirements have been determined must beviewed as highly uncertain and subject to substan-

20. For a discussion of estimates of the costs of cleaning up Super-fund sites, see Congressional Budget Office, The Total Costs ofCleaning Up Nonfederal Superfund Sites (January 1994).

21. Under the Comprehensive Environmental Response, Compensa-tion, and Liability Act, liability for cleaning up a Superfund site isstrict, joint, and several. Any company that has owned or op-erated a site or disposed of wastes on it is potentially responsiblefor the entire cost of cleaning it up, even if that company contrib-uted only a small fraction of the environmental problem. Section107(d) of CERCLA provides protection from liability to a contrac-tor engaged in cleanup, except for damages resulting from thecontractor's negligence, but some contractors have expressedconcerns nonetheless.

tial change. Appendix B describes DOE's processof estimating costs at a site.

Estimating the costs of cleaning up the entirenuclear weapons complex will continue to presentchallenges for some time. Even though estimates ofthe costs of cleaning up each site (or other manage-able unit of observation, such as the ADS) can bemade with increasing confidence, aggregating andcoordinating them poses additional problems. Forexample, experience might suggest that excavating,storing, treating, and disposing of contaminated soilwould cost a certain amount per ton. But if vastquantities of contaminated soil from numerous sitesare excavated at once, they may overwhelm the ca-pacity of storage, treatment, and disposal facilities.DOE's demands for environmental cleanup servicesmay account for such a large share of the total mar-ket that they would create shortages and bottlenecksthat would drive up costs and also cause delays.

As increasing numbers of site assessments arecompleted, DOE will have a firmer grasp of themagnitude of the cleanup actions that lie ahead.And as DOE gains experience with remedial ac-tions, it will be able to refine its estimates of clean-up costs. But cost estimates depend fundamentallyon the cleanup standards DOE is required to meet.

Progress Tracking System

DOE has been hampered in its ability to estimatecosts and show accomplishments-and to understandwhy some cleanup projects appear to be more suc-cessful than others-by the lack of a comprehensivesystem for tracking performance. To fill that need,it has been developing the Progress Tracking Sys-tem (PTS). DOE implemented an initial version ofthe PTS in October 1991, but that version couldperform only a few functions and was not easy toupdate. The PTS has been under continual develop-ment in an attempt to make it the robust manage-ment information system needed to ensure manage-rial and financial control of EM's programs.

As its name suggests, the system is designed totrack EM's progress toward meeting its cleanup ob-jectives. The PTS reports information about costs,schedules, and technical performance and eventually


should enable DOE managers and others to monitorprogress, identify projects that are experiencingproblems meeting cost and schedule objectives, andtake steps to get such projects back on track.

In September 1993, an independent panel ofexperts in the management and control of large-scale programs found that the PTS had made sub-stantial progress and "has the potential to be thepremier DOE-wide reporting system."22 The panelmade a number of recommendations for improvingthe system. On the accuracy and timeliness of thedata, the panel made the following observation:

Quality of data should be one of the mostimportant objectives of the PTS. This es-tablishes the credibility necessary for man-agement use of the system and its reports.Quality and accuracy of the data beginswith the inputs of baseline budgets, sched-ules (milestones), and actual cost and sched-ule data . . . . The field organizations mustfeel that PTS is an extension of their re-porting system and therefore data input andits accuracy is their full responsibility andthey are being held accountable for its va-lidity . . . . Correct data at the time of inputcan add much to the credibility of PTS.23

Among its other recommendations, the panelsuggested that the PTS be linked more closely toother financial and budget reporting systems atDOE. It also commented that, at present, the PTSis on a current year basis, which does not permitlooking forward or reviewing historical trends-capabilities that would provide valuable informationfor developing cleanup plans. The panel also sug-gested measures that would enhance technical capa-

22. Assessing the U.S. Department of Energy Progress Tracking Sys-tem, Report by the Review Panel (September 24, 1993), p. 1-1.

23. Ibid., pp. 5-10 and 5-11.

bilities, reduce the time and cost of entering datainto the system, and make the system easier to use.

The PTS is able to track differences betweenbudgeted cost, work performed, and actual costs.This information is reported on the activity datasheet and presented in an overall summary. Thesystem is not yet able to maintain a history ofchanges to the budgets and schedules, however. Itcan report a variance between the most recent bud-get and actual spending, but it cannot show howmuch the budget has increased or decreased overthe history of the project. Maintaining a baselinehistory would help analysts track changes in costs-both for individual projects and for the program as awhole-and would better identify factors that in-crease costs or cause schedules to slip.

Identifying changes in the scope of a project orchanges in its management history is difficult withinthe present Progress Tracking System. DOE isconsidering incorporating a mechanism for record-ing and managing such changes, including a formalsystem of review and approval. The Department ofDefense (DoD) uses this kind of approach in man-aging acquisitions of major weapon systems. Eventhough requirements, quantities, and other character-istics of a new weapon system may change overtime, there are decision points throughout the acqui-sition process at which estimates of cost and sched-ule are made. DoD's Selected Acquisition Reportsprovide a way of tracking deviations from plans andattributing them to such factors as changes in quan-tities, changes in requirements, and inflation. Thiskind of information would be very useful for DOE'scleanup program because it would help to identifythe factors that drive up costs and delay cleanup.

The Progress Tracking System is a promisingaddition to DOE's data bank. Pulling togethermassive amounts of data, ensuring their reliability,and making them readily available and understand-able to disparate users is a challenging task, butcontinuing to upgrade and improve the system ap-pears well worth the effort.

Chapter Three

Strategic Policy Issues forDOE's Cleanup Program

D ecisions about a series of broad policy is-sues could dramatically change the shapeand cost of the Department of Energy's

program to clean up the nuclear weapons facilities.Reexamining the goals and objectives of the pro-gram could result in a refocusing of cleanup efforts.The basic questions are what to do and when to doit. Should DOE try to eliminate all risks to humanhealth and the environment, or is some risk accept-able? What should be done with the facilities-restore them for industrial, commercial, or residen-tial use, or for recreation or other purposes? Howfast can and should cleanup occur? Given the enor-mous cost of cleaning up the facilities and limitedbudgets for that purpose, what can DOE do toachieve the greatest benefits per dollar spent oncleanup?

Answers to these questions will affect the type,extent, and pace of cleanup. Given budgetary con-straints, stretching out the cleanup program beyondthe 30 years initially envisioned appears inevitable.DOE and its regulators will have to decide whichobjectives, types of problems, or sites to focus onfirst.

Information about risk to human health and theenvironment at DOE's nuclear weapons facilities isscant. Yet understanding these risks is crucial tomaking decisions about what remedial actionsshould be undertaken immediately, what should bedone in the near future, and what would be desir-able, if not essential, over the longer term. Thediscussion below suggests a need for substantialadditional effort to identify and characterize risksfrom environmental contamination.

Intertwined with the goal of reducing risk aredecisions about eventual land use at DOE sites. At

some, there is little choice. Sites where long-livedradioactive wastes have been disposed of will beoff-limits to other uses permanently. Such sitescould be fenced off permanently or until technolo-gies are developed to allow detoxification. Pluto-nium, for example, remains hazardous for more than25,000 years. Other areas also could be fenced off,at least until their value for alternative use becomeshigh enough to justify remediation. Still other sitesmight be made available for industrial use withoutcleaning them up to as high a standard as sitesopened up for general use, provided that all thenecessary precautions are taken to protect workers.Decisions about eventual land use will depend onboth technical and economic feasibility as well aspublic preferences.

Efforts to reduce health, safety, and environ-mental risks and to restore lands to other uses willoccur within the constraints of the federal budget.As discussed in Chapter 2, reliable estimates ofcleanup costs are difficult to make but will inevi-tably be part of the process of determining goalsand setting priorities. This chapter discusses the useof benefit-cost analysis as a way of determiningwhich cleanup projects would offer the greatestbenefits per dollar spent. It concludes with anexample illustrating the trade-offs.

Background

In 1989, DOE announced the goal of completingenvironmental cleanup at the nuclear weapons facili-ties within 30 years. More recently the departmenthas expressed the goal in terms of meeting all appli-


cable legal requirements by 2019. Returning allDOE sites to a pristine state by 2019 is clearly notrealistic, given the presence of such contaminants aslong-lived radioactive materials and substances thatpersist in groundwater. Meeting legal requirementsis also a challenge, especially where technologiesare inadequate to achieve the prescribed level ofcleanliness. Moreover, legal requirements maychange over time, requiring DOE to meet increas-ingly strict standards.

Most people would agree that imminent hazardsto the public should be eliminated promptly.DOE's program for corrective activities is intendedto remedy such hazards. At Hanford, for example,one of DOE's top priorities is a storage tank con-taining a witches' brew of hazardous wastes thatperiodically releases hydrogen gas in concentratedamounts, raising fears of an explosion.1 Much ofthe environmental contamination at DOE facilities,however, does not appear to pose an imminentdanger to the public. No one really knows the risksat the nuclear weapons facilities, but no strongevidence has been presented of imminent dangersthat DOE is not attempting to resolve. Rather, thereis a broad range of potential risks. For example,contaminants in the soil may be migrating towardunderground aquifers, but at such a slow pace thatreaching them would take many years.

Still other potential problems may pose fewerhazards if cleanup is delayed. Hanford's surplusreactors, for instance, contain a number of radioac-tive materials that decay over time. If DOE post-poned decontaminating and decommissioning thesereactors for 75 years, half of the cobalt 60 woulddecay, significantly reducing potential exposure toworkers.

Should DOE aim to eliminate all potential risks?Or to reduce them to some level that weighs costsagainst risks and is acceptable to the general public?Reaching consensus on this matter may be impos-

sible, but informed debate could contribute to a gen-erally acceptable result.

Many people perceive hazardous waste sites tobe extremely dangerous. Such well-publicized con-taminated sites as Love Canal and Times Beachhave raised public consciousness about hazardouswastes. Polls conducted by the Roper Organizationin 1988 and 1990 reported high levels of publicconcern about Superfund and other hazardous wastesites.2 This concern has been reflected in legislationauthorizing environmental cleanup-in particular theResource Conservation and Recovery Act and theComprehensive Environmental Response, Compen-sation, and Liability Act—and in appropriations.3

Public perceptions have also been a driving force inthe development of the cleanup agreements thatDOE has signed with the Environmental ProtectionAgency and state agencies.

Reaching a consensus about the goals of envi-ronmental cleanup at DOE's nuclear weapons com-plex can be facilitated by discussion and debate thatincludes all interested parties. As the following sec-tions suggest, however, experts, policymakers, andthe general public all need more information than isnow available to make informed judgments aboutgoals and priorities.

Understanding Risks

How dangerous are DOE's nuclear weapons facili-ties? How much of a threat do they pose to health,safety, and the environment? How do risks fromhazardous waste sites compare with risks associatedwith other environmental problems? The answers tothese questions can help shape DOE's plans forcleaning up the hazardous waste sites.

In July 1993, workers installed a pump intended to agitate theliquid waste in the tank. They think the agitation will reducepressure inside the tank by preventing the buildup of hydrogenand other gases. See Inside Energy/with Federal Lands (newsletterpublished by McGraw-Hill, Inc., New York, July 12, 1993), p. 6.

2. Environmental Protection Agency, Office of Policy, Planning, andEvaluation, "Environmental Problem Area Profiles" (July 20,1991), pp. 7, 16.

3. In addition to DOE's program, the Department of Defense has alarge and growing budget for cleaning up hazardous wastes, andthe Superfund program now accounts for about one-quarter ofEPA's budget.

CHAPTER THREE STRATEGIC POLICY ISSUES FOR DOE'S CLEANUP PROGRAM 19

The lack of comprehensive measures of risksfrom contamination at its facilities has hamperedDOE's planning efforts. It has some informationthat was collected at various sites and by variousoffices within DOE, but the information has notbeen coordinated into a unified framework thatwould facilitate comparisons. As noted in Chapter1, the Congress has directed DOE to evaluate risksto health and safety and to report on how they areto be addressed under the agreements DOE hasentered into with the Environmental ProtectionAgency and state regulators. DOE has been di-rected to estimate risks based on the best scientificinformation available; it has not been asked to un-dertake the massive task of performing exhaustiveformal risk assessments.4

In response, DOE has launched a risk manage-ment program whose objective is to bring togetherwhat is known about risks, identify the gaps inknowledge, and seek to fill them. (See Box 1 for adiscussion of risk assessment and risk management.)To assist with this effort, DOE has asked the Na-tional Research Council of the National Academy ofSciences (NAS) to form a committee of experts toreview DOE's risk management. NAS sponsored aworkshop on November 3-4, 1993, for the commit-tee to hear the views of facility managers, regula-tors, and stakeholders, and it issued a report on Jan-uary 4, 1994.5

DOE says it has acted to remove all of theimminent threats that it knows about. The depart-ment recognizes that all sites do not pose equalthreats to public health, safety, and the environment.For example, "a small, 'wet' site in a populated areamay have the potential for higher risk due to theproximity of the community and the migration ofcontaminants than . . . contamination in the middleof a larger, 'dry' site that is far from populated

4. U.S. House of Representatives, Making Appropriations for Energyand Water Development for the Fiscal Year Ending September 30,1994, and for Other Purposes, Conference Report 103-305, toaccompany H.R. 2445 (October 22, 1993), pp. 94-95.

5. National Research Council, Building Consensus Through RiskAssessment and Management of the Department of Energy's Envi-ronmental Remediation Program (Washington, D.C.: NationalAcademy Press, 1994).

areas."6 However, at the NAS workshop on risks,DOE's Assistant Secretary for EnvironmentalRestoration and Waste Management acknowledgedthat the department had not yet identified the risks.

Until detailed investigations of each site havebeen completed, one can only draw inferences aboutpotential risks on the basis of information frommore general studies. Information about risks athazardous waste sites in general may have importantimplications for DOE's problems.

Preliminary Assessments by ATSDR

The Agency for Toxic Substances and Disease Reg-istry (ATSDR) has begun to assess public healthrisks associated with DOE facilities.7 It has com-pleted a preliminary survey of 19 DOE sites and hascategorized them by potential public health risks.8

The rankings are preliminary but will aid ATSDR indeveloping plans for in-depth studies of potentialpublic health problems. Over the next severalyears, ATSDR will compile and analyze data onexposure to hazardous substances at DOE sites. Itwill then assess current and likely future impacts onhuman health from the release of hazardous sub-stances into the environment.

In ATSDR's initial screening, only Hanford wasincluded in the highest potential risk category, de-fined as "sites where exposure is known to be oc-curring or has occurred at levels of contaminationthat might cause acute health effects."9 Fernald, Los

6. Department of Energy, Environmental Restoration and WasteManagement Five-Year Plan, Fiscal Years 1994-1998, vol. 1(January 1993), p. v.

7. Section 104(i) of CERCLA created the ATSDR within the PublicHealth Service to study health effects of toxic substances.

8. Department of Health and Human Services, Public Health Service,Agency for Toxic Substances and Disease Registry, Division ofHealth Assessment and Consultation, Federal Programs Branch,Health Assessment Activities at Department of Energy NationalPriorities List Sites for FY 1992 (November 1992). The sitesinclude 16 DOE sites on the National Priorities List (NPL), oneproposed for the NPL, one jointly listed as a DOE/Army site, andone for which ATSDR received a petition for assessment.

9. Ibid., p. 9. ATSDR considered Hanford areas 100, 200, 300, and1100 separately; it put the 100, 200, and 300 areas in the highestrisk category.


Box 1.Risk Assessment and Risk Management

Risk assessment and risk management are commonlyconfused. Indeed, the terms were sometimes used inter-changeably at the risk-management workshop convened bythe National Academy of Sciences on November 3-4,1993. Risk assessment generally includes the developmentand analysis of scientific information; it is an input—butnot the only input—to risk-management decisions. Riskmanagement consists of actions that are intended to alterthe likelihood of certain events or outcomes.

A risk assessment is based on detailed scientific anal-ysis. It involves identifying substances that may be haz-ardous to human health and determining their effects onhealth. In many cases, the latter entails exposing labo-ratory animals to very large doses of the substance, mea-suring the effects, and then interpreting the implications ofsmaller exposures for humans. A risk assessment mayestimate the number of people who might be exposed to aharmful substance and the length and type of exposure.The product of a risk assessment is usually a measure ofadditional risk, such as one additional cancer case over 20years among 10,000 people living within a given distanceof the hazard.

But in many respects, risk assessment is an art, not ascience. The assessor rarely has complete informationabout every aspect; the art includes a judgment about whatcan safely be left out of detailed analysis. Lacking suffi-cient information about key elements of the analysis, theassessor may have to make numerous assumptions in orderto proceed. How uncertainty about the assumptions ishandled can have an impact on the results. Stating clearlythese assumptions can enhance the credibility of a riskassessment and can facilitate testing the sensitivity of theresults.

People not trained in science may be able to provideinformation that would improve the reliability and thecredibility of risk assessments by augmenting the work ofscientists. For instance, a risk assessment may assumethat a certain number of people live within a certain radiusof a contaminated site for a certain number of years. Butthe local community may disagree with the assumptionsand present evidence that many people live there for lon-ger than assumed, or that the community is growing andmore people are moving into the area where they might beexposed to contaminants. A risk assessment should bepresented in such a way that these alternative assumptionscan be factored in, to see what their effect is on the result-ing estimate of risk.

Risk assessments also may make assumptions aboutthe amounts of various foods people eat. If some ethnicgroups consume much larger amounts of fish, in whichcontaminants bioaccumulate, relative to other sources ofprotein, the potential risk may be underestimated.

Managing risk involves combining the results of riskassessments and many other factors to determine an appro-priate set of actions. These other factors may includecompeting demands on the resources available to reducerisks, the availability of technological remedies, and com-parisons with other risks. For example, one might want tocompare the risks existing at a site with the risks of reme-diation. The latter particularly affect workers-those en-gaged at the site in current activities (for example, moni-toring or maintenance) and those who would performcleanup tasks. Digging up contaminants that have beencontained poses additional risks, as does transporting themto storage and disposal facilities. Transporting contami-nants off-site may be especially hazardous, since DOEcannot control safety on the highways or railroads.

Managing risks involves a number of thorny issues.Among them are:

o How to compare risks that are voluntary (skydiving)with those that are involuntary (living in an area withpoor air quality);

o How to factor in effects on the most susceptiblemembers of the exposed population when those ef-fects greatly exceed the effects on the average person;

o How to account for large differences in expectedexposure (for example, if most people living near ahazard stay there only a few years but some spend alifetime there);

o How to deal with risks that are very large but affectonly a small number of people living near a site;

o How to treat risks that affect various age groups orother population groups differently; and

o How to compare scientific studies whose results differ(for example, somehow "average" the results, placethe greatest weight on the most pessimistic, take whatthe majority of "experts" conclude, and so on).

There are no easy answers in evaluating and manag-ing risks. But the presence of subjectivity does not meanthat risk assessments are not useful nor that consensusabout managing risks is unattainable. If estimated risksfrom different environmental problems vary by orders ofmagnitude, for example, it may be clear which ones topursue first. Where risks appear high in some studies butnot in others, additional research may be an importantelement of risk management. And public debate on themost difficult issues that involve individual and societalvalues can provide useful information for officials as theyset policies for managing risks.


Alamos National Laboratory, Oak Ridge, andHanford's 1100 area were included in the secondhighest risk category, defined as sites where expo-sure is probable, chronic health effects are possible,and/or residents have alleged that health effectshave occurred. Rocky Flats and Savannah Riverwere put in the third category—sites for which thereis limited information and where available informa-tion indicates exposure is occurring or has occurred—but will be subject to additional review. IdahoNational Engineering Laboratory was put in a fourthcategory of sites for which limited information indi-cates that exposure is not occurring. ATSDR sur-veyed other DOE sites as well, but the sites listedabove are the ones with the most environmentalcleanup activity.

ATSDR emphasizes that its categorization ispreliminary and may change as additional informa-tion becomes available. The categories do notrepresent conclusions about public health risks.

EPA Studies of Risks

The Environmental Protection Agency has publishedtwo reports about the risk associated with hazardouswaste sites compared with other kinds of environ-mental problems. The reports are by no meansconclusive, nor do they purport to be. They recog-nize that the state of knowledge about a broad rangeof environmental risks is still in its infancy.10 Still,they provide a comprehensive overview of the rela-tive risks of various environmental problems, in thejudgment of experts. The reports discuss in detailthe importance of understanding how different en-vironmental threats-such as pesticide residues infood, automobile emissions, and leaking drums ofhazardous wastes—pose risks to human health, hu-man welfare, and ecology.11

10. Risk assessment is a complicated undertaking, involving an under-standing of a variety of disciplines, including toxicology, epidemi-ology, and statistics. For a summary of issues surrounding riskassessment, see the Carnegie Commission on Science, Technol-ogy, and Government, Risk and the Environment: Improving Reg-ulatory Decision Making (New York: Carnegie Commission, June1993), especially pp. 76-78.

11. EPA included in "welfare effects" such effects as damage to prop-erty and groundwater supplies, aesthetic losses, and loss of recre-ational benefits.

The EPA reports reach the surprising conclu-sion—at least in view of common perceptions-thathazardous waste sites rank considerably lower inrisk than many other environmental problems. Al-though the DOE facilities may have special prob-lems that would make them rank higher than otherhazardous waste sites, the EPA studies indicate thatother types of environmental problems pose greaterrisks to health and safety than do waste sites in gen-eral.12

The first report, Unfinished Business, containsassessments by EPA experts of risks associated witha variety of environmental problems.13 The reportdrew on the expertise of health scientists, engineers,economists, and other specialists in environmentalprograms administered by EPA. The expertsformed four groups, each focusing on one of fourmajor categories of risks: cancer risks, noncancerhealth risks, ecological risks, and welfare risks.Each group of experts compared risks across the fullrange of environmental programs—clean air, cleanwater, toxic substances, pesticides, radiation, solidwastes, hazardous substances, and so on—and rankedrisks from environmental problems as relativelyhigh, medium, or low.14

Because most of DOE's environmental cleanupinvolves hazardous waste sites-active and inactive-the experts' rankings are useful in providing a senseof the relative threat that each site poses to publichealth and the environment. The rankings are by nomeans conclusive, however. Additional specific

12. A motivation for the EPA studies was to help set priorities inenvironmental policy. DOE is more constrained than EPA, havingto adhere to EPA rules, triparty agreements, and laws and regula-tions set by other regulatory agencies.

13. Environmental Protection Agency, Office of Policy, Planning, andEvaluation, Unfinished Business: A Comparative Assessment ofEnvironmental Problems (February 1987). As the title suggests,the risks addressed by EPA are those remaining; many potentialrisks have already been removed or abated through controls inplace under existing laws and regulations.

14. The groups took slightly different approaches in structuring theirranking systems. The experts assessing cancer risks and welfarerisks ranked the environmental problems separately and thengrouped them into clusters of relatively high, medium, and lowrisks. The experts assessing noncancer health risks and ecologicalrisks presented their findings in clusters reflecting degree of rela-tive risk but did not rank individual environmental problems sepa-rately.


information about risks at individual sites would berequired to establish a more definitive ranking.

Cancer Risks. EPA's experts on cancer risksranked inactive hazardous waste sites eighth highestof the 29 environmental problems studied, andactive sites 13th. These individual rankings puthazardous waste sites in the second riskiest group ofcancer-risk problems. The experts considered inac-tive hazardous waste sites to be less of a threat thanthe exposure of workers to chemicals, indoor radon,pesticide residues on foods, indoor air pollutantsother than radon, exposure of consumers to chemi-cals, hazardous and toxic air pollutants, and deple-tion of stratospheric ozone.15

Noncancer Health Risks. EPA's experts on non-cancer risks to human health rated both active andinactive hazardous waste sites relatively low on therisk scale.16 A major reason for the low ranking isthat despite the moderate toxicity of substances athazardous waste sites, relatively few people are ex-posed, and exposures are generally by indirectroutes and low in concentration. These characteris-tics contrast with problems rated as a higher risk,such as air pollutants, pesticides, and discharges tosources of drinking water and water habitats of ed-ible fish and shellfish.17 Exposure of a small num-ber of people, of course, does not mean that a haz-ard can be ignored. It may have a bearing, how-ever, on the best way to reduce or eliminate a risk.Even though lack of hard scientific data led thework group on noncancer health risks to relyheavily on professional judgment, there was a sur-prising amount of consensus about the rankings.

15. Environmental Protection Agency, Unfinished Business, AppendixI, "Report of the Cancer Risk Work Group," Table 1, pp. 4-10.

16. Ibid., Appendix II, "Report of the Non-Cancer Risk Work Group,"Table 2-1, p. 2-2. Because of uncertainties and lack of data, thenoncancer work group did not numerically rank each type of envi-ronmental problem (although they used a quantitative scoringsystem to help in the ranking process). Instead, they used threecategories of risk: low, medium, and high.

17. Ibid., p. 2-3. Municipal and industrial nonhazardous waste siteswere ranked in the medium-risk category (Table 2-1, p. 2-2). Theexperts considered them riskier than hazardous waste sites becauselarger numbers of people were potentially exposed to pollutants,and in the case of industrial sites, exposure concentrations wereexpected to be higher (Table 2-2, p. 2-7).

Although the experts did not consider the healthrisks at hazardous waste sites to be high, they notedthat if hazardous substances at these sites are notcontained-if they seep into groundwater, for in-stance, or pollute the air-the risks may be muchhigher. Risks also may be high for workers at thesites. These factors may raise the relative threat tohealth at some DOE facilities.

Ecological Risks. The ecology experts consideredrisks of damage to entire ecological systems, to geo-graphical regions, and to the biosphere itself. Theyranked active hazardous waste sites in the lowest ofsix risk groups, and inactive hazardous waste sitesin the second lowest group.18 The primary reasongiven for the relatively low ranking is that environ-mental problems caused by hazardous waste sitesgenerally are highly localized. Moreover, even thelocal environmental impacts typically are low. Theexperts noted, however, that the rankings of ecologi-cal risks of Superfund sites must be consideredtentative because they are based on relatively littledata.19

Welfare Risks. EPA's experts on welfare risksconsidered how environmental problems reduce thevalue of a variety of goods, services, and activities.They reported two types of risks: to communitydrinking water supplies, and to property values ofnearby residences. Examples of welfare effects in-clude damage to property by air and water pollut-ants, diminished enjoyment of natural resources be-cause of pollution, and lower yields from farmingand fishing. Of the 23 environmental problemsranked, inactive hazardous waste sites were rankedninth, and active hazardous waste sites were rankedllth. Like the ecological risk group, the welfarerisk group downgraded the risks of waste sites be-cause "most of their effects are localized, and mostwaste sites are located away from areas of highpopulation density."20

18. Ibid., Overview, pp. 48-49.

19. Ibid., Appendix III, "Basis for Ranking Position."

20. Ibid., Appendix IV, "Report of the Welfare Risk Work Group,"p. 7-10.


Criticisms and Follow-up to EPA'sUnfinished Business

EPA's Unfinished Business represented a major stepin systematically examining risks and comparingdifferent kinds of threats to human safety, health,welfare, and ecology. The EPA Administrator sub-sequently asked EPA's Science Advisory Board(SAB) to review the report, "to examine strategiesfor reducing major risks, and to recommend im-proved methodologies for assessing and comparingrisks and risk reduction options in the future."21

Building on the report's findings that the activitiesand media regulated by EPA presented substantiallydifferent risks, the SAB noted the lack of coordina-tion among environmental laws and programs. Itrecommended measures to integrate environmentalpolicies so as to make use of risk comparisons insetting priorities among policy actions.

The SAB found several shortcomings in Unfin-ished Business, most of which were attributed to thelimited information available about risks and otherrelevant factors. The SAB also expressed concernthat the report defined some environmental prob-lems so broadly as to encompass all gradations ofrisks-thus making assignment to a single risk groupmeaningless. The SAB noted that the environmen-tal problems identified in Unfinished Business werean amalgamation of noncomparable items, includingspecific pollutants, sources of pollutants, and typesof exposure.

Still, the SAB did not dispute the report's rank-ings of relative health risks. Instead of reassessinghealth risks, the SAB's Human Health Subcommit-tee devoted its efforts to seeking ways of resolvinga variety of methodological problems. For example,the subcommittee noted the desirability of develop-ing an aggregate ranking that would combine therelative risks of cancer and noncancer health effects.

It also recommended stepped-up efforts to gaininformation about the health effects of various typesand amounts of exposure to various substances.22

Until better scientific information becomes avail-able, the health subcommittee deemed it "not illogi-cal" to rank as most risky the environmental prob-lem areas with the highest probability of exposinghumans to toxic substances-in effect, the problemareas classified as highest risk in Unfinished Busi-ness.23

For nonhealth risks, the SAB's assessment dif-fered somewhat from that of Unfinished Business.The SAB's Ecology and Welfare Subcommittee rec-ommended moving active hazardous waste sites upone risk category from the lowest to the secondlowest category and moving inactive hazardouswaste sites to a medium-risk category. The sub-committee questioned Unfinished Businesses as-sumptions that current regulation is adequate to dealwith all potential threats of migration of contami-nants from waste sites. It also expressed concernabout release of toxic substances from inactive haz-ardous waste sites and recommended continuedmonitoring of such sites.24

The SAB noted the dichotomy between thepublic's perceptions of risk and the risk assessmentsof environmental professionals. Given the nature ofthe political process, the problems the public per-ceives to be serious have received greater attentionin environmental legislation, regulation, and budgetsthan have problems deemed serious by experts.This dichotomy probably also applies to the DOEcleanup program: the public appears to be greatlyconcerned about hazardous wastes, even when theyare properly stored and considered by experts topose relatively little threat to health, safety, or theenvironment (the limited information available aboutactual risks at each site, however, precludes a defin-itive assessment). The SAB recommended improv-ing the public's understanding of environmentalrisks.

21. Environmental Protection Agency, Science Advisory Board, Re-ducing Risk: Setting Priorities and Strategies for EnvironmentalProtection, SAB-EC-90-021 (September 1990), p. ii (cover letterfrom SAB Chair Raymond Loehr and Co-Chair of Relative RiskReduction Strategies Committee Jonathan Lash to EPA Adminis-trator William K. Reilly).

22. Ibid., Appendix B, pp. 7-8.

23. Ibid., p. 93.

24. Ibid., Appendix A, pp. 50-53.


Conclusion

Much work remains to be done in assessing health,safety, and environmental risks. Understanding therisks at DOE's nuclear facilities-as well as thosefrom other environmental problems—is essential fordirecting resources to where they buy the mostprotection from risks. Responsibility for studyingthese risks does not necessarily lie with DOE, how-ever. Such agencies as EPA and ATSDR mostlikely have a comparative advantage in conductingrisk assessments because protecting the public fromhealth and environmental risks is central to theirmissions. The National Academy of Sciences'effort to advise DOE on risk assessment and riskmanagement should make a valuable contribution tounderstanding risks.

Weighing Benefits and Costs

Risk assessment in its current state has many short-comings. What role, then, can it play in formulat-ing public policy? Although in some instances theuncertainties are so great that even trying to rankrelative risks is fraught with peril, the findings ofrisk assessments can help inform the debate on howto proceed with hazardous waste sites. Nonetheless,extremely difficult choices remain to be determinedby the political process, guided by informed publicopinion. How does society value preserving anecological resource? Is a given reduction in risk ofcancer today more valuable than the same reductionin risk 50 years hence? What about one cancer casetoday versus 10 cases in 100 years?

If resources to reduce risks were unlimited,setting priorities would not be a problem. Butchoosing to spend cleanup money to address certainenvironmental problems has the opportunity cost ofnot addressing others that may pose bigger risks.How do the threats of DOE's environmental prob-lems compare with others? What criteria could beused to help decide how much should be spent oncleaning up DOE facilities? How can the nation getthe maximum benefit for the money spent on envi-ronmental cleanup?

Risks and Resources

The amounts of resources the federal government isdevoting to environmental problems do not closelyparallel the risks as assessed by the experts. Forexample, EPA's experts considered risks to humanhealth from indoor air pollutants (including radon,asbestos, and other pollutants) to be high; in con-trast, a Roper poll found public concern about in-door air pollutants to be low. EPA's budget in1992 included just $28 million for indoor radon and$32 million for other indoor air pollutants. In com-parison, EPA's Superfund budget was more than$1.7 billion, the environmental restoration compo-nent of DOE's cleanup budget (the part most likeSuperfund) was $1.1 billion, and the entire DOEcleanup budget was $4.4 billion.

These numbers do not necessarily indicate thatthe federal government is spending the wrongamount on these problems; that conclusion wouldrequire at least two types of additional information.First is the matter of the appropriate role of thefederal government in reducing environmental risks.In some cases, such as radon in homes, the appro-priate federal role may be to make people aware ofa potential problem but to leave testing and mitiga-tion to the homeowner. For many environmentalproblems, the federal government requires compa-nies to reduce pollution using their own resources.For the DOE complex, however, the environmentalresponsibility clearly rests with the federal govern-ment.

The other type of additional information neededconcerns the benefits to be gained from spending onenvironmental problems. Ideally, the federal envi-ronmental budget would be allocated so as toachieve the greatest benefits per dollar spent.

Risk Trade-offs

If reducing existing risks is a primary goal of envi-ronmental cleanup programs, then it also is impor-tant to avoid increasing risks when carrying outcleanup activities. Some environmental cleanupactions present their own set of increased risks. Insome cases, for example, excavating and burning


soil to rid it of contaminants may cause releases ofhazardous air pollutants. Transporting wastes fromdispersed sites to a central disposal facility carrieswith it risks of accidental release of hazardous ma-terials.25 Thus, unless hazardous substances areleaking or otherwise migrating into the environment,it may be safer to leave them where they are. Andit may be safer to contain spills or leakages than totake more active measures to remove or destroy pol-lutants.

Using Benefit-Cost Analysis

Because environmental cleanup places large de-mands on limited resources, cleanup dollars must bespent as effectively as possible. That can be accom-plished by comparing the benefits and costs of alter-native plans to see which is likely to yield the larg-est net benefits.

Benefits of environmental cleanup include re-ducing risks to human life and health, mitigatingharmful effects on plants and animals, improvingthe quality of the environment, maintaining bio-diversity, and so on. Measuring—or even com-pletely identifying—these benefits presents signifi-cant challenges because of the many uncertaintiesabout the way hazardous wastes migrate into theenvironment and affect the health or welfare ofhumans and other living things. Consequently,attempts to quantify the benefits from reducingexposure to risks should make clear the uncertain-ties attached to them.

Once benefits have been identified, they must beevaluated; that is, what value do people place onthem? Estimating values of benefits is not easybecause no active market exists for many types ofbenefits. Perhaps the most difficult challenge isfiguring what value to place on reducing the risk ofpremature death. One method is to observe whatpeople are willing to pay, or give up, to reduce

risks, such as equipping their cars with antilockbrakes or installing smoke detectors in their houses.The offering of higher wages to get workers toaccept dangerous jobs provides additional evidenceabout how people value risks.26 The federal govern-ment implicitly places values on risk reduction inthe regulations it issues. The estimated cost perpremature death averted ranges from $100,000 tomore than $5.7 trillion (see Table 3).

Placing a value on natural resources is alsodifficult. For some lands, private market transac-tions enable comparisons. For instance, the value ofa DOE facility offered for industrial use could bedetermined by the amount that businesses are pay-ing for comparable facilities. But it is harder toestimate how much people are willing to pay for re-storing a site to its pristine condition. Some evi-dence can be derived from observing how muchpeople are willing to spend on national parks or forrecreational resorts. Additional information can begained from surveys in which people have beenasked how much they would be willing to pay for avariety of attributes of environmental quality.27

One can sometimes get around the problem ofplacing a direct value on potential benefits by in-stead comparing the cost-effectiveness of alternativepolicies to environmental cleanup. This approach isbased on how much can be achieved (in terms ofrisks reduced, or any other objective) per dollarspent. Alternative policies can then be compared interms of the benefits they produce. Cost-effective-ness is likely to vary greatly among sites and amongthe types of waste to be cleaned up.28

25. See, for example, Department of Energy, Decommissioning ofEight Surplus Production Reactors at the Hanford Site, Richland,Washington, Draft Environmental Impact Statement, DOE/EIS-0119D (March 1989), p. 3.51.

26. Economists have written many books and articles on this subject.For a recent reference, see Kip W. Viscusi, Fatal Tradeoffs:Public and Private Responsibilities for Risk (New York: OxfordUniversity Press, 1992).

27. These surveys are known as "contingent valuation" studies in theeconomics literature. For a more complete discussion, see A.Myrick Freeman III, The Measurement of Environmental andResource Values: Theory and Methods (Washington, D.C.: Re-sources for the Future, 1993).

28. Cost-effectiveness analysis can also be used to compare optionsfor meeting a narrower set of objectives. Examples include find-ing the least-cost method of reaching a given regulatory standardor the least-cost method of treating a given volume of waste.


Table 3.Implicit Costs per Life Saved of Selected Regulations

Regulation

Unvented Space Heater Ban

Trihalomethane Drinking Water Standards

Aircraft Floor Emergency Lighting Standard

Auto Side Door Support Standards

Hazard Communication Standard

Standards for Radionuclides in Uranium Mines

Benzene Occupational Exposure Limit

Hazardous Waste Listing for Petroleum Refining Sludge

Coke Ovens Occupational Exposure Limit

Diethylstilbestrol (DES) Cattlefeed Ban

Hazardous Waste Land Disposal Ban

Hazardous Waste Listing for Wood-Preserving Chemicals

YearIssued

1980

1979

1984

1970

1983

1984

1987

1990

1976

1979

1988

1990

Agency

CPSC

EPA

FAA

NHTSA

OSHA

EPA

OSHA

EPA

OSHA

FDA

EPA

EPA

Cost per PrematureDeath Averted

(Millions of1990 dollars)

0.1

0.2

0.6

0.8

1.6

3.4

8.9

27.6

63.5

124.8

4,190.4

5,700,000.0

SOURCE: Congressional Budget Office using data from Budget of the United States Government: Fiscal Year 1992, Part Two, p. 370, TableC-2. Cost estimates were based on John F. Morrall III, "A Review of the Record," Regulation, vol. 10, no. 2 (1986), p. 30.

NOTES: CPSC = Consumer Product Safety Commission; EPA = Environmental Protection Agency; FAA = Federal Aviation Administration;NHTSA = National Highway Traffic Safety Administration; OSHA = Occupational Safety and Health Administration; FDA = Food andDrug Administration.

Example of EPA's Use of Benefit-CostAnalysis

Some of these issues are illustrated in an analysis ofa proposed EPA regulation concerning facilities atwhich wastes are treated, stored, and disposed of(see Box 2). The analysis, called a regulatory im-pact analysis, identifies the benefits and costs of aproposed rule governing corrective actions at facili-ties regulated under the Resource Conservation andRecovery Act.29 The analysis is intended to providea framework for evaluating the costs and benefits ofalternative policy approaches to environmental

cleanup at RCRA sites. It points up deficiencies ininformation about costs and benefits, and it identi-fies trade-offs facing policymakers.

Although the regulatory impact analysis did notcover the largest DOE facilities, it provides an ex-ample of the type of analysis that would be usefulin understanding the policy choices and trade-offs at

29. Environmental Protection Agency, Office of Solid Waste, DraftRegulatory Impact Analysis for the Final Rulemaking on Correc-tive Action for Solid Waste Management Units: Proposed Method-ology for Analysis (March 1993).


the nuclear weapons complex. As DOE completesassessments of contaminated sites, a better pictureof risks and potential risk reductions will appear andwill facilitate comprehensive analysis of alternativeenvironmental cleanup activities. As indicated inChapter 2, DOE is working on improving its abilityto estimate costs, although additional informationand refinements are needed there as well.

Deciding When to Clean Up

Benefit-cost analysis can help guide policy decisionsabout both the timing and the extent of cleanup.For each site requiring remediation, DOE will haveto decide whether to clean it up immediately, delaycleanup temporarily, or postpone cleanup indefi-nitely.30 In each case, DOE must consider severalfactors: the cost; the reduction in risk to workers,the public, and the environment; and alternativeuses of land and facilities that are currently notavailable to the public.

In terms of reducing risks and restoring landand facilities to alternative uses, immediate remedia-tion would generally produce the greatest benefit.This benefit must be weighed not only against thecost, however, but also against risks-primarily toworkers—associated with cleanup and disposal.

Benefits and Costs of DelayedRemediation and Permanent Isolation

Delaying cleanup can reduce the risk of exposurecaused by accidental release of radioactive contami-nants to cleanup workers, the general public, andthe environment by allowing time for the naturaldecay of radioactive material. For instance, tritium—a common contaminant at DOE installations--decays at the rate of 50 percent every 12 years.Thus, 24 years from now, 75 percent of the con-taminating tritium will have decayed to a harmless

30. Indefinitely postponing remediation need not be construed aslicense for further contamination of a site. Rather, it could entailstorage at an acceptable level of risk, leaving open the option ofcleanup if new technologies and less constrained budgets permit.

form of helium; after 36 years, 87 percent will havedecayed. Other types of radionuclides decay fairlyquickly also and, although not as widespread astritium, would, at current concentrations, presentsignificant hazards to workers involved in the clean-up. Cobalt 60 and cesium 137, commonly found inreactor cores, are examples of such substances.

Delaying cleanup can also provide time for saferand more effective cleanup technologies to be de-veloped. In addition, cleanup projects that are notaffordable in the near term could become so in thefuture if additional funds become available or ifcost-saving technologies are developed.

Delaying remediation has disadvantages, how-ever, especially if it increases the potential for ex-posing the public to contaminants. Such exposureis particularly likely if the contaminants cannot becompletely contained. Delaying cleanup would alsorequire continued limits on the use of contaminatedlands or facilities.

Permanently isolating a site, especially one thatis particularly difficult to clean up, is a viable op-tion given the remote location of many of DOE'sinstallations. Because of the secrecy of the Manhat-tan Project, for example, the original sites selectedin the 1940s were chosen specifically because theywere far away from population centers. Further-more, to enhance the security of operations, manyof the sites on which DOE facilities were built werevery large, thereby making it difficult for outsidersto observe them. The same characteristics make iteasy to isolate the contaminated sites.

Permanently isolating some sites may be consis-tent with storage requirements for nuclear waste.For the foreseeable future, the nation will need alocation in which to store radioactive materials dis-carded from its military endeavors. Some of thesewastes, though not necessarily highly radioactive,may be rather bulky. Examples of such wastesinclude discarded housings of nuclear reactors fromdecommissioned submarines and the sections ofsubmarine hulls that surrounded the reactors. Theseslightly radioactive components are being stored atHanford. They probably present little immediatedanger to the environment since their low level ofcontamination is unlikely to spread. Ensuring that


Box 2.EPA's Analysis of Hazardous Waste Cleanup

The Environmental Protection Agency (EPA) has recentlypublished a regulatory impact analysis (RIA) that attemptsto identify and estimate the benefits and costs of a pro-posed rule governing corrective actions at facilities regu-lated under the Resource Conservation and Recovery Act(RCRA).1 The analysis is intended to provide a frame-work for evaluating the costs and benefits of alternativepolicy approaches to environmental cleanup at RCRAsites.

EPA's analysis is useful and relevant to the debateabout the Department of Energy's (DOE's) cleanup pro-gram. Although it does not estimate the costs and benefits(risk reductions) of complying with the proposed regula-tion at the most problematic DOE facilities, it sheds lighton the kinds of benefits, the magnitudes of costs, andother implications of the proposed rule.2 Perhaps the mostimportant lessons to be drawn from the RIA are the diffi-

2.

Environmental Protection Agency, Office of Solid Waste,Draft Regulatory Impact Analysis for the Final Rulemakingon Corrective Action for Solid Waste Management Units:Proposed Methodology for Analysis (March 1993). Thediscussion here is not intended to be a comprehensive cri-tique of EPA's analysis, nor an endorsement of its method-ology or findings. Rather, it illustrates the type of analysisthat can help inform policy decisions and the strengths andshortcomings of the current state of knowledge and estimat-ing techniques.

The sample consisted of 359 federal facilities and 5,432nonfederal facilities. But the RIA excluded seven of thelargest DOE facilities—Hanford, Savannah River, RockyHats, Idaho National Engineering Laboratory, Fernald, LosAlamos National Laboratory, and the Nevada Test Site. Italso excluded two large Department of Defense facilities-Rocky Mountain Arsenal and McClellan Air Force Base.

culties of determining costs and benefits and the uncertain-ties about the estimates.

The analysis estimates the total discounted cost ofcorrective actions at the sample facilities to be $18.7 bil-lion. These costs occur over a long period; the estimatedmedian time to remediate contaminated groundwater is115 years for on-site plumes and 90 years for off-siteplumes.3 Most of the costs are associated with removingand treating contaminated soil and groundwater.

The RIA assesses both health and ecological benefits.The expected health benefits over the 128-year modelingperiod include averting 400 to 13,300 cancer cases and100 to 12 million cases in which thresholds for noncancerhealth effects are exceeded. (EPA did not discount thebenefits although it did discount the costs.) In addition tounquantified ecological effects, benefits include cleanup of1.4 million acres contaminated by groundwater and 18million cubic yards of soil. The large range of cancersaverted and the much larger range of noncancer healthbenefits reflect the difficulties of assessing health risks.By the same token, these ranges suggest that devotingmore resources to understanding and evaluating healthbenefits could produce a large payoff in helping to for-mulate cleanup priorities.

Some parts of DOE's cleanup program fall under the pro-posed rule, but other parts come under the jurisdiction of theComprehensive Environmental Response, Compensation, andLiability Act. Most of the same considerations apply, how-ever, regardless of which statute governs.

In the context of environmental cleanup, a plume is ground-water contaminated by flowing through a hazardous sub-stance.

the contaminants at such sites are contained may bepreferable to attempting to remove and treat them.

Deciding against restoring a particular site alsohas other advantages. It would probably minimizerisks to workers since exposure to contaminantscould be limited. In addition, it would avoid thecost of remediation, although those savings wouldhave to be weighed against the costs of maintainingsafety and security at the site.

In addition to the cost of safeguarding and mon-itoring a site, forgoing remediation would have

other drawbacks. The potential for accidentallyexposing people and the environment to the contam-inants would remain. In addition, it would perma-nently preclude other uses of the land and facilities.

The Surplus Reactors at Hanford: AnExample of the Choices Facing DOE

The trade-offs involved in deciding which cleanupactions to undertake and when to do so are illus-trated in DOE's analysis of decommissioning eightsurplus reactors at Hanford. DOE examined several


The problem can be illustrated by considering theimplications of the estimates. Focusing on health benefits,suppose first that the low end of the ranges turns out to becorrect. If all cancer or noncancer health effects result inpremature deaths, premature deaths would total 500. At acost of $18.7 billion, the cost per life saved would cometo $37.4 million.4 If instead the midpoints of the rangeswere correct, then the cost per life saved would be justover $3,000. The latter would be quite a bargain; how-ever, the former would not be such a good deal whencompared with other risk-reducing expenditures (seeTable 3, but note that its estimates are lower because itdiscounted future benefits whereas the RIA did not).

These results leave policymakers little basis for un-derstanding whether the proposed rule would make wiseuse of environmental cleanup dollars, but the RIA offers amodest amount of additional information. For example, itindicates that many of the potential benefits are concen-trated at a relatively small number of facilities. Focusingon those facilities might be a way to make a large prob-lem more tractable.

The RIA suggests another intriguing option. Notingthat the primary benefits of the proposed rule involvereducing contamination in drinking water, the RIA indi-cates that the health effects of the hazardous waste sites inthe sample could be mitigated largely by treating waterthat was destined for use by humans (for drinking, wash-ing, and so on) or by providing water from alternativesources rather than by rooting out all contamination at thesource. Ensuring the safety of water to be used by hu-mans would cost only $4.8 million, according to the RIA.

4. This estimate is calculated as if all the benefits occur im-mediately. In reality, it would be many years before correc-tive actions would avert premature deaths. The estimatealso ignores potential ecological benefits.

As far as protecting human health is concerned, then,the choice would appear to be between spending $4.8million and spending $18.7 billion. Moreover, providingclean water from alternative sources could probably beachieved much more quickly than cleaning up contami-nated groundwater, thereby giving health protectionsooner. Each choice would provide essentially the samereduction in risk to human health. The difference is thatthe $18.7 billion would also buy clean groundwater andsoil—or at least an attempt at cleanliness. As the RIAnotes, some substances in the groundwater cannot be fullycleaned up using existing technologies: pumping and treat-ing is relatively ineffective in removing dense nonaque-ous-phase liquids, for example. If those are the choices,then adopting the more expensive approach would implyplacing a very high value on nonhealth benefits.5

To summarize, the RIA proposes a methodology forevaluating benefits and costs associated with cleaning uphazardous wastes, identifies the kinds of benefits andcosts, and provides some estimates of their magnitude. Itsuggests that knowing how contamination at hazardouswaste sites is likely to affect people's health is essential tomaking sound decisions about remediation. And it pointsout that at least in some cases, there may be less expen-sive ways to reduce risks to human health than undertak-ing massive cleanup efforts. Such alternatives may meetonly the objective of reducing risks to people; they maynot meet the other objective of making sites available forother uses. The choice of objective is ultimately a politi-cal one, but it can be informed by knowledge of the trade-offs.

5. Choosing the option of providing clean water from alter-native sources does not foreclose the option of cleaning upthe contaminated groundwater. Using alternative sourcescould be considered an interim action, pending developmentof technology that would offer cheaper or more effectivetreatment of groundwater.

alternatives: immediate removal of the reactors to adisposal site, delayed removal, and permanent isola-tion with no remedial action.31 As criteria for com-parison, DOE considered the cost of each alternativeover the next 100 years and the number of excesscancer deaths during the next 10,000 years causedby contaminants in the reactors.

31. Department of Energy, Decommissioning of Eight Surplus Produc-tion Reactors at the Hanford Site, Richland, Washington, Adden-dum (Final Environmental Impact Statement), DOE/EIS-0119F(December 1992). DOE considered dismantling the reactor coresas well as one-piece removal, but the latter is most relevant to thecurrent discussion.

Immediate Remediation. DOE considered the op-tion of immediately removing the reactor cores totheir final resting place and dismantling the rest ofthe reactor buildings. DOE estimated that this ac-tion, which would take about 12 years for all eightreactors, would cost about $190 million (in 1990dollars) and that the annual cost to monitor ground-water at the disposal site would be about $400,000.The total cost of this option over the next 100years, therefore, would be about $230 million (seeTable 4). DOE did not take account of the effectsof time in its cost estimates, however. At a dis-count rate of 2 percent, the cost of immediate reme-


Table 4.Costs and Cancer Fatalities Under Three Alternatives for Decommissioning Reactors at Hanford

Alternative

Cost Over 100 Years(Millions of dollars)

In 1990Dollars Discounted8

OccupationalDose

(Person-rem)bCancer Fatalities

Occupational0 Population

Immediate Remediation

Delayed Remediation

Permanent Isolationwith No Remediation

230

190

40

180

60

20

159

51

24

0

0

1

1

20

SOURCE: Congressional Budget Office using data from Department of Energy, Decommissioning of Eight Surplus Production Reactors at theHanford Site, Richland, Washington, Addendum (Final Environmental Impact Statement), DOE/EIS-0119F (December 1992).

NOTE: Remediation includes one-piece removal and disposal of reactor core.

a. Assumes a real annual rate of 2 percent.

b. rem = roentgen equivalent man, a measure of exposure to radiation.

c. Over a period of 100 years.

d. Over a period of 10,000 years.

diation would be $180 million over the next 100years.32

DOE concluded that the public would not beexposed to radiation during the transportation of thereactor cores or dismantlement of the buildings.Nor would the workers be sufficiently exposed tocause cancer fatalities, according to DOE's analysis.Finally, the land on which the reactors are currentlysited would be available for unrestricted use at theend of the 12-year remediation process.

Delayed Remediation. Another alternative DOEexamined was to wait 75 years before removing thereactor cores from their present locations and trans-porting them to the disposal area. By allowing timefor some of the radionuclides to decay, DOE esti-mates that this option would reduce the exposure of

32. At a discount rate of 4 percent, the cost would be $160 million; atthe 7 percent discount rate favored by the Office of Managementand Budget, the cost would be $140 million.

workers by more than two-thirds. Although DOEconcluded that exposure rates would not result inany cancer-related deaths of workers even if remedi-ation was undertaken immediately, it asserted thatthis delay would further reduce the risks.

DOE also determined that a 75-year delay inremediation would not increase risks to the publicfrom the contaminants in the reactors. By continu-ing security and maintenance of the reactors, DOEis confident that those risks would not significantlyincrease.

Finally, the cost to DOE to remediate 75 yearsfrom now and provide security and maintenance ofthe reactors in the meantime would be less than thecost to remediate immediately, when the time valueof money is considered. According to DOE, thecost to guard and maintain the eight facilities beforeremediation would be about $400,000 per year (in1990 dollars). DOE estimates that remediationwould cost about $190 million in 1990 dollarswhether it takes place now or in 75 years. At a


discount rate of 2 percent, however, the cost of thisoption would be $60 million-roughly one-third ofthe cost of immediate remediation.33

Postponing remediation for 75 years would haveone major disadvantage: it would preclude for a sig-nificant period of time any alternative uses of theland on which the reactors are situated. The reac-tors are located on or near the banks of the Colum-bia River—a popular recreation area and prime agri-cultural land—and that land would be unavailable forpublic use for at least 75 years.

Permanent Isolation of the Reactors. DOE alsoevaluated the option of leaving the reactors, withtheir cores and associated other nuclear waste, inplace permanently rather than removing them to adisposal site. Given that DOE estimated an annualcost of about $400,000 to guard and maintain theeight facilities, total undiscounted costs over thenext 100 years would amount to approximately $40million. At a discount rate of 2 percent, the cost ofthis approach would be about $20 million, signifi-cantly less than either immediate or delayed remedi-ation.34

Some important disadvantages, however, areassociated with this option. The first is the potentialfor increased exposure of the public to the reactorsand their contents if the reactor buildings were everallowed to deteriorate. DOE illustrates this risk byassuming that the reactors are abandoned after 100years. Without routine maintenance, the reactorbuildings would deteriorate, leading to the potentialrelease of radionuclides to the environment and thepotential exposure of humans to radioactivity andother safety hazards if people entered the site. As aconsequence, DOE estimates that maintaining the

33. At a discount rate of 4 percent, the cost of delayed remediationwould be about $20 million; at 7 percent, it would be about $7million. DOE's undiscounted cost estimates for the two optionsinvolving remediation are very similar because the annual costs tomonitor the quality of groundwater at the disposal site, which areincurred after immediate remediation, are approximately equal tothe annual costs to guard and maintain the facilities before delayedremediation.

34. At a rate of 4 percent, the discounted cost would be about $10million; at 7 percent, about $6 million.

reactors for 100 years and then abandoning themcould result in 20 excess cancer fatalities over thenext 10,000 years.35

The potential for exposure leading to fatalitiesexists even after 100 years because some of theradionuclides that are present in significant quanti-ties will be radioactive for a very long time. Car-bon 14, an element commonly found in nuclearreactors, has a half-life of 5,730 years. Thus, at theend of 100 years, almost 99 percent of the carbon14 would remain. In addition, lesser amounts ofeven longer-lived radionuclides, such as uranium238 and chlorine 36, would remain.

Finally, this option would preclude using theland or structures for any other purpose, such asrecreation or agriculture. DOE did not include anestimate of the value people place on these uses, butit should be factored in to any decision about ulti-mate land use.

Choosing Among the Alternatives. Choosing oneof these alternatives requires weighing the costs andbenefits. The least costly alternative is permanentisolation with no remediation. The costs in dis-counted dollars at a rate of 2 percent would beequal to $180 million, $60 million, and $20 millionfor immediate remediation, delayed remediation, andno remediation, respectively. On the benefit side,DOE's analysis suggests that no cancer fatalities(among either the public or workers) should resultfrom exposure to contaminants in the reactors dur-ing a period of 100 years. That is, no lives wouldbe saved over the next 100 years by DOE's invest-ment in remedial action, arguing for the cheapestalternative of no remediation.

Performing no remediation appears less attrac-tive, however, if one believes that DOE mightsomeday stop maintaining the reactors. In that case,

35. The estimated dose from exposure to contaminants in reactors ismuch lower than the dose that would result from natural radiation.In fact, DOE estimates that the same population would receive adose 200 times higher from natural radiation over the 10,000-yearperiod than from contaminants in the reactors during the sameperiod.


some fatalities might occur. Based on DOE's anal-ysis, which assumes the reactors are abandoned after100 years, 20 fatalities might result over 10,000years without remediation, compared with one ifremedial action was taken, either immediately orafter 75 years (see Table 4).

Based in part on this reasoning, DOE has de-cided to undertake remedial action within 75 years.36

DOE is currently working on the conceptual designand has not determined when removal of the reac-tors will begin.

Setting Priorities

Because DOE is responsible for cleaning up thou-sands of sites and because available resources willbe constrained, the department, with guidance fromthe Congress and other interested parties, will haveto make decisions about which cleanup tasks totackle first and which to defer. To make the bestuse of limited budgetary resources, DOE mustweigh the benefits and costs of various cleanup pro-jects when establishing priorities among them. Inother words, DOE should focus its attention on pro-jects that would yield the greatest benefit—in termsof deaths averted or land and facilities returned tothe public domain—per dollar invested.

But as noted throughout this chapter, the de-tailed information necessary to make such choices isoften not available. In many cases, DOE does notknow the nature or extent of contamination at itsfacilities, nor the health risks posed.

Additional Information Needed forRational Decisionmaking

Thus, before DOE can make choices about how toclean up its complex, it will need additional infor-mation and tools. First, DOE must know the extentof the task it faces. To gain this knowledge, it

must determine the type and extent of contaminantswithin its complex. DOE is in the process of char-acterizing its sites and will continue to do so for thenext few years.

Second, once the extent of the contaminationhas been determined, DOE, in order to decidewhether the problem requires remediation, mustknow what risk it poses to human health and theenvironment. Expanding the body of knowledgeconcerning health risks posed by pollutants is essen-tial not only to DOE's cleanup program but also toenvironmental policy in general.

Next, DOE must have efficient and effectivetechniques to accomplish cleanup. As discussed inChapter 4, this area might deserve added emphasis.

Finally, DOE needs standards for acceptablelevels of contaminants, to indicate when cleanuptasks are complete. Such standards would be tied inpart to the risks posed to human health, a questiondiscussed above. An additional criterion for deter-mining cleanup standards, however, is the ultimateuse to which the facility or land will be put. If theland is going to be used for a park, it will have tobe cleaned to pristine standards. But if it will beused as a national nuclear waste dump, then it willnot have to be cleaned up at all. Thus, before DOEcan make choices concerning how to conduct vari-ous cleanup projects, it will have to determine, inconjunction with the public and its representatives,the ultimate use of its surplus facilities.

Models for Setting Priorities

DOE has made a number of attempts to develop aprocess for setting priorities for its environmentalactivities. In consultation with state and local offi-cials and other interested groups, DOE developed amodel for setting priorities in the EnvironmentalRestoration program that incorporated risks to hu-man health, uncertainty, environmental impact,socioeconomic impact, regulatory commitments, andfuture costs. The Waste Management program hasbeen working on a similar model.37

36. Department of Energy, "Record of Decision: Decommissioning ofEight Surplus Production Reactors at the Hanford Site, Richland,Washington," Federal Register, vol. 58, no. 178 (September 16,1993), pp. 48509-48513.

37. For a thorough discussion of setting priorities, see CongressionalResearch Service, Setting Priorities for Department of EnergyEnvironmental Activities, Report 91-150 ENR (February 6, 1991).


In the wake of substantial criticism, however,DOE seems to have abandoned these modeling ef-forts. For example, the environmental restorationmodel was used in developing the fiscal year 1992budget request but was set aside during preparationof the 1993 budget. Critics have complained of anumber of shortcomings, especially inadequate con-sultation with affected parties and the use of subjec-tive weights on which there was considerable dis-agreement.38 In reality, moreover, regulatory re-quirements have preempted the results of the model,since DOE has had to devote essentially all of itsnear-term resources to specific actions set forth inits agreements with the Environmental ProtectionAgency and state regulatory authorities.

Public Involvement in SettingObjectives and Priorities

DOE's experience with trying to develop a model toset priorities has underscored the importance of in-volving the public in the process. Consequently,DOE has stepped up efforts to increase public par-ticipation in establishing cleanup policies. The de-partment is participating in an effort led by the En-vironmental Protection Agency and aimed at findingways of improving communications between federalagencies responsible for cleaning up waste sites andinterested individuals and groups. The Federal Fa-cilities Environmental Restoration Dialogue Com-mittee, also known as the Keystone group after itsfacilitators, the Keystone Center, includes represen-tatives of federal agencies, tribal and state govern-ments, and environmental, community, and labororganizations. The committee has recommendedthat federal agencies keep citizens informed aboutenvironmental cleanup plans and solicit their viewsabout objectives and priorities.39 DOE has indicated

a willingness to adopt the Keystone group's recom-mendations to improve the lines of communica-

40tion.

The Keystone process is based at the facilitylevel, not the national level, which has several ad-vantages. First, people living near a facility are theones who would most directly experience the effectsof cleanup policies—effects such as reductions inrisk, availability of facilities and land for other uses,changes in employment, and effects on transporta-tion (especially if wastes are transported off-site).They also may bring valuable knowledge of theneeds and preferences of the local community—itsvalues and heritage as well as socioeconomic fac-tors-which may vary from facility to facility.

In some communities in which DOE has been amajor employer, residents largely support the decon-tamination of facilities to make them available forother government, industrial, or commercial use.Some people may consider keeping jobs and pre-serving the economic base of the community to bemore important than removing every trace of con-tamination—at least in the near term. Others mayconsider it a moral obligation to future generationsto restore expeditiously all lands to their pristinestate. In particular, some DOE facilities are onlands considered sacred by Native Americans.

The Keystone group recognized that federalfunding might fall short of the amount needed tomeet legal requirements at federal facilities. Itproposed that budget shortfalls be allocated on anequal percentage basis among all facilities. This ap-proach points up a disadvantage of focusing onindividual facilities in that national priorities maydiffer from local priorities. If one facility wasfound to pose much more serious risks than others,

38. For a detailed critique, see James D. Werner, "Comments on Be-half of the Natural Resources Defense Council and the Environ-mental Defense Fund Regarding the Proposed Department ofEnergy Priority System for Environmental Restoration" (NaturalResources Defense Council, Washington, D.C., November 21,1991).

39. Federal Facilities Environmental Restoration Dialogue Committee,Interim Report: Recommendations for Improving the Federal

Facility Environmental Restoration Decision-Making Process andSetting Priorities in the Event of Funding Shortfalls (February1993).

40. Statement of Paul D. Grimm, then Acting Assistant Secretary ofEnergy for Environmental Restoration and Waste Management,before the Subcommittee on Energy and Water Development ofthe House Committee on Appropriations, April 26, 1993.


a broader perspective would probably favor shifting41resources to it.

Conclusion

Some people might argue that setting priorities be-yond simply complying with regulations and agree-ments is unnecessary, or at least not productive,since compliance will take most of the resourcesDOE has available for the foreseeable future. Oth-ers worry that if DOE targets its efforts toward

41. The additional resources could come from sources other than theDOE cleanup budget. But raising taxes or reducing federal spend-ing on other services to pay for it would be more difficult, giventhe framework of the budget process.

cleaning up the most severe problems first, it maylose the will or the resources to take care of smallerproblems later. But any cleanup project uses re-sources that could be spent on some other activity,and society as a whole can gain the most by devot-ing those resources to activities that produce themost benefits.

Cleaning up all areas to pristine standards maybe the ultimate goal desired by the public. Butattaining that objective will be very expensive. Se-lectively reconsidering cleanup standards, and ac-cepting some level of risk greater than zero, couldsubstantially lower total costs. That could free pub-lic funds for other programs, such as environmentalcleanup efforts deemed to be of higher priority thansome of DOE's cleanup problems.

Chapter Four

Issues Related to DOE'sCleanup Program

A lthough appreciable uncertainty surroundsthe total cost of the Department of Energy'scleanup effort, the department can make

choices as it proceeds with the cleanup that mayreduce the cost of the program. This chapter ex-plores four areas of DOE's current programs thathave been the subject of Congressional, departmen-tal, or public concern-DOE's efforts to developnew cleanup technologies, the schedule for cleanupactivities dictated by interagency agreements, DOE'shigh overhead costs, and its large number of surplusfacilities.

Investing to Develop BetterTechnologies

Cleaning up the Department of Energy's nuclearweapons complex could, by some estimates, take 30years or more and cost hundreds of billions of dol-lars. To reduce both the cost and duration of thecleanup effort, DOE has initiated several programsto develop new technologies and processes for char-acterizing and cleaning up its sites. Based on re-search and analyses that have been completed, DOEfeels that it can save potentially large amounts ofmoney in the course of its cleanup by using newtechnologies. Yet the share of its budget that DOEdevotes to developing new technologies is lowcompared with its own target and may be insuffi-cient to fund development of some new cleanuptechniques that could reduce costs.

DOE's Funding for TechnologyDevelopment

DOE has already invested considerable amounts ofmoney to develop new techniques for characterizingand cleaning up contaminated sites. Since fiscalyear 1990, when it created a separate Office ofTechnology Development within the EnvironmentalRestoration and Waste Management program, DOEhas allotted about $1.5 billion to research in envi-ronmental technologies, including nearly $400 mil-lion in funds for fiscal year 1994.

Annual funding for DOE's Technology Devel-opment (TD) program will probably increase abovethe level appropriated for fiscal year 1994, thoughthe growth may be slower than envisioned in previ-ous five-year plans. DOE has asked for $426 mil-lion for fiscal year 1995 to fund technology devel-opment and plans to increase funding to $467 mil-lion by 2000 (see Figure 3). Previous budget plans,however, had included higher levels of funding fortechnology development through the end of thisdecade. The five-year plan submitted in August1991 included two budget levels that would haveincreased that funding to about $500 million andabout $540 million, respectively, by 1997. Indeed,the Administration's plans as of June 1993 alsoincluded levels of funding for technology develop-ment that were higher for the years after 1996 thanthose currently planned, reaching $500 million by1999.


Figure 3.Funding for Technology Development Within theOffice of Environmental Restoration andWaste Management

500Millions of Dollars of Budget Authority

400

300

200

100

Planned Budget

Appropriated

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000Fiscal Year

SOURCE: Congressional Budget Office based on data from theDepartment of Energy.

Research and Development Projects. Within thebudget for technology development, the research,development, demonstration, testing, and evaluation(RDDT&E) account commanded slightly more thanhalf of TD funding in 1992 and 1993. This accountincludes funds for research on particular cleanupand assessment techniques in three broad areas-groundwater and soils cleanup, waste retrieval andprocessing, and pollution prevention. In each ofthese categories, DOE is conducting individual pro-jects called integrated programs and integrated dem-onstrations to investigate specific techniques andmethods (see Table 5).

Because detailed plans outlining how the fundsfor technology development will be spent for thenext five years are not available, it is impossible tosay specifically what the funding will buy. Basedon previous spending and the request for fundingfor 1995, however, the two research and develop-ment areas most likely to receive the largest sharesof the research funds are groundwater and soilscleanup, and waste retrieval and processing. Fund-ing for these two categories accounts for more than

two-thirds of the RDDT&E funding from 1992through 1995 (see Table 6).

Other Activities in the Technology DevelopmentProgram. Although more than half of the TDfunds support activities in RDDT&E in the yearsfrom 1992 through 1995, the remaining funds—anaverage of 42 percent of total funding—go to otheraccounts that do not support research activitiesdirectly but that DOE feels are necessary to supportother technical needs within the EM program (seeTable 6).1 One program provides laboratory andanalytic support to characterize waste, water, andsoil samples. Another program seeks to develop ascientific, technical, and educational system to en-sure that an appropriately educated work force willbe available in the future.

The Benefits of New Technologies

DOE feels that it has already benefited from its $1.5billion investment in technology development. In-deed, it often points to two examples of the directbenefits of that effort. The first is a $15 millionsavings realized by capping the radioactive wastestored in a silo at the Fernald installation. (Cappinginvolves covering the waste in the silo with benton-ite to seal it and reduce the amount of radon re-leased to the surrounding environment.) Thesesavings were achieved by using a remotely con-trolled robot to measure accurately the three-dimen-sional contours of the solid waste in the silo. Usingthe three-dimensional map, DOE could then be veryefficient in depositing the bentonite on top of thewaste, resulting in minimum wastage and the mostcomplete seal. The second example DOE oftencites is the anticipated reduction in treatment time-from an estimated 200 years with old techniques toas little as six months-that could result from theintroduction of new techniques for treating a gaso-line spill at Lawrence Livermore National Labora-tory.

The other accounts in the TD budget include Supporting Technol-ogies and Infrastructure Programs, Technology Integration andEducational Development, Program Support, Program Direction,the Environmental and Molecular Sciences Laboratory, and theHazardous Materials Training Center.

CHAPTER FOUR ISSUES RELATED TO DOE'S CLEANUP PROGRAM 37

DOE expects to reap the bulk of the benefits ofnewly developed technologies in the future, how-ever, when it begins to undertake remedial action inearnest. In some cases, new technologies may makepossible the cleanup of contaminants for whichtechnologies are not yet available. In addition,DOE has estimated that savings over the next 30years, compared with the cost of using current tech-nologies to complete the cleanup, could amount toalmost $100 billion if technologies now under in-vestigation can be applied successfully throughoutthe DOE complex.

The department's estimates of the savings thatcould result from introducing new technologies mustbe viewed with a certain amount of skepticism.First, DOE has not been able to show that it canestimate accurately the cost of cleaning up its facili-ties using today's technologies (see Chapter 2).That alone makes it difficult to calculate futuresavings. Second, many of those estimates are basedon multiple assumptions. For example, to estimatethe savings that might result from new remediationtechniques that can be used with contaminants stillin place, DOE calculates the cost to pump out con-

Table 5.Individual Initiatives in DOE's Research, Development, Demonstration, Testing,and Evaluation Program, by Key Problem Area

Type of InitiativeGroundwater

and Soils CleanupWaste Retrievaland Processing

PollutionPrevention

Integrated Demonstration Volatile Organic Compoundsin Nonarid Soils

Volatile Organic Compoundsin Arid Soils

Plutonium in Soils

Uranium in Soils

Mixed Waste Landfill

Buried Waste

Underground Storage Tank

Decontamination andDecommissioning

Waste ComponentRecycling.Treatment,and Disposal

EnvironmentallyConsciousManufacturing

Integrated Program In Situ RemediationTechnology

Characterization, Monitoring,and Sensor Technology

Efficient Separations

Mixed Waste

Program Dynamic Stripping

Resource Recovery

Minimum Additive WasteStabilization

Supercritical WaterOxidation

DOE/U.S. Air ForceMemorandum ofUnderstanding

SOURCE: Congressional Budget Office using data from Clyde W. Frank, "Technology Development Is a Strategic Investment" (presentationby the Department of Energy to Congressional Budget Office staff, February 5, 1993).


laminated groundwater or dig up contaminated soil—costs that would not be incurred using new tech-nologies. In order to arrive at savings that could berealized during the entire cleanup process, DOEmust then apply the estimated savings per gallon ofwater or ton of soil to its estimate of the total num-ber of gallons or tons that would need to be treated.

Calculations of savings based on these methodsmay represent upper bounds and should be viewedwith caution. Rather than predictions of actual sav-ings, these estimates might better be viewed as indi-

cations of the technologies that DOE sees as havingthe greatest potential. The following discussionexamines some of DOE's technology developmentefforts, including both broadly applicable integratedprograms and more narrowly focused integrateddemonstrations.

Integrated Programs. Integrated programs (IPs)are DOE's technology development efforts with thebroadest potential application. They are designed toexplore basic technologies and techniques that canbe applied to many cleanup projects throughout the

Table 6.Summary of Funding for the Office of Technology Development, 1992-1995(In millions of dollars of budget authority)

Activity 1992 1993 1994 1995a

Research, Development, Demonstration,Testing, and Evaluation

Groundwater and soils cleanupWaste retrieval and processingPollution preventionInnovation and supportOther

Subtotal

Supporting Technologies and Infrastructure Programs6

Technology Integration and Educational Development0

Program Support

Program Direction

Environmental and Molecular Sciences Laboratory

Hazardous Materials Training Center

Total

54546

550

169

30

37

35

15

17

0

303

77693

348

191

45

38

44

16

28

0

362

76104

14923

253

57

36

37

15

0

0

398

7299

16120

253

75

29

38

17

0

14

426

SOURCE: Congressional Budget Office using data from Department of Energy, Office of Technology Development, FY 1993 ProgramSummary (February 1994); and Department of Energy, FY 1995 Congressional Budget Request, vol. 5, Environmental Restorationand Waste Management (February 1994).

a. Request.

b. Includes programs for liaison and communications, analytical laboratory management, robotics, decision support, and emergencymanagement.

c. Includes programs for educational development, technology integration, and international technology exchange.


DOE complex and thus have the potential to gener-ate large savings.2

One IP is investigating ways to separate wasteefficiently from its surrounding medium, be itgroundwater or soil. Since, according to DOE, only0.5 percent of its more than 3 million cubic metersof cataloged waste actually consists of radionu-clides, separating that portion from the noncontami-nated portion will greatly reduce the volume ofwaste requiring disposal. DOE is investigating sev-eral processes that are available commercially andin use elsewhere to separate radioactive waste fromits substrate. These processes include leaching andwashing contaminated soil and incinerating combus-tibles in waste that contains both radioactivecomponents and solvents.

Another IP is developing techniques for in situremediation, a process that also can be used to treatwaste in either water or soil. Methods being devel-oped would eliminate the need to extract the con-taminant from the soil or groundwater in order totreat it, thus avoiding the expense and effort associ-ated with pumping out groundwater or digging upsoil. Examples of in situ remediation include theintroduction of microbes into the soil to break downcontaminants and electrothermal means of turningcontaminated soil into glass so that the contamina-tion cannot spread.

DOE's estimates of potential savings from thesetwo integrated programs are based on comparisonsbetween the costs of operations using today's tech-nology and costs that would result from doingthings differently using new technologies. DOEestimates that in situ remediation could save $54billion in cleanup costs throughout the complexcompared with current methods. Similarly, the IPdesigned to develop efficient ways to separate radio-active waste from soil and mixed-waste solutionscould save $40 billion according to DOE estimates;these savings would result from the reduction inprojected volumes of soil and mixed waste requiringdisposal. Thus, the technology developed in these

two IPs together, if applied widely, could save morethan $90 billion during the cleanup.

(j)f course, DOE has not yet proved that thetechnologies under investigation in its integratedprograms can actually yield the results it anticipatesor that the techniques can be applied at all sites.Nor is it possible to know precisely how much itwill tost to develop the technologies envisioned inthe li?s. Since 1990, DOE has allotted about $275million to develop remediation techniques for con-taminated soil and groundwater; its annual expendi-tures in this area have been on the order of $60million to $80 million. Even if a 10-year invest-ment at this level is required to develop appropriatetechnologies and even if DOE has overestimated itssavings by an order of magnitude, savings wouldstill greatly exceed the cost to develop the technol-ogy. DOE therefore is pursuing a number of tech-nology projects that appear to show promise-that is,projects that will yield net savings when the presentvalue of their cost is subtracted from the presentvalue of the savings and other benefits that theymake possible.

Integrated Demonstrations. Integrated demonstra-tions (IDs) are technology development efforts thatare more narrowly focused than integrated pro-grams. They are designed to apply and prove thefeasibility of all technologies needed to conduct acleanup from beginning to end—that is, from charac-terization to postclosure monitoring of a particularsite.3 DOE is currently conducting at least 10 dem-onstrations.

At the Fernald installation, for example, an IDis investigating the application of a technique forseparating wastes, developed in the integrated pro-gram discussed above, to treat the contaminatedsoil. DOE estimates that this technique-referred toas leaching-would reduce by 80 percent the volumeof waste requiring disposal. Including the costs oftreatment, savings of 36 percent in the cost to clean

2. These programs are in some ways analogous to the basic technol-ogy programs within the Department of Defense, such as stealthresearch, that can, in theory, be applied to many weapons pro-grams.

Again using an analogy with the Department of Defense's re-search programs, integrated demonstrations are similar to programsthat develop specific weapons. Thus, if stealth research is analo-gous to an integrated program investigating in situ remediation,then the B-2 bomber development program would be an appropri-ate analogy for the integrated demonstration at Hanford using insitu bioremediation to clean up groundwater contaminated withnitrates and organic matter.


up Fernald would result. Using new excavationtechniques rather than the current method of bull-dozing could again, according to DOE estimates,reduce the amount of soil to be treated by almosttwo-thirds, yielding additional savings. Using bothof these techniques could result in total estimatedsavings of 80 percent in the cost to clean upFernald, compared with the current technology ofexcavation followed by disposal off-site.

Other integrated demonstrations also deal withcleaning up soils (for example, using a "rotomill"machine that monitors soil as it is excavated andremoves less soil requiring subsequent treatment) orwith cleaning up groundwater (for example, usingan air-stripping method that pumps compressed airinto an aquifer to flush out some of the contami-nants). Still others deal with retrieving and process-ing waste: some seek better, cheaper means of iden-tifying buried waste; others seek to reduce the costof stabilizing and storing highly radioactive wastes.Appendix C discusses DOE's integrated demonstra-tions in more detail and presents the department'sestimates of potential savings assuming the demon-strations are successful.

DOE estimates that the technologies under in-vestigation in its integrated demonstrations couldsave a total of $12 billion, much less than the morethan $90 billion in savings it ascribes to just twointegrated programs. The savings associated withthe integrated demonstrations are smaller than thoseof the integrated programs because IDs are morenarrowly focused and their application is not sp uni-versal as technologies resulting from IPs. Th0 esti-mates of savings associated with technologies beinginvestigated in demonstrations might also be morerealistic than those DOE associated with its inte-grated programs, because the demonstrations aretied to specific problems at specific sites and sohave a closer association with actual experiencethan with predicted savings. Furthermore, DOE hasconducted extensive cost-effectiveness analyses ofsome of the specific techniques under investigationin its demonstrations.4 Thus, although these esti-

mates also must be viewed with caution, they sug-gest a potential for savings with some basis inreality.

Possible Increases in Fundingfor Technology Development

The Department of Energy is spending a significant,but still relatively small, portion of its budget on thetechnology development programs designed to helpthe department realize these savings. The level offunding for technology development from 1990through 1994 represented an average of 7 percent ofthe annual EM budget-roughly the same averageshare reflected in the Administration's latest plan,for 1995 through 2000.5

Compared with other major cleanup programs,DOE devotes a substantial share of its cleanup (EM)budget to technology development. For example, itexceeds the share of the Superfund budget that theEnvironmental Protection Agency spends to developcleanup technologies. EPA administers Superfund,a federal program that provides funds for cleaningup pollution in the private sector and then attemptsto recoup all or part of the cleanup cost from theresponsible private parties. EPA allotted 4 percentof its total cleanup funds from 1988 through 1993to research and development compared with 7 per-cent for DOE. In addition, DOE's absolute level offunding for technology development, which wasalmost $400 million in 1994, is appreciably largerthan that of EPA's Superfund. In fact, DOE's 1992allocation of $303 million is more than 4.5 timesSuperfund's investment of about $64 million in thatyear.

The share of funding that DOE allocates totechnology development is, however, significantlybelow that of other government agencies engaged incomplex technical projects. The Department of De-fense invested an average of 13 percent of its totalfunding from 1988 through 1993 in research anddevelopment of new weapons. DOE itself, within

4. Joyce D. Schroeder, Steven R. Booth, and Linda K. Trocki, CostEffectiveness of the Site Characterization and AnalysisPenetrometer System (Los Alamos, N.M.: Los Alamos NationalLaboratory, December 1991).

5. In contrast to the discussions of the EM budget in Chapter 2, thetotal EM budget used as a basis for determining TD's share in-cludes funding for the Uranium Enrichment Decontamination andDecommissioning program.


its weapons-related programs, invested an averageof 24 percent of the funds in research, development,and testing during the same period. The weaponsprojects of both DoD and DOE are, of course, verydifferent undertakings from the DOE cleanup pro-gram. Nonetheless, the difference in shares devotedto technology development is striking.

The planned 7 percent of DOE's cleanup fund-ing earmarked for technology development is alsowell below DOE's own goal and the goal set by theCongress. DOE's first five-year plan, submitted in1989, included a goal of devoting 10 percent of EMfunding to developing new technologies for clean-up.6 The Congress has also stipulated that at least10 percent of EM funding for 1994 be devoted tothat effort.7 Achieving that goal would cost anadditional $200 million to $250 million annually,from 1995 through 2000, based on the Administra-tion's plan for spending levels in the out-years.

Availability of Suitable Projects. If the Congressis going to allocate additional funds for technologydevelopment, that money should be devoted only toprojects that will ultimately result in significant netsavings and other benefits. The Congressional Bud-get Office cannot evaluate candidate projects, butsome evidence suggests that a number of projectswould meet that test.

Various sources have identified several areas inwhich new technology could produce large savings.Representatives of DOE's Office of TechnologyDevelopment have identified the contamination ofsoil by heavy metals as an area that is currently notbeing investigated for lack of funds.8 The Depart-ment of Defense, which often works in partnershipwith DOE to develop cleanup technologies, hasidentified many projects it deems worthwhile butcannot fund at this time. The Office of TechnologyAssessment (OTA) and representatives of the Natu-ral Resources Defense Council have pointed out the

6. Department of Energy, Environmental Restoration and WasteManagement Five-Year Plan (1989), p. 30.

7. House Committee on Armed Services, National Defense Authori-zation Act for Fiscal Year 1993, Report 102-527 (May 19, 1992),p. 342.

8. Personal communication from the Department of Energy, February1993.

Figure 4.Cost of Treating, Storing, and Disposingof Low-Level and Transuranic Waste,With and Without Minimization of Waste

350

300

250

200

150

100

50

Millions of 1988 Dollars

Without Minimization

1990 1992 1994 1996 1998 2000 2002 2004 2006 2008

Fiscal Year

SOURCE: Congressional Budget Office based on data fromDepartment of Energy, Environmental Restorationand Waste Management Five-Year Plan, FiscalYears 1992-1996 (June 1990).

lack of technology for removing dense organic com-pounds that are not soluble in water from soil orgroundwater.9 DOE's latest detailed accounting ofhow it plans to spend its technology developmentfunds does not specify any planned investment intechnology to remediate such contaminants, but newapproaches in this area may offer benefits that ex-ceed the cost to develop them.

OTA identified waste minimization as anotherarea deserving attention. Research in waste minimi-zation or pollution prevention attempts to find waysto change manufacturing processes in DOE's plantsso that they generate less waste requiring treatmentor disposal. Reducing the volume of waste savesthe cost of treating or disposing of it in the future,potentially saving billions of dollars (see Figure 4).During its deliberations on the fiscal year 1993 bud-

Office of Technology Assessment, Complex Cleanup (February1991), p. 177; and statement of the Natural Resources DefenseCouncil before the Western Governors Association Waste TaskForce, Denver, Colo., October 7, 1992, p. 2.


get, the Congress also encouraged increased empha-sis on research into ways to minimize waste.10

DOE's funds to investigate such technologies, how-ever, were reduced in 1993 by 50 percent from theprevious year's level and have been further reducedto $1 million annually for 1994 and 1995.

Possible New Role for National Laboratories.Assuming that a number of projects are worthwhile,DOE would need to have enough trained personnelavailable to warrant an additional investment of$200 million to $250 million annually. That shouldnot be a problem for two reasons. First, DOE'sinvestment in developing weapons technology isdecreasing from $1.7 billion in 1994 to $1.6 billionin 1995, and this reduction in effort might makepersonnel available for projects in other areas. Sec-ond, DOE has been discussing redirecting some ofthe more than 20,000 people employed at its labsfrom weapons research to research on environmentalcleanup.11 Thus, there should be sufficient person-nel to perform the research.

Despite the availability of the labs and associ-ated personnel for research in environmental clean-up, questions arise concerning the suitability of us-ing the labs for this purpose. Transition from re-search in nuclear weapons design to research incleanup technologies may not be all that easy. Nev-ertheless, the national labs are a potential source ofscientists with experience in dealing with nuclearmaterials.

Possible Role for Other Agencies. Finally, DOEneed not be the sole recipient of any additionalfunds made available for research into ways to dealmore efficiently with cleanup problems. Many fed-eral agencies are involved in remediation efforts,including the Environmental Protection Agency andthe Department of Defense. Some cleanup tasks areunique to DOE, but many are common to all threeagencies. Thus, a more productive approach mightbe to encourage collaboration among the three agen-cies, something the Congress promoted when it es-

tablished the Strategic Environmental ResearchDevelopment program in 1993 to coordinate envi-ronmental research and development efforts amongfederal agencies. Another approach would be toprovide additional funds for technology develop-ment efforts at EPA or DoD rather than to DOEalone. A discussion of changes in budgets at EPAand DoD is, however, beyond the scope of thisstudy.

Time Sensitivity of Investingin New Technology

If added investments are to be made in technologydevelopment, DOE argues that they should be madesoon. The time available to develop promising newtechnologies that could reduce the cost of DOE'scleanup is relatively short if the department is tofulfill the terms of its agreements with the statesand the Environmental Protection Agency. Recentstatements by DOE indicate that if it is to realizesignificant returns on its investment in technologydevelopment and still meet its scheduled cleanupobligations, some new techniques must be ready forapplication by the late 1990s.12

This deadline is a consequence of DOE's agree-ments with various states and EPA that includetimetables for initiating and completing cleanuptasks at various sites. An important milestone in thecleanup process is the filing of the records of deci-sion for cleaning up specific sites. These docu-ments outline how the cleanup will be accomplishedand what technology will be used. The optimaltime to develop and prove productive new technolo-gies is therefore before the records of decision havebeen filed.

Based on the filing dates of records of decisionstipulated in its agreements with the states and EPA,DOE has identified windows of opportunity for de-veloping promising technologies in six areas inwhich it is conducting integrated demonstrations(see Table 7). Three windows—those associated

10. Senate Committee on Armed Services, National Defense Authori-zation Act for Fiscal Year 1993, Report 102-352 (July 31, 1992),p. 352.

11. "Environmental Cleanup Role Considered for A-Weapons Lab,"The Washington Post, March 9, 1993, p. A10.

12. Statement of Paul D. Grimm, then Acting Assistant Secretary ofEnergy for Environmental Restoration and Waste Management,before the Subcommittee on Energy and Water Development ofthe House Committee on Appropriations, April 26, 1993, p. 47.


Table 7.Schedule for Records of Decision in Selected Technology Areas

Records of Decision CompletedActivity First Half Last

DOEWindow3

Volatile Organic Compounds in Arid Soil

Buried Waste

Volatile Organic Compounds in Nonarid Soil

Uranium in Soil

Plutonium in Soil

Underground Storage Tanks

1990

1995

1991

1992

1992

1990

1995

1999

1999

1999

2000

2008

2005

2010

2010

2013

2007

2025

1991

1992

1992

1997

1998

1998

SOURCE: Congressional Budget Office based on Clyde W. Frank, "Technology Development Is a Strategic Investment" (presentation by theDepartment of Energy to Congressional Budget Office staff, February 5, 1993).

a. Optimum time by which introducing a new technology for cleanup will have the greatest impact in terms of reducing the costs ofremediation.

with volatile organic compounds in arid soil and innonarid soil, and with buried waste—were identifiedby DOE as opening in 1991 or 1992. Those dateshave already passed, and DOE has stated that forevery year that research is delayed, the benefits thatcan be reaped from the new technologies diminish.According to DOE, the best time to introduce im-proved technologies in the remaining three areas-uranium in soil, plutonium in soil, and undergroundstorage tanks-is by 1997 or 1998.

Improving the Accountability andManagement of DOE's TechnologyDevelopment Program

If additional funds are to be appropriated for tech-nology development, the Congress must have confi-dence that the funds will be spent wisely. ButDOE's Technology Development program has beencriticized for poor management. In an April 1992report, the General Accounting Office (GAO) identi-fied several shortcomings in that program that it felthampered an efficient use of funds devoted to re-search.13 In particular, GAO criticized DOE's lackof performance goals for its development programs.

Furthermore, GAO pointed out that DOE has notestimated likely costs for its development projectsor established schedules for program milestones orcompletion. In conclusion, GAO stated that the de-velopment of new technologies was important forDOE to accomplish its goals but felt that the depart-ment lacked the management tools to conduct anefficient program.

The General Accounting Office and others havemade several recommendations for ways in whichDOE could improve the management of its Technol-ogy Development program. To correct the prob-lems of overall vagueness and lack of accountabil-ity, GAO suggested that DOE establish measurableperformance goals, timetables, and key decisionpoints for each technology development project. Inits report, GAO noted that effective program man-agement requires that estimates of the overall costof each project be made early and updated and vali-dated as the project matures. GAO felt that estab-lishing a structure for each project, with cost and

13. General Accounting Office, Cleanup Technology: Better Manage-ment for DOE's Technology Development Program, GAO/RCED-92-145 (April 10, 1992).


performance goals and prescheduled evaluationpoints, would focus DOE's technology developmentefforts and help to identify both projects that suc-ceed and those that fail.

Detailed Budget Reporting. As a first step towardmaking the Technology Development program moreaccountable, the Congress could reiterate its requestthat DOE provide more detailed information on howthe department plans to spend its annual allocationof funds for research into new cleanup methods.The EM program has provided this information tothe public and the Congress in the past-in its 1991annual report to the Congress and its August 1991five-year plan.14

The Congress, however, apparently has not beengiven detailed data on spending for individual re-search projects since 1991, and certainly not on aroutine annual basis. DOE has failed to provide thisinformation even though the Congress establishedthe requirement for such a report in the NationalDefense Authorization Act for Fiscal Years 1990-1991. The report was to detail DOE's efforts todevelop new techniques designed to reduce environ-mental hazards and contamination resulting fromdefense waste and to expedite environmental resto-ration efforts at inactive waste disposal sites. TheCongress has obviously long recognized the needfor additional information from DOE in order toperform its oversight function properly. Given thenumerous concerns expressed by the Congress inreviewing DOE's cleanup budget, it may be time forthe Congress to ask again for additional data to sup-port DOE's request for research and developmentfunds.

A New Management System. To provide struc-ture, direction, and accountability for DOE's devel-opment programs, the Congress could also encour-age or require the department to create an evalua-tion and reporting system for major projects that issimilar to the one employed by the Department ofDefense for its major weapon systems. Under thescheme used at DoD, all systems requiring more

than $200 million to develop, or $1 billion to buy-both figures in 1980 dollars—are subject to internalreview at four specific points in their maturationprocess: initiation, concept validation, initiation offull-scale development, and before entering produc-tion.15 (See Appendix D for a detailed descriptionof DoD's weapons acquisition process.) The statusof each of these major systems is summarized annu-ally in Selected Acquisition Reports, which DoDthen submits to the Congress.

DOE and the Congress could establish a similarsystem for major projects in the cleanup arena, par-ticularly those that require the development of newtechnologies to enable their completion. The sys-tem would cover an entire cleanup project—fromdefinition of the site or operational unit, develop-ment of new technologies, and characterization andassessment, through completion of remedial action.This system would be particularly applicable to thenarrowly focused integrated demonstrations, each ofwhich would be only one part of an overall cleanupeffort.

Including the development of technologies aspart of a specific task to be completed, such ascleanup of the tanks at Hanford, would focus theresearch efforts and put them on rigorous schedules.Indeed, a report prepared for DOE in January 1993on technology needs recommended that milestonesin technology development projects be linked toschedule requirements of specific problems in envi-ronmental restoration.16

Possible Decision Scheme. To set up such a re-porting and evaluation system, DOE would firstneed to establish milestones for the major decisionpoints. For its cleanup activities, DOE could usefour milestones:

1. establishing a cleanup project,

2. justifying the need for a new technology andinitiating the applied research and developmentwork,

14. Department of Energy, Office of Environmental Restoration andWaste Management, Office of Technology Development, AnnualReport to Congress, Fiscal Year 7997; and Department of Energy,Environmental Restoration and Waste Management Five-YearPlan, Fiscal Years 1993-1997 (August 1991).

15. These thresholds correspond to about $350 million in developmentfunds and about $1.8 billion in procurement funds when convertedto 1995 dollars.

16. Chem-Nuclear Geotech, Inc., Technology Needs Crosswalk Report,DOE/ID/12584-117, 1st ed. (prepared for the Department ofEnergy, January 1993), p. xv.


3. beginning assessment and characterization, and

4. filing a record of decision and beginning reme-dial action.

In justifying the need for new technologies atthe second milestone, DOE would need to establishcriteria against which to judge the new methods. Inparticular, new techniques should save money ortime compared with current techniques. To focusresources on the most promising technologies, esti-mated savings in cost or time resulting from use ofthe new technology should exceed thresholds estab-lished by DOE. The department would also need toestablish and approve preliminary schedules andcost estimates at the second milestone.

Reaching the third milestone-initiation of as-sessment and characterization—would depend oncompleting the research and development efforts forprojects that require new techniques. For example,cleaning up the highly radioactive waste in theHanford tanks requires new assessment techniquesthat should be developed and proved before charac-terization of the site begins. Furthermore, it is atthe third milestone that DOE should establish finalcost and schedule baselines, so most of the develop-ment work should be completed before these base-lines are set.

The fourth and final milestone would involvefiling the record of decision and selecting cleanuptechniques. At this point, the characterization of theproblem and development of cleanup techniquesshould be essentially completed.

DOE would need to establish a panel of seniormanagers within the EM program to approve transi-tion of a project from one milestone to the next inthe scheme outlined above. Projects making insuffi-cient progress, demonstrating substantial cost over-runs, or meeting unexpected technical problemswould be restructured before proceeding. In thisway, problems could be identified before they be-came too costly.

Reporting on major projects using a milestoneformat would also permit the Congress to judgetheir progress. For example, at some point (prob-ably at the third milestone) estimates of the cost

required to complete a project should be available.Once such baseline estimates are in place, the Con-gress can compare subsequent cost estimates withthe baseline and focus on projects whose costs aregrowing.

The Congress has already mandated that parts ofthis management system be established. In its billauthorizing appropriations for national securityfunctions for 1994, the Congress directed DOE tosubmit reports that provide schedules for and theestimated cost to complete many of the projectswithin the EM program. The initial reports-orbaselines—for environmental restoration activitiesare due by March 1, 1995; similar reports for allactivities for waste management, transition of opera-tional facilities to safe shutdown status, and researchand development activities are due by June 1,1995.17 The Congress also directed DOE to submit,following the initial report, annual status and vari-ance reports that would inform the Congress of theamount of funds expended for any project duringthe prior fiscal year, as well as any growth in costsor schedule slippages with respect to the initialreport.

These Baseline Environmental ManagementReports, as they are called in the authorization bill,would be similar to the reporting requirements envi-sioned in the decision scheme described above.They would not, however, require DOE to establishan internal review system, nor would they integrateresearch and development efforts with cleanup tasks.But they attempt to address the same problem as thepostulated management scheme—that is, the lack ofjustification, continuity, and visibility within the EMprogram of individual projects designed to developnew and beneficial technologies for cleanup.

Not all projects in the EM program need to besubject to this management system. Some TDprojects are not tied to a specific cleanup problembut instead represent efforts to develop techniques

17. The Department of Energy has made a partial response to therequest in the authorization act with the information contained inits report Environmental Management 1994 (February 1994). Thatreport contains some data on funding, milestones, and performanceby installation. The report does not, however, fulfill DOE's obli-gation to submit the Baseline Environmental Management Reports,as specified in the authorization act.


that are applicable to many environmental programs.The integrated programs, which represented about10 percent of TD funds in 1991, fit this descriptionand are more analogous to DoD's basic researchprograms that are not tied to specific weapons pro-grams. Some or all of those programs might there-fore be excluded from this management system, asmight small projects.

DOE and the Congress will need to set thresh-olds for cleanup tasks so that only those of suffi-cient total or annual cost are subject to enhancedscrutiny. DoD's experience in this area might beused as an example. Acquisition programs that aresubject to extensive reporting requirements repre-sented historically about half of DoD's annual pro-curement funding and 15 percent of the researchand development budget. By comparison, all ofDOE's integrated demonstrations together receivedabout 20 percent of EM's total funds for technologydevelopment in 1991.

A scheme similar to DoD's may not be the finalanswer for DOE's EM program. But it does repre-sent a system for establishing goals, schedules, andspecific decision points that is currently lacking inDOE's Technology Development program. By set-ting up a mechanism for keeping the Congressabreast of the progress and problems that arise in itscleanup program, DOE could help to answercharges of poor management and lack of account-ability within that program. Addressing these con-cerns would be particularly important if the Con-gress decided to increase funding for technologydevelopment in an effort to reduce the total cost andduration of the cleanup.

Delaying Cleanup Until NewTechnologies Are Available

Even with additional funding, developing new tech-nologies that expedite the cleanup effort or reduceits costs will not always be possible. Some researchand evaluation efforts take a certain amount of timeto come to fruition, and some techniques must beevaluated over fixed amounts of time and so cannotbe hurried. Thus, in certain cases, no amount ofadded resources will yield earlier results.

In such cases, a weighing of costs and benefits—the method discussed in Chapter 3 as a means forsetting priorities among cleanup tasks-may favorrenegotiating the agreements DOE has with thestates. Otherwise, these agreements will force DOEto undertake difficult projects on a stringent sched-ule rather than devote additional time and resourcesto accelerate research efforts. If all parties to anagreement concede that no promising technology islikely to be available before the deadline to com-mence cleanup, they could arrange an extension,assuming, of course, that the problem in questionposes no immediate threat to human health or theenvironment. By renegotiating these agreements,DOE could avoid the expense of either additionalresearch and development work or inefficient clean-up efforts.

Budgetary Effects of Delay

The Environmental Restoration (ER) program withinthe Office of Environmental Restoration and WasteManagement is responsible for the actual cleanup ofDOE's inactive sites. ER's activities include identi-fying and measuring contamination at DOE facili-ties, taking subsequent remedial action, and contin-uing to monitor a site after cleanup has been com-pleted. In DOE's request for funds for 1995, ER'sshare was $1.8 billion, representing 29 percent ofthe total EM request. According to the Administra-tion's plan, annual funding for the ER programwould grow to $2 billion by 2000.

Delaying projects that are difficult to executewith today's technology would lessen the need forER funds during the next six years. If delays oftechnically difficult projects were judged appropri-ate, how many projects might be affected? Andhow would such delays affect the budget?

Many of the projects that DOE is undertaking inits ER program are difficult or expensive to com-plete with today's technologies. According to theOffice of Technology Assessment, cleaning upgroundwater and soil may be extremely expensiveor require a long time even if contaminants can beremoved with current technology.18 Almost all of

18. Office of Technology Assessment, Complex Cleanup.


DOE's major installations suffer from some sort ofgroundwater contamination. And cleaning up build-ings contaminated with highly radioactive materials,referred to as decontamination and decommission-ing, may present technical difficulties because cur-rent methods are costly, inefficient, and producelarge quantities of waste material requiring disposal.Furthermore, DOE has already identified 500 sur-plus facilities that must be cleaned up before theycan be demolished or released for other uses, andexpects to add at least 1,000 more to its list. Thechallenge in decontamination and decommissioningis to develop methods that minimize the exposure ofthe workers to radiation and hold down costs.

To determine how many of DOE's projectswould be difficult to perform with today's technol-ogy, the Congressional Budget Office (CBO) exam-ined almost 1,000 activity data sheets in the Envi-ronmental Restoration program in the August 1991five-year plan.19 CBO designated the followingtypes of projects in that plan as "technically diffi-cult":

o characterizing very large land areas or highly ra-dioactive waste in tanks,

o cleaning up contaminated groundwater or soil,and

o decontaminating and decommissioning buildingsthat contain radioactive waste.

That designation reflects the fact that DOE is con-ducting projects to develop new techniques in thoseareas. In the process, CBO also determined theportion of DOE funds that is slated for technicallydifficult projects.

Cost of Technically Difficult Projects. CBO'sexamination of DOE's August 1991 five-year planidentified 134 projects that were technically diffi-cult. Results for the two budget levels included inthat plan were similar; the share of ER funds de-voted to technically difficult projects during the

19. None of DOE's five-year plans submitted since August 1991include a detailed breakdown of ER funding planned for five yearsby individual cleanup project. Therefore, CBO had to rely on thefive-year plan submitted in August 1991 for its analysis.

five-year period rose from about one-quarter in1993 to about one-third by 1997. Total funding forthese projects for the 1993-1997 period accounted inboth budget levels for an average of nearly 30 per-cent of funding for the Environmental Restorationprogram. Such consistent results for both budgetlevels suggest that DOE would allocate the sameproportion of ER funds planned for the 1995-2000period to technically difficult cleanups. Based onthe assumption that 30 percent of ER funding isallotted to such projects, DOE could spend up to$3.4 billion from 1995 through 2000 on character-izations and remediations that are difficult to con-duct with today's technologies.

Illustrative Budgetary Effects. Delaying the exe-cution of these projects until newer and better waysof carrying them out have been developed couldreduce the funds needed for the Environmental Res-toration program over the next six years. It is im-possible for CBO to determine which of the 134technically difficult projects should be pursued us-ing today's technology and which should be de-layed. To illustrate the budgetary effects, however,CBO assumed that half of these projects were de-layed. In that case, funding for environmental res-toration would be reduced by 15 percent over thenext six years. Annual savings associated with sucha slowdown would increase from almost $270 mil-lion in 1995 to almost $300 million in 2000. Thesesavings would exceed the additional amount neededto increase funds for technology development fromthe current 7 percent of the Office of EnvironmentalRestoration and Waste Management's total budgetto 10 percent, should that be judged a desirable op-tion.

Reduced Funding WouldAffect Agreements

Any delay in the cleanup process would cause prob-lems. DOE feels compelled to move ahead, evenwith technically difficult projects, because it hasentered into agreements with various states and theEnvironmental Protection Agency that dictate theschedule for cleaning up sites within the complex.Some schedules call for beginning the cleanup pro-cess soon without regard to the availability of tech-nologies that would facilitate its execution. Indeed,


for projects involving many of the types of cleanuptasks for which DOE is attempting to develop newtechnologies, the first record of decision, whichspecifies how the cleanup is to be accomplished, hasalready been submitted (see Table 7). For projectsinvolving the cleanup of volatile organic compoundsin arid soils, for example, half of the records ofdecision are due by 1995. DOE therefore plans toproceed apace with many efforts in the next fiveyears using current technologies to clean up somesites as specified in various agreements, eventhough that might not be the best or cheapest wayto conduct the cleanup in the long run.

DOE is also required by law to request fundingfor its programs that is sufficient to meet the mile-stones included in the agreements to which it is aparty. That would apply primarily to funding forthe Environmental Restoration program, since mostof the deadlines stipulated in agreements pertain tothe beginning and completion of feasibility studies,assessments, and remedial actions. For this reason,officials at DOE feel that they have little discretionto delay or terminate ER projects. Failure to meetthe terms of the agreements may result in the levyof fines by the states against DOE, something thatthe state of Ohio has already done.

Not only is DOE bound by many triparty agree-ments to timetables that dictate the start and finishof specific cleanup tasks, but it is also legally re-sponsible for preventing the spread of contaminationfrom its facilities to the surrounding environment.Thus, in some instances, DOE feels it must containits waste and prevent it from entering water sources,either above or below ground, even though currenttechniques for containing such wastes are expensive.

Risks of Delay. A decision to delay some of itsprojects would mean that DOE would have to rene-gotiate its agreements with various states and EPA.If renegotiations were not successful and DOEbreached existing agreements, it could face bothfinancial and political costs. States could levy fineson the federal government that could, cumulatively,be substantial. Indeed, an Administration officialtestified in 1991 that subjecting federal facilities tothe same environmental regulations as the privatesector could create an atmosphere that would causethe federal government to commit its limited clean-

up resources at the courthouse rather than in thefield.20 The state of Ohio levied a $750,000 fineagainst DOE in 1989, and DOE could face similarfines if it does not comply with state agreements orrenegotiate them.

Benefits of Delay. Nevertheless, fines of this mag-nitude are small compared with the annual savingsof approximately $300 million that DOE could real-ize by delaying some cleanup projects. Moreover,many of the concerned parties are becoming in-creasingly aware that some of the agreements areuntenable. Many agreements were signed in the late1980s and early 1990s before DOE knew muchabout the scope of its environmental problems. Inthe past few years, DOE and its contractors haveconducted numerous exploratory samplings and in-vestigations and now have a better idea of the mag-nitude of the problem facing them. In light of thisnew information, some of the goals and deadlinesestablished in the agreements may be unrealistic.

Delaying some cleanup projects could also buyDOE time to gather the information needed to makeinformed decisions concerning the management anddirection of its cleanup program. Recent statementsby Administration officials indicate their awarenessthat DOE needs more information to manage itscleanup program efficiently.21 The types of infor-mation needed include data concerning health haz-ards posed by the pollutants at its sites and theultimate use of its surplus land and facilities.

Need to Reevaluate Priorities. Perhaps for thesereasons, several parties, including Members of Con-gress and DOE officials, have suggested that DOEreevaluate its priorities, which could ultimately re-sult in the need to renegotiate DOE's agreements.22

20. Statement of Thomas E. Baca, Deputy Assistant Secretary ofDefense for the Environment, before the Environmental Restora-tion Panel and Nuclear Facilities Panel of the House Committeeon Armed Services, June 6, 1991.

21. "OMB Provides No Figures on Environmental Cleanup," CongressDaily, September 21, 1993, p. 4.

22. Senator J. Bennett Johnston raised this issue during a hearing heldby the Senate Committee on Energy and Natural Resources on theDepartment of Energy's cleanup budget on July 29, 1993. It isalso mentioned in U.S. House of Representatives, Making Appro-


The Assistant Secretary for EM, Thomas Crumbly,has testified that in some cases, the department haslearned through site investigations that the problemsuncovered are larger than anticipated or have noeffective long-term technical solutions.23 He as-serted that in these cases, the best course of actionmay be to stabilize the site and invest in appropriateresearch to solve the problem more effectively inthe long run. He emphasized that the department iscommitted to complying with its agreements. Buthe held open the door to conducting the cleanup inthe most effective way possible through mutual re-negotiation of those agreements.

Events at the Hanford installation provide anexample of such a renegotiation. The Departmentof Energy, the Environmental Protection Agency,and the Washington State Department of Ecologyrecently updated and revised the original agreementthey had signed in 1989. Under the renegotiatedagreement, the date for beginning the cleanup of thehighly radioactive waste has been delayed 10 years,and priorities for various cleanup tasks have beenreordered. In particular, DOE has agreed to givehigher priority to the treatment of low-level wasteand to convert the waste into a form that is moredurable and therefore poses less risk over the longrun.

In their conference report on appropriations for1994, the appropriations committees provided fundsto implement the revised accord at Hanford and en-couraged the department to review all of its compli-ance agreements.24 The Congress acknowledgedthat DOE needs to develop a mechanism for estab-lishing priorities among its cleanup tasks. To thatend, the Congress directed DOE to submit to theappropriations committees a report evaluating the

priations for Energy and Water Development for the Fiscal YearEnding September 30, 1994, and for Other Purposes, ConferenceReport 103-305, to accompany H.R. 2445 (October 22, 1993); andSenate Committee on Armed Services, National Defense Authori-zation Act for Fiscal Year 1994, Report 103-112 (July 27, 1993),pp. 239-240.

23. Statement of Thomas P. Crumbly before the Subcommittee onEnergy of the House Committee on Science, Space, and Technol-ogy, July 15, 1993.

24. See U.S. House of Representatives, Making Appropriations forEnergy and Water Development for the Fiscal Year Ending Sep-tember 30, 1994.

risks to public health and safety posed by conditionsat the complex's sites that are addressed by require-ments in the compliance agreements.

Problems Associated with TargetingSpecific Cleanup Tasks for Delay

Both the Administration and Members of Congresshave stated that the current regulatory environment,and particularly the large number of complianceagreements to which DOE is party, limit DOE'sability to carry out its cleanup program in the mosteffective way possible. The agreements create a setof priorities and demands that may be at odds withthe limitations presented by technological and fiscalrealities.

The Keystone process described earlier in thisstudy was developed in recognition of these con-straints and is intended to provide a framework fordealing with fiscal limitations. The Keystone re-port, which was prepared by a committee composedof representatives from the environmental commun-ity, recommended an approach for addressing fund-ing shortfalls caused by Congressional appropria-tions that fell below DOE's request. If the approachrecommended by the Keystone group is adopted-and it has been endorsed by DOE—then reducingfunding for difficult projects may not have the de-sired effect of delaying them.

The Keystone group recommended that whenthe Congress appropriates less than what DOE re-quested for its environmental restoration activities,the funding shortfall be distributed equally amongall sites. Thus, if the Congress reduced ER fundingby 15 percent, as discussed in the section on illus-trative budgetary effects, then each of DOE's 15major installations would receive 15 percent lessfunding for ER activities. The committee furtherrecommended that if DOE could absorb fundingcuts without affecting the scope or schedule of es-tablished, legally binding agreements, then it coulddo so without consulting the regulators, affectedtribes, and other stakeholders. But if DOE couldnot meet its obligations after absorbing a significantcut, the group strongly recommended that the statesand EPA renegotiate DOE's obligations and mile-stones rather than take punitive actions.


The process laid out by the Keystone groupprovides DOE and other federal agencies with aframework for accommodating fiscal limitationswhile minimizing confrontation with regulators andaffected parties. But rigid adherence to a propor-tional sharing of budget shortfalls among all af-fected sites could hamper the Congress's and DOE'sability to set nationwide priorities for its cleanupprogram.

Under the process outlined above, a reduction infunding for the Environmental Restoration programwould be shared equally by all of DOE's installa-tions. But distributing the cuts equally could defeatthe purpose of targeting reductions toward projectsthat are difficult to conduct with today's technolo-gies, particularly if the more challenging tasks areconcentrated at a small number of installations.Thus, if the Congress or DOE wishes to target spe-cific projects for delay, it may have to repudiate theprocess that the Keystone group proposed and thatDOE has subsequently endorsed.

Conclusion

Reducing funding for technically difficult projectsduring the next six years, thereby delaying them atleast temporarily, could ultimately reduce the timearid money that DOE needs to clean up its complex,particularly if the funds and time that would becomeavailable to DOE were used to develop productivenew technologies. The resulting delays would re-quire renegotiating some agreements with EPA andstate regulators. But once the ultimate benefits ofdelay have been made clear, existing agreementsmay be easier to renegotiate.

Spending for Administrationand Support

The Department of Energy's cleanup effort, as em-bodied in the EM program, is now five years old.Since its creation in 1989, the program has receiveda total of $23 billion, and more than $6 billion ofthat amount was appropriated for 1994 (includingfunding for the Uranium Enrichment Decontamina-

tion and Decommissioning program). Some criticshave claimed that very little cleanup has actuallyresulted from this multibillion-dollar investment,however, and have questioned whether DOE isspending the taxpayer's money efficiently. In 1992,Leo Duffy, then Assistant Secretary for EM, statedpublicly that waste and inefficiency chewed up 40cents of every dollar in the cleanup program.25

Several reviews of DOE's cleanup budget havehighlighted concerns about the costs of administra-tion and support.

Organization of the NuclearWeapons Complex

Defining and estimating costs for administration andsupport require an understanding of how DOE oper-ates the nuclear weapons complex. Management ofand operations at DOE's 15 major installations havebeen the responsibility of contractors to DOE andits predecessors. At all sites still producing weap-ons or their components, the same contractor isresponsible for both managing production and clean-ing up the pollution that has resulted from past orcurrent production.

DOE oversees the performance of its contractorsfrom its headquarters in Washington, D.C., and 10operations offices (Rocky Flats and nine field of-fices). There are many more people employed bythe contractors, however, than there are DOE per-sonnel supervising their performance. In the EMarea alone, more than 49,000 contractor personnelare engaged in cleanup, overseen by fewer than1,800 DOE employees. Because most contractorpersonnel are located at various DOE sites, the ratioof contractors to DOE staff is particularly high inthe field. For instance, only about 400 DOE em-ployees work at the Richland field office, whichoversees operations at the Hanford installation, incontrast to the approximately 12,000 contractorpersonnel employed at Hanford.

Because of the way it does business, DOE in-curs administrative and support—or indirect—costs at

25. Douglas Pasternak, "A $200 Billion Scandal," U.S. News andWorld Report, December 14, 1992, p. 34.


two levels. The first occurs at the project levelwhere the contractor charges for administrativecosts. These indirect costs are generally referred toas overhead rates. The second layer occurs at theinstallation or program level and includes DOE'scosts to direct and manage its programs as well ascosts to provide services-such as security, road

Figure 5.Portion of Environmental Management FundsDevoted to Administrative and Support Activities

Percent

Corps(Narrow a

Analysis)

Corps(Broad b

Analysis)

ICE"

SOURCE:

b.

c.

d.

Source of Data

Project Overhead Programwide Support

Congressional Budget Office based on data fromvarious analyses.

Army Corps of Engineers, Supplemental Report on Cost Esti-mates (Report to the Associate Director for Natural Re-sources, Energy and Science, Office of Management andBudget, April 29, 1992). The narrow analysis focused on 55projects at eight installations, representing about 20 percentof the total cleanup budget.

Army Corps of Engineers, Supplemental Report on Cost Esti-mates. The broad analysis expanded the narrow analysis byincluding all the projects at two additional installations-Fernald and Hanford.

The percentage devoted to project overhead was calculatedby DOE's Independent Cost Estimating (ICE) team. SeeGilbert/Commonwealth, Inc., Independent Cost Estimate forthe Environmental Restoration and Waste Management Five-Year Plan, Fiscal Years 1993-1997 (Reading, Pa.: Gilbert/Commonwealth, Inc., November 22, 1991). The percentagedevoted to programwide support is a CBO estimate.

The percentage devoted to project overhead is from Indepen-dent Project Analysis, Inc., Project Performance Study(Reston, Va.: IPA, Inc., November 1993). The percentagedevoted to programwide support is a CBO estimate.

maintenance, utilities, and even laundry service—atsome of its installations. These costs are referred toas programwide costs.

Estimates of EM Spending onAdministration and Support

Funds devoted to administration and support activi-ties in the EM program-a combination of spendingon overhead and programwide support-account forabout 40 percent of EM's total budget. Several re-views have concluded that DOE spends at least one-fifth of the funds for each project for contractoroverhead.26 Analyses by CBO and the Army Corpsof Engineers found that DOE devoted a similarshare to programwide support. The combinedspending on administrative and support functionswithin the EM budget—at both the project and pro-gram levels-ranges from 43 percent to 48 percentbased on the Corps's analyses, and from 39 percentto 45 percent based on combining the overheadrates determined by DOE's Independent Cost Esti-mating team and Independent Project Analysis, Inc.,with CBO's estimates of programwide spending (seeFigure 5). These resulting estimates of total spend-ing of about 40 percent or more on administrationand support represent a proportion that is signifi-cantly higher than the share spent by some othergovernment agencies that may be performing com-parable tasks. Reductions in this category of spend-ing may therefore represent one means of reducingEM budgets.

Overhead Costs at the Project Level. The Depart-ment of Energy has commissioned several studies ofthe overhead costs of its projects.27 These studies

26. DOE's Independent Cost Estimating team estimated 26 percent,the Army Corps of Engineers estimated 27 percent to 28 percent,and Independent Project Analysis, Inc., estimated 20 percent to 23percent.

27. Interagency Review Group, Interagency Review of the Departmentof Energy Environmental Restoration and Waste ManagementProgram (April 29, 1992); Army Corps of Engineers, Supplemen-tal Report on Cost Estimates (Report to the Associate Director forNatural Resources, Energy and Science, Office of Managementand Budget, April 29, 1992); Independent Project Analysis, Inc.,Project Performance Study (Reston, Va.: IPA, Inc., November1993); Gilbert/Commonwealth, Inc., Independent Cost Estimatefor the Environmental Restoration and Waste Management Five-Year Plan, Fiscal Years 1993-1997 (Reading, Pa.: Gilbert/Commonwealth, Inc., November 22, 1991).


have found overhead rates charged by the contractorto be in the range of 20 percent to 28 percent.

DOE's Independent Cost Estimating team re-viewed in detail the approximately 1,600 projectsreceiving funds in the EM program in the 1993-1997 five-year plan and evaluated actual overheadrates for subcontracts, labor, and material based oninformation provided by the department. Using thisdefinition it found overhead rates of 26 percent,which it considered high.

The Army Corps of Engineers examined a morelimited number of projects~55, representing slightlyless than $1 billion, or about 20 percent, of the EMbudget-in detail in order to evaluate the ability offield offices to estimate costs. In its analysis, theCorps defined overhead to include project manage-ment, supervision of construction at a site, the pro-rated cost of administrative functions, and thecontractor's profit. The Corps found that 27 percentof the funds requested were earmarked for overhead.Compared with its overhead rate of 18 percent forsimilar projects, the Corps felt the overhead chargedby DOE's contractors was high.

In its report, the Corps acknowledged that somefactors might cause DOE's overhead rates to behigher than its own. Since DOE must deal withradioactive substances, certain activities that resultfrom regulatory requirements unique to nuclearwaste might lead to higher overhead costs. TheCorps noted, however, that many activities in theplanning, design, and construction phases of projectsare common to the cleanup of both hazardous wasteand radioactive waste. Since the Corps claims thatit has extensive experience in construction and haz-ardous waste management, it felt that costs for mostof the environmental activities undertaken by DOEand the Corps should be comparable.

A more recent study of DOE's cleanup programfound that the EM office spent more than either theprivate sector or other government agencies forequivalent work. This study, conducted by Indepen-dent Project Analysis, Inc., for the Office of Envi-ronmental Restoration and Waste Management, wasintended to assess the status of EM projects andprovide a baseline from which to measure improve-ment. The study's authors compared key parame-

ters of EM projects—such as management costs, costgrowth, and schedule slippage—with comparabledata from environmental remediation and wastemanagement projects completed by the private sec-tor and other government agencies.

The authors of the study concluded that costsfor work performed for DOE were significantlygreater than the industry norm, in part because ofhigh project management costs. The study's moststriking conclusion was in the area of remediationprojects; the total costs of DOE's environmentalrestoration projects were found to be 15 percentgreater than those of other government agencies and32 percent greater than those of the private sector(see Figure 6). The costs were higher, in part,because DOE's project management—or overhead-costs for its restoration projects, which consumed 23percent of the costs of individual projects, weremore than double those of other government agen-cies and nearly four times those of the private sec-tor. These findings support those of the Corps andthe ICE team and confirm that DOE is spending toomuch for project overhead.

Administration and Support Costs at the Pro-gram Level. In addition to the indirect costs thatDOE incurs as overhead on each project, it mustalso pay for the cost of managing and directing itsoverall cleanup program and for costs at each instal-lation to support all the cleanup projects there.Programwide support activities would include over-sight of compliance, preparation and review ofdocuments, program support or direction, technicalsupport, litigation, quality assurance, and liaisonwith local groups such as Native American tribes.

To estimate the total costs of these activitieswithin the EM budget, the Congressional BudgetOffice reviewed the almost 2,000 projects includedin the 1993-1997 five-year plan.28 Based on thetitle of the activity data sheet that describes each ofthe projects funded by the EM budget, CBO identi-fied projects that involved programwide administra-

28. In estimating the cost of administration and support within the EMbudget, CBO used the lower of the two budgets included in thefive-year plan submitted in August 1991. More recent five-yearplans lack the detail needed to estimate these costs. As a conse-quence, the August 1991 five-year plan is the most recently sub-mitted plan that presents the necessary detailed data.


Figure 6.Comparison of DOE's Environmental RestorationCosts with Those of Other Sectors

Total Cost of Projects Relativeto Private-Sector Norm

2.0

1.5

1.0

0.5

Relative Cost

PrivateSector

OtherGovernment

DOEEnvironmental

Restoration

30

25

20

15

10

Project Overhead Rates

Percentage of Total Project Cost

PrivateSector

OtherGovernment

DOEEnvironmental

Restoration

SOURCE: Congressional Budget Office based on data fromIndependent Project Analysis, Inc., Project Perfor-mance Study (Reston, Va.: IPA, Inc., November1993).

a. Normalized based on average performance of private-sectorcompanies.

tion and support activities, as defined above. CBOincluded as programwide support those projectswithin the Technology Development program thatdid not contribute directly to research; that is, fundsnot in the RDDT&E account were considered to beprogramwide costs.

Using this definition, CBO found that more than360, or 18 percent, of EM's projects were devotedsolely to programwide support activities. Overall,funding for these projects represented 25 percent ofthe total EM budget for 1993. Based on the August1991 five-year plan, that share would remain rela-tively constant, averaging 24 percent during thefive-year period from 1993 through 1997. As in allother projects, a portion of the funds allotted tothese projects performing programwide support ac-tivities would be spent on overhead. Assuming aproject overhead rate of 20 percent, total EM fund-ing devoted solely to programwide support—and notto project overhead-would be 19 percent.

The findings of a study by the Army Corps ofEngineers support CBO's findings that a significantportion of DOE's cleanup budget pays for program-wide management and support functions. TheCorps, in its detailed analysis of one-fifth of the EMbudget, found that 16 percent of the funds were de-voted to such activities. These costs, according tothe Corps, fall mainly into the categories of prog-ram management and program direction. The Corpsalso reviewed an additional third of the EM budgetin less detail and found that 22 percent of it wasdevoted to programwide support activities. Whenthis additional spending is taken into account, aboutone-fifth of the EM budget examined by the Corpsappears to be devoted to administration and supportactivities. This level of spending-20 percent-ismore than twice the 8 percent that the Corps says itspends for programwide support.

Illustrative Savings Realizedfrom Reducing Spending onAdministration and Support

Based on the analysis by CBO and others, theroughly 40 percent of EM funds spent on adminis-trative and support functions is high compared withspending by other agencies—both government and


private—engaged in cleanup activities and could bereduced. If DOE could reduce spending on admin-istrative and support activities by roughly 25 per-cent, for example, it could save $630 million in1995 based on the Administration's request. An-nual savings would increase to $710 million by2000, based on the levels outlined in the Adminis-tration's out-year targets. An illustrative 25 percentcut in overhead would result in a 10 percent overallreduction in total EM spending, which would fallwithin the range of cuts proposed by DOE's Inde-pendent Cost Estimating team and the Army Corpsof Engineers.

Because the two reviews used different defini-tions of overhead, the cuts in total EM spendingthat would result from the reductions in overheadrecommended by the ICE team and the Corps arenot proportional to the reductions in overhead. TheICE team recommended a 25 percent cut incontractors' overhead only, which would result in a7 percent reduction in total spending. The Corpssuggested a decrease of roughly 60 percent in totaladministrative and support spending-both overheadand programwide—which would yield a 25 percentoverall reduction. Compared with the reductionsrecommended in these two reviews, the 10 percentcut in the overall budget contemplated in this illus-tration is well below that recommended by theCorps and is somewhat higher than the reductionproposed by the ICE team.

How would savings in administration and sup-port costs be realized? Although identifying man-agement changes is not the focus of this study,other organizations have suggested ways in whichDOE could better manage its EM program andthereby reduce administrative and support costs.They include increasing the program's oversight ofcontractors and reforming the process by which itmakes contracts.

Increase Oversight of Contractors. Both internaland external reviews of DOE's operations havehighlighted the lack of government oversight ofcontractors' performance. The interagency reviewof the EM budget found that the program, which isincreasing rapidly in scope and funding, had insuffi-cient staff both to budget effectively and to overseethe programs for which it was responsible.29 At

many field offices, for example, it is the contractors,not government personnel, who prepare and reviewtechnical documents and cost estimates.

The Corps recommended that DOE add person-nel in order to provide better oversight of contrac-tors' performance. Assistant Secretary Crumbly, inhis statement of July 15, 1993, before the Subcom-mittee on Energy of the House Committee on Sci-ence, Space, and Technology, outlined his intentionto respond to this recommendation by shifting es-sential project management responsibilities at eachsite to DOE personnel. For example, Grumblywould like DOE personnel to prepare internal costestimates rather than rely on contractors to preparethem.

Greater reliance on DOE personnel would re-quire an increase in the number of DOE employees.In a period when the federal government is commit-ted to reducing the size of its work force, such anincrease could be difficult to achieve and couldincrease the cost of the EM program in the shortrun. It might be possible, however, to shift somepersonnel within the national defense function.That budget function includes both the Departmentof Defense (with about 900,000 federal civilianemployees) and the weapons and associated cleanupactivities of the Department of Energy (about 6,000federal civilian employees). A shift of personnelfrom DoD to DOE could benefit DOE's cleanupeffort by providing personnel who have extensivemanagement experience.

Enlisting more government staff, either fromDOE or DoD, to perform more of the managerial,planning, and budgeting functions that contractorsnow perform, though possibly costing more in theshort run, could save money in the long run. Forexample, increased oversight might result in lessduplication of effort among contractors or in loweroverhead rates. Indeed, Grumbly predicted thatadding 400 federal workers to do jobs now handledby contractors would save $360 million in 1995 andmore in later years. He also predicted that in-

29. Interagency Review Group, Interagency Review of the Departmentof Energy Environmental Restoration and Waste ManagementProgram.


creased oversight, in conjunction with contract re-form, could result in a 10 percent to 20 percent in-crease in contractors' cost efficiency over the nextfour years.

Reform the Contracting Process. DOE has tradi-tionally contracted with a single entity (referred toas a management and operations, or M&O, contrac-tor) to manage operations at an installation. TheM&O contractor may then hire subcontractors toperform specific functions, such as drilling wells orproviding security.

Both GAO and Assistant Secretary Crumblyhave endorsed the need to reform the contracts thatDOE has with its M&O contractors. The agenciesthat preceded DOE used special incentives in theircontracts because they believed such incentiveswere needed to attract and retain contractors toconduct work and research on nuclear weapons.Since then, DOE's contractors have been reluctantto negotiate contracts with more stringent clauses.One example of a nonstandard contract is DOE'sexclusion of the standard procurement clause fromits contract with the University of California foroperating the three national laboratories. As a con-sequence, one of the labs leased vehicles at com-mercial rather than government rates, costing DOEan additional $600,000.

Another example concerns nonstandard indemni-fication clauses in some of DOE's contracts thathave grown out of DOE's historical practice ofindemnifying, or reimbursing, almost all of acontractor's costs to compensate for the unique risksinherent in producing nuclear weapons. Suchclauses have had unforeseen consequences, how-ever. In one case, DOE could not prove that a con-tractor's costs had been caused by bad faith orcorporate mismanagement and so was forced toreimburse the contractor's loss of $420,000 inmoney and materials that employees had stolen.GAO applauded DOE's attempt to avoid a recur-rence of these types of expenses by moving to de-lete clauses from its contracts that do not reflectstandard practice.

Both GAO and Crumbly have highlighted theneed for DOE to establish requirements for theperformance of its contractors that would specify

the product to be delivered, cost targets, and sched-ules. Because some of DOE's contracts have noestablished criteria for determining management oraward fees, contractors have no idea on what basistheir performance is evaluated. Some contractorshave been awarded significant fees-nearly $2 mil-lion at Rocky Flats—even though a DOE reviewboard initially recommended no award at all; andsome contractors receive substantial managementfees, which increase automatically every year, tocover "indirect costs" and "complementary andbeneficial activities" that were never specified.30 Toremedy this deficiency, both GAO and Crumblyrecommended replacing the subjective or, in somecases, nonexistent award criteria with ones that canbe measured.

In his testimony of July 15, 1993, AssistantSecretary Crumbly provided additional ideas onhow to improve the performance of contractors. Hesuggested that fixed-price contracts that are solicitedcompetitively for some tasks, such as remedialactivities and landlord functions, could be appropri-ate at some installations. In this way, the singlecontract between DOE and the M&O contractor ateach site and its attending subcontracts could bereplaced by several smaller contracts made directlywith the EM program. Such a system would engen-der more competition and, as a consequence, lowerprices. Assuming that DOE hired more governmentpersonnel to monitor the contracts, this systemwould also give the EM program more control overthe performance of its contractors. Crumbly alsorecommended that DOE make contractors moreresponsive to its needs by limiting the length of allcontracts for services to no more than five years.

Separate contracts for weapons production andcleanup might also enhance efficiency. Eventhough many managers within the EM programacknowledge that the contracts they monitor carryhigh overhead rates, they point out that they do notalways have direct control of those rates. At instal-lations where both production and cleanup are stilloccurring-Savannah River and Oak Ridge, forexample—the overall contract for M&O is not under

30. In a case cited by GAO, one M&O contractor's annual manage-ment fee started at $12 million and increased automatically everyyear by $250,000.


EM's control. Since the M&O contractor is respon-sible for both production and cleanup, as contractingpractices now stand, the terms of the contract arecontrolled by whichever program is bigger, which istraditionally weapons production and not cleanup.To remedy this situation, Assistant Secretary Crum-bly, along with other observers, has suggested thatall environmental restoration work at installationswith production activities be contracted separately.EM staff would then be directly responsible foroverseeing the work and contracts for environmentalrestoration at all installations.

It is impossible to quantify with confidence thesavings that could result from each of these initia-tives. Assistant Secretary Crumbly, however, haspredicted that the amount the EM program must paycontractors to perform tasks at DOE installationswill drop 10 percent to 20 percent in four years. Asa consequence, DOE is reducing its budgets in an-ticipation of these savings. In its authorization billfor 1994, the Senate Committee on Armed Servicesrecommended a $40 million reduction in EM's bud-get to reflect management efficiencies. In the samevein, the renegotiated agreement at Hanford speci-fied that a savings of $1 billion be achieved overthe five years from 1994 through 1998, whichwould represent slightly more than a 10 percent re-duction in the Administration's proposed funding of$8.4 billion for the same period.

Of course, some risk is associated with cuttingbudgets in anticipation of management efficiencies.If the efficiencies are not realized, then cleanup pro-grams will be delayed.

Nevertheless, those risks may be worth taking.Budget cuts targeted toward management costs mayincrease DOE's incentives to reduce those costs ex-peditiously. And if administration and support costscan be reduced, then DOE could devote a largershare of its funding to cleanup activities.

Maintaining DOE's SurplusFacilitiesThe scope of DOE's mission has decreased drasti-cally since the end of the Cold War. The nuclear

weapons complex, designed to produce and main-tain some 23,000 nuclear warheads, will shrink overthe next decade along with the size of the U.S. nu-clear arsenal. As a consequence, many facilitiesthat have been producing nuclear weapons or thematerials to build them, some for as long as 50years, will no longer be needed. Accelerating theprocess of putting the surplus facilities in a state oflow maintenance would require additional fundinginitially, but it could save money in the long run.

As facilities cease operations, responsibility fortheir security and maintenance will be transferredfrom the Defense Program in DOE to the Environ-mental Restoration and Waste Management pro-gram. In July 1992, DOE created a new officewithin the EM program—Facility Transition andManagement (FT)-to monitor these facilities afterthey have been declared surplus and as they aremade ready for final cleanup and disposition. TheFT program eventually could be responsible forremoving as many as 7,000 facilities from defenseproduction in the next 30 years.

Some of the facilities that DOE decides it nolonger needs for production may have to go throughseveral steps to prepare for final disposition. Thefirst step is a decision by DOE that the building issurplus. This decision results in a transfer of thefacility from the control of the Defense Program tothe FT program. The next step is to remove allhazards that need to be eliminated so that the facil-ity can be maintained cheaply and safely until it canbe turned over to the Environmental Restorationprogram for decontamination and decommissioning.After D&D, the building can be demolished orturned over to another government or private agencyfor other uses.

While the facilities are in transition from pro-duction to cleanup, a process that can take manyyears, they must be maintained and guarded so thatno harm comes to either the public or the environ-ment from the hazardous material inside. If most ofthe hazardous material and contaminated productionequipment is removed from the building, then therequirements for inside utilities, such as ventilatorsand radiation detectors, and security measures canbe minimized. This state—known as safe shut-down-means that the facility cannot be restarted for


production but can be left in a state of low mainte-nance. Achieving safe shutdown requires an invest-ment of time and money. But if many years elapsebefore the facility is cleaned up or disposed of—asin the case of the surplus reactors at Hanford dis-cussed in Chapter 3—then placing it in a state thatrequires low annual investments for maintenance isworth the initial effort.

In a recent report, GAO concluded that the FTprogram faces problems concerning maintenance,safety, and costs.31 In general, DOE's inactive facil-ities are deteriorating physically. GAO reported thatthe upkeep of inactive facilities is not a high pri-ority among maintenance jobs and that such work isnot generally required by environmental regulationsor covered by interagency agreements. As a result,many projects for repairing surplus buildings aredeferred in favor of higher-priority work elsewhereat the sites. As a consequence, conditions at thosebuildings are often in violation of regulations estab-lished by DOE and the Occupational Safety andHealth Administration. These violations have re-sulted in accidents; in one such accident, a workeron the roof of a 48-year-old reactor building waskilled when the roof panel collapsed.

GAO has concluded that the failure to properlymaintain inactive facilities can increase the dangersand costs associated with cleaning them up. Slip-shod or undocumented work performed while shut-ting down a facility can lead to unanticipated prob-lems or accidents during subsequent decontamina-tion and decommissioning. An explosion that oc-curred during the cleanup of a nuclear research fa-cility at Hanford, for example, was the direct resultof a contractor's decision to eliminate an interimwork step earlier in the project.

Improper maintenance can also affect the cost ofcleanup projects. When the plutonium fuel facilityat Savannah River was put on standby in 1983, theequipment was not fully decontaminated, nor has itbeen cleaned up in the subsequent 11 years of inac-tivity. The internal equipment is now so badly de-

31. General Accounting Office, Cleaning Up Inactive Facilities WillBe Difficult, GAO/RCED-93-149 (June 1993).

teriorated that it can no longer be used to removethe plutonium that remains in the facility. DOEestimated in January 1992 that an additional $115million will therefore be needed to decontaminateand decommission the facility. Had the facilitybeen adequately cleaned initially, DOE could haveavoided the subsequent higher cleanup costs.

In general, the cost to maintain surplus facilitiesawaiting cleanup is substantial and could grow be-cause of problems of the sort just noted. Increasingnear-term funding designed to attain safe shutdownstatus at surplus facilities could produce long-termsavings in the DOE budget by reducing annual se-curity and maintenance costs in later years.

Hanford's N-reactor as an Example

The N-reactor at the Hanford installation in westernWashington provides a good example of the typesof costs that can be incurred and the potential forsavings that exists. The reactor, which was used toproduce plutonium for nuclear weapons, has beenmaintained since October 1990 on a "cold standbystatus," which means that no production is occurringand the reactor is not operating but could be re-started if necessary. Even though DOE has decidedthat the reactor will never be restarted, 130 full-timeemployees are needed to maintain the reactor at acost of about $25 million annually in 1995 dollars.

Because large amounts of highly radioactivematerials are still stored on the site, these personnelare needed primarily to maintain security and tomeet safety standards. In particular, over 500 tonsof radioactive debris such as fuel carts, fuel baskets,and process tubes are stored in a water-filled basinadjacent to the reactor building. In addition, threeunderground silos contain highly radioactive spacersthat were placed inside the fuel rods. Before thefacility can be decontaminated and decommissioned,all of these wastes must be removed and disposedof. In the meantime, DOE must guard and maintainbuildings, monitor radiation levels, and completeroutine upgrades to utilities on schedule. In fact,DOE contends that costs to maintain the facility willrise over time as it ages if steps are not taken toplace it in safe shutdown.


Placing the N-reactor in safe shutdown wouldinvolve removing superfluous materials from thesite, including all the highly radioactive debris fromthe storage basin and the storage silos, as well as7,000 cubic feet of documents. Because DOE hasnot made funds available for this purpose, however,current plans call for maintaining the N-reactor incold standby status through 2000, at the cost ofabout $25 million a year, before beginning efforts toachieve safe shutdown. Once safe shutdown isattained, however, annual maintenance costs woulddrop by 90 percent, to about $2 million, based onestimates made by Westinghouse~the contractorresponsible for most of the operations at Hanford(all costs are in 1995 dollars). The process of at-taining safe shutdown of the N-reactor could takeup to six years and require an investment of almost$290 million. Overall, attaining safe shutdownsooner rather than later—starting in 1995 rather than2000—would reduce total costs, adjusted for thetime value of money, by about 30 percent.32

Other Surplus Facilities

Detailed data on other DOE facilities are not avail-able, but the situation represented by Hanford's N-reactor is probably not unique. The DOE complexcontains many large facilities, some with high levelsof radiation. More and more of these facilities arebeing idled as the need to produce nuclear weaponsdiminishes. Hanford alone, in addition to the N-reactor, contains several large plants that at sometime in the past 50 years housed operations to sepa-rate plutonium from irradiated fuel cells. Otherinstallations across the DOE complex contain sur-plus buildings with significant radioactive contami-nation, including plants designed to machine pluto-nium at Rocky Flats and the gaseous diffusion plantat Oak Ridge.

DOE will eventually have to make investmentsat many of these facilities to attain safe shutdown,with decommissioning to follow. Making these in-

vestments now could reduce maintenance costs, re-sulting in substantial annual savings. But determin-ing how much might be saved at facilities through-out the complex is difficult for several reasons.

First, it is difficult to ascertain how much DOEis spending to maintain surplus facilities before safeshutdown. The 1995 request for surveillance andmaintenance in the FT budget totals $319 million,although that figure is unlikely to include the entirecost of maintaining the facilities. In fact, a recentpress article reported that DOE was spending $240million annually at Hanford alone to maintain itsidle facilities.33 But if the figure of $319 millionrepresents a conservative estimate of the cost of sur-veillance and maintenance throughout the complex,and if reductions in maintenance costs similar tothose associated with the N-reactor could berealized generally, then investing to achieve safeshutdown of DOE facilities could yield savings inmaintenance costs of hundreds of millions of dollarsa year.

Second, accurate estimates of the investmentrequired to achieve these savings are not currentlyavailable. In one sense, the size of the investmentis irrelevant, since at most sites the investment willhave to be made eventually because of the presenceof long-lived radioactive debris. But the magnitudeof the cost would provide some indication of thefeasibility of making that investment sooner—givencurrent budgetary constraints-in order to achievesavings in maintenance costs.

In its 1995 budget, the Office of Facility Transi-tion and Management has requested a total of $109million to place facilities in safe shutdown. As withsurveillance and maintenance costs, however, it isimpossible to know, without additional data, exactlywhat that request included and what activities arebeing funded elsewhere in the EM budget. Even ifthe $109 million represented an appropriate annualtotal for the investment required to achieve safeshutdown, it is not clear how long DOE would haveto sustain that level of investment. Given the scopeand schedule of other cleanup activities, DOE might

32. The reduction of 30 percent results from differences in discountedpresent values, an economic calculation that takes into account thetime value of money. A savings of 30 percent is based on anannual real discount rate of 2 percent. Savings fall to 24 percentat a discount rate of 4 percent and rise to 35 percent without dis-counting.

33. Matthew L. Wald, "At an Old Atomic-Waste Site, the Only SureThing Is Peril," New York Times, June 21, 1993, p. Al.


have to continue investing that much for a substan-tial period.

Even though DOE realizes that these trade-offsexist, it does not have the budgetary data needed toevaluate the benefits or drawbacks of either defer-ring or accelerating the deactivation of surplusbuildings. Part of this deficiency may stem fromthe fact that DOE's shift in mission—from produc-tion to cleanup, dismantlement, and storage—hasbeen recent and abrupt. DOE is in the preliminarystages of planning and conducting cleanup work forall of its inactive facilities. It does not know thenumber of facilities that are inactive but not yettransferred to the EM program, the full extent of thedangers they pose, or the cost to maintain them

safely until they can be turned over to the Envi-ronmental Restoration program for ultimate cleanupand disposal. Some facilities that were preparingfor modernization or restart are now being closedpermanently.

The newly created program for overseeing thesetransitions may facilitate the collection and stan-dardization of budgetary data and may make theevaluation of the benefits of accelerating or defer-ring attainment of safe shutdown possible in thefuture. Once the necessary data are available, DOEshould be able to determine where increases infunding to accelerate safe shutdown should bemade.

Appendixes

Appendix A

Profile of Major Facilities andRemedial Action Programs

T he Department of Energy's budget for theOffice of Environmental Restoration andWaste Management supports a number of

facilities and activities across the nation. This ap-pendix briefly describes the budgets and environ-mental challenges for the department's major facili-ties and remedial action programs. Table A-l pres-ents their 1994 appropriations and 1995 budgetrequests; Table A-2 shows the growth in their bud-gets since 1990.1

Hanford Site

The Hanford Site occupies 560 square miles insoutheastern Washington State, near Richland. Oneof the original facilities of the Manhattan Project, itwas established in 1943 as the site of the first full-size reactor to produce plutonium. The historic B-reactor produced the plutonium used in "Fat Man,"the nuclear bomb dropped on Nagasaki in August1945. After World War II, Hanford continued toproduce and process plutonium for nuclear weapons.

The activities at Hanford's nuclear reactors andchemical separation facilities have generated largequantities of dangerous wastes-radioactive materi-als, heavy metals, volatile organic compounds, and

The budget totals for 1994 and 1995 differ from those given inChapter 2 because they include the budget of the Uranium Enrich-ment Decontamination and Decommissioning program.

other hazardous chemicals. Over the years, wasteshave been disposed of in large underground storagetanks. The oldest tanks, with only single shells, areof greatest concern because many are leaking orfeared to be leaking. The reactors also presentdisposal problems because of their radioactivity.

Hanford's environmental cleanup budget for1994 is nearly $1.5 billion, 24.1 percent of DOE'scleanup budget. DOE anticipates that Hanford willrequire approximately that share of the cleanupbudget for the foreseeable future.

Savannah River Site

The Savannah River Site is located on approxi-mately 325 square miles of land along the SavannahRiver in south-central South Carolina near Aiken.DOE considers it one of the department's greatestenvironmental challenges. Its five reactors, twochemical separation facilities, and reactor fuel man-ufacturing facility have produced tritium and pluto-nium. As a consequence of these activities, it hasvarious kinds of radioactive and mixed wastes todispose of. Savannah River's Defense Waste Pro-cessing Facility is intended to stabilize high-levelradioactive waste.

Savannah River's environmental cleanup budgetfor 1994, $757.4 million (12.3 percent of the total),is about 3 percent lower than its peak of $779.0million in 1993.


Oak Ridge Reservation

The Oak Ridge Reservation, located near OakRidge, Tennessee, contains several discrete facilitiesthat together account for 10.6 percent of DOE'scleanup budget in 1994.

The K-25 site was one of the original Manhat-tan Project facilities, with a mission to enrich ura-nium by gaseous diffusion. Following shutdown ofthe uranium enrichment process in 1987, the K-25site has been used for various environmentalcleanup functions. Its cleanup budget in 1994 is$270.2 million.

Table A-1.Environmental Cleanup Budgets for DOE's Facilities and Remedial Action Programs, 1994 and 1995

Facility or Program

1994 AppropriationIn Millions As a Percentage ofof Dollars DOE's Cleanup Budget

1995 Budget Request(Millions of dollars)

Washington, D.C.

HanfordSavannah RiverOak RidgeRocky FlatsIdaho National Engineering LaboratoryFernaldWaste Isolation Pilot PlantLos Alamos National LaboratoryLawrence Livermore National LaboratorySandia National Laboratory (Albuquerque)Mound PlantPantex PlantNevada Test SiteKansas City PlantPinedas PlantOther

Headquarters

726.7

Facilities

1,490.0757.4652.7477.2408.7304.4185.3185.189.573.047.435.718.014.111.1

551.7

Remedial Action Programs

Formerly Utilized Sites Remedial Action ProgramUranium Mill Tailings Remedial Action Project

42.7104.1

Total8

11.8

24.112.310.67.76.64.93.03.01.41.20.80.60.30.20.28.9

0.71.7

995.0

1,591.6743.6648.3639.7392.4294.2184.6180.080.251.945.045.623.113.29.0

167.9

74.1100.9

All Facilities and Programs 6,174.8 100.0 6,280.3

SOURCE: Congressional Budget Office based on data from Department of Energy, Environmental Management 1994, DOE/EM-0119(February 1994).

a. Includes funds for the Uranium Enrichment Decontamination and Decommissioning program.

APPENDIX A PROFILE OF MAJOR FACILITIES AND REMEDIAL ACTION PROGRAMS 65

Table A-2.Funding by Facility, 1990-1995 (In millions of dollars)

1990 1991Appropriations

1992 1993 1994a1995

Request8'5

Headquarters

Washington, D.C.

HanfordSavannah RiverOak RidgeRocky FlatsIdaho National

Engineering LaboratoryFemaldWaste Isolation

Pilot PlantLos Alamos

National LaboratoryLawrence Livermore

National LaboratorySandia National

Laboratory (Albuquerque)Mound PlantPantex PlantNevada Test SiteKansas City PlantPinellas PlantOther

60.5

441.3471.1282.7139.7

185.684.4

104.6

47.9

33.8

16.319.15.4

13.012.03.0

353.7

196.3

828.6644.6353.3173.0

323.2263.6

164.0

82.1

52.7

37.730.719.7n.a.17.44.7

260.6

416.5

Facilities

1,060.4550.5448.6181.8

248.4214.3

141.0

120.5

77.8

58.542.226.213.727.54.6

463.5

447.2

1,481.4779.0553.1291.2

372.9293.9

150.7

172.9

107.6

73.744.541.020.716.99.2

484.2

726.7

1,490.0757.4652.7477.2

408.7304.4

185.3

185.1

89.5

73.047.435.718.014.111.1

551.7

995.0

1,591.6743.6648.3639.7

392.4294.2

184.6

180.0

80.2

51.945.045.623.113.29.0

167.9


Formerly Utilized SitesRemedial Action Program

Uranium Mill TailingsRemedial Action Project

All Facilities and Programs

0

n.a.

2,274.1

29.2

119.6

3,601.0

49.0

141.9

Total

4,286.9

40.9

139.3

5,520.3

42.7

104.1

6,174.8

74.1

100.9

6,280.3

SOURCE: Congressional Budget Office based on data from Department of Energy, Environmental Management 1994, DOE/EM-0119(February 1994).

NOTE: n.a. = not available.

a. The 1994 and 1995 budgets include funding for the Uranium Enrichment Decontamination and Decommissioning program.

b. The 1995 budget request allocates all funding for technology development ($426.4 million) and transportation management ($20.7 million)to headquarters.


The Y-12 plant also was built during WorldWar II. Its original mission was to separate ura-nium isotopes by an electromagnetic process. Afterthe war, it was used for manufacturing and develop-mental engineering and for treatment, storage, anddisposal of radioactive and hazardous wastes. Itsenvironmental budget for 1994 is $97.8 million, adecline of about $10 million from 1993.

The Oak Ridge National Laboratory conductsresearch on fusion, fission, and many other energytechnologies. Its site is contaminated with radioac-tive and hazardous wastes. The budget for environ-mental cleanup in 1994 is $163.1 million, a declineof $17 million from 1993.

A new responsibility of the Oak Ridge office in1994 is the Uranium Enrichment Decontaminationand Decommissioning program. This program,which is funded through taxes paid by utilities tosupport enrichment of uranium used at nuclearpower plants, is not part of the weapons-relatedwork described in the main body of this study. Butits budget of $286.3 million in 1994 accounts for alarge share of the budgetary growth at Oak Ridge($199.1 million) and at the gaseous diffusion plantsat Portsmouth ($47.8 million) and Paducah ($39.4million).

Rocky Flats

The Rocky Flats Plant is located about 16 milesnorthwest of Denver, Colorado. The plant itselfcovers about 400 acres on an 11-square-mile site.From 1952 to 1992, Rocky Flats produced weaponscomponents fabricated from plutonium and othermetals. As Denver has grown, suburban develop-ment has pushed closer to Rocky Flats, and con-cerns about migration of contaminated groundwateras well as other environmental risks have increased.

The environmental cleanup budget for RockyFlats in 1994 is $477.2 million, 7.7 percent of thetotal. The growth of its budget by $186 million (arate of 63.9 percent) from 1993 to 1994 made it thefastest-growing part of the Environmental Restora-tion and Waste Management program.

Idaho National EngineeringLaboratory

The Idaho National Engineering Laboratory (INEL)covers about 890 square miles in southern Idaho,about 40 miles northwest of Idaho Falls. Its re-search, development, and operations activities havegenerated radioactive, hazardous, and mixed wastes.The Idaho Chemical Processing Plant (ICPP), lo-cated at the same site, handles and stores spent fuelfrom naval reactors. It treats wastes to stabilizethem and reduce their volume.

The environmental cleanup budget for INEL (in-cluding ICPP) in 1994 is $408.7 million, 6.6 percentof the total.

Fernald

The Fernald Environmental Management Project islocated on 1,050 acres near Fernald, Ohio, about 17miles northwest of Cincinnati. Formerly a producerof uranium metal ingots and uranium oxides, itbecame in 1991 the first DOE facility to be turnedover entirely from production to environmentalrestoration. One of the major problems is ground-water contaminated by radionuclides, metals, inor-ganic compounds, and volatile organic compounds.

Fernald's environmental restoration budget for1994 is $304.4 million, 4.9 percent of the total.

Waste Isolation Pilot Plant

The Waste Isolation Pilot Plant (WIPP), locatedabout 25 miles east of Carlsbad, New Mexico, isintended to be the long-term disposal site for trans-uranic wastes from Hanford, Savannah River, INEL,Rocky Flats, and other DOE facilities. Develop-ment of the facility has been held up by regulatoryhurdles. One major obstacle was overcome withpassage in 1992 of the WIPP Land Withdrawal Act,which withdrew from public use lands surroundingWIPP, allowing DOE to proceed with testing once itobtains approval from regulators. DOE still facesthe challenge of convincing regulators and the gen-

APPENDIX A PROFILE OF MAJOR FACILITIES AND REMEDIAL ACTION PROGRAMS 67

eral public that it will ensure safe storage of wastes.WIPP's 1994 budget of $185.3 million is about 3percent of DOE's cleanup total.

Los Alamos National Laboratory

The Los Alamos National Laboratory, located about25 miles north of Santa Fe, New Mexico, is the siteof nuclear weapons research and development dat-ing back to World War II. Its activities have re-sulted in a variety of hazardous and radioactivewastes. Its environmental budget of $185.1 millionin 1994 is about 3 percent of the total.

Lawrence Livermore NationalLaboratory

The Lawrence Livermore National Laboratory is inLivermore, California, 40 miles southeast of SanFrancisco. It has interim status under the ResourceConservation and Recovery Act for treatment, stor-age, and disposal of hazardous, mixed, and low-level radioactive waste. It also has nonnuclearexplosive test facilities at a site about 15 milessoutheast of the main site. Testing activities haveresulted in contamination of both soil and ground-water. The 1994 environmental cleanup budget is$89.5 million, about 1.4 percent of the total.

Sandia National Laboratory

The Sandia National Laboratory performs research,development, and testing of nonnuclear componentsof nuclear weapons at sites near Albuquerque, NewMexico, and Livermore, California. This work hasresulted in hazardous, radioactive, and mixedwastes. The environmental cleanup budgets are$73.0 million at Albuquerque and $5.7 million atLivermore in 1994. Together they account for 1.3percent of the total.

Mound Plant

The Mound Plant in Miamisburg, Ohio, conductedresearch and development and produced nonnuclear

and tritium-containing components for nuclear wea-pons. In 1993, its mission changed from productionto environmental cleanup. Its environmental budgetof $47.4 million is a little less than 1 percent of thetotal.

Pantex Plant

Nuclear weapons are assembled and disassembled atthe Pantex Plant, near Amarillo, Texas. The envi-ronmental cleanup budget is $35.7 million, about0.6 percent of the total cleanup budget.

Nevada Test Site

The Nevada Test Site covers about 1,350 squaremiles in Nevada, about 65 miles northwest of LasVegas. Above-ground and underground tests ofnuclear weapons have contaminated both surfaceand subsurface soils. In addition, transuranic wasteand mixed waste are stored at the site. The cleanupbudget for 1994 stands at $18 million.

Kansas City and Pinellas Plants

The Kansas City Plant manufactures nonnuclearweapons components. These operations result inhazardous and toxic wastes. The environmentalcleanup budget in 1994 is $14.1 million.

The Pinellas Plant, near St. Petersburg, Florida,developed and produced special electronic and me-chanical equipment. Production has ceased, havingbeen consolidated at the Kansas City Plant. Theenvironmental cleanup budget is about $11.1 millionin 1994.


In addition to its facilities, DOE has two majorprograms for cleaning up contaminated sites.

The Formerly Utilized Sites Remedial ActionProgram is intended to clean up radioactive contam-ination at 33 sites in 13 states resulting from early


activities of the atomic energy program. Its 1994 tailings at 24 sites and approximately 5,000 proper-budget is about $42.7 million. The program is ties in 10 states and on two Indian tribal lands,overseen by the Oak Ridge field office. Cleanup of surface contamination is nearly com-

plete, but cleanup of groundwater is still at an earlyThe Uranium Mill Tailings Remedial Action stage of assessment. Its 1994 budget of $104.1

Project is intended to stabilize and control mill million is overseen by the Albuquerque field office.

Appendix B

Estimating the Costs ofCleaning Up DOE Sites

E stimators of costs at the site or project levelface many of the same problems as thoseworking at the program level. Information

yet to be acquired about the type, magnitude, loca-tion, and toxicity of contaminants and the likelihoodof their migrating to populated areas will affect theestimates of remediation costs. But one advantageat the site level is that the project typically is welldefined, as are the tasks required to accomplish it.Cost engineers break down the work needed toaccomplish each task into its basic components,known as the work breakdown structure (WBS).The Department of Energy typically aggregatesseveral tasks within a WBS into an activity datasheet (ADS), the basic building block of the depart-ment's five-year plan. The ADSs set forth thespecific assessment or cleanup tasks that serve asthe basis for cost estimates.

Estimating the costs of assessing and cleaningup a site typically begins by identifying all the basiccomponents of work, such as drilling wells or dig-ging up soil. Cost estimators determine the amountof labor of different types and the equipment andmaterials needed to complete the task. After multi-plying the inputs by wage rates and other unit costs,estimators add in overhead and other allowancesthey deem necessary to arrive at a cost estimate.The job of estimating the costs of completing indi-vidual cleanup tasks throughout the nuclear complextypically falls on contractors at DOE's facilities, inconsultation and coordination with DOE staff at thefacilities. But headquarters staff also become in-volved, providing general guidance on preparingcost estimates and monitoring and coordinating thework done at the field offices.

Cost estimators use a variety of analytical tools.The choice of tools and the degree of confidence in

the estimate depend on the stage and type of activ-ity. DOE's environmental restoration work consistsof two principal types of activities: assessment andremediation (cleanup). The costs of each are esti-mated separately.

Estimating the Costsof Assessing Sites

Assessing a site to determine the type and extent ofcontamination is perhaps the more challenging ac-tivity to estimate because so many factors are un-known. In this respect, assessment is akin to re-search and development: at the start of a project,researchers do not know exactly what they willencounter and therefore cannot confidently predictcosts, schedules, or outcomes. For example, beforebeginning assessment, workers may not know howmany monitoring wells will be needed to obtain anaccurate picture of subsurface contamination.

DOE makes three types of estimates of assess-ment costs: planning estimates, preliminary esti-mates, and detailed estimates.1 Planning estimatesare made during the preliminary assessment/site in-vestigation (PA/SI) stage of a cleanup project. Atthis stage, cost estimates are based on what littleinformation is available about a site, such as loca-tion and history of use. DOE uses analogy, simplecost-estimating relationships, and statistical tools to

Department of Energy, Environmental Restoration and WasteManagement Cost Assessment Team, Cost Estimating Handbookfor Environmental Restoration (n.d.; updated periodically), p. 2-2.


make cost estimates.2 Because of uncertainties, con-tingency factors must be built in to the cost esti-mates.

After the PA/SI is completed, DOE makes apreliminary estimate of costs as it develops a workplan for the remedial investigation/feasibility study(RI/FS). By this time, cost estimators have someidea about the kinds of assessment activities to becarried out—for example, drilling, sampling, labora-tory analysis—and can make rough estimates basedon past unit costs of such activities.

During the RI/FS stage, DOE makes detailedestimates of assessment costs as additional informa-tion becomes available. "Detailed" estimates can bebased on engineering data and drawings, specifica-tions, the contract schedule, and other factors spe-cific to each project. By this point, cost estimatescan be made with much greater confidence thanbefore. From experience, cost estimators knowroughly how much labor—and what skill levels-willbe required to complete certain tasks. Applyinghourly rates gives an estimate of labor costs. Simi-larly, they can estimate the amount of equipmentand materials. The estimators add an overhead rateto take into account management resources devotedto the task. But none of these tasks is strictlymechanical. All require judgment and "guessti-mates" about the amount of resources needed: theless well defined a task, the more subjective the costestimates.

Estimating the Costs ofCleaning Up Sites

For long-range planning and budgeting purposes,DOE makes preliminary estimates of cleanup costseven before assessment has proceeded far enough toreduce uncertainty about what the eventual cleanupprocess will entail. These "planning estimates" arelike those for assessment in that they come so earlyin the process that there is little solid information onwhich to base them and therefore considerable un-certainty in the estimate.

As part of the RI/FS stage, the ComprehensiveEnvironmental Response, Compensation, and Liabil-ity Act (which established Superfund) requires anal-ysis of alternative remedies and their costs. The"feasibility estimates" of cost, schedule, and designthat are prepared at this stage serve as a basis forselecting a remedy in the record of decision. Inprinciple, DOE considers the estimates of cost andschedule made at this point as forming the firstformal baseline for measuring and evaluating clean-up performance.3 In practice, however, very fewprojects at major DOE facilities have reached therecord-of-decision stage.4

As DOE progresses with cleanup, cost estimatescan become more detailed, following the samepattern as estimates of assessment costs.

Tools for Estimating Costs

Estimating costs of environmental cleanup, if notstill in its infancy, can hardly be called a maturescience. The Superfund program is just 13 yearsold, and given the time required to identify and as-sess contaminated sites before proceeding to reme-dial action, the amount of cleanup experience onwhich to base estimates of new projects is limited.As cleanup efforts proceed around the nation—notonly at DOE facilities but also at defense and otherfederal facilities and at private Superfund sites-pro-fessional cost estimators gain new information toadd to their data bases that they can then use in es-timating costs of new projects. The lessons learnedfrom these experiences are shared through inter-agency task forces as well as professional journalsand conferences.

DOE's Environmental Restoration and WasteManagement Cost Assessment Team has preparedthe Cost Estimating Handbook for EnvironmentalRestoration to assist the DOE offices that estimatethe cleanup costs at various sites. The handbook

2. Ibid.

3. Ibid., p. 2-5.

4. The DOE cleanup program that is farthest along is the UraniumMill Tailings Remedial Action Project. The nature of cleanupwork at these sites is so different from cleanup at the major weap-ons facilities, however, that it provides only limited guidance tocost estimators.

APPENDIX B ESTIMATING THE COSTS OF CLEANING UP DOE SITES 71

describes different types of cost estimates and pro-vides guidance on how to develop them. It makessuggestions on how to deal with uncertainties, con-tingencies, cost escalation, and other factors, and itdescribes data bases, tools, and techniques that canbe useful to cost estimators in the field offices.

Building the cost estimates from the bottom up--estimating the cost of each component of work re-quired to complete a task-is just one approach tothe problem of estimating the environmental cleanupcosts facing DOE. Another approach is to developparametric cost-estimating models that expresscleanup costs as a function of such variables as thetype, volume, and concentration of contaminants;the medium (soil, groundwater, air); and the re-quired remedy. Both DOE and the EnvironmentalProtection Agency have developed such models tohelp estimate Superfund and similar cleanup costs.

DOE's Models for Estimating Costs

Cost analysts from the firm Independent ProjectAnalysis, Inc. (IPA), have prepared studies on esti-mating costs and schedules of assessment and clean-up projects for the Department of Energy.5 Thesestudies contribute to the understanding of assess-ment and cleanup costs by developing statisticalrelationships between costs and a number of vari-ables and between growth in costs and a number ofvariables. They have found that about three-quar-ters of the variance in assessment costs can be ex-plained by an equation with five independent vari-ables: the number of borings and new wells, an in-dex of site complexity, the threat posed to the sur-rounding community, the number of previous clean-up efforts at a site, and whether assessment tasks ata site are occurring in sequential phases or concur-rently.6 Growth in costs can largely be explainedby variables describing the complexity of the site,the complexity of the media, and the stage of the

project relative to other work at the site (referred toas project definition).7

IPA has developed similar models for theremediation of hazardous waste sites. It finds thatcleanup costs can be expressed as a function of sixkey variables: the volume of waste excavated, tech-nological complexity, whether the site is a landfill,whether there is mixed debris at the site, the com-plexity of the waste, and whether the primary threatis groundwater contamination.8 Together, thesevariables explain 96 percent of the variance incleanup costs. Sources of error in cost estimates arethe lack of complete information about the project,the type and complexity of remedial technologies,the complexity of the media, and the complexity ofwastes at the site.9

IPA's studies drew from experiences at a varietyof hazardous waste sites-not only DOE sites butalso EPA Superfund and private industrial sites.But since the DOE cleanup program is still rela-tively young, there is not much experience withsites containing radioactive wastes. As more DOEsites are assessed and cleaned up, IPA plans to enterthem into its data base and reestimate the relation-ships between costs and other factors.

EPA's CORA Model

The EPA uses the Cost of Remedial Action(CORA) model to estimate the costs of cleanup atindividual Superfund sites.10 EPA also aggregatesthe estimates generated by the CORA model in de-veloping the overall budget for its Superfund pro-gram. To use the CORA model, one enters data onsite characteristics, the kinds and amounts of con-taminants, the cleanup technologies, and other perti-

5. Brett R. Schroeder and others, The HAZRISK Assessment Study,prepared for the Department of Energy by Independent ProjectAnalysis, Inc., WD-90-04-HAZ (December 21, 1990); and B. R.Schroeder and J.B. Hartung, The HAZRISK Cleanup Report, pre-pared for the Department of Energy by Independent Project Anal-ysis, Inc. (Draft, February 1991).

6. Schroeder and others, The HAZRISK Assessment Study, p. 23.

7. Ibid., pp. 38-39.

8. Schroeder and Hartung, The HAZRISK Cleanup Report, pp. 24-25.

9. Ibid., pp. 45-46.

10. Environmental Protection Agency, Office of Solid Waste andEmergency Response, The Cost of Remedial Action Model, QuickReference Fact Sheet, Publication No. 9375.5-06a/FS (May 1991).In addition to estimating costs, CORA contains a module thatsuggests the kind of remedial action that should be taken at eachsite.


nent information; the model uses this information to up costs at mining sites or sites with radioactiveestimate cleanup costs. Tests of the model suggest waste, nor is it able to estimate costs associatedthat actual costs run from 30 percent below to 50 with such emerging technologies as in situ vitrifica-percent above the cost estimate it generates. Al- tion. It could be expanded, however, to cover addi-though the CORA model deals with most types of tional technologies or types of pollution,hazardous contaminants, it does not estimate clean-

Appendix C

Description of Specific IntegratedDemonstrations and Estimates of Savings

P reliminary results from the Department ofEnergy's research efforts have identifiedtwo areas in which the department feels

technology can yield an early payoff-groundwaterand soils cleanup, and waste retrieval and process-ing. DOE is conducting several integrated demon-strations in these two areas; they represent the twobiggest items in the portion of the budget for tech-nology development that funds research and devel-opment (see Chapter 4). Furthermore, one ofDOE's major technical challenges involves cleaningup contaminants in the soil or groundwater.

initially as well as methods of treating the contami-nated soil to reduce the volume of waste ultimatelyrequiring disposal. By monitoring soil as it is exca-vated using a special "rotomiH" machine, DOE canremove less material requiring subsequent treatment.The department is also attempting to further concen-trate plutonium in the soil using such techniques asmagnetic separation and centrifugation. Accordingto DOE estimates, rotomilling and then concentrat-ing the plutonium could reduce the amount of soilto be disposed of by a factor of five; the savings atthe Nevada Test Site would be $2 billion, or 80percent of the cost of using current technology.

Cleaning Up Soils

New technologies could save significant sums ofmoney by reducing the amount of contaminated soilrequiring disposal. Conventional treatment of con-taminated soil involves excavating all soil thatmight contain waste, treating it if appropriate, andthen disposing of it at an approved site. Withoutthe ability to precisely locate contaminants withinthe soil, current techniques using bulldozers oftenrequire excavating at least twice the estimated vol-ume of contaminated soil in order to ensure that allcontamination has been removed. DOE estimatesthat excavating and disposing of soil from the Ne-vada Test Site alone could cost $2.5 billion usingcurrent techniques.

DOE is investigating techniques that wouldreduce the amount of soil that must be excavated

Cleaning Up Groundwater

Cleaning contaminated groundwater is anotherthorny and major problem facing DOE. First,groundwater contamination has been detected atalmost all of the major installations in DOE's nu-clear weapons complex, although the extent andtypes of contamination at individual sites have notyet been fully characterized. Second, cleaning upcontaminated groundwater may be very difficult,expensive, and time consuming. The most commontechnique used for eliminating contaminants fromgroundwater is to extract the water from the aquiferand then treat it. This pump-and-treat process cantake a very long time because contaminants candiffuse or be absorbed into the material in the aqui-fer and be slowly released back into the water as itis being treated. For all of these reasons, cleaning


up the contaminated groundwater at DOE sitesusing current techniques could be a very difficultand costly, if not impossible, task.

To reduce the time and money needed to reme-diate the groundwater at its installations, DOE islooking at new methods for extracting contaminantsfrom groundwater that are more efficient than sim-ply pumping groundwater up to the surface to treatit. Again, DOE has identified significant potentialsavings in terms of cost, time (from years tomonths), or both, using various techniques underinvestigation. The department estimates that thesenew methods could save about $3 billion over thecleanup period.

One method, called air stripping, pumps com-pressed air down into the aquifer containing thegroundwater to flush out some of the contaminants,which are then collected in a vacuum grid locatedabove the water table. This method works particu-larly well for extracting volatile organic compoundsfrom groundwater. The vaporized contaminants arethen pumped from the vacuum grid up to the sur-face where they can be treated. DOE estimates thatair stripping can reduce the cost of cleaning upgroundwater contaminated with volatile organiccompounds by 65 percent compared with conven-tional pump-and-treat methods. DOE is conductingan integrated demonstration at Savannah River usingair stripping and feels that such a technique mightbe applicable to half of its sites and of those sitesbelonging to the Department of Defense (DoD) thathave groundwater contaminated with volatile or-ganic compounds.

As part of that demonstration, DOE is alsoinvestigating technologies that would destroy con-taminants while still in the aquifer. Such methods,called in situ remediation, would obviate the needfor any above-ground treatment of vapors or water.Techniques such as air stripping, by pumping gasinto the aquifer, would also allow the introductionof beneficial microbes for bioremediation or chemi-cals to break down contaminants in the ground-water. DOE estimates that such techniques couldreduce costs by 70 percent compared with conven-tional pump-and-treat methods and that bioremedia-tion might be applicable to 15 percent of DOE andDoD sites with contaminated groundwater.

Waste Retrieval andProcessing

The Department of Energy has large amounts ofwaste in various types of storage-more than 1 mil-lion 55-gallon drums, some of which are buried, andmore than 300 underground tanks. Many of thesedrums and tanks are decades old and are, or couldbe, leaking. A major task facing DOE involveslocating the waste and retrieving, characterizing, anddisposing of it in a safe and stable manner. DOE isinvestigating several techniques to address theseproblems, including ways to identify and separatecontaminated soil from clean soil and to stabilizeextremely radioactive waste stored in undergroundtanks and provide for its ultimate disposal. Inte-grated demonstrations are being conducted at theIdaho National Engineering Laboratory on buriedwaste, and at five sites including Hanford on under-ground storage tanks.

Buried Waste

The integrated demonstration at the Idaho NationalEngineering Laboratory is exploring and developingways to locate buried radioactive waste more pre-cisely and to retrieve the contaminated soil andseparate it from clean soil. By knowing where thewaste is buried, the amount of soil to be excavatedcan be reduced by 20 percent, according to DOEestimates. New and better ways to retrieve the soilcould eliminate the need for workers to wear protec-tive clothing, which limits their work time to two orthree hours per day. Improved methods would alsoincrease efficiency by a factor of at least two. Thecost of cleaning up waste buried in trenches usingcurrent technology ranges from $14,000 to $26,000per cubic meter. New characterization and retrievalprocesses could reduce costs to an estimated $700per cubic meter. DOE estimates that at least 59,000cubic meters of transuranic waste and surroundingsoil need to be cleaned up at the Idaho site. If thecost to retrieve and dispose of a cubic meter ofburied waste can be reduced to the levels estimatedby DOE, then savings at that site alone could reach$1.5 billion.

APPENDIX C DESCRIPTION OF SPECIFIC INTEGRATED DEMONSTRATIONS AND ESTIMATES OF SAVINGS 75

Underground Storage Tanks

The Department of Energy has more than 332 un-derground tanks at five installations that store wastein various forms with varying levels of radioactiv-ity. Some tanks are now 50 years old; because theywere intended originally only as temporary storagefacilities, some have degraded over time and someof their contents have leaked into the surroundingsoil.

DOE has chosen Hanford as one of the installa-tions at which to investigate and demonstrate newtechnologies for dealing with the problem of under-ground storage tanks. Hanford alone has more than170 underground storage tanks, 149 of which are ofthe obsolete, single-shell design most prone to leaks.The first task is to determine exactly what is in the

tanks, since some were filled more than 40 yearsago and no one knows precisely what chemicalsthey contain. Most of the material remaining in thetanks is a highly radioactive sludge or cake that isrelatively hard and would require drilling by specialtools to attain cores for sampling. Such a proce-dure, within the tank, would be difficult and costly.DOE estimates that it needs five samples from eachtank to determine the contents using current tech-niques. DOE also predicts that two fewer sampleswould be needed using newer techniques such aslaser range-finders deployed inside the tank on aremotely operated robotic arm to identify the bestlocations for taking samples. At $1 million persample, reducing the number of samples needed by40 percent could result in a savings of $300 millionat Hanford alone, according to DOE.

Appendix D

The Acquisition Process fora Major Weapon System

C hapter 4 included a discussion of ways inwhich the Department of Energy's manage-ment of its projects to develop new technol-

ogies might be improved. One possibility wouldinvolve establishing a framework for periodic re-view and decision points during a project. Theprocess used to monitor Department of Defense(DoD) programs was presented as a possible modelfor such a framework. This appendix provides abrief description of that process.

The Department of Defense has established aprocess to manage the development and productionof major weapon systems. This process is laid outin Department of Defense Instruction 5000.2, whichhas been revised many times in the past 20 years.Nevertheless, the basic concept remains the same-toprovide a basis for comprehensive management anddecisionmaking associated with shepherding aweapon from its inception until it rolls off a produc-tion line.

The design and development of major systemssuch as aircraft, missiles, and ships is a long pro-cess, in many cases taking 12 to 15 years. Aweapon system proceeds through a series of devel-opment stages, from identifying alternative conceptsfor the system to initial operational capability, de-ployment, and support. These stages are paralleledby a series of technical and management decisions,called milestones, made by the Under Secretary ofDefense for Acquisition, the head of the relevantservice (for example, the Secretary of the AirForce), or the service's delegated acquisition execu-tive.

First, the military services, the Office of theSecretary of Defense, or the Joint Chiefs of Staffdetermine that a particular mission requires a newweapon system to add operational capabilities orimprovements that will enhance the effectiveness ofexisting equipment. The originator prepares a mis-sion need statement (MNS) that is then reviewed bythe appropriate DoD component. Milestone 0 is thedecision point at which the Defense AcquisitionBoard (DAB) validates the need for a new weaponto meet the threat and grants permission to proceedto the next phase.

Following favorable review at Milestone 0, thenew weapon system proceeds into the concept ex-ploration and definition phase. During this rela-tively short period-typically one to two years-activity focuses on selecting the best alternative tofulfill the mission needs stated in the MNS. At thenext milestone, Milestone I, the service seeks ap-proval to initiate a new program and enter the dem-onstration and validation phase. The DAB estab-lishes baselines for cost, schedule, and performancecharacteristics to be met at the next milestone.

During the demonstration and validation phase,the program office responsible for the weapon sys-tem directs preliminary engineering and designwork-typically performed by a defense contractor-with an emphasis on reducing the risk of incorporat-ing new and emerging technologies into the finalweapon system. The contractor may develop earlyprototypes to demonstrate the feasibility of systems,subsystems, and components. Also during this


phase, which usually lasts two to three years but canextend for five or more years in the case of compli-cated systems, the contractor may conduct somepreliminary tests to demonstrate that the system isready to enter the next phase.

Milestone II marks the entry into engineeringand manufacturing development. At this decisionpoint, the review board must be convinced that theproduction of the weapon system, and its perfor-mance up to standards, are feasible. The program'scost, schedule, and performance characteristics,initially established at Milestone I, are updated.The new thresholds serve as development baselinesfor reports to the Congress. The DAB also reviewsand updates the plans for testing, acquisition, andsupport and logistics.

Following favorable review at Milestone II, theweapon system enters the engineering and manu-facturing development phase in which the finaldesign for the system is established. Tests are con-ducted to determine that design and performancecriteria are met and that the weapon system willperform as desired in an operational setting. Anyfinal design and engineering changes needed toready the system for production are made.

Milestone III marks approval for production. Atthis final major decision point, the review boardexamines results of tests conducted during the previ-ous phase and establishes the acquisition strategyand production baseline. Before the system mayenter full-rate production, the Secretary of Defensemust certify to the Congress that all operationaltesting has been completed successfully. The re-view board may approve initial production (at Mile-stone IIIA) before testing has been completed, witha proviso for subsequent review and approval forfull-rate production at Milestone IIIB following thecompletion of all tests.

Once the system has been produced and de-ployed to the field, management responsibility forthe system is transferred to the service and therelevant subordinate command. The military per-sonnel who use and maintain the weapon continueto monitor its performance so that problems can beidentified and fixed. Some weapons, after beingdeployed for a number of years, require major mod-ifications to address a different mission, correct anoperational deficiency, or incorporate new technol-ogy. If the modification is sufficiently expensive,its execution may generate another milestone (IV)and be subject to reporting requirements similar tothose associated with new weapons.

U.S. GOVERNMENT PRINTING OFFICE: 1994 - 301-967 - 814/17259

CONGRESSIONALBUDGET OFFICESecond and D Streets, S.W.

Washington, D.C. 20575

OFFICIAL BUSINESS

PENALTY FOR PRIVATE USE, $300

THIRD-CLASS MAILPOSTAGE & FEES PAID

C.B.O.WASHINGTON, D.C.

PERMIT No. G-70

cleaning up the department of energy's nuclear weapons complex

Documents

volatile organic compounds

pumps compressed air

environmental protection agency

nuclear weapons complex

producing nuclear weapons

acting assistant secretary

state regulatory authorities

world war ii