Nuclear Cleanup of Rocky Flats: DOE Can Use Lessons Learned to	 
Improve Oversight of Other Sites' Cleanup Activities (10-JUL-06, 
GAO-06-352).							 
                                                                 
In 2001, when GAO reported on the cleanup of the Department of	 
Energy's (DOE) Rocky Flats site, a former nuclear weapons	 
production facility, the cleanup was behind schedule and over	 
cost. In October 2005, the contractor declared that it had	 
completed the cleanup much earlier and at less cost than DOE and 
the contractor had anticipated 5 years earlier. GAO was asked to 
determine the (1) factors that contributed to the cleanup's early
completion, (2) remaining work and total costs, (3) measures to  
assess whether the cleanup achieved a level of protection of	 
public health and environment consistent with the cleanup	 
agreement, and (4) lessons the Rocky Flats cleanup may offer for 
other DOE cleanup projects.					 
-------------------------Indexing Terms------------------------- 
REPORTNUM:   GAO-06-352 					        
    ACCNO:   A56599						        
  TITLE:     Nuclear Cleanup of Rocky Flats: DOE Can Use Lessons      
Learned to Improve Oversight of Other Sites' Cleanup Activities  
     DATE:   07/10/2006 
  SUBJECT:   Contract oversight 				 
	     Contract performance				 
	     Cost analysis					 
	     Decontamination					 
	     Environmental assessment				 
	     Environmental cleanups				 
	     Lessons learned					 
	     Nuclear waste disposal				 
	     Performance measures				 
	     Rocky Flats (CO)					 

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO Product.                                                 **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
******************************************************************
GAO-06-352

     

     * Report to Congressional Requesters
          * July 2006
     * NUCLEAR CLEANUP OF ROCKY FLATS
          * DOE Can Use Lessons Learned to Improve Oversight of Other Sites'
            Cleanup Activities
     * Contents
          * Table
          * Results in Brief
          * Background
          * Four Key Factors Contributed to Early Cleanup
               * DOE and Contractor Overcame Previously Identified Cleanup
                 Challenges
                    * Completing Decontamination and Demolition of Structures
                    * Overcoming Limitations on Shipping and Transportation
                      Capabilities
                    * Identifying the Overall Scope of the Cleanup
                    * Problems with Automated Plutonium-Packaging System
                    * Preventing Safety Problems
               * Accelerated Cleanup Process Allowed Faster, More
                 Collaborative Work
               * Site-Specific Characteristics Aided Cleanup Effort
               * DOE Provided the Contractor with a Financial Incentive to
                 Finish Early and Under Cost
          * Cleanup of Rocky Flats Is Complete at a Cost of about $10
            Billion, but Key Steps Remain Before the Planned Wildlife Refuge
            Will Open
               * Regulatory Steps Remain Before the Site Can Open as a
                 Wildlife Refuge
               * Rocky Flats Cleanup Cost about $10 Billion, Including
                 Long-Term Costs
          * Numerous Measures Were Taken to Assess the Cleanup's Sufficiency,
            but DOE Could Improve Its Oversight of Data Quality and Clarify
            Its Verification Policy
               * DOE Did Not Independently Review the Quality of Cleanup Data
                 Crucial to the Regulatory Process
                    * Soil Cleanup Actions Followed a Four-Step Process
                    * Draft Comprehensive Risk Assessment Concluded That
                      Residual Risk Is Acceptable
                    * Rocky Flats Cleanup Data Quality Procedures
                      Comprehensive, but DOE's Oversight Was Lacking
               * DOE's Planned Cleanup Verification Not Completed
                    * DOE Policy Unclear on Application and Purpose of
                      Verification
                    * DOE Did Not Fully Implement Its Planned Verification or
                      Explain Its Reasons for Not Doing So
               * Independent Consultants and Other Federal Agencies Conducted
                 Additional Cleanup Reviews, and Some Influenced DOE's Final
                 Cleanup Strategy
                    * Several Recommendations from Independent Consultants
                      Were Incorporated into Site Remedies, but Disagreements
                      and Concerns Remain
                    * Other Federal Reviews Concluded Minimal Risk or Are
                      Still Under Way
          * DOE Implemented Some Lessons Learned at Rocky Flats but Has No
            Requirement to Ensure That Lessons Learned Are Implemented at
            Other Sites
               * DOE Identified Lessons from Rocky Flats and Assessed Their
                 Applicability to Other Sites
               * DOE Lacks a System for Ensuring Continued Collection and
                 Implementation of Lessons Learned at Its Cleanup Sites
               * Rocky Flats Offers Additional Lessons That May Be Applicable
                 to Other DOE Sites
          * Conclusions
          * Recommendations for Executive Action
          * Agency Comments and Our Evaluation
     * Objectives, Scope, and Methodology
     * GAO Survey Methodology, Response Rates, and Summary Responses
          * Survey Methodology
          * Survey Response Rates
          * Summary Responses
     * Additional Information on Decontamination and Demolition Activities at
       Rocky Flats
     * Additional Information on the Quality of Remediation Data at Rocky
       Flats and on DOE's Planned and Actual Verification
          * Information on Controls over Data Quality
          * Additional Information on DOE's Planned and Actual Verification
            Activities
               * DOE Reduced Size and Scope of Scanning and Sampling Portion
                 of Planned Verification
               * DOE Revised ORISE's Portion of Planned Verification to
                 Incorporate MARSSIM
               * Results of the Contractor's Final Scanning and Sampling
                 Cleanup Verification Plan and ORISE's Independent
                 Verification
     * Comments from the Department of Energy
     * Comments from the Department of the Interior
     * Comments from the Colorado Department of Public Health and Environment
     * Comments from Kaiser-Hill Company L.L.C.
     * GAO Contact and Staff Acknowledgments

                 United States Government Accountability Office

Report to Congressional Requesters

GAO

July 2006

NUCLEAR CLEANUP OF ROCKY FLATS

    DOE Can Use Lessons Learned to Improve Oversight of Other Sites' Cleanup
                                   Activities

                                       a

NUCLEAR CLEANUP OF ROCKY FLATS

DOE Can Use Lessons Learned to Improve Oversight of Other Sites' Cleanup
Activities

  What GAO Found

Four factors contributed to the early completion of Rocky Flats' cleanup:

(1) DOE's and the contractor's ability to overcome numerous challenges,
(2) the use of an accelerated cleanup process, (3) site-specific
characteristics that limited the scope of the contamination, and (4) the
contractor's financial incentive to finish the work quickly and safely.

Although the cleanup is complete, its sufficiency has not yet been
ascertained; key steps remain before the planned Rocky Flats National
Wildlife Refuge that will occupy the site can open to the public. For
example, in about November 2006, the regulatory agencies-the Environmental
Protection Agency (EPA) and the Colorado Department of Public Health and
Environment-plan to issue their joint final decision on the sufficiency of
the cleanup and any risk posed by residual contaminants.

The total cost of the cleanup, since 1995, is about $10 billion in
constant 2005 dollars. This cost includes contract costs of about $7.7
billion (including contractor fees of about $630 million), long-term
stewardship and pension liabilities estimated at about $1.3 billion, and
other costs of nearly $1 billion.

Although numerous measures in place to assess the cleanup appear adequate
to judge the sufficiency of the cleanup, DOE did not effectively carry out
some aspects of its oversight responsibilities. Among the assessment
measures are completion of the regulatory process, activities undertaken
to verify remedial actions, and reviews by independent and federal
entities. The regulatory agencies have approved the cleanup of 360 areas
of known or suspected contamination at the site. Data supporting the
cleanup of these areas form the basis of regulatory decisions regarding
the cleanup's sufficiency. Accordingly, we reviewed the contractor's
controls intended to ensure the quality of these data and found them to be
robust. However, DOE lacked assurance that the controls were working as
intended because it did not independently assess the quality of these key
data. One official told us that DOE was involved daily in reviewing
documents and discussed with the contractor any data quality issues that
arose.

DOE has identified and implemented at other sites some lessons from Rocky
Flats, but DOE has not systematically tracked lessons learned at all of
its cleanup sites, thus potentially losing the benefits of such lessons.

Source: DOE.

United States Government Accountability Office

                                    Contents

Letter                                                                   1 
                             Results in Brief                               4 
                             Background                                     8 
                             Four Key Factors Contributed to Early         10 
                             Cleanup                                      
                             Cleanup of Rocky Flats Is Complete at a Cost 
                             of about $10 Billion,                        
                             but Key Steps Remain Before the Planned      
                             Wildlife Refuge Will                         
                             Open                                          26 
                             Numerous Measures Were Taken to Assess the   
                             Cleanup's                                    
                                   Sufficiency, but DOE Could Improve Its 
                                                Oversight of Data Quality 
                             and Clarify Its Verification Policy           39 
                             DOE Implemented Some Lessons Learned at      
                             Rocky Flats but Has No                       
                               Requirement to Ensure That Lessons Learned 
                                                       Are Implemented at 
                             Other Sites                                   56 
                             Conclusions                                   62 
                             Recommendations for Executive Action          63 
                             Agency Comments and Our Evaluation            63 
Appendixes                                                             
                 Appendix I: Objectives, Scope, and Methodology            67 
                Appendix II: GAO Survey Methodology, Response Rates, and  
                             Summary                                      
                             Responses                                     70 
                             Survey Methodology                            70 
                             Survey Response Rates                         70 
                             Summary Responses                             71 
              Appendix III:  Additional Information on Decontamination    
                             and Demolition                               
                             Activities at Rocky Flats                     90 
                Appendix IV: Additional Information on the Quality of     
                             Remediation Data                             
                             at Rocky Flats and on DOE's Planned and      
                             Actual                                       
                             Verification                                  95 
                             Information on Controls over Data Quality     95 
                             Additional Information on DOE's Planned and  
                             Actual Verification                          
                             Activities                                    99 
                 Appendix V: Comments from the Department of Energy       107 
                Appendix VI: Comments from the Department of the Interior 110 
              Appendix VII:  Comments from the Colorado Department of     
                             Public Health                                
                             and Environment                              113 
              Appendix VIII: Comments from Kaiser-Hill Company L.L.C.     114 

Contents

:GAO Contact and Staff Acknowledgments
:Survey Response Rates, by Subgroup

Table

Figures

:Workers in Protective Suits Cut Plutonium-Contaminated Equipment 10
:Workers Use Cerium Nitrate to Decontaminate a Glovebox; Workers Load a
Dismantled Glovebox into a Shipping Container 12
:Use of Explosives in the Demolition of Rocky Flats' Building 881 13
: Transuranic Waste Packaged for Shipment to the Waste Isolation Pilot
Plant 15
:GAO Staff Being Briefed on the Prepackaging of Wet
Combustibles, July 2002 and October 2000 18
:Waste Removed and Contaminated Water Treated at Rocky Flats 29
:Rocky Flats, before Cleanup, in 1995, and after Cleanup (in 2005) 30
:Key Steps in Rocky Flats' Transition to a Wildlife Refuge 31
:The Lindsay Ranch 33
:Plan for Public Access Trails and Facilities at the Rocky Flats National
Wildlife Refuge 34
:Four-Step Remediation Process for Individual Areas of Known or Suspected
Contamination 43
:Treatment of Contaminated Soils by Thermal Desorption; Installation of a
Passive Treatment System at the Mound Storage Site, Rocky Flats 45
:Terminal Pond in the B Series of Ponds 46
: Super-Compacter without and with InstaCote 90
: Hydrolase System at Work Removing Contaminated Paint and Underlying
Surface Contamination 92
: Workers Cut Contaminated Concrete from Building 771's `Infinity Room' 93
: A High-Purity Germanium Detector Can Detect Certain Radionuclides 100
:Felled Electrical Poles Allow Lower Flight; Helicopter Scans Rocky Flats
for Residual Plutonium in Surface Soil 102

Contents

Figure 19: Confirmation and Independent Verification Samples Taken at the
903 Lip Area

  Abbreviations

Advisory Board Rocky Flats Citizens Advisory Board

ATSDR            Agency for Toxic Substances and Disease Registry          
CERCLA                 Comprehensive Environmental Response, Compensation, 
                    and Liability Act                                         
Coalition        Rocky Flats Coalition of Local Governments                
Colorado         Colorado Department of Public Health and Environment      
DOE              Department of Energy                                      
EPA              Environmental Protection Agency                           
FWS              Fish and Wildlife Service                                 
HPGe             high-purity germanium                                     
MARSSIM          Multi-Agency Radiation Survey and Site Investigation      
                    Manual                                                    
ORISE            Oak Ridge Institute for Science and Education             
PARCC            precision, accuracy, representativeness, completeness,    
                    and comparability                                         
pCi/g            picocuries per gram                                       
RCRA             Resource Conservation and Recovery Act                    
Refuge Act       Rocky Flats National Wildlife Refuge Act                  
RI/FS            remedial investigation/feasibility study                  
TRC              total recordable cases                                    
VOC              volatile organic compound                                 
WIPP             Waste Isolation Pilot Plant                               

This is a work of the U.S. government and is not subject to copyright
protection in the United States. It may be reproduced and distributed in
its entirety without further permission from GAO. However, because this
work may contain copyrighted images or other material, permission from the
copyright holder may be necessary if you wish to reproduce this material
separately.

A

United States Government Accountability Office Washington, D.C. 20548

July 10, 2006

The Honorable Jeff Sessions Chairman Subcommittee on Strategic Forces
Committee on Armed Services United States Senate

The Honorable Wayne Allard United States Senate

Once a bustling nuclear weapons production complex employing thousands of
workers, the Department of Energy's (DOE) Rocky Flats site, near Denver,
Colorado, is being transformed into a wildlife refuge that will be managed
by the Department of the Interior's Fish and Wildlife Service (FWS). For
about 40 years, the Rocky Flats plant produced plutonium triggers, or
"pits," for nuclear weapons. That activity resulted in radiological and
chemical contamination of some of the site's buildings, soil, and water.
Cleanup of the site began in 1995 under a contract between DOE and
Kaiser-Hill Company L.L.C.,1 and included decontaminating and demolishing
several plutonium-processing buildings, one of which had been labeled "the
most dangerous building in America," along with hundreds of other
contaminated buildings and structures. The cleanup required the contractor
to remove large volumes of nuclear material and radioactive debris, and
investigate and remediate as necessary contamination at 360 individual
hazardous substance sites that were identified through historical records
and extensive sampling and analysis at the site in the early 1990s. In
total, according to GAO's calculations, the amount of cleanup waste that
had to be removed from the site was equivalent to a 65-story building the
length and width of a football field.

Cleanup of the site was conducted under the Rocky Flats Cleanup Agreement,
the legally binding agreement that provided the framework for the cleanup
effort. The cleanup agreement implements the provisions of the applicable
statutes, including the Comprehensive Environmental Response,
Compensation, and Liability Act of 1980, as amended

1Kaiser-Hill Company L.L.C. is a joint venture between Kaiser Group
Holdings Inc. and CH2M Hill.

(CERCLA);2 the Resource Conservation and Recovery Act of 1976, as amended
(RCRA); and the Colorado Hazardous Waste Act. The cleanup agreement
specified the roles of DOE and the two regulatory agencies for the site:
the Environmental Protection Agency (EPA) and the Colorado Department of
Public Health and Environment (Colorado). Pursuant to the cleanup
agreement, EPA had lead regulatory authority over the cleanup of the
site's buffer zone, which includes about 5,900 acres of undeveloped land
around the center of the site, while Colorado had lead regulatory
authority over the cleanup of the core industrial area-which includes
about 385 acres in the center of the site, where the plutonium-processing
activities occurred.

When GAO reported on the cleanup effort in February 2001, the project was
behind schedule and over cost.3 However, in October 2005, the contractor
declared that it had finished the physical cleanup of the site, as defined
in the contract (e.g., demolishing buildings and shipping contaminated
waste and soil off site), much earlier and at less cost than DOE and the
contractor had anticipated 5 years earlier. Although the cleanup is
complete, its sufficiency has not yet been ascertained. EPA's and
Colorado's joint final decision on the sufficiency of the cleanup and the
final remedy for the site is expected to be published around November
2006.4 With the exception of one pedestrian trail, most of the planned
wildlife refuge will not open until at least 5 years after the refuge is
established (depending on the availability of funding).

In this context, we determined the (1) factors that contributed to the
physical cleanup's early completion; (2) work remaining to be done as well
as total project costs, including long-term costs; (3) measures in place
to assess whether the cleanup achieved a level of protection of public
health and environment consistent with the Rocky Flats Cleanup Agreement;
and

2Rocky Flats was added to CERCLA's National Priorities List in October
1989. The National Priorities List is EPA's list of national priorities
among the known or threatened releases of hazardous substances,
pollutants, or contaminants throughout the United States and its
territories. As of April 2006, the final list comprised 1,244 sites, with
another 59 sites proposed for listing. Another 309 sites that were
formerly on the final list have been deleted.

3GAO, Nuclear Cleanup: Progress Made at Rocky Flats, but Closure by 2006
Is Unlikely, and Costs May Increase, GAO-01-284 (Washington, D.C.: Feb.
28, 2001).

4The final decision document will be a joint corrective action decision
under RCRA and record of decision under CERCLA.

(4) lessons the Rocky Flats project may hold for other DOE cleanup
projects.

In conducting our work, we visited the Rocky Flats site several times and
reviewed documents and data prepared by DOE, EPA, Colorado, FWS, the
contractor, and various scientific organizations. We reviewed and analyzed
agency and contractor documents on decontamination and demolition
accomplishments and techniques. We also analyzed documents related to the
cost-plus-incentive-fee contract, including contract negotiation
documents, the contract cost and fee structure, and contract
modifications, although we did not evaluate the cost-effectiveness of the
contract. We reviewed and analyzed cleanup verification strategies and
results, and reviewed the content of scientific analyses, including a
study of how actinides5 migrate through soil and water and an assessment
of the public health risk posed by contaminants remaining at the site, but
did not evaluate the science underlying them. In addition, because
decisions about the sufficiency of the cleanup have been and will be based
on remediation data, we assessed the soundness of the agencies' and
contractor's processes and procedures for ensuring the quality of these
data. We also reviewed the results of audits of the key laboratories used
by the contractor to analyze samples of radioactively contaminated soil.

For further review, we selected a nonprobability sample of four of Rocky
Flats' 360 cleanup areas (including areas referred to as individual
hazardous substance sites, potential areas of concern, and under-building
contamination sites), using criteria such as the location of the cleanup
area and the severity of its contamination.6 For these four individual
cleanup areas-building 771, the 903 pad, the 903 lip area, and trench
T-7-we assessed the closeout reports and the data supporting them to
determine the extent to which the data collection and laboratory analyses
adhered to data quality standards and procedures. For example, we reviewed
records of laboratory analyses of contaminated soil samples from the
areas, and documents showing the extent to which the projects adhered to
stated data quality objectives. We did not evaluate the remedial data or
laboratory analyses themselves.

5Actinides are a series of 15 radioactive chemical elements with atomic
numbers 89 (actinium) and greater.

6Results from nonprobability samples cannot be used to make inferences
about a population, because in a nonprobability sample some elements of
the population being studied have no chance or an unknown chance of being
selected as part of the sample.

Page 3 GAO-06-352 Nuclear Cleanup of Rocky Flats

                                Results in Brief

Further, to obtain citizen and local government views about community
input to the cleanup and the effect of various cleanup verification
activities, we surveyed 58 current and past members of the two local
community groups that served in an advisory capacity to DOE: the Rocky
Flats Citizens Advisory Board and the Rocky Flats Coalition of Local
Governments. We also attended monthly meetings of these two groups. We
interviewed officials of DOE, EPA, Colorado, FWS, the contractor, and
various scientific organizations. We conducted our work in accordance with
generally accepted government auditing standards from March 2005 through
May 2006. In September 2005, we briefed you and your staff and reported on
the preliminary results of our review.7 Details on our objectives, scope,
and methodology are in appendix I; summary survey results are in appendix
II.

Officials of DOE, EPA, Colorado, and the contractor identified four key
factors that contributed to Rocky Flats' early cleanup:

     o DOE and the contractor overcame significant challenges that we
       identified in our 2001 report, such as decontaminating and demolishing
       hundreds of structures and packaging and shipping vast quantities of
       radioactive and hazardous wastes. For example, innovative techniques,
       such as the use of cerium nitrate to decontaminate gloveboxes so they
       could be shipped whole rather than cut into pieces, enabled the
       contractor to proceed with cleanup much faster and at less expense
       than anticipated.
     o An accelerated cleanup process allowed cleanup actions to proceed much
       more quickly and collaboratively than a traditional cleanup process
       would have allowed. As the cleanup progressed, DOE, the contractor,
       EPA, and Colorado staff often worked side by side in the field,
       participating in or observing soil removal actions and sampling
       procedures.

7GAO, Nuclear Cleanup: Preliminary Results of the Review of the Department
of Energy's Rocky Flats Closure Project, GAO-05-1044R (Washington, D.C.:
Sept. 22, 2005).

Page 4 GAO-06-352 Nuclear Cleanup of Rocky Flats

     o Site-specific characteristics (e.g., climate, geography, the robust
       construction of the buildings, and the chemical nature of the key
       contaminants) physically limited the extent of the contamination. For
       example, the dry Colorado climate and the alluvial fan8 on which the
       site is situated helped minimize erosion, thereby inhibiting off-site
       migration of contaminants. Also, the thick shale and claystone that
       underlie the site prevented contaminants from seeping into the deep
       drinking-water aquifer.
     o The contractor had a large financial incentive, shared throughout the
       contractor workforce, to complete the work economically, quickly, and
       safely. The earlier the contractor finished its work, and the lower
       its cost in doing so, the more the contractor stood to earn-as much as
       $560 million in incentive fees. This financial incentive drove site
       workers to seek creative cleanup solutions.

Although DOE's contractor completed the physical cleanup of the site in
October 2005, several additional actions need to be completed before the
site can open to the public as a wildlife refuge. In July 2006, DOE
expects to issue for public comment its proposed plan, which will include
a summary of the three proposed future monitoring and remediation
alternatives for the site and will identify the preferred alternative.
After public comments on the proposed plan have been received and
considered, EPA and Colorado will jointly determine whether any further
cleanup is needed at the site or whether the cleanup actions already taken
are sufficient to render the site safe for its intended purpose. A final
decision by EPA and Colorado is expected in late 2006. If the joint
decision is that no further action is required, and EPA certifies that the
cleanup and closure of Rocky Flats is complete, DOE will transfer primary
administrative jurisdiction over the majority of the site to the
Department of the Interior for use as a wildlife refuge. If funds are
available, FWS plans to open one pedestrian trail in the first year after
the transfer and to open the remainder of the public trails and facilities
between the fifth and 15th years following transfer. DOE expects to retain
possession of the former industrial area, which will contain shallow
groundwater treatment systems and monitoring wells, and will be
responsible for the long-term stewardship of these systems. To protect
these systems, DOE will restrict public access to this portion of the
site.

8An alluvial fan is a fan-shaped wedge of sediment that typically
accumulates in arid or semiarid climates on land where a stream emerges
from a steep canyon onto a flat area.

Page 5 GAO-06-352 Nuclear Cleanup of Rocky Flats

The total cost of the cleanup since 1995 is about $10 billion.9 This cost
includes DOE's 1995 and 2000 contracts with Kaiser-Hill, which total
nearly $8 billion; the cost of DOE's Rocky Flats Field Office, which
totals about $760 million; and the cost of other DOE sites' support of the
Rocky Flats project, which totals about $138 million; estimated future
costs such as pension and benefits liabilities, estimated at approximately
$1 billion over 60 years; long-term stewardship of the site, projected to
cost $303 million through 2080; and the potential acquisition of some
private mineral rights at Rocky Flats, for which the Congress authorized
$10 million in January 2006.10

Numerous measures have been and are being taken to assess the cleanup.
Although these measures appear adequate to judge the sufficiency of the
cleanup, DOE did not effectively carry out some aspects of its oversight
responsibilities. The three key cleanup assessment measures are (1)
completion of the regulatory process-EPA's and Colorado's review and
approval of the cleanup actions already taken and the results of other
ongoing assessments under CERCLA and RCRA; (2) a DOE-initiated
verification of the contractor's cleanup actions; and (3) reviews
conducted by other federal agencies and by consultants hired by local
community groups:

o  Completion of the CERCLA and RCRA process will culminate in EPA's and
Colorado's final decision on the cleanup's sufficiency and the selection
of a final remedy. The contractor's procedures to ensure the quality of
the data supporting the accelerated cleanup actions appeared sound and
comprehensive. However, DOE did not assess the contractor's data quality
assurance process, as required by the protocols agreed to with the
regulatory agencies.11 A DOE official explained that DOE officials'
day-to-day review of data enabled them to detect data issues as they
arose.

9Unless otherwise noted, all costs cited in this report have been adjusted
for inflation and are expressed in constant 2005 dollars.

10National Defense Authorization Act for Fiscal Year 2006, Pub. L. No.
109-163, S: 3112, 119 Stat. 3136, 3540 (2006).

11The review requirement was specified in the industrial area and buffer
zone sampling and analysis plan, which contained the sampling, analysis,
and documentation protocols agreed to between DOE and the regulatory
agencies.

     o A DOE-initiated verification of the contractor's cleanup was
       inconsistent with the verification plan. DOE did not complete an
       independent review of the accuracy of contractor-conducted scans for
       remaining radiological contamination. DOE officials said they had
       decided that these activities would not provide sufficient additional
       information to justify their completion, but had no documentation to
       support this decision. Further, existing DOE policy and guidance lack
       clarity about how they apply to the Rocky Flats cleanup and what the
       verification goals are.
     o Federal agency reviews and community-driven independent reviews
       assessed aspects of the cleanup. For example, a CERCLA-required public
       health assessment concluded that no public health hazard exists for
       the communities surrounding Rocky Flats.

Although DOE has identified and implemented at other cleanup sites some
lessons learned at Rocky Flats, DOE does not require that lessons learned
at one cleanup site be implemented at other sites, even though they may be
applicable. In 2002, DOE identified a number of lessons from the Rocky
Flats cleanup, including implementing a performance-based contracting
strategy and encouraging the innovative application of technology. DOE
then assessed the applicability of these lessons at some DOE sites and
directed these sites to implement applicable lessons. Although DOE
conducted follow-up reviews at the sites through 2005, these reviews were
focused on other aspects of the cleanup program and did not assess whether
the sites had implemented the lessons learned from Rocky Flats. During our
review, we noted additional lessons that could also be useful for other
DOE sites planning or undergoing cleanup, such as involving the future
site manager in remedial decisions and taking a consultative approach with
the regulatory agencies on cleanup decisions. However, DOE has no process
for ensuring that all lessons are captured and implemented as appropriate
at other DOE sites. As a result, DOE may be losing the chance to save both
time and money in its planned or ongoing site cleanup efforts.

We are making a number of recommendations aimed at strengthening DOE's
oversight of data quality, its cleanup assessments, and its process for
ensuring that lessons learned at cleanup sites are analyzed and
implemented as appropriate at other DOE sites.

We provided a draft of this report to DOE, Interior, EPA, Colorado, and
Kaiser-Hill for their review and comment. DOE, Interior, Colorado, and
Kaiser-Hill provided written comments (see apps. V, VI, VII, and VIII,
respectively), and generally agreed with the contents of the report. EPA
did not provide official written comments but did provide editorial and
technical suggestions, as did the other agencies, that we incorporated, as
appropriate. DOE concurred with one of our recommendations but did not
state whether it concurred with the other two.

For nearly 40 years, the Rocky Flats plant, located about 16 miles

  Background

northwest of downtown Denver, served as a nuclear weapons production
facility. Afterward, the site bore the scars of that role: soil,
groundwater, surface water, and many of the buildings at the site were
contaminated with radioactive materials, such as plutonium and uranium;
toxic metals, such as beryllium; and hazardous chemicals, such as cleaning
solvents and degreasers. Accordingly, the site became one of DOE's
priorities for environmental cleanup. While most of the approximately
6,300 acres that make up the Rocky Flats site served through the years as
an undeveloped buffer zone, about one-half of a square mile (385 acres) in
the center of the site constituted the industrial area, where for decades,
plutonium was shaped into plutonium triggers or "pits" for use in nuclear
weapons. About one-fourth of the site's more than 800 original structures
(e.g., buildings and storage tanks) were radiologically or chemically
contaminated-some severely so-by site operations over the years.

The site was cleaned up under the terms of a contract between DOE and
Kaiser-Hill. The first contract took effect July 1, 1995. In late 1999,
DOE and Kaiser-Hill negotiated a new contract, which took effect on
February 1, 2000. The cleanup work was done predominantly by
subcontractors, under the contractor's management.

The 2000 contract specified both the contractor's and DOE's
responsibilities. The contractor was responsible for processing,
packaging, and shipping off site all of Rocky Flats' nuclear materials and
radioactive and hazardous wastes;12 cleaning up and demolishing more than
700 structures that remained on site in February 2000; and cleaning up the
site's contaminated soil and groundwater. DOE was required to provide a
variety of services and items to support the project. Essentially, the
contract required DOE to arrange receiver sites for all the materials and
wastes and obtain the necessary certifications for the containers in which
the materials and wastes had to be packed and shipped. Many DOE sites
played a significant role in Rocky Flats' cleanup and closure, especially
those sites that received materials or wastes from Rocky Flats, such as
the Savannah River Site in South Carolina, the Waste Isolation Pilot Plant
(WIPP) in New Mexico,13 the Nevada Test Site, and the Hanford Site in
Washington.

Much of the cleanup work at Rocky Flats was labor intensive and tedious.
Plutonium is dangerous to human health, even in minute quantities,
especially if inhaled or ingested. Accordingly, workers handling
plutoniumcontaminated materials and equipment had to wear cumbersome
protective suits with enclosed respiratory systems, and sometimes had to
wield heavy and ungainly tools. Also, the equipment being worked on had to
be enclosed within plastic or glass to prevent airborne contaminants from
reaching unprotected workers or surfaces. Figure 1 shows workers in
protective clothing cutting contaminated materials to fit shipping
containers.

12These wastes-items such as clothing, gloves, equipment, rags, paper,
filters, and plastic- included low-level radioactive waste, transuranic
waste, and mixed waste. Low-level radioactive waste contains radioactive
constituents measuring 100 or fewer nanocuries of transuranic isotopes
(described below) per gram of waste (a nanocurie is 1 billionth of a
curie; a curie is the amount of radioactivity in 1 gram of radium).
Transuranic waste is radioactive waste contaminated with transuranic
isotopes (i.e., isotopes of elements heavier than uranium, such as
plutonium), with half-lives greater than 20 years, in concentrations above
100 nanocuries per gram of waste. Mixed waste is radioactive waste-either
low level or transuranic-that also contains hazardous wastes such as toxic
metals, cleaning solvents, degreasers, and paint thinners.

13WIPP is DOE's deep geologic repository for transuranic and transuranic
mixed wastes, located in an underground salt formation near Carlsbad, New
Mexico.

  Four Key Factors Contributed to Early Cleanup

Figure 1: Workers in Protective Suits Cut Plutonium-Contaminated Equipment

                                  Source: DOE.

Four key factors contributed to the early completion of the physical
cleanup of Rocky Flats: (1) DOE and the contractor overcame several major
challenges identified in GAO's 2001 report on the Rocky Flats cleanup, (2)
DOE and the site's regulatory agencies agreed to use an accelerated
process to clean up the site, (3) a number of site-specific
characteristics combined to limit the scope and complexity of the cleanup
effort, and (4) DOE offered the contractor $560 million in total incentive
fees to finish the cleanup ahead of schedule and under cost.

    DOE and Contractor Overcame Previously Identified Cleanup Challenges

Completing Decontamination and Demolition of Structures

The first key factor that contributed to the early completion of the
physical cleanup of Rocky Flats was that DOE and the contractor resolved
four of the five major challenges we identified in our 2001 report. These
challenges included (1) completing the decontamination and demolition of
the site's structures; (2) overcoming limitations on the available number
of transportation casks and on the loading capability for transuranic
waste;

(3) identifying the overall scope of the cleanup project-specifically, the
extent of contamination, the eventual use of the site, and the level of
cleanup that eventual use would require; (4) getting the automated
plutonium-packaging system to reliably perform at the rate needed for
timely completion; and (5) preventing safety problems, which can result in
work shutdowns and delay cleanup work.

One major challenge DOE and the contractor overcame was completing the
decontamination and demolition of hundreds of structures at the site. At
the end of fiscal year 2000, the contractor had completed only about 10
percent of the predemolition work, which included activities such as
removing plutonium and other nuclear materials from furnaces, pipes, and
other locations within buildings; draining and removing plutonium- or
uranium-laden liquids or residues from process pipes and tanks;
dismantling plutonium-processing furnaces; stripping out contaminated
process pipelines; and cutting up and removing hundreds of contaminated
gloveboxes.14 Further, at the time of our 2001 report, the contractor had
demolished only 81 of the 802 structures that existed at the site when
cleanup began. That accomplishment equated to about 10 percent of the
total number of structures and only about 5 percent of the total square
footage. Remaining to be demolished at the end of fiscal year 2000 were
721 structures, encompassing about 3.4 million square feet.

DOE and the contractor overcame challenges to decontamination and
demolition primarily through innovation, as workers continuously sought
innovative cleanup technologies that would accomplish tasks more safely,
quickly, and cheaply. For example, the contractor found that it could
chemically decontaminate large pieces of equipment, such as tanks and

14A glovebox is a closed glass, plastic, or metal chamber for handling
hazardous or radioactive material. The operator handles the material
through gloves sealed to the chamber's wall. Gloveboxes range in size from
a few square feet to thousands of square feet.

Page 11 GAO-06-352 Nuclear Cleanup of Rocky Flats

gloveboxes, by wiping their surfaces with a liquid cerium nitrate
solution.15 By doing so, the contractor could reduce the contamination
enough that the contaminated object could be shipped as low-level
radioactive waste rather than transuranic waste, which requires a more
costly shipping and disposal process. The Rocky Flats buildings contained
approximately 1,475 gloveboxes, ranging from shoe box size to the size of
an entire room. By reducing the level of contamination, the contractor
could ship even large gloveboxes whole, rather than having to cut them
into pieces small enough to fit into standard waste boxes used for
shipping transuranic waste. Figure 2 shows workers using cerium nitrate to
decontaminate a glovebox and loading a dismantled, decontaminated glovebox
into a container for shipping.

Figure 2: Workers Use Cerium Nitrate to Decontaminate a Glovebox; Workers
Load a Dismantled Glovebox into a Shipping Container

Source: DOE.

15After being used to decontaminate tanks and gloveboxes, the cerium
nitrate solution was processed and disposed of as low-level liquid waste.
According to Kaiser-Hill, cerium nitrate was one of several
decontamination solutions used at the site.

Page 12 GAO-06-352 Nuclear Cleanup of Rocky Flats

The use of explosives in building demolition was another important
timesaving demolition technique. Contractor officials said they had
learned lessons from another DOE site at which DOE had doubted the
demolitions experts and reduced the amount of explosives used, resulting
in an explosion that failed to bring the structure down. At Rocky Flats,
accordingly, the contractor and DOE brought in demolitions experts and,
while closely reviewing the demolition plans and overseeing the work, let
the experts' opinions prevail. As a result, the use of explosives saved
time and money for the contractor and DOE. For example, in bringing down
building 881, one of the four original manufacturing facilities at Rocky
Flats, a contractor official estimated that the use of explosives saved
about 4 months, and reduced risks to workers by removing hazards
associated with heavy equipment and the falling debris that typically
results from their use in lieu of explosives. After being completely
decontaminated, building 881 was wired with explosives and, as shown in
figure 3, collapsed into its basement.

Figure 3: Use of Explosives in the Demolition of Rocky Flats' Building 881

Source: DOE.

In addition to building 881, the contractor used explosives to bring down
smaller facilities, such as the guard and water towers, and to separate
the concrete from the rebar (the steel reinforcing bar) on the walls of
building 886.

Additional information on decontamination and demolition activities at
Rocky Flats is contained in appendix III.

Overcoming Limitations on   A second challenge DOE and the contractor      
Shipping and Transportation overcame was limitations on the available      
Capabilities                number of transportation casks and on the      
                               loading capability for transuranic waste. At   
                               the time of our 2001 report, the contractor's  
                               ability to ship the total volume of waste off  
                               site to DOE's WIPP facility by                 
                               the target closure date was in doubt because   
                               of uncertainties about                         
                               whether DOE could obtain and provide the       
                               number of transportation casks                 
                               necessary to meet the contractor's projected   
                               shipping schedule, especially                  
                               for fiscal years 2002 and 2003, and because of 
                               uncertainties about whether                    
                               the three loading facilities (two of which     
                               were still under construction at               
                               the time of our 2000 review) would be able to  
                               meet the shipping schedule                     
                               for the site's peak shipping years. That       
                               shipping schedule had been                     
                               compressed by various events, including delays 
                               in the opening of WIPP.                        
                               DOE and the contractor overcame this challenge 
                               primarily through                              
                               cooperation and perseverance. Whenever a DOE   
                               site could not use a                           
                               shipping container as scheduled, WIPP would    
                               send the container to Rocky                    
                               Flats instead. Also, DOE headquarters          
                               maintained a high priority on Rocky            
                               Flats throughout the cleanup and made sure     
                               that it received the                           
                               government-furnished services and items it     
                               needed. Figure 4 shows special                 
                               transportation casks used to transport         
                               transuranic waste.                             

Identifying the Overall Scope of the Cleanup

Figure 4: Transuranic Waste Packaged for Shipment to the Waste Isolation
Pilot Plant

Source: DOE.

Also, while buildings at Rocky Flats were being demolished, the contractor
built the two additional loading facilities, as previously mentioned, to
ensure that loading and shipping schedules could be supported. The new
facilities resolved the challenge of inadequate loading capability and
allowed the contractor to increase its transuranic waste-shipping
capability to 15 shipments a week.

A third major challenge DOE and the contractor overcame was identifying
the overall scope of the cleanup-specifically, the extent of
contamination, the eventual use of the site, and the level of cleanup that
eventual use would require. At the time of our 2001 report, the
postcleanup use of the site-and, therefore, the level of cleanup
required-was uncertain. Moreover, the extent of soil contamination on the
site and the depth and degree of contamination under many of the former
production buildings was not fully understood, particularly in the former
industrial area, because parts of the site had not been fully sampled.
Finally, at the time of our 2001 report, DOE, the regulatory agencies, and
the community stakeholders had not yet reached agreement on an appropriate
level of soil cleanup, although various studies were under way. These
challenges were resolved through congressional intervention and community
collaboration.

Congressional action ended the uncertainty about the end use of the site
with passage of the Rocky Flats National Wildlife Refuge Act of 2001
(Refuge Act).16 Knowing the site's end use enabled DOE and the regulatory
agencies to then make assumptions about site users and their risk of
exposure to residual contamination. The risk posed to a person by exposure
to a contaminant is measured by a combination of the contaminant's
toxicity, together with the frequency, pathway, and duration of exposure.
Thus, a more stringent cleanup level is required for a site that will be
used constantly, such as for residential purposes, than for a site that
will be used sporadically, such as for recreational purposes. Once the
Refuge Act was passed, DOE and the regulatory agencies assumed that the
principal site users would be wildlife refuge workers, who would spend
about 8 hours a day at the site (4 hours indoors and 4 hours outdoors),
250 days a year, for 18.7 years; and refuge visitors, who would spend
about 2.5 hours a day at the site, 100 days a year, for 30 years (6 years
as a child and 24 years as an adult). Soil cleanup levels could then be
set at a level that would protect both types of users.

DOE collaborated with the community in determining the level of
remediation the cleanup would achieve. Before passage of the Refuge Act,
the Rocky Flats Cleanup Agreement had specified an interim soil cleanup
level (action level) of 651 picocuries of plutonium per gram (pCi/g) of
soil,17 meaning that a soil cleanup action18 would be triggered by a soil
sample

16Pub. L. No. 107-107 S:S: 3171-3182, 115 Stat. 1012, 1379-1388. Prior to
enactment of the Refuge Act, DOE and the regulatory agencies proceeded
under the assumption that the land would be used in the future for open
space.

17A picocurie is a trillionth of a curie, which is the amount of
radioactivity in a gram of radium. The higher the soil action level, the
more plutonium can be left in the soil. A lower action level means that
less plutonium can be left, so more soil must be removed.

18Remediation of radioactively contaminated soil at Rocky Flats generally
meant removal. In general, it is more cost effective to remove
plutonium-contaminated soil than to treat and return it.

indicating the existence of plutonium in excess of 651 pCi/g.19 This
action level was considered unacceptable by the two stakeholder community
groups: the Rocky Flats Citizens Advisory Board and the Rocky Flats
Coalition of Local Governments; these groups argued for a much more
stringent action level. DOE funded a community-directed, independent
scientific assessment of soil action levels. This independent assessment
was overseen by a panel of community representatives. DOE also funded a
workshop, organized by the Rocky Flats Citizens Advisory Board, featuring
experts who provided presentations on radiological assessments, risk
analyses, and various computer-modeling scenarios and assumptions. Working
with local governments and members of the public, DOE and the regulatory
agencies ultimately agreed on a soil action level of 50 picocuries of
plutonium per gram (pCi/g) of surface soil (i.e., the top 3 feet of soil),
meaning that a remedial action would be triggered by a soil sample
indicating the existence of plutonium in excess of 50 pCi/g. A plutonium
level of 50 pCi/g translates to a risk level of 1 in 500,000 (that is,
this level of contamination could result in one more case of cancer than
otherwise would have been expected in every 500,000 persons).20 DOE and
the regulatory agencies decided that, with a surface soil action level of
50 pCi/g, higher concentrations of plutonium in soil could remain at
greater depth. Accordingly, DOE and the regulatory agencies agreed to set
an action level of 3,000 pCi/g for depths of 3 to 6 feet below the
surface, depending on the extent of the contamination. Once a remedial
action was triggered, cleanup would continue until contamination greater
than 1,000 pCi/g had been removed.

In addition to collaborating on decisions about soil cleanup levels,
community groups provided comments and advice on many other aspects of the
cleanup that influenced DOE's decisions in some cases. Our survey of the
Rocky Flats Citizens Advisory Board and the Rocky Flats Coalition of Local
Governments found that 21 of the 25 respondents were very or somewhat
satisfied with how DOE engaged the group in the cleanup process. Also, 19
of the 25 respondents said they were very or somewhat

19According to Colorado officials, the action level (651 picocuries per
gram) that was specified in the 1996 cleanup agreement was based on EPA's
draft proposed rule, which at the time was a dose-based, rather than the
current risk-based, calculation.

20Initially, according to Colorado officials, cleanup action levels for
all contaminants at the site, including uranium and americium, were set at
a level equivalent to a risk level of 10-5 (i.e., an excess cancer risk of
1 in 100,000). For plutonium, that level would be about 116 pCi/g.
However, through collaboration with the community stakeholders, a more
stringent level was set for plutonium: 50 pCi/g.

Problems with Automated Plutonium-Packaging System

satisfied with how DOE used the group's input. Nearly all (23 of 25) said
they believed the groups' input had a very or somewhat positive effect on
the cleanup process.

The unreliable functioning of the automated plutonium-packaging system is
the only one of the five challenges we identified in 2001 that DOE and the
contractor never fully resolved. It was only through persistence and
"brute force" that the packaging was finally completed in July 2003, more
than a year late and at least $50 million over budget. The system never
worked reliably, and many of the processes that were designed to be
automated had to be done manually. It required continuous maintenance and
modification, and broke down as many as two or three times a day,
requiring a "triage" repair team on call 24 hours a day. Nevertheless,
according to DOE officials, the pursuit of any alternative to the system
would probably have resulted in even greater cost and schedule delays.
Figure 5 shows GAO officials and staff being briefed on the glovebox line
behind the automated plutonium-packaging system; this glovebox line
handled the prepackaging of wet combustibles.

Figure 5: GAO Staff Being Briefed on the Prepackaging of Wet Combustibles,
July 2002 and October 2000

Source: DOE.

Preventing Safety Problems The final challenge that DOE and the contractor
overcame was safety problems, which can result in shutdowns and delay
cleanup work. At the time of our 2001 report, DOE was concerned about the
number and severity of safety violations that had occurred since the
inception of the 2000 contract. In response to those concerns, the
contractor developed a comprehensive plan to improve its safety and
compliance performance.

Although some safety incidents occurred after our 2001 report, the
contractor's overall safety performance improved. In two key measures of
safety, for example, Rocky Flats' performance was considerably better than
the average performance in the construction industry. One of these
measures is total recordable cases (TRC). Recordable cases, in general,
are any that require "more than a Band-Aid,TM" according to a contractor
safety official-for example, an injury that requires stitches,
prescription medication, or 1 or more days away from work. At Rocky Flats,
the 12month TRC rolling average at the end of 2004 was 0.9 per 100
full-time workers. This was a significant improvement over the 7.6 rate
that existed at the site in July 1995, when Kaiser-Hill took over from the
previous contractor. By comparison, the TRC average in the construction
industry for calendar year 2004 was 6.4 per 100 full-time workers. The
contractually established limit for TRCs at Rocky Flats was a 12-month
rolling average of

3.5 per 100 full-time workers. Using the other key measure of safety, lost
workdays, Rocky Flats also performed well. At the end of 2004, the site's
12-month rolling average rate for lost workday cases was 0.2, much lower
than the site's July 1995 rate of 4.6 and well below the construction
industry rate of 2.4 for calendar year 2004. The contractually established
limit for lost workdays at Rocky Flats was a 12-month rolling average rate
of 2.0.

The contractor's safety director reported that these safety improvements
were achieved through a combination of actions, including getting the
Chief Executive Officer, Chief Operating Officer, and other top management
officials involved; improving the work-planning process to incorporate
job-hazards analyses; involving the workers early in safetyplanning and
job-hazards analyses; holding first-line supervisors (i.e., foremen)
responsible for informing workers of top management's emphasis on safety;
and improving the training provided to workers. The site developed a
course, commonly known as Rocky Flats 101, that contained a hands-on,
simulated work environment in which workers learned how to use ladders and
various other tools in contaminated spaces. The course was very effective,
according to the safety director.

    Accelerated Cleanup Process Allowed Faster, More Collaborative Work

The second key factor that contributed to the early completion of the
physical cleanup of Rocky Flats was the use of an accelerated cleanup
process. The Rocky Flats cleanup began in the late 1980s and early 1990s
with extensive characterization (sampling of soil, groundwater, surface
water, and air). But after several years of characterization, public
frustration was building at the lack of visible progress in cleaning up
the contamination at the site. According to DOE officials, the extensive
sampling that was done at the site in the late 1980s and early 1990s,
together with historical documents about the use and disposal of materials
and wastes, facilitated use of the accelerated process.21 Instead of
proceeding under the full CERCLA/RCRA process, the Rocky Flats Cleanup
Agreement authorized DOE to perform most of the cleanup through removal
actions. In a 1996 report, we recommended that DOE make greater use of
removal actions to increase the efficiency and cost-effectiveness of
cleanups.22 Under the accelerated process, cleanup actions, including soil
removal, occur early and throughout the process rather than at the end,
with confirmation sampling at each cleanup area to ensure that the
remediation was sufficient.23 DOE officials report that the accelerated
process allowed more timely removal of contaminants from the site.

The pace of work in reviewing and approving cleanup documents was
"frantic," according to Colorado officials. Colorado, DOE, and the
contractor agreed on a "review template" of documents Colorado needed to
see for each building and area; the template allowed Colorado officials to
provide immediate review by working closely with the contractor.

21The Rocky Flats cleanup implements both CERCLA and RCRA. Although the
terminology used by each program differs, the CERCLA and RCRA processes
are similar. The typical full remedial processes under CERCLA and RCRA
include (1) determining the nature and extent of contamination; (2)
assessing potential threats to human health and the environment; (3)
establishing risk-based action levels; (4) identifying potential cleanup
technologies and processes, and evaluating cleanup alternatives and
corrective measures; (5) identifying the preferred remedy and obtaining
public comment on it; (6) selecting and implementing the final remedy; and
(7) monitoring and maintaining the remedy.

22GAO, Nuclear Waste: Greater Use of Removal Actions Could Cut Time and
Cost for Cleanups, GAO/RCED-96-124 (Washington, D.C.: May 23, 1996).

23The cleanup actions, called "accelerated actions," were planned and
executed to satisfy, among other things, the requirements of the
Environmental Restoration Rocky Flats Cleanup Agreement Standard Operating
Protocol for Routine Soil Remediation. Notification of the planned
activities was approved by EPA or Colorado, depending on where the
accelerated action was located. (EPA was the lead regulatory agency for
the buffer zone; Colorado, for the industrial area.)

Although the cleanup agreement specified that Colorado had 14 days (30
days in some cases) to review and provide comments on cleanup-related
documents, use of the review template shortened the review time
considerably. For example, the contractor would send draft documents to
Colorado officials, who would provide comments directly to the contractor;
the contractor then would respond by making corrections, providing
additional data or documents, or doing more work if necessary. By working
closely with the contractor along the way to resolve questions or
concerns, Colorado was able to respond quickly-sometimes on the same
day-to DOE's faxed requests for official approval of cleanup-related
documents. Colorado officials pointed out that had they not been willing
to work in this way with the contractor, but had instead taken the 14 or
30 days provided for review and approval under the cleanup agreement, the
contractor would not have been able to complete the cleanup as early as it
did. (DOE noted that Colorado's work with the contractor was not done
without DOE's involvement; as mentioned throughout the report, the entire
process was collaborative.)

DOE, EPA, and Colorado agreed that the accelerated process and the
collaboration it fostered were essential to the cleanup's early
completion. According to EPA and Colorado officials, their agencies had
staff on site "continuously," working in the field with DOE and contractor
staff to discuss and observe cleanup operations, and requiring "real time"
changes as necessary. Throughout the process, the regulatory agencies were
intimately involved in reviewing data and otherwise contributing to the
development of studies, plans, and decision documents. As a result, the
"back end" of the review process was very streamlined, according to
contractor officials.

While the accelerated cleanup actions were under way, and as data from the
completed cleanup actions became available, DOE concurrently conducted the
remedial investigation and feasibility study (RI/FS),24 which summarizes
site conditions at closure and evaluates necessary long-term remedies,
such as site monitoring. Based on all of the data collected at the site,
including the results of the individual accelerated cleanup actions and

24Throughout this report, our use of either a general descriptive term or
a CERCLA term for a specific regulatory document is intended to include
the RCRA equivalent as well. For example, if we refer to the final
decision document, we mean both the CERCLA record of decision and the RCRA
corrective action decision. Similarly, if we refer to the feasibility
study (a CERCLA term), we mean that to include the corrective measures
study (a RCRA term).

Page 21 GAO-06-352 Nuclear Cleanup of Rocky Flats

    Site-Specific Characteristics Aided Cleanup Effort

the RI/FS, DOE will identify its preferred final remedy in the proposed
plan, which will be released for formal public comment once the final
RI/FS has been approved by EPA and Colorado. DOE shared the draft RI/FS
with the public and obtained EPA's and Colorado's comments on it; these
comments were considered during preparation of the final RI/FS. After
considering comments received on the proposed plan during a 60-day public
comment period, DOE will select and document its preferred remedy in the
final decision document. Once EPA and Colorado approve the final decision
document, it will serve as both the corrective action decision (under
RCRA) and the record of decision (under CERCLA).

The third key factor that contributed to the early completion of the
physical cleanup of Rocky Flats was a confluence of
circumstances-climatic, geologic, chemical, structural and economic-that
confined the scope and complexity of the cleanup effort. Climatically and
geologically, the semiarid climate and the alluvial fan upon which the
site is located contributed to slow erosion of soils over time, thus
inhibiting the off-site migration of contaminants. The site also benefited
from layers of shale and claystone- hundreds of feet thick in places-that
prevented contamination from seeping into the deep drinking-water aquifer
that underlies part of the site.

Chemically, the cleanup was aided by the nature of the two key
contaminants-plutonium and americium, which are essentially insoluble in
water and, instead, tend to bind to soil and sediment particles, according
to a study undertaken in the late 1990s.25 At Rocky Flats, there is little
groundwater movement, and no groundwater was contaminated with plutonium
and americium.26 Instead, soil was the predominant environmental medium
requiring cleanup. Remediation of contaminated soil generally consists of
digging it up and shipping it to a disposal site, a simpler task than
remediating groundwater. The contaminants move around the site, and
potentially off site, by surface-soil erosion, wind, or surface water. For
example, westerly winds spread contaminated soil to off-site lands east of
Rocky Flats. In a 1997 combined corrective action decision/record of
decision about these lands,27 EPA and Colorado determined that the lands
did not pose an unacceptable or significant risk to human health or the
environment, notwithstanding the low levels of plutonium and americium
that existed in the soil-in portions of the lands-from Rocky Flats
activities.28

Structurally, the robust construction of the plutonium-processing
buildings, with their thick concrete floors and walls, contained many
leaks and spills of radioactive and hazardous materials, thereby limiting
the extent and severity of contamination under the buildings. Building
371, for example, was a complex, extremely strong, and rigid structure
that was built to withstand accidents such as earthquakes, tornadoes,
winds, and fires. The building was a partially buried structure of
reinforced concrete with an extensive foundation of concrete caissons up
to 6 feet in diameter, drilled

25Kaiser-Hill, Actinide Migration Evaluation Pathway Analysis Summary
Report, ER-108 (2002).

26Some contaminated groundwater seeps to the surface, particularly during
periods of rain or snow, and then trickles into ditches and streams. The
primary groundwater contaminants are volatile organic compounds (e.g.,
degreasers and solvents such as trichloroethene, carbon tetrachloride, and
chloroform). These contaminants are largely controlled by the groundwater
treatment systems on site, which were installed to protect surface water.

27EPA, Superfund Record of Decision: Rocky Flats Plant (USDOE), EPA ID:
CO7890010526, OU 3, Golden, CO, EPA/ROD/R08-97/196, 1997 (Golden, Colo.,
June 3, 1997).

28In 1990 a class action suit was filed in federal court against two
corporations that formerly operated the Rocky Flats Plant under contract
to DOE. Among other things, the plaintiffs claimed that their land was
damaged by the release of hazardous substances from Rocky Flats. In
February 2006, a jury found for the plaintiffs and awarded compensatory
and punitive damages. However, the court has not yet entered a judgment,
and the litigation is ongoing. Cook v. Rockwell International, No. Civ.A.
90-CV-181-JLK (D. Colo.).

    DOE Provided the Contractor with a Financial Incentive to Finish Early and
    Under Cost

into bedrock, and concrete wall and slab thicknesses beyond typical
industrial use and code requirements. Similarly, building 771 featured
concrete thicknesses ranging from 6.5 inches to more than 2 feet.

Finally, the local economy was a factor in the success of the cleanup,
according to Colorado officials. The local economy was not heavily
dependent on the cleanup project to provide jobs, so closing the site did
not meet large-scale community opposition.

The fourth key factor that contributed to the early completion of the
physical cleanup of Rocky Flats involved certain features of the cleanup
contract, including a substantial financial incentive for the contractor
to finish the cleanup work ahead of schedule and below budget. The
contract provided for an incentive fee tied partly to schedule but
predominantly to cost. The contractor would earn the target fee of $340
million if it completed its work between December 16, 2006, and March 31,
2007, at a cost between $4 billion and $4.2 billion.29 If the contractor
completed the cleanup outside of these schedule and cost ranges, the fee
would rise or fall accordingly-to as much as $460 million for less costly,
earlier completion, or as little as $130 million for more costly, later
completion. In 2003, when officials saw that the contractor was confident
of earning the maximum fee and was no longer working aggressively to
further shorten the completion timeline or reduce costs, DOE approved a
$100 million increase in the maximum incentive fee. In the end, DOE
awarded the contractor $49 million of the additional available fee, based
on a reduction in the total project cost of an additional $129 million.

29In this paragraph, the cost and fee amounts are those cited in the
contract and have not been adjusted to constant 2005 dollars; the contract
contains no provision for inflation.

Page 24 GAO-06-352 Nuclear Cleanup of Rocky Flats

DOE also encouraged the contractor to continue its practice of sharing
incentive fees among its workforce by offering employees financial
incentives to continue to work safely and generate ideas for ways to save
costs and time. In its initial contract bid for the 1995 contract for the
Rocky Flats cleanup, Kaiser-Hill committed to share 20 percent of its
profits with employees. The contractor's president explained that
incentives were offered across the workforce, although the incentive type
varied by work group. Salaried employees were individually evaluated based
on their project and safety performance; the value of their incentive pay
was based on the final total project cost. For hourly employees, incentive
pay was determined though collective bargaining but was primarily based on
schedule performance. Steelworkers received an annual incentive bonus,
based on schedule performance; workers in the building trades received an
annual hourly increase, also based on schedule performance. The contractor
also had a safety incentive program that included spot cash awards. Over
the life of the project, the contractor distributed approximately $100
million of company earnings (about 20 percent of its $510 million
incentive fee) to its workers, and DOE contributed an additional $30
million that was used for incentive pay.30

According to the contractor, financial incentives for early or
under-budget completion of a project are a powerful motivator; however, it
is important to note that the financial incentive offered at Rocky Flats
was only one of many factors contributing to the early completion of the
cleanup. In GAO's experience, such incentives are not always this
successful. As we reported in December 2005, the Department of Defense has
not fared well at using incentive-fee contracts to improve cost-control
behavior or meet program goals.31 Specifically, about half of the 27
incentive-fee contracts included in GAO's review failed or were projected
to fail to meet a key measure of program success-completing the objective
(i.e., delivering the product or service specified by the contract) at or
below the target price. Research on incentive fees by GAO, Harvard
University, and the RAND Corporation going back decades has concluded that
incentive fees do not consistently motivate contractors to control cost.

30DOE's contribution was specified in the contract, as 4 percent of salary
cost.

31GAO, Defense Acquisitions: DOD Has Paid Billions in Award and Incentive
Fees Regardless of Acquisition Outcomes, GAO-06-66 (Washington, D.C.: Dec.
19, 2005).

Page 25 GAO-06-352 Nuclear Cleanup of Rocky Flats

  Cleanup of Rocky Flats Is Complete at a Cost of about $10 Billion, but Key
  Steps Remain Before the Planned Wildlife Refuge Will Open

Additionally, according to the contractor, enhanced workforce transition
benefits could be considered important to worker motivation and
productivity. Enhanced benefits included pension and post-retirement
medical benefits and outplacement services such as relocation and
education benefits.

Another unusual and critically important feature of the Rocky Flats
contract was consistent funding. An underlying premise of the contractor's
ability to finish the job early and under budget was that a stable source
of funding would be provided throughout the cleanup. As part of the
contract negotiation process, DOE and the contractor assumed an annual
funding level of $657 million per year. Both DOE headquarters and Congress
helped each year to ensure that the site received a consistent funding
level. The contract also stipulated the services and items that the
government would provide, including making a number of shipping containers
available and arranging treatment and disposal sites for radioactive waste
at the site. If DOE did not meet its contractual obligation, it would be
grounds for changing the contract. In that way, DOE bore the liability for
any role it had in increasing contract costs.

The physical cleanup at Rocky Flats is complete, at a total cost
(including long-term costs) of about $10 billion;32 however, several
regulatory steps remain before land can be transferred to the Department
of the Interior for establishment of the wildlife refuge planned for the
site. After DOE issues and considers public comments on its proposed plan,
identifying its preferred alternative for the site's future monitoring and
remediation, EPA and Colorado will jointly determine whether any further
cleanup is needed at the site or whether the cleanup actions already taken
are sufficient to render the site safe for its intended purpose. A final
decision by EPA and Colorado is expected in late 2006. If they decide that
no further action is needed to remediate the site, EPA will delete
portions of the site from the CERCLA National Priorities List33 and
certify that the cleanup and closure of Rocky Flats has been completed;
the Secretary of the Interior can then accept administrative jurisdiction
of designated lands and establish the

32Unless otherwise noted, all costs cited in this section have been
adjusted for inflation and are expressed in constant 2005 dollars.

33Portions of the site being retained by DOE for long-term stewardship may
not be removed from the list.

Page 26 GAO-06-352 Nuclear Cleanup of Rocky Flats

    Regulatory Steps Remain Before the Site Can Open as a Wildlife Refuge

refuge, in accordance with the Rocky Flats National Wildlife Refuge Act of
2001.

Although the contractor has finished the cleanup at Rocky Flats, DOE, EPA,
and Colorado must complete the remaining steps of the regulatory process
before Rocky Flats can be removed from the National Priorities List. On
October 13, 2005, the contractor declared that it had successfully
completed the Rocky Flats cleanup-more than 1 year before the target
cleanup completion date cited in the contract and more than 3 years before
the completion date the contractor thought likely in our 2001 report.34

In declaring the cleanup complete in October 2005, the contractor stated
that it had met all of the cleanup requirements in the contract, which
included: (1) demolishing all buildings; (2) investigating and remediating
as necessary 360 individual cleanup areas; (3) removing wastes as
specified in the cleanup agreement; (4) installing closure caps for two
landfills; (5) covering appropriate areas such as building foundations,
utilities, paved roads, and parking lots with a minimum of 3 feet of fill
(e.g., Rocky Flats alluvium) after final grade; (6) ensuring that on-site
surface water meets health-based standards for open-space use; and (7)
ensuring that water leaving the site via two creeks will meet Colorado
water quality standards. The contractor reported its major cleanup
accomplishments in an October 2005 report.35 Among these accomplishments
were that the contractor

     o deactivated, decontaminated, removed, and cut up 1,475 gloveboxes, and
       disposed of them off site;
     o cleaned up and removed more than 800 structures, including more than 1
       million square feet associated with five major plutonium facilities
       and two major uranium facilities;
     o shipped to other DOE facilities 21 tons of special nuclear materials
       (plutonium and highly enriched uranium), including metals, oxides, and
       over 100 tons of plutonium residues;

34 GAO-01-284.

35Kaiser-Hill Company L.L.C., Executive Summary, Draft RCRA Facility
Investigation- Remedial Investigation/Corrective Measures
Study-Feasibility Study Report for the Rocky Flats Environmental
Technology Site, prepared for the Department of Energy, October 2005.

Page 27 GAO-06-352 Nuclear Cleanup of Rocky Flats

Major cleanup accomplishments at Rocky Flats

     o Shipped off site an amount of cleanup waste equivalent to a 65-story
       building the length and width of a football field, including more than
       15,000 cubic metersof transuranic and other radioactive waste, more
       than 500,000 cubic meters of low-level and other radioactive waste,
       more than 800,000 cubic meters of sanitary waste, and more than 4,300
       cubic meters of hazardous waste (see fig. 6).
     o Treated more than 11 million gallons of contaminated groundwater and 5
       million gallons of contaminated seep water-enough to fill 24
       Olympic-size swimming pools (see fig. 6).
     o deactivated, decontaminated, removed, and cut up 690 tanks, and
       disposed of them off site;
     o installed covers at two landfills;
     o investigated and remediated, as necessary, 421 areas of known or
       suspected contamination, including 360 cleanup areas such as
       individual hazardous substance sites and contamination located under
       buildings; and
     o installed three barriers and a seep collection system to treat
       contaminated groundwater plumes; installed passive treatment systems
       to protect surface water quality.

Figure 6 illustrates the equivalent amounts of waste removed and
contaminated water treated at Rocky Flats during the cleanup.

     Figure 6: Waste Removed and Contaminated Water Treated at Rocky Flats

        Waste volume

65-story building (809 feet) the size of a football field (160 x 360 feet)

        Ground and seep water volume

24 Olympic-size swimming pools (660,253 gallons each)

                                 Sources: GAO.

Figure 7 shows the site as it was in 1995, when the contractor began its
cleanup job, and on October 13, 2005, when the contractor declared its job
done.

  Figure 7: Rocky Flats, before Cleanup, in 1995 (left), and after Cleanup (in
                                     2005)

Source: DOE.

The contract provided that when the contractor declared the cleanup
finished, DOE would have 90 days to either accept the project as complete
or provide the contractor with a list of items to address. In November
2005, DOE provided the contractor with the final list. In December 2005,
after reviewing cleanup documentation, doing a final walk-through of the
site, and ensuring that the contractor had completed the remaining items,
DOE agreed that the contractor had fulfilled all of the cleanup actions
specified in the contract and the cleanup was complete.

The next steps include the review, revision, and final approval of the
remaining regulatory documents. Figure 8 shows the key steps in Rocky
Flats' transition from a cleanup site to a wildlife refuge.

      Figure 8: Key Steps in Rocky Flats' Transition to a Wildlife Refuge

Sources: GAO, based on information provided by DOE, EPA, and FWS.

Those steps are as follows:

     o July 2006: DOE expects to issue its proposed plan, specifying its
       preference for future monitoring and remediation of the site from
       among the following alternatives outlined in an RI/FS: (1) no further
       action, with monitoring; (2) monitoring plus both institutional and
       physical controls; and (3) monitoring and both institutional and
       physical controls, plus additional targeted removal of
       plutonium-contaminated soil in an isolated area to reduce the excess
       cancer risk for wildlife refuge workers from 1 in 500,000 to 1 in
       1,000,000. The proposed plan will also specify the area to be retained
       by DOE rather than transferred to FWS. As agreed between DOE and the
       regulatory agencies, the plan will be issued for a 60-day public
       comment period, which will include a public hearing.
     o November 2006: After considering public comments on the proposed plan,
       EPA and Colorado expect to finalize and issue the final joint decision
       document.36 The final decision document will specify what additional
       cleanup, if any, is necessary at the site and what monitoring and
       institutional and physical controls, if any, will be necessary. The
       document will also include a "responsiveness summary" that responds to
       questions from the public and other stakeholders on the proposed plan.
     o December 2006: EPA expects to certify that the cleanup is complete. It
       may then remove portions of the site from the National Priorities
       List.37
     o Early 2007: DOE expects to transfer jurisdiction of the majority of
       the site (approximately 5,400 acres) to FWS. The specific acreage to
       be transferred will be specified in the final decision document and
       will be contingent upon DOE's purchasing privately owned "essential
       mineral rights" (i.e., the right to mine sand and gravel) affecting
       about 700 acres. Under the legislation authorizing DOE to purchase
       these essential mineral rights, the refuge will not include land that
       is subject to sand and gravel mining or is being actively mined by
       private parties.38
     o 2012 through 2022: FWS expects to open public access trails and
       facilities in the wildlife refuge, as described in the agency's Final
       Comprehensive Conservation Plan.39 FWS's ability to implement its
       conservation plan is contingent on funding availability. In the first
       year following land transfer and refuge establishment, if funds are
       available, FWS will open one 1.75-mile pedestrian trail, which will
       follow the

36This final decision document will be a joint CERCLA record of decision
and RCRA corrective action decision.

37In the future, some part of the DOE-retained lands may also be deleted
from the National Priorities List. A site may be deleted from the list
even though operation and maintenance of a remedy continue. Accordingly,
at some point EPA may agree to delete the surface of the DOE-retained
lands but not the subsurface or the groundwater.

38Such lands are excluded from the refuge until the essential mineral
rights are purchased or the surface land is mined and reclaimed by the
mineral rights holder(s). National Defense Authorization Act for Fiscal
Year 2006, Pub. L. No. 109-163, S: 3112(b)(7)(B), 119 Stat. 3136, 3542.

39U.S. Fish and Wildlife Service, Rocky Flats National Wildlife Refuge:
FINAL Comprehensive Conservation Plan and Environmental Impact Statement
(Denver, Colo., Sept. 16, 2004).

existing road to the Lindsay Ranch homestead site, which dates from the
early 1940s (see fig. 9).

                          Figure 9: The Lindsay Ranch

Source: DOE.

Before opening other trails, FWS officials said they will concentrate on
restoring wildlife and plant habitat, controlling noxious weeds, removing
roads and culverts, and restoring the prairie grasslands. As funding
permits, between 2012 and 2022, they plan to phase in the remaining 15
miles of trail as well as other public-use facilities, such as parking
lots, a visitors' kiosk, wildlife-viewing blinds, and scenic overlooks.
Figure 10 depicts FWS's plan for visitor use at the refuge, including
about 16 miles of trails, parking, and developed overlooks.

                     Source: U.S. Fish and WildlifeService.

Even after the wildlife refuge opens, some stewardship activities will
likely continue at the site for an indeterminate time. For example, DOE
will maintain groundwater monitoring wells and treatment facilities as
long as necessary, subject to approval of the regulatory agencies.
Additionally, as required by CERCLA,40 DOE will review the site at least
every 5 years to determine whether remedies continue to be fully
protective of human health and the environment. DOE, EPA, and Colorado
reviewed parts of the site in 2002; the next review is scheduled for 2007.

Further, DOE negotiated an environmental covenant with Colorado for one of
the landfills at the site, called the Present Landfill. The purpose of
this covenant is to ensure protection of human health and the environment
by preventing intrusions into the landfill or damage to the various
engineered structures on adjoining lands and preventing exposure to
hazardous wastes. The covenant, which runs with the land in perpetuity, is
binding on DOE and all other interested parties (including persons using
the land), as well as their heirs and successors. Activities prohibited by
the covenant include digging, drilling, tilling, grading, and excavating,
as well as any activities that could damage or impair the proper
functioning of the landfill cap and runoff controls, the passive seep
intercept and treatment system, or the groundwater monitoring wells.
According to a DOE official, DOE and the state anticipate that the
covenant for the Present Landfill will be supplanted by a broader covenant
for the DOE-retained lands; this broader covenant will become effective
with execution of the final decision document (the combined corrective
action decision/record of decision).

The long-term care of the site is a subject of continuing debate and, for
some, concern. The long-term condition and care of the site were the
concerns most mentioned by community group survey respondents. Of the 17
(of 25) survey respondents who said they had remaining concerns about the
cleanup, 15 mentioned concerns about what would happen at the site in the
future.

Rocky Flats Cleanup Cost The total cost of the cleanup-about $10
billion-comprises costs incurred from 1995 through 2005 and estimated
long-term costs through about 2080.

    about $10 Billion, Including

Long-Term Costs Costs incurred to date include the following:

40CERCLA S: 121(c); 42 U.S.C. S: 9261(c). EPA and Colorado also have a
significant role in the 5-year review process.

Page 35 GAO-06-352 Nuclear Cleanup of Rocky Flats

     o The 2000 closure contract cost DOE over $4.1 billion through October
       2005-nearly half the total cost of cleanup through closure. This cost
       includes the $510 million incentive fee that DOE paid the contractor.
     o The 1995 cleanup contract cost DOE nearly $3.6 billion through early
       2000, including a $120 million fee to the contractor.
     o The cost of DOE's Rocky Flats Field Office from fiscal year 1995
       through October 2005 was nearly $760 million. This cost included staff
       salaries, site utilities, litigation support, regulatory oversight,
       and other expenses.
     o The cost incurred by other DOE sites in support of the Rocky Flats
       closure was approximately $138 million, for such activities as
       certifying shipping containers, providing transportation for nuclear
       materials and wastes, and receiving and storing Rocky Flats' materials
       and wastes.

In addition, in January 2006, Congress authorized DOE to spend up to $10
million on the purchase of designated "essential mineral rights" at the
Rocky Flats site.41 The law provided that these mineral rights at the site
can be purchased only from willing sellers and that the cost must not
exceed fair market value. In January 2006, DOE shared the results of an
appraisal of the mineral rights with the three ownership groups at Rocky
Flats. DOE will meet with various natural resources trustees, including
the Department of the Interior, to determine how to facilitate the sale of
the designated essential mineral rights.

DOE anticipates that long-term costs will exceed $1.3 billion. The bulk of
these costs will be the cost to DOE for pensions and postretirement
benefits (primarily medical and life insurance coverage) for the
contractor's Rocky Flats employees. These costs will likely exceed $1
billion, including about $822 million in medical benefits alone. DOE is
liable for such costs under the provisions of not only the 2000 contract,
but also previous site management contracts with Kaiser-Hill and its
predecessors. Postretirement benefits and pensions are part of the total
allowable compensation for DOE contractor employees, and DOE considers
them to be necessary to attract the most qualified employees. DOE Order
350.1 provides that when operations at a DOE facility are terminated and
no

41National Defense Authorization Act for Fiscal Year 2006, Pub. L. No.
109-163, S: 3112, 119 Stat. 3136, 3540 (2006).

Page 36 GAO-06-352 Nuclear Cleanup of Rocky Flats

other work is to occur under the contract (as in the case of Rocky Flats),
pension and postretirement health benefit continuation will be provided by
DOE for those contractor employees who earned retirement benefits in these
plans. According to actuarial estimates prepared for DOE by a consultant,
payments for contractor employees at Rocky Flats will continue until about
2064.

At Rocky Flats, 2,815 (43 percent) of 6,616 contractor employees were
eligible for postretirement benefits.42 These employees worked for either
the prime contractor (Kaiser-Hill) or one of its major subcontractors, and
retired between January 1, 1995, and December 31, 2005. The estimated cost
of over $1 billion does not include pensions and postretirement benefits
for employees who would have become eligible if the site had closed on its
original target closure date of December 31, 2006. When the site closed
more than a year ahead of schedule, the early closure date preceded the
eligibility date for full pensions and postretirement benefits of 34
employees, according to a contractor official.

Costs for the long-term stewardship of the site are estimated at $303
million. Long-term stewardship includes site surveillance and maintenance,
as well as management of site records and information systems. For many
decades to come, DOE will need to monitor environmental conditions at the
site and maintain the systems and structures that remain there (e.g.,
groundwater treatment systems, ponds and surface water control features,
and landfills). According to draft DOE guidance, surveillance and
maintenance refers to "all activities necessary to ensure protection of
human health and the environment following cleanup at a site, in
perpetuity." Thus, while long-term stewardship costs are estimated through
about 2080, some costs will continue beyond that, depending on the extent
of stewardship needed.

42Plan benefits were different for the eligible salaried employees (1,546
employees), hourly employees (1,187 employees), and security policy
officers (82 employees).

Page 37 GAO-06-352 Nuclear Cleanup of Rocky Flats

Our 2001 report noted that the total cost could rise if any claims for
monetary damages were brought against DOE to compensate for injuries to
natural resources. Under CERCLA, federal, state, and Indian tribal
officials who have been designated as trustees can file claims for
monetary damages for injuries to natural resources (including wildlife,
fish, and lakes) resulting from releases of hazardous substances. Damages
are usually for injuries that were not rectified by the cleanup and are to
be used to restore, replace, or acquire equivalent resources. In a
November 2005 report,43 DOE's Inspector General said that DOE had not yet
conducted the required Natural Resource Damage Assessment at Rocky Flats
and that only the completion of that assessment would determine whether
additional costs will be ultimately incurred.

But speculation about the potential cost of natural resource damage claims
at the site was laid to rest with passage of the 2006 National Defense
Authorization Act. Under the act, a natural resource damage liability
claim would be considered to be satisfied by the purchase of mineral
rights for $10 million, as authorized by the act, or the payment of the
authorized $10 million to the natural resource trustees, or a combination
of the two, for a total payment of $10 million.44 Although DOE had agreed
with the Inspector General's recommendation to immediately initiate the
damage assessment, the subsequent provisions of the Defense Authorization
Act rendered the damage assessment requirement moot.

43DOE, Management Controls over Assessing Natural Resource Damage at Rocky
Flats, OAS-M-06-02 (Washington, D.C., Nov. 25, 2005).

44Pub. L. No. 109-163, S: 3112(b)(4), 119 Stat. 3136, 3540-41. For the
Rocky Flats site, according to the cleanup agreement, the designated
natural resource trustees are the Secretary of Energy, the Secretary of
the Interior, the Executive Director of the Colorado Department of Public
Health and Environment, the Colorado Attorney General, and the Deputy
Director of the Colorado Department of Natural Resources.

  Numerous Measures Were Taken to Assess the Cleanup's Sufficiency, but DOE
  Could Improve Its Oversight of Data Quality and Clarify Its Verification
  Policy

Numerous measures were and are being taken to assess the sufficiency of
the cleanup; although these measures appear adequate, DOE did not carry
out some aspects of its oversight responsibilities. Cleanup assessment
measures include (1) the completion of the CERCLA and RCRA regulatory
process, (2) DOE's verification of the contractor's remediation of
radiologically contaminated soil, and (3) other reviews by federal
agencies and independent consultants. The first assessment
measure-completion of the CERCLA and RCRA regulatory review and approval
process-relies on remediation data collected throughout the cleanup. Our
review showed that the contractor appeared to have comprehensive quality
controls for data. Nevertheless, DOE did not independently review the
quality of these data. A DOE official said that, while he had no
explanation for DOE's not conducting the required reviews, DOE officials'
day-to-day review of data did enable them to detect data issues as they
arose. The second assessment measure-DOE's verification of the
contractor's actions to remediate radiologically contaminated surface
soil-was not completed. DOE decided to eliminate parts of the planned
independent review of the accuracy of contractor-conducted scans for
remaining radiological contamination because DOE officials decided that
the likely results would not justify the completion of an independent
review. The third assessment measure-outside reviews-included a consultant
review that convinced DOE to revise a key water management strategy at the
site, and a federal review that found no public health hazard for the
communities surrounding Rocky Flats.

    DOE Did Not Independently Review the Quality of Cleanup Data Crucial to the
    Regulatory Process

DOE did not independently review the quality of the data obtained
throughout the cleanup-data crucial to the first assessment of the
cleanup's sufficiency, the CERCLA and RCRA process. As discussed earlier,
the final steps of this regulatory process are still in progress,45 but
much of the process is complete. Key data were obtained through a
four-step process of remediating 360 individual areas at Rocky Flats, and
EPA and Colorado documented their approval of these cleanups in closeout
reports, giving DOE officials confidence that the site will meet the final
cleanup requirements. The closeout reports also include data on residual
contamination (i.e., the contamination that remains following completion
of the cleanup action). The residual contamination data were analyzed in a

45As discussed earlier, the regulatory agencies anticipate issuing the
final decision document at the end of 2006.

Page 39 GAO-06-352 Nuclear Cleanup of Rocky Flats

Soil Cleanup Actions Followed a Four-Step Process

draft comprehensive risk assessment that was part of the remedial
investigation/feasibility study, which determined that the site's overall
risk falls within EPA's acceptable risk range. Because the data from the
accelerated actions are crucial to the regulatory agencies' final decision
on the sufficiency of the cleanup, we reviewed the controls in place to
ensure the quality of these data. We discussed these data quality policies
and procedures with DOE, EPA, Colorado, and contractor and subcontractor
officials and determined that, although the contractor's data quality
controls appeared robust and comprehensive, DOE could improve its
oversight of them. Further, our review of several closeout documents
revealed that the policies and procedures were generally followed.

At Rocky Flats, accelerated soil cleanup actions were generally completed
and approved through a four-step process.46 As of May 2006, EPA and
Colorado officials said they had approved the closeout reports on all 360
areas of known or suspected contamination at Rocky Flats and had concluded
that no further accelerated cleanup actions were necessary for those
areas.47 Closeout reports are the last step of a regulatory approach that
documented the accelerated cleanup actions taken, quantified contaminants
remaining after the cleanup actions were completed, and documented whether
the project goals had been met. The remediation of each area of soil
contamination at Rocky Flats followed a four-step process:
characterization, contaminant remediation, confirmation sampling, and
clean fill and project documentation.

The first step-characterization-identified whether contamination had
occurred. The contractor had to design a sampling and analysis approach
for each specific cleanup area and receive DOE's approval before
submitting the approach to the regulatory agencies for their review and

46Much of the contamination at Rocky Flats was confined to the soil.
Hazardous and radioactive waste was buried in trenches, and soil in areas
where spills or leaks had occurred was contaminated with solvents and
other hazardous chemicals.

47Our references to closeout reports include data summary reports, which
were prepared and approved in lieu of closeout reports if no cleanup was
necessary at a cleanup area.

Page 40 GAO-06-352 Nuclear Cleanup of Rocky Flats

approval.48 Once a potentially contaminated area was identified, the
contractor would propose one of several approaches to collect the
necessary samples. For example, the contractor might choose a standard
statistical sampling approach, which entails designing a sample grid
capable of identifying an elevated area of contamination with 90 percent
confidence, then applying that grid to the actual cleanup area. Sample
collection involved taking small scoops of dirt, which were analyzed in a
laboratory for a variety of contaminants such as metals, volatile organic
compounds, and radionuclides.

The second step-contaminant remediation-cleans up any contaminants that
characterization found to exceed acceptable levels. Again, the contractor
would prepare a remediation plan, based on a standardized approach for
removing soil,49 in consultation with EPA, Colorado, and members of the
public, for EPA and Colorado to approve. Once approval was granted, the
contractor could proceed with the actual cleanup. Cleanup typically
involved digging out the contaminated soil, packaging it, and shipping it
to a licensed treatment or disposal facility. The depth to which soil was
removed varied depending on the type and severity of contamination. In
some areas, less than a foot of soil needed to be removed; in other areas,
20 feet of soil needed to be removed. The contractor took samples
throughout the remediation to identify when the remediation goal had been
met and could then move on to the third step. According to officials of
both EPA and Colorado, they frequently directed their staff to observe the
cleanup actions to ensure they were being implemented as agreed. Colorado
officials said that, while the frequency and duration of their oversight
varied by project, daily interaction and observation by Colorado staff was
typical.

The third step-confirmation sampling-required taking small scoops of dirt
from areas where the contractor believed remediation was complete, and
having them analyzed for the presence of previously identified
contaminants. The samples could be analyzed at an on-site laboratory or

48The characterization approach is documented in a Sampling and Analysis
Plan Addendum. Occasionally, when schedule concerns existed and the
planned work was routine in nature, DOE allowed the contractor to send
draft documents to the regulatory agencies while DOE was reviewing them.

49Depending on the complexity and the time required to complete the
remediation, the remediation plan was set forth in an interim
measure/interim remedial action, a proposed action memorandum, or a Rocky
Flats Cleanup Agreement standard operating protocol.

sent off site, depending on the types of contaminants. If the laboratory
confirmed that the contaminants had been removed to an acceptable level,
the contractor could move on to the next step. If the laboratory results
showed an unacceptable level of contaminants, then additional soil would
be removed, followed by additional confirmation sampling. Additional soil
removal and sampling rarely occurred, however, because field measurements
and analyses were used to guide the remediation work before the final
confirmation samples were taken.

The fourth and final step-clean fill and project documentation-included
filling the excavated area and preparing a closeout report that documented
that the cleanup had met the stated goals and quantified any residual
contamination. The contractor generally filled the excavated area with
Rocky Flats alluvium, a mixture of sand, dirt, and gravel. Documentation
of residual contamination was important to the regulatory process because
these data were instrumental in completing the remedial investigation and
feasibility study, including the comprehensive risk assessment.50 After
DOE approved the closeout report, it was delivered to the regulatory
agencies for final approval. The regulatory agencies' approval meant that
the cleanup was sufficient and that no further accelerated action was
needed. This four-step remediation process is illustrated in figure 11.

50The comprehensive risk assessment, prepared under CERCLA and included in
the draft remedial investigation and feasibility study, is an evaluation
of the potential adverse impacts to human health and the environment that
may exist from contaminated environmental media associated with
site-related activities. See 40 C.F.R. S: 300.430(a)(2).

Page 42 GAO-06-352 Nuclear Cleanup of Rocky Flats

Although soil remediation was the primary cleanup focus at Rocky Flats,
some groundwater required remediation for elevated levels of uranium,
nitrates, and volatile organic compounds. Three passive systems were
installed to treat the contaminants in the groundwater. At one former
waste-drum storage area, for example, remediation of a carbon
tetrachloride plume, along with other volatile organic compounds, involved
excavating about 700 cubic yards of contaminated soils, treating them
through thermal desorption,51 and then constructing a passive treatment
system that uses iron filings to cleanse contaminants from the
groundwater. Figure 12 shows the cleanup of a plume of carbon
tetrachloride and other volatile organic compounds (VOC) that originated
from a waste-drum storage area at Rocky Flats.

51Thermal desorption is a remedial technology that uses heat to physically
separate (desorb) volatile organic compounds (VOC) from excavated soils.
As the soils are subjected to high heat (e.g., 900 degrees Fahrenheit),
the VOCs change into gas (vapors) and evaporate out of the soil.
Essentially, the process involves collecting the vapors and further
heating them (to as much as 1,800 degrees), which causes oxygen to react
with the hydrogen and carbon in the VOC vapors, forming water and carbon
dioxide. Hydrochloric acid is also formed from the chlorine in the VOCs.
The carbon dioxide, water, and hydrochloric acid are then neutralized,
forming water and salt. The treated soils are tested to ensure that they
meet cleanup objectives; the soils can then be returned to the excavation
area.

Source: DOE.

Because of the tendency of plutonium and americium particles to cling to
soil, a primary concern among stakeholders was the potential off-site
transport of soil contaminated with plutonium and americium via surface
water. Contaminated soil washes into ditches and streams when it rains or
snows. Accordingly, cleanup efforts have focused on the surface soils to
reduce the source of radionuclides that could be susceptible to erosion
and eventual transport off site via Walnut and Woman Creeks. Also, surface
water is managed and monitored both on site-through a series of holding
ponds-and at the points where it leaves the site. The ponds, which DOE
constructed decades ago to retain any major release during operations, are
used to ensure that water leaving the site meets all applicable water
quality standards.52 DOE operates the final pond in each series as a
"batch and release" system; that is, releases from those ponds occur only
after water sampling has confirmed that the water meets water quality
standards. The

52The ponds were not constructed as a remedial action; rather, they are
historical features of the site that were left in place and serve as
insurance that contaminated surface water does not leave the site.
Further, they serve as excellent flood control mechanisms. One of the
objectives of the cleanup is for all surface water on site to meet
Colorado water quality standards. Rocky Flats currently has a waiver-for
nitrate, nitrite, and organic parameters-in one stream segment; that
waiver will expire in 2009.

Page 45 GAO-06-352 Nuclear Cleanup of Rocky Flats

Draft Comprehensive Risk Assessment Concluded That Residual Risk Is
Acceptable

efficacy of the accelerated cleanup actions taken to protect water quality
will be evaluated in the remedial investigation/feasibility study. Figure
13 shows the terminal pond in one series.

Source: GAO.

The draft sitewide comprehensive risk assessment, one of several documents
that will contribute to the final regulatory decision on the sufficiency
of the Rocky Flats cleanup, concluded that the cumulative remaining risk
to human health and the environment at Rocky Flats is acceptable. The
draft comprehensive risk assessment-part of the larger remedial
investigation and feasibility study-concluded that the cumulative risk to
human health, after completion of the accelerated cleanup actions, falls
at the more protective end of EPA's "excess cancer rate" risk range of
between 1 in 10,000 and 1 in 1,000,000.53 This is the range that EPA
considers to be adequately protective of human health.

As discussed previously, the Rocky Flats cleanup was designed so that any
residual contamination after completion of the accelerated cleanup actions
would be documented in closeout reports and subsequently analyzed in
accordance with the exposure assumptions in the sitewide comprehensive
risk assessment. The comprehensive risk assessment incorporated
approximately 2 million data records and examined both human and
ecological risk. Specific tasks included

     o identifying remaining contamination at Rocky Flats and determining
       whether it can adversely affect humans, animals, or plants;
     o determining whether pathways exist whereby human or ecological
       receptors may come into contact with these contaminants, for example,
       through inhalation or ingestion of surface water;
     o assessing the contaminants' potential effects, including cancer, on
       the human and ecological receptors, where complete pathways exist; and
     o calculating the potential risk to the human and ecological receptors,
       based on the pathways and the levels and toxicity of the contaminants.

The human health risk assessment calculated the risk that residual
contaminants posed to wildlife refuge workers and visitors. The assessment
identified five residual contaminants at the site that could affect human
health, including plutonium located adjacent to the most contaminated
precleanup areas of the site. The risk model assumed that the primary
pathways for plutonium were either through inhalation or through skin
contact. Based on assumptions about how frequently a wildlife refuge
worker would be exposed to residual plutonium at the site, the model
calculated an EPA acceptable risk of 1 in 500,000 excess cancers, and an
even lesser risk for a refuge visitor.

The ecological risk assessment estimated the risk that exposure to any
residual contaminants would pose to species of concern, including the

53For example, an excess cancer rate of 1 in 1,000,000 means that, given
exposure to the residual contamination at Rocky Flats at the end of the
cleanup, one would expect only one additional (excess) cancer than would
normally be expected in a million people.

Page 47 GAO-06-352 Nuclear Cleanup of Rocky Flats

Rocky Flats Cleanup Data Quality Procedures Comprehensive, but DOE's
Oversight Was Lacking

Preble's Meadow jumping mouse, a protected species. After modeling risks
to selected terrestrial and aquatic species, the risk assessment concluded
that the residual contaminants at Rocky Flats do not represent a
significant risk of adverse ecological effects to these species.

Nearly all respondents to our community group survey expressed confidence
in the cleanup's outcome. Specifically, 11 of 25 respondents said they
were very or extremely confident that the cleanup will be protective of
public health and the environment, and another 12 said they were somewhat
confident.

One of the most important aspects of the cleanup process was ensuring the
validity of the data used to determine whether the site had been
remediated to the agreed-upon levels. However, DOE did not complete the
independent and management assessments required by the cleanup agreement
to ensure that these data quality controls were working as intended. We
identified four key data quality controls: establishing data quality
objectives, using data quality parameters, verifying and validating data,
and auditing laboratories that analyzed samples for Rocky Flats. Our
review showed that the contractor generally followed these data quality
controls and documented its compliance with them. Appendix IV contains
additional information on the four data quality controls and our review of
the contractor's compliance with them.

Although the data quality controls the contractor had in place appeared to
be robust and comprehensive, DOE did not conduct independent assessments
to ensure that this was the case. Specifically, DOE did not conduct the
management and independent assessments called for by the quality assurance
project plan.54 A quality assurance project plan describes the planning,
implementation, and assessment procedures for a particular project, as
well as any specific quality assurance and quality control activities.55
These independent assessments, according to DOE's Management Assessment
and Independent Assessment Guide, are intended to measure the adequacy of
work performance, among other things. Among the essential areas that
should be assessed, according to the

54The Rocky Flats Quality Assurance Project Plan is included in the
Industrial Area and Buffer Zone Sampling and Analysis Plan.

55The required assessments are based on DOE Order 414.1A on Quality
Assurance, which directs field managers to perform independent assessments
of their contractors to evaluate the adequacy and implementation of their
quality assurance plans.

Page 48 GAO-06-352 Nuclear Cleanup of Rocky Flats

guide, are data quality controls such as calibration controls, computer
software controls, and environmental management systems.

DOE's failure to conduct independent assessments is particularly troubling
because of the importance of the cleanup and residual contamination data.
These data were not only the basis for EPA's and Colorado's approvals of
the accelerated cleanup actions, but also the foundation for EPA's and
Colorado's pending decisions about the overall sufficiency of the site's
cleanup. Without independent assessments of the contractor's data quality
control measures, DOE had no assurance that the controls were working as
intended. Also troubling was that EPA and Colorado-the regulatory agencies
that jointly approved the site's quality assurance project plan and are
responsible for ensuring its implementation-were unfamiliar with these
assessment requirements. When we discussed with EPA officials DOE's
failure to conduct independent or management assessments, they
acknowledged that their confidence in the data quality would have been
increased had DOE completed these assessments. A DOE official said he had
no explanation for DOE's not conducting the required assessments, other
than that DOE officials had reviewed sampling and analysis plans,
remediation plans, and closeout reports, and discussed with the contractor
any data quality issues that arose.

Colorado officials said they reviewed the data and the controls provided
by the contractor, which is the state's standard procedure for determining
the usability of data provided it, and that DOE's failure to conduct
assessments of the data does not affect the decisions made by the state.
Colorado noted that it did not have significant issues with the data
provided, and that any data issues were resolved using the consultative
process, including bringing in experts to discuss and resolve specific
issues. In addition, Colorado said, it performed routine independent
sampling and analysis of water samples, as well as occasional building
samples, which confirmed the data results provided to the state.
Accordingly, although DOE assessments might have provided another check on
data quality and adequacy, the lack of these assessments had little
adverse effect on the quality or usability of the data and does not cloud
the results or appropriateness of the site's cleanup.

DOE's Planned Cleanup The second assessment of the cleanup's
sufficiency-DOE's planned verification of the contractor's remediation of
radiologically contaminated

    Verification Not Completed

soil-was inconsistent and not completed as planned. DOE policy currently
requires radiological cleanup to be verified, but the policy is

Page 49 GAO-06-352 Nuclear Cleanup of Rocky Flats

DOE Policy Unclear on Application and Purpose of Verification

unclear about how and why verification should be done. DOE's planned
verification for the cleanup at Rocky Flats was twofold: First, DOE asked
the contractor to develop a final scanning and sampling plan, and second,
DOE asked its Oak Ridge Institute for Science and Education (ORISE) to
develop a separate verification plan that included a review of
contractorconducted scans for remaining radiological contamination.56
However, DOE chose not to complete several of the plan's objectives,
including part of ORISE's review. A DOE official said they had decided
that these activities would not provide sufficient additional information
to justify their completion, but he had no documentation to support this
decision. As a result, DOE lost the opportunity to independently verify
the sufficiency of several aspects of the cleanup.

The respondents to our community group survey had mixed views on the
degree to which the verification activities-in the aggregate-affected
their confidence in the site's cleanup. Of the 21 (of 25) who provided
responses, 13 said that the verification activities greatly or somewhat
increased their confidence in the site's cleanup. Another 8 said the
verification activities had no effect or a negative effect on their
confidence. The remaining four said they either did not know or had no
basis to judge.

Although DOE has a general policy on radiological cleanup verification,
its guidance is unclear on how the policy applies and what the
verification goals are. In January 2001, the Secretary of Energy issued a
memorandum that contained guidance on the release of radiological property
and directed DOE field offices to "establish independent verification
programs" that "should be commensurate with the potential for
contamination, as well as the complexity and hazard."57 Additional DOE
guidance on verification is found in other documents, including draft
guidance called "Control and Release of Property with Residual Radioactive
Material;" this document implements guidance for DOE Order 5400.5, which
includes requirements

56ORISE is a DOE facility operated by Oak Ridge Associated Universities, a
multiuniversity, nonprofit consortium established in 1946. Over the past 2
decades, ORISE has performed radiological surveys and environmental
assessments at sites contaminated with hazardous or radioactive materials.
ORISE staff include health physicists, environmental specialists,
radiochemists, and analytical chemists.

57Memorandum 2001-001288, January 19, 2001, Managing the Release of
Surplus and Scrap Materials.

DOE Did Not Fully Implement Its Planned Verification or Explain Its
Reasons for Not Doing So

for the release of radiologically contaminated property.58 The draft
guidance, which was approved for interim use, includes a section on
verification that states that "the DOE organization responsible for the
release of property should verify or provide for independent verification
of the radiological condition of the property before release." However,
the draft neither contains specific guidance about verification techniques
nor specifies the goals of verification. Accordingly, it was little help
to Rocky Flats officials and may have contributed to the inconsistent
implementation of the verification plan.

Lacking clear guidance, DOE's project manager at Rocky Flats said he took
a common sense approach that, in his view, fulfills the intent of DOE's
policy by cleaning the site up through the CERCLA and RCRA process. That
is, he believes that the regulatory agencies' approvals of the
radiological cleanup actions at the site constitute independent
verification. However, the official acknowledged that there is room for
interpretation and disagreement on this issue. Another area of confusion
was what the goals of the independent verification should be. At Rocky
Flats, DOE officials commissioned a cleanup verification plan that used a
different methodology than the one used to implement the cleanup. The
result was that while the cleanup verification confirmed that an area had
met the standards of the cleanup agreement, it also identified "hot spots"
that caused alarm among the public at the end of the 10-year cleanup. The
Rocky Flats manager stated that clear guidance on independent verification
would have been helpful, especially as to how it related to CERCLA
cleanups.

Although DOE agreed to ORISE's final verification plan, DOE did not fully
implement it and did not offer a public explanation for its decision.
Specifically, ORISE never completed two of its objectives-an assessment of
the aerial and ground-based scanning and an assessment of the contractor's
investigations of the results of the aerial and targeted groundbased
scans. According to a DOE official, DOE decided in mid-September 2005 that
it needed to re-evaluate the need for ORISE's work. DOE decided that
because the contractor's aerial survey had failed to find any anomalies,
ORISE's remaining work was unnecessary. DOE remained unsure about the
final outcome of the ground-based scanning but still decided not to ask
ORISE to complete the remaining objectives. A DOE official explained that

58Draft DOE G 441.1-XX, Control and Release of Property with Residual
Radioactive Material, for use with DOE 5400.5, Radiation Protection of the
Public and the Environment.

Page 51 GAO-06-352 Nuclear Cleanup of Rocky Flats

    Independent Consultants and Other Federal Agencies Conducted Additional
    Cleanup Reviews, and Some Influenced DOE's Final Cleanup Strategy

Several Recommendations from Independent Consultants Were Incorporated
into Site Remedies, but Disagreements and Concerns Remain

completing the work was not warranted, given the results of the aerial and
ground-based scanning. According to an ORISE official, DOE's Rocky Flats
Project Office notified ORISE on November 11, 2005, that it would not be
asked to complete the remaining objectives. Additional information on
DOE's independent verification is contained in appendix IV.

Reviews conducted by independent consultants and other federal agencies
commented on cleanup actions, and some reviews influenced the final
cleanup strategy, thereby providing additional assurance to DOE and the
regulatory agencies that their remedial decisions were correct.
Independent consultants, hired by local stakeholder organizations,
commented on several cleanup remedies, including surface and groundwater
cleanup actions and landfill remediation. Some of these influenced the
final cleanup; for example, a study on surface water management called for
a specific pond discharge strategy, with which DOE concurred. On the other
hand, DOE did not incorporate all comments and suggestions made by
independent contractors, such as recommendations on how to close a
landfill. Other federal agencies, including the Agency for Toxic
Substances and Disease Registry and FWS, also conducted studies, one of
which concluded that no health hazard existed for surrounding communities.
Another study is still under way.

Independent consultants, hired by local communities or through the Rocky
Flats Coalition of Local Governments, conducted technical reviews of
cleanup actions and provided recommendations, some of which DOE
incorporated into its cleanup plan. The Rocky Flats Coalition of Local
Governments sponsored these independent reviews to assure the local
governments and the public that the cleanup would meet the regulatory
guidelines. Overall, members of the Coalition initiated four technical
reviews, which examined groundwater, surface water and pond management,
and landfill remediation. The Coalition hired a consultant to comment on
DOE's independent verification process.

One instance in which DOE incorporated a recommendation from a consultant
involved surface water management. During Rocky Flats' operational days, a
series of ponds was constructed as part of a surface water management
system to ensure that no contaminated surface water left the site.59 As a
safeguard to ensure that all water leaving the site meets the state's
water quality standards, the water is tested prior to its release. An
independent consultant's study raised the concern that DOE was allowing
one of the terminal ponds to fill to a high level, thereby diminishing its
ability to store large quantities of water should a heavy rainfall occur.
Such an event could result in water being released without being tested.
The independent review recommended that DOE maintain the pond at a lower
capacity as a precautionary strategy, and DOE concurred. Another example
was DOE's concurrence with the Coalition consultant's recommendation that
ORISE include in its verification plan a 100 percent scan of certain areas
to detect any residual contamination that might exceed established cleanup
levels.

However, several of the independent consultants' reviews identified points
of contention with DOE's ultimate cleanup approach. One area of contention
involved the landfill that had been used at Rocky Flats in the 1950s and
1960s. DOE's characterization of the landfill suggested that uranium and
volatile organic compounds were present.60 DOE and the regulatory agencies
subsequently agreed to implement an accelerated cleanup action that
included buttressing the landfill to prevent the waste from slumping into
Woman Creek; regrading parts of the landfill, which is located on a
hillside, to prevent erosion; placing 2 feet of soil atop the surface to
isolate the contaminants; installing storm water management berms to
divert surface flows to perimeter channels; and seeding the entire cover,
buttress, and channels with native grass species. An independent review
disagreed with several aspects of this accelerated action, including that
it did not include a biointrusion layer; these layers inhibit the ability
of burrowing animals like mice and prairie dogs to bring contaminants up
to the surface, where people at the site might be exposed to them. DOE and
Colorado felt that a biointrusion layer was unnecessary because the
accelerated cleanup action met the legal requirements for landfill
closure, and the observed environmental conditions indicated that the
landfill posed only a minimal risk. EPA added that the issue of
biointrusion was examined during the remedy's design, with the conclusion
that the steep slope (18 percent), combined with the native grass cover,
would not be attractive habitat for burrowing animals. In the end, a
biointrusion layer was not included in the landfill cover. Monitoring of
Woman Creek and

59These ponds were not constructed as part of the CERCLA remedy.
60According to the contractor, four uranium hot spots were removed in July
2004.

Page 53 GAO-06-352 Nuclear Cleanup of Rocky Flats

Other Federal Reviews Concluded Minimal Risk or Are Still Under Way

wells will continue at locations both upgradient and downgradient of the
landfill to ensure that no contaminants are escaping from the landfill.

The Agency for Toxic Substances and Disease Registry (ATSDR), part of the
Department of Health and Human Services, conducted a public health
assessment that concluded that Rocky Flats poses no health hazard.61
Specifically, ATSDR concluded that the Rocky Flats data present a
consistent picture that local residents' current and future exposures to
contaminants from Rocky Flats are below levels associated with adverse
health effects. ATSDR officials conducted their own data reliability tests
on the site's data and concluded that the data were adequate to make
public health decisions. The officials then assessed the contaminant
pathways at Rocky Flats, including soil, air, and surface water and
groundwater and concluded that environmental contamination at Rocky Flats
posed no apparent public health hazard to surrounding communities.
(However, ATSDR did not evaluate the health implications for people within
the boundaries of Rocky Flats.) ATSDR offered several recommendations on
how to ensure that the site's contamination will not pose a future risk to
residents of surrounding communities, including continued monitoring of
surface water along the eastern boundary of the site and groundwater
wells. DOE is not required to officially respond to the ATSDR
recommendations, but reported that the recommendations had already been
included, or were under discussion with the regulatory agencies for
inclusion, in the postclosure monitoring and maintenance plans.

Additionally, as discussed previously, the 5-year reviews required by
CERCLA will continue to compile information about whether remedies at the
site continue to fully protect human health and the environment. The first
5-year review, completed in 2002,62 included areas of the site for which
final decision documents (i.e., corrective action decisions/records of
decision) had been completed, as well as areas where accelerated cleanup
actions had been completed as of September 30, 2001. At that time, final

61ATSDR was established by CERCLA in 1980-CERCLA S: 104(i); 42 U.S.C. S:
9604-and, since 1986, has been required to conduct public health
assessments of sites on CERCLA's National Priorities List- CERCLA S:
104(i)(6); 42 U.S.C. S: 9604(i)(6). In conducting a public health
assessment, ATSDR officials review environmental data and assess pathways,
or how individuals might come into contact with the contaminants, and
whether such contact would result in any health effects.

62Department of Energy, Rocky Flats Field Office, First Five-Year Review
Report for Rocky Flats Environmental Technology Site, Golden, Colorado
(Golden, Colo., July 2002).

decision documents had been completed for two areas (the 881 hillside and
the off-site areas), and accelerated cleanup actions had been completed at
several areas. Among the completed accelerated actions were the removal
and, as necessary, treatment of contaminated debris and soil at several
trenches; the installation of three groundwater treatment systems; the
removal of contaminated sludge from solar evaporation ponds; and emptying
and treating the contents of six underground storage tanks. The review
concluded that the remedies for these two areas were protective and that
the accelerated actions had addressed immediate hazards and were generally
functioning as intended. The review of the off-site areas, though not
required under EPA guidance,63 was nonetheless conducted because of the
substantial public interest in those areas. The off-site areas did not
have a defined boundary, but rather referred to off-site contamination
emanating from Rocky Flats in general, including surface contamination of
lands to the east of the site, along with the Great Western Reservoir,
Standley Lake, and Mower Reservoir. The review concluded that all
calculated excess cancer risks in the entire unit were well within or
below EPA's acceptable risk range of 1 in 10,000 to 1 in 1,000,000.

Also, FWS sampled areas at Rocky Flats that are likely to become part of
the future wildlife refuge. Prior to any transfer of land management
authority to FWS, FWS typically surveys the property to identify any
potential hazardous substances that pose a threat to fish and wildlife. In
May 2006, consistent with its survey plan, FWS took 45 soil samples
(including four duplicate samples), mostly along proposed trails as
indicated in the Comprehensive Conservation Plan for the refuge. The
samples will be examined for a range of potential contaminants, including
metals, radionuclides, organics, and polychlorinated biphenyls, commonly
referred to as PCBs. FWS took an additional 12 vegetation samples
(including one duplicate sample), principally to ensure that any future
prescribed burns will be safe. An FWS official said that the results of
the laboratory analyses of the samples are due in early July, at an
estimated cost of $70,000. If the results identify contaminants that
concern them, FWS officials will notify DOE, EPA, and Colorado; the
agencies will then determine what steps would need to be taken.

63EPA's 2001 Comprehensive Five-Year Review Guidance requires 5-year
reviews of remedial actions resulting in any hazardous substances,
pollutants, or contaminants remaining above levels that allow for
unlimited use and unrestricted exposure. The contaminants remaining in the
off-site areas were at low enough levels to allow unlimited use and
unrestricted exposure.

Page 55 GAO-06-352 Nuclear Cleanup of Rocky Flats

  DOE Implemented Some Lessons Learned at Rocky Flats but Has No Requirement to
  Ensure That Lessons Learned Are Implemented at Other Sites

Although DOE has identified and implemented at other cleanup sites some
lessons learned at Rocky Flats, DOE does not require applicable lessons
learned at one cleanup site to be implemented at others. The Rocky Flats
project offers many lessons about innovative techniques, risk- and
costsharing contract provisions, accelerated cleanup processes,
involvement of community groups, oversight of contractor controls over
data quality, and cleanup verification processes. However, DOE has not
developed a system to ensure that all lessons are captured and implemented
as appropriate at other DOE sites. As a result, DOE may be missing
valuable time- and costsaving opportunities at other sites that are
planning or undergoing cleanup.

    DOE Identified Lessons from Rocky Flats and Assessed Their Applicability to
    Other Sites

DOE has gathered and disseminated to some other DOE sites numerous lessons
learned at Rocky Flats. These lessons included the following:

     o Clearly define government oversight of the contractor, and limit the
       number of DOE personnel providing direction.
     o Conduct external reviews of the project baselines to build credibility
       and provide objective recommendations for project improvement.
     o Use employee incentives to reward high-performing individuals.
     o Use a flexible project management approach that allows the contractor
       to complete the project in the safest and most cost-effective manner.
     o Establish a clear "end state" vision and risk-based cleanup defined in
       conjunction with specific future land/site use.
     o Develop and use an integrated project baseline schedule and budget.
     o Use government-furnished services and items to integrate and manage
       the delivery of items not within the contractor's control.
     o Implement new technology that significantly accelerates the schedule
       and reduces total costs, such as techniques for reducing the number of
       radioactive waste shipments off site for disposal.

In October 2002, DOE authorized a corporate review team to determine the
effectiveness of DOE cleanup efforts. In 2003, the team reviewed work

    DOE Lacks a System for Ensuring Continued Collection and Implementation of
    Lessons Learned at Its Cleanup Sites

activities, management processes, and contract administration practices at
selected sites, and used a checklist of more than 50 lessons learned at
Rocky Flats to assess their applicability and potential benefits to each
site. The review team issued nine reports with findings and
recommendations. With the exception of one report, which was part of the
preliminary review effort, each report also included the team's
determination of whether the lessons on the checklist applied to the
sites, and the progress the sites had made toward implementing applicable
lessons learned.

During 2004 and 2005, DOE's Office of Performance Assessment conducted
follow-up reviews of many of the sites the corporate review team had
reviewed earlier. These follow-up reviews, however, did not assess whether
the sites had implemented the lessons learned from Rocky Flats. According
to the DOE official responsible for tracking the status of these reviews,
DOE does not require sites to implement applicable lessons from Rocky
Flats (or from other cleanup sites). Instead, he said, each site is
responsible for tracking its implementation of these lessons. Although he
said he believes the sites are taking steps to implement those lessons and
are continuing to improve their systems for managing cleanup, he had no
documentation to support his conclusions.

DOE's Office of Performance Assessment reviewed lessons learned at Rocky
Flats in the March 2004 and June 2005 Rocky Flats baseline performance
review reports. In the 2004 report, DOE noted that valuable information
and processes from Rocky Flats should be available to other sites that are
beginning the cleanup process. The report noted that DOE's Rocky Flats
office had begun a program to archive project cost information that could
be used to prepare cost estimates for future contracts at other sites, and
recommended that the office develop a database that identifies the number
of personnel and time durations required to perform specific cleanup
tasks. According to the Rocky Flats DOE manager, the costarchiving program
continued to provide periodic information updates, and provided the final
download of project data to DOE's Office of Engineering and Construction
Management in December 2005. The 2005 Rocky Flats baseline performance
review report included 30 additional lessons learned at Rocky Flats. Among
these were improved contract language that established a close working
relationship between DOE and the contractor, made measurement of progress
easier, and reduced the need for contract changes; improved safety
processes at the site; consolidated procurement functions; and streamlined
methods for handling and shipping radioactive waste and disposing of
excess property. One of the lessons identified in the 2005 report actually
resulted from a request from another cleanup site. A DOE official in the
Office of Performance Assessment told us that officials at DOE's Office of
Environmental Management visited Rocky Flats in 2005 to document lessons
learned from the demolition of a plutoniumprocessing building. They
visited Rocky Flats at the request of DOE officials at the Savannah River
site who were designing a similar facility and sought information from the
Rocky Flats demolition that might improve the design of their building.

Although DOE has a database of environmental safety and health lessons, it
does not have a database of broader lessons learned across the DOE
complex. DOE has a nationwide database managed by the Society for
Effective Lessons Learned Sharing, a volunteer DOE organization dedicated
to identifying, sharing, and using lessons learned in order to improve the
safety, efficiency, and effectiveness of DOE work processes. These lessons
focus primarily on individual safety incidents and how to prevent their
recurrence. However, the database generally does not capture lessons
dealing with broader issues such as contract management, pricing, and
working with regulatory agencies. Consequently, DOE may not be maximizing
the use of valuable information that could save time and money.

DOE's Office of Engineering and Construction Management commissioned a
study in the fall of 2003 to correlate components of project performance
with project success, and identify best practices to improve DOE project
performance. The July 2004 report on the study's findings identified many
factors that influenced the success of DOE projects, including the
importance of an effective lessons learned program. The report did not
examine the effectiveness of DOE's lessons learned programs, but it noted
examples at DOE where lessons were effectively transferred from one
project to another. It concluded that

"DOE Headquarters has a responsibility to assure that such lessons are
being transferred across sites with similar facilities. Sharing lessons
learned needs aggressive attention; without it valuable savings are lost
and frustrations compounded. Lessons learned are useful for mitigating
risk and providing training material for project directors/managers.64

64Civil Engineering Research Foundation, Independent Research Assessment
of Project Management Factors Affecting Department of Energy Project
Success (Reston, Va., July 12, 2004).

Page 58 GAO-06-352 Nuclear Cleanup of Rocky Flats

The report also recommended that DOE proactively encourage the sharing of
lessons learned. According to a DOE headquarters official, a national
database of lessons learned from closure sites would allow DOE to
proactively share experiences that would benefit future closure
operations.

Rocky Flats Offers During our review, we gathered additional lessons
learned at Rocky Flats that could be useful for other DOE sites. These
lessons include the

    Additional Lessons That May Be Applicable to Other following:

DOE Sites  o  Safety is a priority. According to DOE officials, this
lesson was reinforced throughout Rocky Flats' cleanup. Early in the
cleanup, DOE officials recognized that a significant investment in hazard
identification, safety planning, and safety implementation (i.e., the
integrated safety management system) during the actual work would ensure
that the work was performed without unacceptable risks or unnecessary
delays. Later, DOE officials said they came to understand that this focus
on safety not only helped work progress, but also facilitated efficiency
by building trust and engaging the workforce. Safety was both consistent
with, and essential to, effective project execution.

     o Performance-based contract incentives improve results. According to
       DOE officials, the first contract they had with Kaiser-Hill
       demonstrated that attaching incentives to clearly defined performance
       measures vastly improved actual results. The 2000 contract took the
       concept to the next level, providing large incentives to the
       contractor and the workers to safely and compliantly complete the
       cleanup within the target schedule and cost. The additional incentives
       for schedule and cost savings resulted in closure more than a year
       ahead of schedule and $530 million under budget. However, as
       previously noted, such incentives are not always this successful, and
       the financial incentive offered at Rocky Flats was only one of many
       factors contributing to the cleanup's early completion.
          * Take a consultative approach to cleanup decisions. As previously
            discussed, the collaborative process was essential to the
            cleanup's early completion. According to DOE officials, the
            cleanup was successful because all of the stakeholders were
            engaged in the process and supportive of the ultimate goal. The
            input of numerous key figures, including members of Congress,
            senior DOE managers, state and local elected officials, and
            officials of federal and state regulatory agencies,
          * was actively solicited and ultimately met with the convergence of
            the cleanup agreement, the contract, and the desired end state.
            DOE officials said they communicated openly and often with
            stakeholders to seek the best solutions, and they came to value
            input from formerly dogmatic opponents. Moreover, Colorado, EPA,
            DOE, and the contractor worked closely together on cleanup
            decisions from the beginning through the end. As cleanup
            proceeded on an area, Colorado had access to meetings and
            detailed information about the cleanup. In the event that
            Colorado or EPA considered a particular activity to be unsafe,
            the Rocky Flats cleanup agreement granted them the authority to
            stop work.
     o Don't let unresolved issues delay progress. DOE officials said that if
       they had focused on what they could not do or delayed work until all
       questions were answered, the project would not have been completed and
       the target completion date would probably still be in question.
       Colorado officials emphasized, however, that this does not mean that
       work progressed without the appropriate approvals from Colorado and
       EPA. It is Colorado's understanding that the consultative process,
       with all of its oversight and meetings to discuss activities and
       issues, ensured that all issues were addressed and resolved in a
       timely manner before activities occurred.
     o Obtain stable project funding. As discussed earlier, consistent
       project funding was a key factor in the cleanup's early completion.
       According to DOE and EPA officials, congressional and departmental
       commitment to stable funding over the life of the project aided
       tremendously in project planning and execution. Also, the stable
       funding helped in regulatory interactions and credibility with the
       general public.
          * Involve the future site manager in remedial decisions. According
            to FWS, when a former DOE site is considered for conversion to a
            wildlife refuge, it is "vastly preferable" for FWS to be involved
            early in the cleanup process and have a decision-making role. DOE
            officials agreed and said they have actively involved FWS since
            passage of the Refuge Act. According to DOE officials, FWS
            substantially influenced cleanup decisions, even in areas of the
            site that will not become part of the future refuge. For example,
            they said, FWS's input effected a major change in the cover
            design for the present landfill. Additionally, FWS influenced the
            composition of the revegetation seed mixes sitewide and the
            methods of protecting sensitive habitat and endangered species.
            But
          * DOE did not incorporate all of FWS's suggestions, such as
            characterizing and disposing of waste in the original landfill,
            incorporating barriers around landfills to discourage intrusion
            by burrowing animals, and using irrigation and soil amendment to
            enhance the success of seeded native vegetation. According to DOE
            officials, the two agencies' different perspectives on cleanup
            methods highlight the difficulty of coordination between two
            federal agencies that have different missions. These difficulties
            can persist even when staff of the two agencies are collocated
            and interact daily, as was the case at Rocky Flats. In
            retrospect, DOE officials said, it would have been better to
            recognize that difficulty early and obtain an executive-level
            consensus on the vision for the outcome.
     o Be aware of potential beryllium contamination. According to a senior
       contractor official, beryllium was "ubiquitous" at the site; it was
       found even in containers of metals and oxides. In terms of worker
       safety, beryllium was more of a predemolition challenge than other
       contaminants because there is no effective way to monitor beryllium on
       a real-time basis.
     o Question accepted technologies. The contractor was able to save
       substantial money and time at Rocky Flats by questioning the usual
       disposal and remediation processes. For example, the accepted practice
       was that a glovebox or any part of one was, by definition, transuranic
       waste. However, the contractor found a way to decontaminate gloveboxes
       and dispose of them much more easily and less expensively as low-level
       waste. According to a senior contractor official, this particular
       technical lesson learned at Rocky Flats may be applicable to DOE's
       cleanup work at the Hanford site.
     o Contain contaminated water. According to an official with the Colorado
       Department of Public Health and Environment, lessons learned at Rocky
       Flats included the importance of ensuring that there are no
       underground drainage systems or conduits when large amounts of water
       are used during a cleanup. This lesson stemmed from an unfortunate
       experience in the cleanup of building 771, when water used in
       decontamination efforts seeped into underground conduits that had not
       been adequately plugged. The water flushed contaminants through the
       conduits, resulting in elevated levels of americium in one series of
       ponds at the site. This turned out to be an expensive lesson, as the
       contractor had to bring in equipment and treat about 26 million
       gallons of americium-contaminated water.

                                  Conclusions

Although DOE captured and implemented at other sites some of the lessons
it learned at Rocky Flats, others risk being lost. For example, contractor
and DOE officials said that at Rocky Flats, and at other cleanup sites,
many lessons that could be gleaned from records and data will be lost if
not recorded, summarized, or otherwise captured and shared. According to
the Rocky Flats manager, an effort termed the "Legacy Project" was begun
in 2001 to start collecting the project knowledge and lessons before the
institutional memory was lost. This effort continued intermittently
through the summer of 2005, drawing upon record documents, working papers,
and personal experience. Also, DOE Rocky Flats officials said that
officials of DOE's Legacy Management office in Grand Junction, Colorado,
had shared with them lessons learned from the Grand Junction office's
experience working with the long-term management of sites formerly
contaminated by uranium mill tailings. According to the Rocky Flats
manager, these lessons will be brought to Rocky Flats as Legacy Management
executes its responsibilities for site management and maintenance. As the
mission scope of Legacy Management expands, it will continue to carry
experience from Rocky Flats to other sites that are transferred to it for
long-term care. Also, although DOE officials told us in March 2005 that
they were planning a workshop on lessons learned at Rocky Flats in
decontamination and demolition, that workshop has not been held. According
to the Rocky Flats manager, several technical assist visits occurred with
contractor and DOE personnel from sites in Idaho, Ohio, and Washington.
These visits were believed to be more focused and efficient than a
workshop format. At this time, a general workshop is not planned; however,
technical assist and assessment visits to facilitate the sharing of
lessons are continuing.

DOE officials at Rocky Flats have drafted the Rocky Flats Closure Legacy
report, a lengthy document that captures the 4-year effort of the Legacy
Project. It is currently at DOE headquarters for review and comment and is
expected to be released in June 2006.

Strong DOE oversight of data quality is important because accurate and
complete data are paramount to DOE's, EPA's, and Colorado's decisions
about the sufficiency of the Rocky Flats cleanup. At Rocky Flats, DOE
placed undue reliance on the contractor to have appropriate data quality
controls, and did not complete the required management and independent
assessments of the data's quality. Our review showed that the contractor
at Rocky Flats appeared to have comprehensive data quality controls, but
this does not negate DOE's responsibility to independently monitor and
assess those controls throughout the cleanup. Also, although DOE policies
call for independent verification of cleanup results, the applicability of
the policies to cleanups conducted under CERCLA or RCRA is unclear.
Lacking clear and specific guidance on cleanup verification, DOE officials
at Rocky Flats undertook what they thought was a reasonable approach.
However, DOE did not complete all of the cleanup verification activities
it had planned; it also did not publicly explain its rationale for not
doing so. As a result, DOE lost a valuable opportunity to increase public
awareness of, and confidence in, the verification results. Ironically,
although the verification activities were undertaken to increase public
confidence in the cleanup, the results sparked additional questions from
the public. Finally, although DOE has implemented at other cleanup sites
some of the lessons learned at Rocky Flats, DOE does not require that
lessons learned at one site be implemented, where applicable, at other
sites. As a result, DOE stands to lose the benefits that such lessons have
to offer.

To improve DOE's oversight of cleanup activities, its conduct of cleanup

  Recommendations for

verification activities, and its monitoring of lessons learned at DOE
cleanup sites, we recommend that the Secretary of Energy take the
following three actions:

     o Ensure that DOE independently assesses contractors' controls over data
       quality.
     o Clarify guidance on whether and how to conduct cleanup verification
       activities.
     o Assess the costs and benefits of developing a method to track the
       lessons learned from cleanup activities at DOE sites across the
       nation, including methods for determining whether lessons are being
       applied at applicable locations.

  Agency Comments and Our Evaluation

We provided a draft of this report to DOE, Interior, EPA, Colorado, and
Kaiser-Hill for their review and comment. DOE, Interior, Colorado, and
Kaiser-Hill provided written comments (see apps. V, VI, VII, and VIII,
respectively). EPA did not provide official written comments but did
provide editorial and technical suggestions, as did the other agencies,
that we incorporated, as appropriate.

DOE commented that it found the report to be comprehensive, generally
thorough, clear, and well structured. DOE agreed that a robust and
effective lessons learned program would be beneficial and said it intends
to follow through and revitalize such a program within DOE's Office of
Environmental Management. DOE reiterated the importance of teamwork and
agency support in accelerating the cleanup of the Rocky Flats site, noting
that the regulatory agencies, community groups, and local government
organizations worked tirelessly along with DOE and contractor
organizations to overcome obstacles. DOE also emphasized that the high
priority given the Rocky Flats cleanup within the DOE complex and at the
congressional level, together with the provision of level annual funding,
was critical to the success of the project.

DOE did not agree or disagree with our recommendations that it ensure
independent assessments of data quality and clarify its guidance on
cleanup verification. In its specific comments, however, DOE noted that
there is direction and guidance on independent verification and other
independent assessments, and referenced several documents that provide
such guidance. We were aware of these documents and had discussed several
of them with DOE officials at the Rocky Flats Project Office, who said the
documents were not helpful in guiding their decisions about whether to
conduct verification activities or how to ensure compatibility between
available verification strategies and the cleanup strategies undertaken at
the site.

In its specific comments, DOE also stated that there is an additional
regulatory step DOE must take before releasing the site to the Department
of the Interior, pursuant to the requirements of DOE Order 5400.5
governing the release of real property with residual radioactive material.
DOE suggested that we reference this regulatory requirement in the report
section dealing with additional steps to be taken. We did not adopt this
suggestion because our report's discussion of regulatory steps focused on
those that remain to be taken by EPA and Colorado, the regulatory agencies
at the site. We do, however, reference DOE Order 5400.5 and its draft
implementing guidance in our discussion of cleanup verification
activities. We noted that the draft implementing guidance neither contains
specific guidance about verification techniques nor specifies the goals of
verification and, consequently, was little help to Rocky Flats officials
and may have contributed to the inconsistent implementation of the
verification plan.

The Department of the Interior generally agreed with the information that
pertains to the Fish and Wildlife Service. The department said that at
this point, it is unable to determine when Rocky Flats will be open to the
public, due to the uncertainty of the completion of the transfer of the
site. We incorporated the department's suggestion that we focus on the
transfer of administrative jurisdiction from DOE to the department rather
than focusing on when the public may access the Rocky Flats National
Wildlife Refuge. Also, as the department suggested, we clarified the
discussions of actions EPA must take to facilitate the transfer and the
acquisition of privately owned mineral rights.

Colorado commented that the report provides an appropriate recognition of
issues and actions occurring during the remedial activities at Rocky
Flats. Colorado emphasized that it has been an integral participant in
Rocky Flats' cleanup and closure; it provided continual in-depth
regulatory oversight of the investigative and remedial activities for both
building decontamination and demolition and environmental restoration.
Colorado cited the dedicated efforts of its staff, along with the
consultative process, as expediting completion of an accelerated cleanup
that resulted in significant cost savings for DOE and for U.S. taxpayers.

Kaiser-Hill commented that the report was thorough, well-written, and
accurate in its description of Kaiser-Hill's role in the cleanup of Rocky
Flats. Kaiser-Hill agreed with the report's focus on the contribution of
Kaiser-Hill's workforce, and stated that the incentive contract played an
important role in the success of the cleanup project. Kaiser-Hill noted
that the success of the project was also due to the cooperative
decision-making process that evolved among DOE, Kaiser-Hill, EPA,
Colorado, and local communities, aided by a firm political commitment to
accelerate the cleanup.

Kaiser-Hill said it believes that the report's calculation of the total
cost of the cleanup is misleading because it includes tangential costs
such as health and pension benefits, which alone add about $1 billion to
the cost reported. Kaiser-Hill noted that a significant portion of these
benefit costs were accrued by previous contractors. Kaiser-Hill
acknowledged, however, that such costs represent financial obligations to
the taxpayer. For our purposes, inclusion of these costs was necessary to
report total costs of the cleanup since 1995, including long-term costs.

Additionally, Kaiser-Hill commented that because the final disposition of
trench T-7 (the fourth individual cleanup project we reviewed in depth)
required only minimal cleanup, it was completed through the standard "no
further accelerated action" justification process. According to the
contractor, the justification documentation for such areas (those
requiring minimal cleanup) should include information on data quality and
adequacy, but not at the same level of detail as was required for other
cleanup areas. We recognize that areas closed out through the "no further
accelerated action" process may not necessitate the same level of data
quality documentation as other areas. Nevertheless, the Rocky Flats
Cleanup Agreement requires that the justification documentation for "no
further accelerated action" areas include information on data quality and
usability. The documentation we reviewed for trench T-7 did not include
such information.

We are sending copies of this report to the Secretaries of Energy and the
Interior, the Administrator of the Environmental Protection Agency, the
Executive Director of the Colorado Department of Public Health and
Environment, the President and Chief Executive Officer of Kaiser-Hill
Company, interested congressional committees, and other interested
parties. We will also make copies available to others upon request. In
addition, the report will be available at no charge on the GAO Web site at

http://www.gao.gov.

If you or your staff have any questions about this report, please contact
me at (202) 512-3841 or [email protected]. Contact points for our Offices of
Congressional Relations and Public Affairs may be found on the last page
of this report. GAO staff who made major contributions to this report are
listed in appendix IX.

Gene Aloise

Director, Natural Resources and Environment

Appendix I

                       Objectives, Scope, and Methodology

Our review objectives were to determine the (1) factors that contributed
to the early completion of the physical cleanup at Rocky Flats; (2) work
remaining as well as total project costs, including long-term costs; (3)
measures in place to assess whether the cleanup achieved a level of
protection of public health and environment consistent with the Rocky
Flats Cleanup Agreement; and (4) lessons the Rocky Flats project may hold
for other Department of Energy (DOE) cleanup projects.

In conducting our work, we visited the Rocky Flats site and reviewed
documents and data prepared by DOE, the Environmental Protection Agency
(EPA), the Colorado Department of Public Health and Environment
(Colorado), the Department of the Interior's Fish and Wildlife Service
(FWS), the contractor, and various scientific organizations. We also
interviewed officials of these entities.

To determine the factors that contributed to the physical cleanup's early
completion, we interviewed DOE, EPA, Colorado, and contractor officials
and reviewed agency and contractor documents on cleanup accomplishments
and techniques, project activity reports, and decontamination and
demolition accomplishments. Information on innovative cleanup techniques
is presented in appendix III.

To determine the work remaining as well as total project costs, including
long-term costs, we reviewed documents and data prepared by DOE, EPA,
Colorado, FWS, and the contractor, and interviewed officials of these
entities. We reviewed documents and discussed issues pertaining to the
cleanup requirements of the Comprehensive Environmental Response,
Compensation, and Liability Act of 1980, as amended (CERCLA), and the
Resource Conservation and Recovery Act of 1976, as amended (RCRA). We also
reviewed documents and discussed issues pertaining to the plans for the
Rocky Flats National Wildlife Refuge, cost records and estimates, and
pension plan provisions. We also analyzed documents related to the
cost-plus-incentive-fee contract, including contract negotiation
documents, the contract cost and fee structure, and contract
modifications, but we did not evaluate the cost-effectiveness of the
contract.

To determine the measures in place to assess whether the cleanup would
achieve a level of protection of public health and environment consistent
with the Rocky Flats Cleanup Agreement, we reviewed documents and
discussed issues pertaining to cleanup verification strategies and
results. We also reviewed scientific analyses, such as a study of how
contaminants migrate through soil and water and an assessment of the
public health risk

Appendix I Objectives, Scope, and Methodology

posed by contaminants remaining at the site. We reviewed the general
content of these analyses but did not evaluate the science underlying
them.

In addition, because decisions about the sufficiency of the cleanup have
been and will be based on remediation data, we assessed the soundness of
the agencies' and contractor's processes and procedures for ensuring data
quality. We visited the subcontractor that performed data verification and
validation activities, and we reviewed the results of audits of the key
laboratories used by the contractor to analyze samples of radioactively
contaminated soil.

We also selected a nonprobability sample of four accelerated cleanup
actions, based on several criteria, including whether EPA or Colorado was
the lead regulatory agency and whether the contamination was in the
surface soil or under contaminated buildings.1 Because radionuclide
contamination was the primary concern at the site, we selected cleanup
actions of plutonium-contaminated soil. For each selected cleanup action,
we reviewed data quality controls pertaining to data quality objectives,
data quality parameters, and data verification and validation. We did not
evaluate the remedial data or analyses themselves. Information on our data
quality review results and details of DOE's verification activities are
presented in appendix IV.

To determine the lessons the Rocky Flats project may hold for other DOE
cleanup projects, we reviewed information and interviewed officials of
DOE, EPA, FWS, Colorado, and the contractor about lessons learned at Rocky
Flats and other sites. We reviewed documents related to and reports from
DOE's Society for Effective Lessons Learned Sharing database, and we
interviewed DOE officials involved in various efforts to capture and
disseminate lessons learned (from Rocky Flats and elsewhere throughout the
DOE complex).

Further, to obtain citizen views about issues such as community input to
the cleanup and the effect of various cleanup verification activities, we
attended monthly meetings of the two local community groups that served in
an advisory capacity to DOE: the Rocky Flats Citizens Advisory Board and
the Rocky Flats Coalition of Local Governments. We also surveyed

1Results from nonprobability samples cannot be used to make inferences
about a population, because in a nonprobability sample, some elements of
the population being studied have no chance or an unknown chance of being
selected as part of the sample.

Page 68 GAO-06-352 Nuclear Cleanup of Rocky Flats Appendix I Objectives,
Scope, and Methodology

current and past members (and current staff) of these two groups to obtain
their opinions on questions such as how DOE engaged the groups in the
cleanup process, how DOE used the groups' input, how satisfied or
dissatisfied they were with the cleanup's results, and the degree to which
DOE's cleanup verification activities affected their confidence in the
cleanup. We pretested the survey twice and revised the questions
accordingly. We documented the responses and verified 100 percent of the
documentation. Response rates and summary responses are presented in
appendix II.

We conducted our work in accordance with generally accepted government
auditing standards from March 2005 through May 2006.

Appendix II

GAO Survey Methodology, Response Rates, and Summary Responses

                               Survey Methodology

To obtain citizen and local government views on the Rocky Flats cleanup,
we surveyed current (as of December 2005) and past members and current
staff of the two local community groups that served in an advisory
capacity to DOE: the Rocky Flats Citizens Advisory Board and the Rocky
Flats Coalition of Local Governments.1 The survey included questions about
how DOE engaged the groups in the cleanup process, how DOE used the
groups' input, what level of confidence group members had in the cleanup's
results, the degree to which DOE's cleanup verification activities
affected their confidence in the cleanup, and any remaining concerns they
might have.

We designed the survey questions to elicit clear and unbiased responses.
We pretested the survey twice and revised questions accordingly. We
transferred the survey responses into a spreadsheet and the comments into
a narrative document; we then verified that 100 percent of the information
was transferred accurately from the individual surveys.

In total, we sent surveys to 58 members (current and past).2 We obtained
group members' e-mail addresses (or mailing addresses, when e-mail
addresses were not available) from the directors of the Rocky Flats
Citizens Advisory Board (Advisory Board) and the Rocky Flats Coalition of
Local Governments (Coalition). After e-mailing the survey, we e-mailed two
reminder notices to encourage members to respond. Additionally, we
attended both groups' January 2006 meetings and personally encouraged
members to submit their surveys.

We received a total of 25 responses, or 43 percent of the total surveys
sent.

  Survey Response Rates

By subgroup, however (e.g., current members of one group, past members of
one group), response rates varied widely. For example, the response rate
for current members of the Advisory Board was 71 percent, whereas the

1The Rocky Flats Citizens Advisory Board and the Rocky Flats Coalition of
Local Governments began phasing out their activities in early 2006; as of
March 2006, the Rocky Flats Stewardship Council took over as the local
oversight group.

2Our references to group members include the groups' staff (two per
group), whom we included in our survey because of their experience and
knowledge about the Rocky Flats cleanup.

Page 70 GAO-06-352 Nuclear Cleanup of Rocky Flats

Appendix II GAO Survey Methodology, Response Rates, and Summary Responses

response rate for current members of the Coalition was 35 percent. Table 1
shows the response rates by subgroup and in total.

                  Table 1: Survey Response Rates, by Subgroup

                                   Number of          Number of Response rate 
Subgroup                   surveys sent   responses received     (percent) 
Current Advisory Board                 14                 10            71 
Current Coalition                      23                  8            35 
Subtotal, current members              37                 18            49 
Past Advisory Board                    14                  6            43 
Past Coalition                          7                  1            14 
Subtotal, past members                 21                  7            33 
Total                                  58                 25            42 

                                  Source: GAO.

Following is a summary of the survey responses, by question, along with

  Summary Responses

selected illustrative comments by respondents. Narrative responses are not
provided in their entirety because we did not wish to introduce a
perceptual bias. For some questions, that is, a majority of respondents
offered positive and concise comments, but one or two respondents provided
negative and lengthy comments. To present all of the comments in their
entirety would thus give an unfair perception of emphasis, at least by
sheer volume of narrative, to the negative minority and would diminish
(again by volume) the positive majority. Accordingly, we summarized the
comments to reflect the preponderance of responses, whether positive or
negative.

The summary responses begin with question 5, as questions 1 through 4
sought information about whether the respondent was a current or former
member of either the Advisory Board or the Coalition and about the length
of time the respondent had served as a member of the group.

Appendix II GAO Survey Methodology, Response Rates, and Summary Responses

    5. How satisfied or dissatisfied are you with how DOE engaged the [group-the
    Board or the Coalition] in the cleanup process?

Very satisfied 8 Somewhat satisfied 13 Neither satisfied nor dissatisfied
1 Somewhat dissatisfied 2 Very dissatisfied 1 Don't know/no basis to judge
0

    6. Please explain or provide examples of how DOE engaged the [group] in the
    cleanup process.

Comments from 6 of the 21 respondents who reported being "very satisfied"
or "somewhat satisfied" (in response to question 5) included the
following:

     o "DOE actively engaged the Coalition and local communities in a variety
       of ways. . . [DOE] provided opportunities for the Coalition to
       participate in certain decision-making processes, provided [a] regular
       series of technical meetings keeping Coalition staff informed on
       various remediation activities, [and] . . . basically interacted with
       the Coalition on an almost daily basis during the course of the site's
       cleanup."
     o DOE's engagement "varied from one manager to another but was generally
       good. More important, the CDPHE [i.e., Colorado], EPA and Kaiser-Hill
       consistently solicited the advice of the Board. In particular, the
       Board was given access to draft documents and decisions while they
       [were] still being discussed. This permitted the Board to actively
       participate in the decision process and influence the decision instead
       of simply responding to a done deal."
     o The "most important three items [were] board makeup, board
       decisionmaking process, and board meeting time. The members include a
       couple of ex-nuclear submarine personnel, a couple of college
       professors, a groundwater expert, a mining expert, peace activists,
       etc.; in addition, the attendance and participation of DOE,
       regulators, cleanup contractors, FWS, and the general public provides
       technical, emotional, and general interest review on items. Board
       decisions require that all agree on all formal Board decisions, which
       forces constructive give and

Appendix II GAO Survey Methodology, Response Rates, and Summary Responses

take in the decision making process. Meeting in the evening not only
permits the general public to attend, it also permits regulators, DOE,
contractors, FWS, etc., etc., to all attend and discuss items-[it] may be
the only time they all get together and discuss some of the items. DOE
attends all [Board] meetings, presents data on requested topics, answers
questions, and provides follow-up data as requested. We may not always
agree but at least we get the data out there and discuss the issues.
Operating as [a] federal board is very important."

     o "I believe that the DOE did a good job in answering reasonable
       requests from Board members. [DOE] did not always jump through every
       hoop presented nor should they have. Some requests were unreasonable;
       some would have been impossible, e.g., cleaning up soil to levels that
       technology couldn't accomplish."
     o DOE "seemed to listen to the Board and take their recommendations
       seriously, when it fit their plan. . . ."
     o "For the most part, DOE provided copies of documents for review and
       comment and provided presentations on cleanup and closure issues and
       decisions. But, the closer to completion the site became, the harder
       it was to get documents in a timely fashion. As an example, we were
       usually provided documents at the same time as the regulators, but at
       the end we received them much later than the regulators. Legacy
       Management [LM] completely ignored the city's input to LM's Public
       Involvement Plan."

    7. How satisfied or dissatisfied are you with how the [group's] input was
    used by DOE?

Very satisfied 3 Somewhat satisfied 16 Neither satisfied nor dissatisfied
2 Somewhat dissatisfied 2 Very dissatisfied 2 Don't know/no basis to judge
0

Appendix II GAO Survey Methodology, Response Rates, and Summary Responses

    8. Please explain or provide examples of how the [group's] input was used by
    DOE.

Comments from 5 of the 16 respondents who reported being "somewhat
satisfied" (in response to question 7) included the following:

     o "The RSAL [radionuclide soil action level] issue was one on which the
       DOE listened to the [Advisory Board] and other public groups."
     o "In some cases DOE added information to documents or adjusted
       processes based on the Board's input."
     o "The Board's concerns about the original soil action levels were met
       with DOE's agreement to fund the independent review."
     o "Our input was used on a number of fronts-pond management, ensuring
       [that] the site would not be a disposal site, and [the] strategy for
       remediating the original landfill, to name [a] few. Most important is
       [DOE's] revising the RFCA [Rocky Flats Cleanup Agreement] to better
       reflect community priorities. The reason I checked `somewhat
       satisfied' instead of `very satisfied' is that we could never get DOE
       to incorporate stewardship planning into the cleanup process in a
       substantive manner. For DOE and the regulators, stewardship was an
       afterthought that got bucked to the end of the project and into
       regulatory closure space. We are now at the end of the project and we
       are still left wondering how DOE will address key questions concerning
       institutional controls."
     o "In the Independent Verification process, the Coalition pushed, and
       the DOE made some changes to their original proposal."

9. Overall, what kind of effect, if any, do you believe the [group's]
input had on the cleanup process?

Very positive effect 6 Somewhat positive effect 17 Neither negative nor
positive effect 1 Somewhat negative effect 0 Very negative effect 0 Don't
know/no basis to judge 1

Appendix II GAO Survey Methodology, Response Rates, and Summary Responses

10. Please explain or provide examples of how the [group's] input affected
the cleanup process.

Comments from 7 of the 23 respondents who believed that their group's
input had a "very positive" or "somewhat positive" effect (in response to
question 9) included the following:

     o "Opposition of the [Board] and other entities to the 651 pCi/g RSAL
       for plutonium adopted for [Rocky Flats] in the 1996 [cleanup
       agreement] resulted in DOE funding a citizen oversight panel to hire
       independent scientific specialists to calculate radionuclide soil
       action levels (RSALs) for the site. Out of this study came the
       recommendation that the plutonium RSAL be reduced by 95% from 651
       pCi/g to 35 pCi/g. While DOE did not accept this recommendation, it
       did work with the regulators to reduce the action level for
       [plutonium] in surface soil (top 3 feet) from 651 to 50 pCi/g. DOE and
       the regulators, however, pushed a tradeoff, in that in exchange for
       this better surface cleanup (top 3 feet) the public would have to
       accept having larger quantities of plutonium left in the subsurface
       environment-between 1000 and 7000 pCi/g at a depth of 3 to 6 feet
       below the surface, with no limit on the concentration allowed below 6
       feet."
     o "The Board has had an impact on all areas of the cleanup process . . .
       from how the landfill caps were constructed, to the overall level of
       cleanup, to how buildings were safely removed. . . ."
     o "I understand that in the beginning the Board had a tremendous impact
       on the cleanup. Because of the Board and community members, the
       surface soil cleanup levels were changed to be more conservative, from
       651 pCi/g to 50 pCi/g. I feel [that] the Board had less of an
       influence later in the process. However, because of the Board, DOE was
       aware that the community was watching, and I believe the community got
       a better cleanup because of that."
     o "DOE solicited a large amount of input from the Coalition over the
       course of the site's cleanup. Although not all of the Coalition's
       input was incorporated into cleanup plans, a substantial amount of
       Coalition positions were incorporated into the cleanup. As a result,
       local community buy-in into the process was enhanced. DOE,
       Kaiser-Hill, the regulators, and the Coalition were successful
       partners in the site's closure."

Appendix II GAO Survey Methodology, Response Rates, and Summary Responses

     o "We were able to influence the level of cleanup. We were also able to
       get more cleanup in areas like the 903 pad in exchange for leaving
       building foundations in place-771 & 371."
     o "By the Coalition engaging in depth on matters of technical
       feasibility, schedule, cost, worker safety, D&D [decontamination and
       demolition] design and implementation, contaminant control and
       monitoring, it forced the DOE, regulators and Kaiser-Hill to focus and
       impart progress and planning to the Coalition board and public."
     o "The Coalition affected the cleanup on a number of levels. We were the
       best supporters and the most effective critic. We helped maintain
       congressional support for the project and held DOE, Kaiser-Hill and
       the regulators accountable. We were central to determining the future
       use of the site and to revising the cleanup priorities to better match
       the needs of the community. We brought DOE to the table in a public
       forum which, along principally with the [Citizens Advisory Board], was
       central to ensuring the dialogue remained open. We pushed for a free
       flow of information."

11. Please explain or provide examples of what, if anything, DOE might
have done differently to change the effect of the [group's] input on the
cleanup process.

Respondents offered a number of suggestions and comments, including the
following:

     o "The only change would be on the future monitoring of the site i.e.,
       LSO [the Local Stakeholder Organization] which has been completely
       controlled by politics and not by DOE and the board. . . ."
     o "I always believed that if DOE had approached it more as a partnership
       with their end customer the cleanup process would have been done
       better."
     o "[DOE] did continue with a very active public participation process to
       address [the soil action levels] and all aspects of the cleanup.
       However, rather than doing this process via the [Board, which was]
       established in 1993 to advise DOE on the cleanup, DOE created a new
       body, the RFCA [Rocky Flats Cleanup Agreement] Focus Group . . . The
       value of the Focus Group was that it allowed more intense discussion
       of details than was usually possible in a [Board] meeting. But
       shifting the principal

Appendix II GAO Survey Methodology, Response Rates, and Summary Responses

discussion of the cleanup to the Focus Group . . . had the effect of
undermining and marginalizing the [Board]."

     o "DOE could have engaged the Board better in terms of budgeting. We
       were never asked to partner with the site in requesting funds for our
       participation through the federal budget process. . . . We also will
       not be funded for what we saw as a valuable project to develop risk
       communication tools to help alleviate concerns about the residual
       contamination at the site. We have heard from citizens who live near
       the site that they are concerned that contamination has been left
       behind. There is no way that these citizens are going to read and
       understand the immense multi-volume Remedial Investigation study and
       Comprehensive Risk Assessment. If there are any further water quality
       exceedences, even if they are not life threatening, they are going to
       call into question the protectiveness of the cleanup. DOE had better
       hope this doesn't happen, because they appear to have no plan or the
       tools to address concerns that might be raised by the less
       knowledgeable in the community."
     o "DOE at times declined to allow members of the Board to observe
       internal meetings between DOE, [Kaiser-Hill], and the regulators. This
       made it more difficult for the Board to thoroughly understand the
       technical details of a cleanup decision and respond in an educated
       way."
     o "One recurring Coalition theme was the lack of long-term stewardship
       considerations in remedial action planning and documents. Instead of
       incorporating the long-term stewardship considerations into the
       documents, it was left to post-physical closure documents. . . . The
       Coalition position has always been that long-term stewardship planning
       should be an integral part of the remedial activities and not
       relegated to post-closure status."
     o "As the clean up was underway in early 2005, DOE seemed to agree that
       ORISE should conduct a [Multi-Agency Radiation Survey and Site
       Investigation Manual, or MARSSIM]-based final clean up verification.
       Since this approach was apparently not specified in the contract,
       Kaiser-Hill balked and DOE was left to fund an over-flight survey by a
       low flying helicopter to, for PR [public relations] purposes, attempt
       to pick up (detect) major hot spots. The overflight technology was not
       sufficiently sensitive to detect exceedences of the clean up level. In
       addition due to soil moisture and shielding in the industrial area,
       the overflight detection approach was of little credible value. By
       omitting this final

Appendix II GAO Survey Methodology, Response Rates, and Summary Responses

verification Kaiser-Hill was able to `complete' physical closure early and
collect an added 100 million dollar award."

We now have a few questions about the results of the cleanup. As you know,
since 1995, when it was awarded the Rocky Flats cleanup contract,
Kaiser-Hill has decontaminated and demolished hundreds of structures,
installed groundwater treatment systems, and removed contaminated soil,
among other cleanup tasks specified by the contract and the Rocky Flats
Cleanup Agreement.

12. How confident, if at all, are you that the cleanup will be protective
of public health and the environment?

Extremely confident 5 Very confident 6 Somewhat confident 12 Not very
confident 1 Not at all confident 1 Don't know/no basis to judge 0

13. Please explain or provide examples that illustrate your view.

Comments from 4 of the 11 respondents who were "extremely confident" or
"very confident" (in response to question 12) in the protective nature of
the cleanup included the following:

     o "I believe that phenomenal effort was very effective in cleaning up
       those sites that needed remediation, e.g., the 903 Pad and uranium
       trenches."
     o "I am personally confident that the government will retain possession
       of the land and keep up the controls now in place so that visitors are
       safe and protected. I am convinced we got the best cleanup possible
       for the conditions, political climate, and money spent."
     o "This process has been open and has benefited from intense public
       oversight, talented and committed regulators (especially at CDPHE
       [Colorado Department of Public Health and Environment]), and
       independent reviews of key assumptions and decisions such as the
       Actinide Migration Panel and the soil action level review. Further,
       the

Appendix II GAO Survey Methodology, Response Rates, and Summary Responses

surface water quality standards are extremely stringent and with rare

exception they are getting met. That increases my confidence."

        * "The main reason I am extremely confident that the cleanup will be
          protective of public health and the environment is because of the
          tremendous amount of excellent oversight work provided by the
          regulators (CDPHE and EPA). I can't say enough of how impressed I
          am with the technical staff of the regulatory agencies. They were
          everywhere during the course of the cleanup, ensuring that the
          remedial actions performed at the site were properly planned,
          implemented, and executed."
        * Comments from 4 of the 12 respondents who were "somewhat confident"
          (in response to question 12) in the protective nature of the
          cleanup included the following:
     o "We don't know everything that is underground (e.g., landfills) and
       might seep out-cleanup is probably the most cost-effective that could
       be done . . ."
     o "I think most of the contamination has been removed. However, I am
       concerned about isolated hot spots."
     o "We were very disappointed that DOE changed course on the verification
       of cleanup. We were led to believe that the verification would be a
       MARSSIM based approach; what we really received was a process that had
       never been used for verification purposes (the aerial flyover); an
       ORISE review that was constrained by DOE and a limited scope review."
          * "I believe the majority of the buffer area will be very low risk
            to the public. The ponds, the industrial area, the 903 pad, 903
            lip and 903 wind blown area could expose the unsuspecting public
            to a higher risk depending on individual sensitivities even
            though the residual Pu [plutonium] level is at or below 50
            pCi/g."
          * One of the 2 respondents who were "not very" or "not at all"
            confident (in response to question 12) about the cleanup's
            protectiveness said,
     o "I don't believe that the cleanup was done to be protective of public
       health and the environment. Otherwise, DOE would not have pursued the
       wildlife preserve aspect. Basically, the wildlife preserve is an

Appendix II GAO Survey Methodology, Response Rates, and Summary Responses

"administrative" control to limit the future land use, limit public

exposure to the existing contamination at the site, and allowed DOE to

leave behind a lot of contamination."

14. What, if anything, might have been done differently during the cleanup
that would have increased your confidence?

Respondents offered a number of suggestions and comments, including the
following:

     o "If the original landfill and the deep basements had been removed, my
       confidence would be greatly increased. I do understand the reasons for
       this not occurring."
     o "A 100 percent scan of the DOE-retained lands would give me a complete
       idea of what is out there. . . ."
     o "The implementation of our consultants' recommendations and a "TRUE"
       MARSSIM based final survey."
     o "Only after the Focus Group had been meeting for a full year was it
       finally made clear that the better cleanup sought by some participants
       was not in the cards because decisions placing a cap on what could be
       spent had already been made. To some of us involved in this enormously
       time consuming process, this announcement revealed that the public
       participation process was in large measure a sham. Some DOE personnel
       insist that DOE had been open about the fiscal cap, but the DOE
       official who made the announcement at the Focus Group later said that
       its effect on the meeting was `like throwing a dead rat on the
       table.'"
     o "I really do not have any concern about the site per se. I am
       concerned that the legal folderol . . . will continue forever."
     o "Really we are relying on the best science known and so I am not sure
       that anything could have been done better."
     o "The verification could have been performed earlier so that it could
       have been used as a tool in the cleanup."
     o "With unlimited funds a total cleanup could have been accomplished- in
       our real world we have better than could have been expected. FWS now
       needs the resources to convert the site."

Appendix II GAO Survey Methodology, Response Rates, and Summary Responses

o  "DOE and the regulators should have required independent [MARSSIM]
based protocols to verify complete and comprehensive surface soil clean up
levels site wide. This should have been done by an independent
contractor."

15. At this point in time, do you have any remaining concerns about the
cleanup?

Yes 17

No 8

16. Please provide examples or details of your remaining concerns.

Respondents offered a number of concerns, including the following:

     o "The underground "stuff"-landfills and basements. Who knows what, when
       and how it might affect us in the future."
     o "I am not certain how we will maintain control of [the] area for the
       centuries needed given that we did not know for certain that all
       contaminants were removed."
     o "DOE needs to continue water quality monitoring on Woman Creek
       permanently!"
     o "My remaining concerns with the cleanup have to do with how well the
       revegetation and erosion control measures function in the next few
       years. The site needs several good growing seasons to help
       re-establish vegetation covering the remediated portions of the
       DOE-retained land. Without adequate vegetative cover, erosion problems
       can arise which could ultimately result in the mobilization of
       actinides from the surface soil into surface water."
     o "My concerns now lie in the importance of creating a robust
       stewardship plan and assuring that monitoring will continue well into
       the future as well as knowing that if we find that a remedy has failed
       the DOE will do what it takes to address this failure."
     o "I worry about the magnitude of the long-time stewardship
       responsibilities. There are numerous examples at other sites of where
       controls fail because governments forget that there are institutional

Appendix II GAO Survey Methodology, Response Rates, and Summary Responses

controls which carry with them use restrictions. How we will protect Rocky
Flats over the long-term remains somewhat of a mystery for me."

     o "What will be done in the coming years, and by whom, to keep the
       public out of the contaminated areas?"
     o "Plutonium and uranium take hundreds of thousands of years to decay.
       Even in 100 years, the site can change sufficiently so that subsurface
       contamination is exposed. Also, physical controls may well decay into
       dust by the time the next century rolls around and institutional
       controls could be forgotten. In the future, the site could be used for
       subsistence farming. Who knows what will happen? The uncertainty of
       the future contributes the largest concern I have about the cleanup."

In the summer of 2005, DOE arranged for a number of cleanup verification
activities, including Kaiser-Hill's ground-based scans along the
boundaries of previously remediated areas where contamination was once
known to exist; Bechtel Nevada's3 aerial scan of the site; and the Oak
Ridge Institute of Science and Education's (ORISE) sampling and
ground-based scanning of the 903 pad and lip area. We would like your
views on the overall verification process, as well as your views on each
of these three specific verification activities.

17. How, if at all, did the overall verification process affect your
confidence in the site's cleanup?

Greatly increased my confidence 3 Somewhat increased my confidence 10
Neither increased nor diminished my confidence 5 Somewhat diminished my
confidence 2 Greatly diminished my confidence 1 Don't know/no basis to
judge 4

3Bechtel Nevada conducted the aerial survey.

Appendix II GAO Survey Methodology, Response Rates, and Summary Responses

18. Please explain or provide examples that illustrate how the overall
verification process affected your confidence in the site's cleanup.

Comments from 5 of the 13 respondents who believed that the overall
verification process "greatly increased" or "somewhat increased" their
confidence (in response to question 17) included the following:

     o "Although I felt there was a low probability of contamination being in
       the buffer zone, given the level of suspicion and the recent
       publication of several false and misleading books and articles I felt
       it was important. This is a good time to state that although almost
       all the public fear involved radiologic contamination, it is not the
       most hazardous."
     o "Additional testing always adds to the vote of confidence."
     o "The aerial scan was the most effective. The other parts of the
       verification were too confusing for the majority of the community to
       understand. When the hot spots were discovered in the 903 Pad Lip
       Area, those of us more familiar with statistical sampling methodology
       could understand that it was not that grave of a situation, but those
       who read newspapers I am sure were not comforted."
     o "ORISE, although paid by DOE, is an independent [entity]. As such, I
       trust their results."
          * "The targeted independent verification areas helped to assure my
            confidence level but did not greatly increase it. I do think the
            verification process was important to other board members and
            most of all to the general public. It probably did very little or
            nothing to assure already skeptical critics of the site cleanup."
          * Comments from 2 of the 5 respondents who said the activities
            "neither increased nor diminished" their confidence (in response
            to question 17) included the following:
     o "I feel what was done was a waste of time and money and did not tell
       us anything new."
     o "Was glad it was in place, but didn't give me 100% confidence."

Appendix II GAO Survey Methodology, Response Rates, and Summary Responses

And of the 3 respondents who said the activities "somewhat diminished" or
"greatly diminished" their confidence in the cleanup (in response to
question 17), 1 said,

o  "It is by no means clear that the full site has ever been well
characterized. Former workers are known to say that unknown areas of
random dumping were never found by the characterizing methods used,
especially the . . . method that took a very limited number of samples in
relatively large areas and estimated the contamination in that area based
on this very limited sampling. Further, . . . the verification methods
used could not detect all hot spots or areas of unknown subsurface
contamination."

19. How, if at all, did Kaiser-Hill's targeted ground-based scanning
activities (using a high-purity germanium, or HPGe, detector mounted on a
tripod) affect your confidence in the site's cleanup?

Greatly increased my confidence 4 Somewhat increased my confidence 7
Neither increased nor diminished my confidence 7 Somewhat diminished my
confidence 3 Greatly diminished my confidence 0 Don't know/no basis to
judge 4

20. Please explain or provide examples that illustrate how Kaiser-Hill's
targeted ground-based scanning activities affected your confidence in the
site's cleanup.

Comments from 4 of the 11 respondents who said that Kaiser-Hill's targeted
ground-based scanning activities "greatly increased" or "somewhat
increased" their confidence in the cleanup (in response to question 19)
included the following:

     o "It showed that there was basically no contamination above the cleanup
       levels left on the site."
     o "I felt that the areas surveyed had a low risk of contamination. While
       it only somewhat increased my personal confidence I feel it was
       critical in increasing the public confidence."

Appendix II GAO Survey Methodology, Response Rates, and Summary Responses

     o "They targeted areas of concern and proved through testing that
       cleanup levels had been achieved."
          * "Showed that the area `covered' by the scan met cleanup
            criteria."
          * Comments from 2 of the 7 respondents who said the activities
            "neither increased nor diminished" their confidence (in response
            to question 19) included the following:
     o "There wasn't enough sampling conducted, and the sampling should have
       been conducted in more areas after remediation was done and before
       backfill took place."
          * "Kaiser-Hill's scanning activities were part of the cleanup, not
            part of an independent review. So, I have confidence in what they
            did, but the question suggests that their scans would increase my
            confidence in their work. Independent review would have increased
            my confidence."
          * And of the 3 respondents who said the activities "somewhat
            diminished" their confidence in the cleanup (in response to
            question 19), 1 said,
     o "DOE and [the contractor] have continued to approach the cleanup using
       methods designed to not find problems. An approach designed to verify
       that no problem exists cannot be used to discover problems."

21. How, if at all, did Bechtel Nevada's aerial scan of the site (using a
helicopter-mounted array of sodium iodide detectors) affect your
confidence in the site's cleanup?

Greatly increased my confidence 3 Somewhat increased my confidence 6
Neither increased nor diminished my confidence 8 Somewhat diminished my
confidence 3 Greatly diminished my confidence 1 Don't know/no basis to
judge 4

Appendix II GAO Survey Methodology, Response Rates, and Summary Responses

22. Please explain or provide examples that illustrate how Bechtel
Nevada's aerial scan of the site affected your confidence in the site's
cleanup.

Comments from 3 of the 9 respondents who said that Bechtel Nevada's aerial
scan of the site "greatly increased" or "somewhat increased" their
confidence in the cleanup (in response to question 21) included the
following:

     o "The [aerial scan] addressed the entire site, not just where one
       anticipated that contamination could exist."
     o "The aerial scan just confirmed what DOE had been telling us all
       along."
          * "Detection capabilities of aircraft scanners was not as sensitive
            as I would have liked. System was designed to detect higher
            concentrations of radionuclides than were expected at Rocky
            Flats."
          * Comments from 2 of the 8 respondents who said the activities
            "neither increased nor diminished" their confidence (in response
            to question 21) included the following:
     o "While the technology, if properly applied, might have been
       sufficient, the results were less than reliable because of ground
       moisture, shielding in the industrial area by equipment, debris, etc.,
       and over-lot grading, which would have buried contamination. If
       properly applied, the aerial scan could detect major hot spots but is
       not designed to detect small hot spots that exceed the cleanup level.
       The aerial scan was more a public relations ploy than a good science
       valid verification of site safety and reliable cleanup."
          * [The aerial scan was] "a method that is suited for determining
            what has happened in a Three Mile Island [or] Chernobyl type
            event, not to verify [that] an action did what it was supposed to
            do."
          * And of the 4 respondents who said the activities "somewhat
            diminished" or "greatly diminished" their confidence in the
            cleanup (in response to question 21), 1 said,
     o "The [remedial soil action levels] established for Rocky Flats set
       standards for plutonium and other materials that may remain in place
       according to depth. The aerial scan can only detect what is present on

Appendix II GAO Survey Methodology, Response Rates, and Summary Responses

the surface (not surface defined as the top three feet, but surface as
limited to a small fraction of this amount). Further, the aerial scan
detects hot spots only if they are quite large in extent and relatively
high in concentration. The aerial scan is a very valuable tool but also a
very limited one."

23. How, if at all, did ORISE's verification activities (e.g., soil
sampling and surface scans using sodium iodide "FIDLER" scintillation
detectors) affect your confidence in the site's cleanup?

Greatly increased my confidence 5 Somewhat increased my confidence 11
Neither increased nor diminished my confidence 2 Somewhat diminished my
confidence 2 Greatly diminished my confidence 0 Don't know/no basis to
judge 5

24. Please explain or provide examples that illustrate how ORISE's
verification activities affected your confidence in the site's cleanup.

Comments from 6 of the 16 respondents who said that ORISE's verification
activities "greatly increased" or "somewhat increased" their confidence in
the cleanup (in response to question 23) included the following:

     o "Unlike the two [other verification activities], the ORISE
       verification was essential in verifying that [Kaiser-Hill] did in fact
       perform the work they were being paid to do."
     o [ORISE] "did find some hot spots that [Kaiser-Hill] then addressed, so
       a second check did help to increase belief in [the] cleanup."
     o "This is the verification activity that turned up the `hot spots'
       which were eventually remediated. I understand that according to the
       risk assessment formulas that were developed and the averaging nature
       of these formulas these `hot spots' technically did not have to be
       remediated. The decision to remediate these, however, was important to
       help assure public confidence in the site cleanup."
     o "The ORISE work ended up raising more questions and concerns by the
       public. Also, we are left with the impression at the end that their
       work is

Appendix II GAO Survey Methodology, Response Rates, and Summary Responses

incomplete because DOE did not provide them with funding to do the final
review of the aerial survey. Also, trying to make MARSSIM fit into what
was needed to verify the cleanup at the site was just too difficult to
follow. . . . The ORISE work started out with great expectations, but
things just got too confusing at the end. For the average citizen, I don't
think the verification made any difference at all."

     o "The ORISE budget was very limited-about $250,000-and was not
       sufficient to do an adequate job of site verification because the
       protocol they had to follow was disjointed."
          * "ORISE'S work showed that for the area they reviewed, the
            statistical approach Kaiser-Hill used was valid. However, given
            that DOE limited ORISE's work it is hard to extrapolate their
            findings over a larger area of the site. So, their work was of a
            limited value."
          * Of the 2 respondents who said that the activities "neither
            increased nor diminished" their confidence (in response to
            question 23), 1 said:
          * "The review became less and less than promised as time went by:
            from a MARSSIM-based approach, to an approach greatly restrained
            by DOE as to what would be reviewed and how the review would be
            conducted."
          * And of the 2 respondents who said that the activities "somewhat
            diminished" their confidence in the cleanup (in response to
            question 23), 1 said,
     o "I believe the FIDLER device is a very good way to detect
       contamination in the surface soil. But its use by ORISE again was
       limited to areas of known contamination rather than to characterize
       portions of the site that have not been adequately characterized."

    25. Do you have any additional comments you would like to provide?

Among the respondents' additional comments were the following:

o  "[The Board] provides the only forum that has provided for the public,
regulators, DOE, and contractor to discuss the issues and for everyone to
have an equal voice. Hope there is someplace in the future (for a couple
of years) for such activities to continue. The proposed LSO [longterm
stewardship organization] appears to be a waste of government

Appendix II GAO Survey Methodology, Response Rates, and Summary Responses

funds-too large of a staff, restricted membership, and too little
technical expertise. . . ."

     o "I believe the lessons learned at Rocky Flats could be put to good use
       at other sites, especially Hanford."
     o "In the future on any similar rad[iologically] contaminated sites, DOE
       should require as a contractual condition that an independent
       verification of the clean up be implemented using [MARSSIM]
       principles. In the 903 pad area where a [MARSSIM] sampling
       verification was implemented-hot spots substantially exceeding 50
       pCi/g were found and picked up. It would seem reasonable to conclude
       that other areas of the 903 pad and the industrial area would have
       similar exceedences."
     o "DOE, other state agencies, and communities should study what we did
       at Rocky Flats because we were on the cutting edge of how to work
       within the regulations but allow great latitude in engaging the
       community and for remediating a complex nuclear site."
     o "Standley Lake Cities still have serious concerns about Woman Creek's
       risk from [the original landfill and the fact that] Woman Creek has no
       terminal testing control pond!"
     o "ORISE should have been allowed to survey more of the area and to do
       some surveying of the Buffer Zone."

Appendix III

Additional Information on Decontamination and Demolition Activities at Rocky
Flats

The contractor used a number of innovative techniques to accomplish
decontamination and demolition tasks at Rocky Flats. One of these, as
discussed in the report, was the use of a cerium nitrate solution to
decontaminate large pieces of equipment. Another innovative
decontaminating technique was the contractor's use of InstaCoteTM to
encapsulate large surface-contaminated tanks and other equipment.
InstaCote is a polyurea plastic coating that was sprayed on equipment. The
InstaCote not only sealed in the contamination, but also served as the
shipping conveyance and disposal package. According to DOE officials, use
of the InstaCote saved "easily thousands of hours" and increased worker
safety, as it eliminated the difficult and dangerous job of cutting
contaminated equipment and packaging it for shipment. The InstaCote could
be used even with extremely large pieces of contaminated equipment, such
as a super-compacter used to reduce waste volume by crushing drums
containing radioactive waste and drums that were empty but had formerly
contained radioactive waste. Figure 14 shows the supercompacter before
InstaCote was applied and with the InstaCote awaiting transport to a
disposal facility.

Source: DOE.

One of the most difficult decontaminating challenges DOE and the
contractor overcame was decontaminating concrete. When concrete became
contaminated with liquid radioactive waste (e.g., by spills during nuclear
material processing or waste disposal), the past practice was to fix

Appendix III Additional Information on Decontamination and Demolition
Activities at Rocky Flats

the contamination in place using sometimes several layers of an epoxy
paint. Before a building containing contaminated concrete could be
demolished, this paint had to be removed down to the base material, and
the base material had to be decontaminated. The contractor removed the
contaminated paint using various methods such as scabbling, shaving, or
hydolasing. Scabbling chipped away the first layer of the surface with a
pneumatic hammer; however, this option created dust and increased the
potential for the release of contaminants into the air. Shaving removed
layers of concrete with a machine, and was considered more efficient and
less hazardous than hydrolasing for decontaminating concrete surfaces with
deeper contamination. Hydrolasing, the preferred method for removing paint
and shallow surface contamination, involved blasting away paint and the
initial layer of the surface material with water, then selfcontaining the
water and resulting debris. Figure 15 shows the hydrolase system in use in
building 886 at Rocky Flats, a former nuclear laboratory known to have
light contamination under painted surfaces.

Appendix III Additional Information on Decontamination and Demolition
Activities at Rocky Flats

Source: DOE.

But in some cases, according to a senior contractor official, contaminated
liquid had seeped so deeply into the concrete that it could not be removed
using any of these three methods. This was the case, for example, in the
"infinity room" (so called because the level of radioactive contamination
in the room was greater than instruments could measure) of building 771,
which had often been referred to as "the most dangerous building in
America." When concrete was contaminated to the core, the only option

Appendix III Additional Information on Decontamination and Demolition
Activities at Rocky Flats

was to cut it out and remove it. The entire "infinity room" floor had to
be cut out and shipped as transuranic waste, as shown in figure 16.

Source: DOE.

In other cases, when the contaminated concrete was far underground (i.e.,
in building foundations), it was decontaminated to the lowest levels
reasonably achievable and left in place. DOE and the regulatory agencies
agreed to decontaminate and leave in place the foundations of buildings
371 and 771/774, and cover them with clean fill dirt. According to DOE and
contractor officials, leaving the foundations in place poses little risk
to human health and safety because the foundations and the soil around
them were decontaminated to specified levels before being covered.1 Also,
they

1Specifically, contamination on the surface of the foundations' concrete
had to be remediated to less than 100 nanocuries per gram, and
contamination by volume (measured through calculations of the foundations'
depth) had to be remediated to less than 7 nanocuries per gram. If
contaminated sections of the concrete could not be remediated to these
levels, those sections had to be removed.

Page 93 GAO-06-352 Nuclear Cleanup of Rocky Flats Appendix III Additional
Information on Decontamination and Demolition Activities at Rocky Flats

said, leaving the foundations in place was safer for the workers at the
site than removing them, which would have required workers wearing
respiratory protection systems that constrain movement and visibility to
winch up massive, multiton pieces of concrete from a confined space. A
Colorado official said he had agreed to the decision to leave the
foundations in place after carefully considering the balance between the
potential future risk of exposure with the real and immediate risk of
worker safety. In lieu of removal, the contractor treated the foundations
with a fixative to prevent contaminants from migrating during demolition
activities, then covered them with a cushion of sand and a layer of clean
fill dirt. According to DOE and contractor officials, plutonium should not
migrate out of the foundations, and if it does, it should not migrate
beyond the surrounding soil. This contention is supported by the results
of the actinide migration study, which concluded that fixed actinide
contamination does not readily move in the environment. Also, neither
humans nor wildlife would be exposed unless a future major excavation
activity were to take place. However slight it may be, the potential risk
to human health and safety posed by leaving the foundations in place is
the reason that controls such as groundwater monitoring wells and
treatment systems will remain in place, according to Colorado officials.

Another innovative decontaminating and demolition strategy the contractor
developed in conjunction with DOE and the regulatory agencies involved
dividing the site into areas and completing most or all of the work in one
area-from building decontamination to demolition-so that environmental
remediation could begin in that area while decontamination and demolition
work was being done in the next area. This strategy enabled the cleanup to
proceed more quickly by allowing the earlier start of environmental
restoration work and by allowing workers more time to address any
unforeseen circumstances. The strategy also achieved economies of scale,
as site services-such as water, steam, and power- could be eliminated
throughout an entire area, rather than on a building-by-building basis.

Appendix IV

Additional Information on the Quality of Remediation Data at Rocky Flats and on
DOE's Planned and Actual Verification

  Information on Controls over Data Quality

This appendix presents additional information on our review of the
processes and procedures in place at Rocky Flats to evaluate the quality
of remediation data. It also presents additional information on DOE's
planned and actual verification of the contractor's cleanup work.

A critical aspect of the cleanup process was ensuring the validity of the
data used to assess whether the site had been remediated to the agreedupon
levels. Accordingly, we reviewed the processes and procedures in place to
ensure data quality, and we identified four key controls: (1) establishing
data quality objectives, (2) using data quality parameters, (3) verifying
and validating data, and (4) auditing laboratories that analyzed samples
for Rocky Flats. Our review showed that, overall, the contractor generally
followed these data quality controls and clearly documented its compliance
with them for the four accelerated cleanup actions we reviewed in depth.
For these four cleanup actions-at building 771, the 903 pad, the 903 lip
area, and trench T-7-we assessed the closeout reports and the data
supporting them to determine the extent to which the data collection and
laboratory analyses adhered to data quality standards and procedures. We
did not, however, evaluate the remedial data or laboratory analyses
themselves.

To assess the use of the first key control-establishing data quality
objectives-we identified the criteria documented in the Industrial Area
and Buffer Zone Sampling and Analysis Plan. We then reviewed the cleanup
project sampling plan addenda, remedial action plans, and cleanup project
closeout reports for each of the four projects to determine whether the
data quality objectives had been established and were considered during
the projects' cleanup. To assess the use of the second control, we
reviewed the four projects' closeout reports to verify that the data had
been checked against the data quality parameters in accordance with the
criteria located in both the Rocky Flats Implementation Guidance Document
and the Industrial Area and Buffer Zone Sampling and Analysis Plan. These
documents establish the guidelines for evaluating analytical data and
address the overall quality of the data quality control. We examined the
closeout reports for each of the four projects to ensure that the projects
met the criteria for the third control-data verification and validation.
And finally, to assess the use of the fourth control, we reviewed reports
on audits of laboratories that analyzed data for the site's cleanup.
Specifically, we reviewed source documentation from audits conducted by
DOE's consolidated audit program and DOE's mixed analyte performance
evaluation program. We concentrated our review on laboratory audit

Appendix IV Additional Information on the Quality of Remediation Data at
Rocky Flats and on DOE's Planned and Actual Verification

results relevant to detecting plutonium for the 3-year period during which
cleanup was under way for the four projects.

Establishing data quality objectives. DOE and the contractor established
data quality objectives to act as planning tools for collecting data and
for making decisions. The data quality objectives process is specified in
the site's quality assurance project plan.1 The process is intended to
provide a systematic procedure for defining criteria for data collection,
including when and where to collect samples, what level of decision error
is tolerable, and how many samples to collect. In our review of the four
selected cleanup areas, we found general adherence to the data quality
objectives process. For three of the four projects we selected, we
verified that both the characterization and confirmation sampling were
planned according to the data quality objective process as defined in the
Industrial Area and Buffer Zone Sampling and Analysis Plan. Because the
fourth project required only minimal cleanup, it was completed through the
standard "no further accelerated action" justification process. According
to the contractor, the justification documentation for such areas (those
requiring minimal cleanup) should include information on data adequacy,
but not at the same level of detail as was required for other cleanup
areas. We recognize that areas closed out through the "no further
accelerated action" process may not necessitate the same level of data
quality documentation as other areas. Nevertheless, the Rocky Flats
Cleanup Agreement requires that the justification documentation for "no
further accelerated action" areas include information on data quality and
usability. The documentation we reviewed for trench T-7 did not include
such information.

Applying PARCC parameters. EPA's policy requires it to apply data quality
parameters for precision, accuracy, representativeness, completeness, and
comparability-known as the PARCC parameters. Because radionuclides were
involved, an additional data quality parameter-sensitivity-was used at
Rocky Flats. These parameters provided information to the contractor and
the regulatory agencies about the acceptability or utility of the data.
For example, precision measures whether using the same analytical
technique will result in the same results for a particular sample (i.e.,
analyzing the same sample twice and

1The Rocky Flats Quality Assurance Project Plan is included in the
Industrial Area and Buffer Zone Sampling and Analysis Plan.

Page 96 GAO-06-352 Nuclear Cleanup of Rocky Flats Appendix IV Additional
Information on the Quality of Remediation Data at Rocky Flats and on DOE's
Planned and Actual Verification

comparing the results).2 According to the sampling and analysis plan,
assessments of the data according to the PARCC parameters are documented
in the closeout reports for the accelerated cleanup actions and, according
to EPA officials, are checked through the verification and validation
process discussed next. Our review of the four selected cleanup areas
indicated that the PARCC parameters were consistently applied. For three
of the four projects we selected, we verified that the data were reviewed
in accordance with each of the data quality parameters. The fourth project
required only minimal cleanup and was completed through the standard "no
further accelerated action" justification process for which no formal data
quality discussion was required.

Verifying and validating laboratory data. Data verification was required
for 100 percent of the laboratory data at Rocky Flats. Data validation was
required for 25 percent of the data, but because this requirement
pertained to the Rocky Flats site as a whole, the percentage of data
validated could vary considerably from one cleanup area to another. Data
verification is a review of a laboratory's electronic data package summary
to evaluate the extent to which the laboratory met specified quality
control and reporting criteria. Data validation is a more in-depth review
that includes not only verifying the data, but also examining the raw data
and manually verifying calculations done by the laboratory. After a
laboratory analyzed samples, it sent the results to a subcontractor that
reviewed the data according to either the verification or the validation
criteria and assigned codes indicating the data's quality. For example,
one code would be assigned to data for which no problems were observed,
whereas another code would be assigned to data that did not meet the
quality control requirements.

Our review of the closeout reports on the four cleanup projects we
selected indicated general adherence to the verification and validation
requirements. Specifically, for three of the four projects, at least 98
percent of the overall data had been verified, and between 23 percent and
50 percent had been validated. The contractor provided several
explanations for why 100 percent of the data had not been verified. For
example, at the 903 lip area, because a number of samples were collected
close to the time that the closeout report was submitted, verification and
validation occurred after the report was completed. Additionally, the
large number of

2At Rocky Flats, the contractor was required to take duplicate samples for
5 percent of all field samples.

Page 97 GAO-06-352 Nuclear Cleanup of Rocky Flats Appendix IV Additional
Information on the Quality of Remediation Data at Rocky Flats and on DOE's
Planned and Actual Verification

samples collected in September 2004 (from the 903 lip area and other areas
around the site) overloaded the verification and validation staff. Most of
these records were verified or validated in late November or early
December 2004, and all had been verified or validated by early 2005,
according to the contractor.

Auditing laboratories. The 32 laboratories that analyzed samples collected
from Rocky Flats were required to undergo annual technical audits to
ensure the accuracy of their analytical results. Since 2000, contractor
officials said they have largely satisfied the audit requirement by
voluntarily participating in DOE's consolidated audit program,3 with the
exception of one on-site laboratory that the contractor audited. A typical
DOE laboratory audit, according to the audit program manager, includes a
3-day visit to the laboratory by a team of five to six people. DOE
certifies laboratory auditors in particular areas, including chemistry
(e.g., radiochemistry or organic chemistry), and uses a standardized
checklist to verify that laboratory practices meet DOE standards. When
auditors identify problems, they can provide comments or issue priority I
or priority II findings. Priority I findings are the most severe; they
represent a serious breakdown in management controls that could render the
laboratory unacceptable for use or unfit to perform services for DOE.
Laboratories that receive priority I and II findings are required to
develop and implement corrective action plans, and the contractor's policy
stipulated that laboratories receiving priority I findings could do no
more work for Rocky Flats until they had rectified the problem. Of the 118
audits conducted of laboratories analyzing Rocky Flats cleanup samples,
only 10 priority I findings were issued, according to the subcontractor
official responsible for laboratory audits at Rocky Flats.

We reviewed the performance of the two primary laboratories that analyzed
confirmatory samples for plutonium 239 for the four selected cleanup areas
and found no areas of concern with the laboratories' quality. To limit our
review, we focused on two performance criteria: (1) laboratory audit
findings from DOE's consolidated audit program documents on alpha
spectroscopy, which is used to analyze confirmatory samples for plutonium;
and (2) results from DOE's mixed analyte performance evaluation program,
which assesses a laboratory's analytical abilities by

3DOE created the consolidated audit program in response to mid-1990s audit
reports that cited inefficiency, redundancy, and ineffectiveness regarding
DOE audits of analytical laboratories.

Page 98 GAO-06-352 Nuclear Cleanup of Rocky Flats

 Appendix IV Additional Information on the Quality of Remediation Data at Rocky
               Flats and on DOE's Planned and Actual Verification

sending the laboratory a sample of a known quantity of a contaminant and
comparing the laboratory's results with the actual quantity. Our review
showed that the audit findings for alpha spectroscopy resulted in one
audit observation.4 Our review of the second performance criteria of the
tests of the two laboratories' analytic ability found that, from 2002
through 2004, the two laboratories analyzed a combined total of 16 samples
of plutonium 239, or alpha samples, and received acceptable test results
in all but 3. Of those 3 samples, 2 samples were considered acceptable but
were off by 20 percent to 30 percent, and 1 sample was not acceptable.

DOE revised two of the planned verification activities at Rocky Flats.
First, due to time and cost constraints, DOE reduced the scope of the
contractor's planned scanning and sampling activities. Second, DOE revised
the Oak Ridge Institute for Science and Education's (ORISE) planned
assessment of the contractor's scans of the 903 lip area. The revision
incorporated an approach consistent with the Multi-Agency Radiation Survey
and Site Investigation Manual (MARSSIM). The actual verification
activities conducted by both the contractor and ORISE identified "hot
spots" of radionuclide-contaminated soil; these hot spots were
subsequently remediated.

  Additional Information on DOE's Planned and Actual Verification Activities

    DOE Reduced Size and Scope of Scanning and Sampling Portion of Planned
    Verification

The first part of DOE's cleanup verification effort-the
contractorconducted scanning and sampling5-was reduced in scope due to
time and cost constraints. In September 2004, the Rocky Flats site manager
requested that the contractor develop a sitewide surface radiological
survey using the Multi-Agency Radiation Survey and Site Investigation
Manual (MARSSIM). MARSSIM is the outcome of a joint effort by DOE, EPA,
the Department of Defense, and the Nuclear Regulatory Commission to
develop uniform guidance for conducting final surveys to demonstrate
compliance with specified radiological cleanup levels. It recommends 100
percent scanning of areas most likely to be contaminated with
radionuclides.

4DOE was unable to provide source documentation for 1 year for one of the
two laboratories.

5Our use of the term "scanning" refers to both stationary in situ
measurements taken via high-purity germanium detectors and mobile scans
taken via sodium iodide scintillation detectors.

Page 99 GAO-06-352 Nuclear Cleanup of Rocky Flats Appendix IV Additional
Information on the Quality of Remediation Data at Rocky Flats and on DOE's
Planned and Actual Verification

Scanning can be conducted through use of devices known as high-purity
germanium (HPGe) detectors, as shown in figure 17, or sodium iodide
scintillation detectors, known as FIDLERs. These devices can scan entire
areas to determine whether certain radionuclides are present in a given
area.

Source: DOE.

This scanning approach had been used several times by ORISE at Rocky Flats
to demonstrate the sufficiency of radiological cleanup of buildings prior
to their demolition. However, according to the contractor, DOE had limited
experience using this approach to characterize surface soil and sediment.
In its October 2004 letter responding to DOE, the contractor concluded
that applying MARSSIM across the site would cost approximately $7.5
million and would take 19 months after the site cleanup had been
completed. Further, the contractor said that implementing

Appendix IV Additional Information on the Quality of Remediation Data at
Rocky Flats and on DOE's Planned and Actual Verification

MARSSIM at the site could constitute a "significant contract change," and
could postpone completion of the cleanup. In essence, the difficulty was
that the MARSSIM approach was inconsistent with the cleanup strategy. The
approved site cleanup strategy was to remove contamination to a 90 percent
confidence level, meaning confidence that at least 90 percent of the
contamination had been remediated to agreed-upon levels. In contrast,
MARSSIM, which was developed years after the Rocky Flats cleanup strategy
had been approved and begun, recommends applying a 100 percent
verification strategy-that is, scanning areas most likely to contain
residual radionuclide contamination. Accordingly, MARSSIM was sure to find
"hot spots"-particularly at the 903 pad and lip area, the extent and
severity of which would then require investigation and potentially cleanup
action. Not having anticipated the implications of applying MARSSIM at the
site, in terms of both cost and schedule, DOE subsequently rejected that
approach.

In March 2005, the contractor proposed a less expensive, less
timeconsuming plan for demonstrating the cleanup's compliance with the
cleanup agreement. The plan's objectives were to verify, with reasonable
certainty, that no unknown areas of radiological contamination remained at
the site, that all remedial actions were complete, and that existing
sampling data remained valid. The verification approach included scanning,
sampling, and reviewing historical and existing data. Through such
activities, the contractor intended to increase DOE's and the community
groups' confidence that the cleanup objectives had been achieved. The plan
purported to exceed the guidance laid out in MARSSIM, although an ORISE
official refuted this claim. Further comments provided by a contractor
hired by the Rocky Flats Coalition of Local Governments to provide input
on the process added to DOE's concerns and resulted in DOE's rejecting the
contractor's sampling methodology.

The contractor's final scanning and sampling plan, issued in May 2005,
included two actions to meet the verification objectives. First, the
entire site would be scanned with sodium iodide detectors affixed to a
helicopter. According to the contractor, the detectors would detect,
within the top several inches of surface soil, gamma rays indicating the
existence of any residual plutonium that exceeded 50 picocuries per gram
of soil, over a

Appendix IV Additional Information on the Quality of Remediation Data at
Rocky Flats and on DOE's Planned and Actual Verification

range of 80 square meters-roughly the size of a two-car garage.6 Figure 18
shows electrical poles that were felled to allow the helicopter to fly as
low as possible over the site and the aerial scan of the site for residual
plutonium in surface soil.

Sources: GAO; DOE.

Of the 21 community group survey respondents who expressed an opinion on
the aerial scan, 9 said the scan greatly or somewhat increased their
confidence in the site's cleanup. Another 8 said the scan neither
increased nor diminished their confidence, and the other 4 said the aerial
scan greatly or somewhat diminished their confidence in the site's
cleanup.

Another component of the contractor's final plan was targeted groundbased
scanning around the perimeter of previously remediated areas, including
the perimeter of the major plutonium buildings and the 903 lip area. These
scans were conducted with a high-purity germanium detector, as described
earlier. Of the 21 community group survey respondents who expressed an
opinion on the contractor's ground-based scans, 11 said these scans
greatly or somewhat increased their confidence in the site's cleanup.

6The sodium iodide detectors on the helicopter can detect only gamma
radiation, which does not directly measure plutonium 239/240
concentrations. However, it does measure americium-241, which can be
multiplied by 5.7 to estimate plutonium 239/240. This conversion factor is
based on a Rocky Flats Cleanup Agreement Contact Record (an agreement
among DOE, EPA, and Colorado).

Page 102 GAO-06-352 Nuclear Cleanup of Rocky Flats

 Appendix IV Additional Information on the Quality of Remediation Data at Rocky
               Flats and on DOE's Planned and Actual Verification

    DOE Revised ORISE's Portion of Planned Verification to Incorporate MARSSIM

Another 7 said the scans neither increased nor diminished their
confidence, and the other 3 said the ground-based scans greatly or
somewhat diminished their confidence in the site's cleanup.

The second part of DOE's cleanup verification effort-ORISE's review of
contractor-conducted scans for remaining radiological contamination- was
revised to apply MARSSIM to the 903 lip area. ORISE's initial plan did not
incorporate MARSSIM. However, the Rocky Flats Coalition of Local
Governments' contractor suggested using MARSSIM at the 903 lip area
because it had been one of the most contaminated areas. ORISE's proposed
final plan included assessing the performance of the contractor's aerial
and targeted ground-based scanning and the adequacy and completeness of
the contractor's closeout reports for the 903 lip area using MARSSIM.

    Results of the Contractor's Final Scanning and Sampling Cleanup Verification
    Plan and ORISE's Independent Verification

Both the contractor's and ORISE's verifications identified
radionuclidecontaminated soil. At a September 1, 2005, public meeting
hosted by the Rocky Flats Citizens Advisory Board, DOE and the contractor
presented the results of the contractor's and ORISE's verifications. DOE
and the contractor reported that the aerial scan had identified known
areas of radionuclide contamination, such as areas where contaminated soil
was stored pending its shipment off site. (The scan also identified an
off-site area, but subsequent investigation found no contamination there.)
However, the aerial scan did not achieve its projected sensitivity.
Whereas detection of 50 picocuries of plutonium per gram of soil over 80
square meters was the criterion for success, according to the contractor's
final plan, the scanner achieved this detection level for only 25 percent
of the site, owing to the moisture content in the ground and the altitude
of the helicopter's flyover. They also reported that the contractor's
ground-based scan found five areas on the perimeter of the 903 lip area
that contained contamination in excess of 50 picocuries per gram of soil,
which were subsequently remediated.

In addition, DOE and the contractor reported that ORISE's MARSSIM sampling
had identified 13 "hot spots," or areas with contamination above the
cleanup agreement's action level of 50 picocuries of plutonium per gram of
soil. Contamination in these 13 hot spots ranged from 65 to 425 picocuries
of plutonium per gram of soil. However, a DOE official said that DOE would
not remediate those areas because the regulatory agencies had already
approved the accelerated cleanup actions, and those actions had been
conducted in accordance with the cleanup methodology agreed to by

Appendix IV Additional Information on the Quality of Remediation Data at
Rocky Flats and on DOE's Planned and Actual Verification

DOE and the regulatory agencies. The cleanup methodology required that at
least 90 percent of any given area be remediated to the required levels;
accordingly, there was a 10 percent chance that areas of elevated
contamination remained. Even though the cleanup had been conducted in
accordance with the approved methodology (which allowed for the chance of
such elevated contamination remaining), DOE's initial refusal to remediate
the hot spots caused concern for the Rocky Flats Coalition of Local
Governments, according to the Coalition's executive director. Within 2
weeks, after consulting with EPA and Colorado, DOE agreed to have the
contractor remediate the hot spots.

Overall, hundreds of samples were taken at the 903 pad and lip area,
including initial and confirmatory samples. Figure 19 shows
postremediation (confirmatory) samples taken by the contractor, as well as
verification samples taken by ORISE.

 Appendix IV Additional Information on the Quality of Remediation Data at Rocky
               Flats and on DOE's Planned and Actual Verification

 Figure 19: Confirmation and Independent Verification Samples Taken at the 903
                                    Lip Area

Source: DOE.

Note: This figure does not show the samples taken on the pad itself (the
white square at the left). According to a contractor official, so many
samples were taken within the pad that a depiction of them would appear as
a solid mass.

Although DOE had agreed to ORISE's final independent verification plan,
DOE did not have ORISE fully implement the plan, and it did not publicly
explain its reasoning. Specifically, DOE did not have ORISE complete the
planned assessment of the aerial and ground-based scanning or the planned
assessment of the contractor's investigations of the results of the aerial
and

Appendix IV Additional Information on the Quality of Remediation Data at
Rocky Flats and on DOE's Planned and Actual Verification

targeted ground-based scans. According to a DOE official, DOE decided that
the results of these remaining assessments would not justify their
completion, particularly because the aerial survey had failed to find any
anomalies.

Of the 20 community group survey respondents who expressed an opinion on
ORISE's soil sampling and surface scans (using a FIDLER-a sodium iodide
scintillation detector), 16 said the scan greatly or somewhat increased
their confidence in the site's cleanup. Another 2 said the scan neither
increased nor diminished their confidence, and the other 2 said the aerial
scan greatly or somewhat diminished their confidence in the site's
cleanup.

                                   Appendix V

                     Comments from the Department of Energy

Appendix V Comments from the Department of Energy

Page numbers in the draft report may differ from those in this report.

Appendix V Comments from the Department of Energy

Appendix VI

Comments from the Department of the Interior

Appendix VI Comments from the Department of the Interior

Page numbers in the draft report may differ from those in this report.

Appendix VI Comments from the Department of the Interior

Appendix VII

Comments from the Colorado Department of Public Health and Environment

                                 Appendix VIII

                    Comments from Kaiser-Hill Company L.L.C.

Appendix VIII Comments from Kaiser-Hill Company L.L.C.

Appendix IX

                     GAO Contact and Staff Acknowledgments

Gene Aloise, (202) 512-3841 or [email protected]

GAO Contact

  Staff Acknowledgments

In addition to the contact named above, Dan Feehan, Assistant Director;
Keith Rhodes, GAO's Chief Technologist; Claire Cyrnak; Glenn Fischer; Tom
Kingham; and Pam Tumler made significant contributions to this report.
Others who made important contributions included Kevin Bailey, Michael
Derr, Doreen Eng, Tom Denomme, Tim DiNapoli, Alison O'Neill, Judy Pagano,
Kim Raheb, and Omari Norman. Also, Brandon Haller, Mehrzad Nadji, and Judy
Pagano made important contributions to the methodologies used in this
report.

  GAO's Mission

The Government Accountability Office, the audit, evaluation and
investigative arm of Congress, exists to support Congress in meeting its
constitutional responsibilities and to help improve the performance and
accountability of the federal government for the American people. GAO
examines the use of public funds; evaluates federal programs and policies;
and provides analyses, recommendations, and other assistance to help
Congress make informed oversight, policy, and funding decisions. GAO's
commitment to good government is reflected in its core values of
accountability, integrity, and reliability.

The fastest and easiest way to obtain copies of GAO documents at no cost
is through GAO's Web site ( www.gao.gov ). Each weekday, GAO posts GAO
Reports and newly released reports, testimony, and correspondence on its
Web site. To

have GAO e-mail you a list of newly posted products every afternoon, go to
www.gao.gov and select "Subscribe to Updates."

                             Order by Mail or Phone

The first copy of each printed report is free. Additional copies are $2
each. A check or money order should be made out to the Superintendent of
Documents. GAO also accepts VISA and Mastercard. Orders for 100 or more
copies mailed to a single address are discounted 25 percent. Orders should
be sent to:

U.S. Government Accountability Office 441 G Street NW, Room LM Washington,
D.C. 20548

To order by Phone: Voice: (202) 512-6000 TDD: (202) 512-2537 Fax: (202)
512-6061

Contact:

To Report Fraud, Web site: www.gao.gov/fraudnet/fraudnet.htm

  E-mail: [email protected]

Federal Programs Automated answering system: (800) 424-5454 or (202)
512-7470

Gloria Jarmon, Managing Director, [email protected] (202) 512-4400 U.S.
Government Accountability Office, 441 G Street NW, Room 7125 Relations
Washington, D.C. 20548

Paul Anderson, Managing Director, [email protected] (202) 512-4800

  Public Affairs

U.S. Government Accountability Office, 441 G Street NW, Room 7149
Washington, D.C. 20548
*** End of document. ***