Nuclear Waste: Department of Energy's Hanford Tank Waste
Project--Schedule, Cost, and Management Issues (Letter Report, 10/08/98,
GAO/RCED-99-13).

Pursuant to a congressional request, GAO provided information on the
Department of Energy's (DOE) revised contracting approach for its
Hanford Tank Waste Project, focusing on: (1) how DOE's current approach
has changed from its original privatization strategy; (2) how this
change has affected the project's schedule, cost, and estimated savings
over conventional DOE approaches; (3) what risks DOE is now assuming
with this change in approach; and (4) what steps DOE is taking to carry
out its project oversight responsibilities. DOE contracted with BNFL,
Inc. to treat about 10 percent of Hanford's tank waste.

GAO noted that: (1) the project as currently envisioned is substantially
different from DOE's 1996 initial privatization strategy; (2) although
the project award was made on the basis of a fixed-price contract,
further competition between contractors and short-term demonstration
facilities have been eliminated in favor of more permanent facilities
that could operate for 30 years or more and, therefore, would be
available to treat additional tank waste; (3) the design phase as well
as the date when DOE and BNFL are to reach agreement on final contract
price have been extended by 2 years to August 2000; (4) BNFL's specific
project financing arrangements which were to be established in May 1998
have been deferred until August 2000; (5) to ensure that BNFL can obtain
affordable private financing, DOE has agreed to repay much of the
project debt if BNFL defaults on its loans and DOE terminates the
contract; (6) this is an unusual feature of a fixed-price contract
because the government normally does not agree to pay a contractor's
debt as an allowable cost; (7) the revised approach extends the
completion date for processing the first portion of the waste from 2007
to 2017, and total costs rise from $4.3 billion to $8.9 billion; (8) the
increased costs are mainly the result of DOE's decision to build
permanent facilities that will take longer and cost more to design and
build, and the higher financing costs and contractor profits involved in
operating these facilities over a longer period of time; (9) DOE
estimated that this approach would save 26 to 36 percent over
contracting approaches it has used in the past; (10) because of
questions about DOE's methodology for estimating savings, considerable
caution is needed in assuming how much the revised approach will save;
(11) the contract now calls for DOE to pay BNFL for most of the debt
incurred in building and operating the facility if BNFL should default
on its loans; (12) thus, DOE faces a financial risk not initially
contemplated on the project that could be in the billions of dollars;
(13) DOE agreed to assume this risk because it did not think BNFL would
be able to obtain affordable financing unless the government provided
some assurance that the loans would be repaid; (14) given that the
project still has a number of technical uncertainties, DOE's financial
risks are significant; (15) DOE has identified additional expertise it
needs and has developed several management tools to strengthen its
oversight of the project; and (16) the success of the project will
depend on how well DOE implements these plans.

--------------------------- Indexing Terms -----------------------------

 REPORTNUM:  RCED-99-13
     TITLE:  Nuclear Waste: Department of Energy's Hanford Tank Waste 
             Project--Schedule, Cost, and Management Issues
      DATE:  10/08/98
   SUBJECT:  Privatization
             Nuclear waste storage
             Nuclear waste management
             Waste treatment
             Hazardous substances
             Radioactive waste disposal
             Fixed price contracts
             Cost control
IDENTIFIER:  DOE Hanford Multi-Function Waste Tank Facility Project
             Hanford (WA)
             
******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO report.  Delineations within the text indicating chapter **
** titles, headings, and bullets are preserved.  Major          **
** divisions and subdivisions of the text, such as Chapters,    **
** Sections, and Appendixes, are identified by double and       **
** single lines.  The numbers on the right end of these lines   **
** indicate the position of each of the subsections in the      **
** document outline.  These numbers do NOT correspond with the  **
** page numbers of the printed 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.                                         **
**                                                              **
** A printed copy of this report may be obtained from the GAO   **
** Document Distribution Center.  For further details, please   **
** send an e-mail message to:                                   **
**                                                              **
**                                            **
**                                                              **
** with the message 'info' in the body.                         **
******************************************************************


Cover
================================================================ COVER


Report to Congressional Requesters

October 1998

NUCLEAR WASTE - DEPARTMENT OF
ENERGY'S HANFORD TANK WASTE
PROJECT-- SCHEDULE, COST, AND
MANAGEMENT ISSUES

GAO/RCED-99-13

Nuclear Waste

(141217)


Abbreviations
=============================================================== ABBREV

  BNFL - British Nuclear Fuels, Ltd. 
  DOE - Department of Energy
  M&O - management and operations
  OSHA - Occupational Safety and Health Administration

Letter
=============================================================== LETTER


B-281093

October 8, 1998

The Honorable Thomas J.  Bliley, Jr.
Chairman, Committee on Commerce
The Honorable Joe L.  Barton
Chairman, Subcommittee on Oversight and Investigations
Committee on Commerce
House of Representatives

Cleaning up waste produced as a by-product of 50 years of supplying
the nation's nuclear materials for weapons is a formidable challenge. 
The Department of Energy (DOE) spends over $5 billion per year on its
program to clean up radioactive and hazardous waste at its nuclear
weapons production sites.  The Hanford Site, located in southeast
Washington State, has one of the greatest concentrations of
radioactive waste in the world.  One of the most difficult cleanup
challenges at Hanford involves the 177 underground storage tanks
holding highly radioactive liquid waste, sludge, and other materials. 
Cleaning up this waste is important because it poses a significant
risk to the environment and to surrounding communities.  Recently,
DOE disclosed that waste leaking from some of the tanks has reached
the groundwater and threatens the nearby Columbia River. 

In 1996, DOE decided it would purchase waste treatment services
through competitively awarded, fixed-price contracts to demonstrate
treatment technologies and treat at least 6 percent of the waste. 
Under these contracts, competing contractors would finance, design,
build, and operate temporary waste processing facilities and be paid
on a per-unit basis if they successfully immobilized the waste for
storage.  DOE referred to this approach as its privatization
strategy.  However, on August 24, 1998, DOE signed a contract with
only one contractor--BNFL, Inc.  (BNFL),\1

a subsidiary of British Nuclear Fuels, plc., to design, build, and
operate permanent facilities to treat about 10 percent of Hanford's
tank waste. 

In view of the billions of dollars that the government will spend
treating this waste, at your request we assessed the implications of
DOE's new approach.  This report discusses (1) how DOE's current
approach has changed from its original privatization strategy; (2)
how this change has affected the project's schedule, cost, and
estimated savings over conventional DOE approaches; (3) what risks
DOE is now assuming with this change in approach; and (4) what steps
DOE is taking to carry out its project oversight responsibilities. 

As a starting point for our review, we used the report DOE issued to
the Congress in July 1998 outlining the proposed shift in its
contracting approach.  We supplemented this with on-site reviews at
Hanford; reviews of DOE and contractor documents and plans; and
discussions with DOE, contractor, and industry officials at Hanford
and in Washington, D.C. 


--------------------
\1 DOE and BNFL signed a modification to an existing contract.  For
simplicity, we refer to this as a contract throughout this report. 


   BACKGROUND
------------------------------------------------------------ Letter :1

Hanford's aging underground tanks contain about 54 million gallons of
highly radioactive waste.  Over the years, more than 1 million
gallons of waste have leaked into the soil.  DOE currently estimates
the total cost of cleaning up the tank waste at more than $50 billion
(in actual year dollars).  Since 1989, DOE has spent about $3.5
billion on this effort.  To convert the waste into a form for more
permanent storage, the waste will be separated into high-level and
low-activity components\2 and then, through a process called
vitrification, converted into a glass-like material that can be
poured into steel containers where it will harden.  The immobilized
high-level waste will be stored on-site for eventual shipment to a
national repository, while the low-activity waste will be permanently
disposed of on the Hanford Site. 

DOE plans to use private contractors to conduct the vitrification
work and over the last several years has developed a contracting
approach.  In February 1996, DOE issued a Request for Proposals for
the tank waste project.  DOE envisioned that under this approach, two
contractors would build and operate demonstration facilities that
would treat at least 6 percent of the waste.  DOE referred to this
part of the waste treatment effort as phase I.  DOE estimated that
phase I would last at least until 2007 and cost about $3.2 billion
for the two fixed-price contracts and another $1.1 billion in
contract support costs, for a total of about $4.3 billion.  In
September 1996, DOE awarded a fixed-price contract for $27 million to
each of the two contractor teams to begin phase I by developing
preliminary facility designs and other preliminary project plans. 
One team was led by BNFL and the other team was led by Lockheed
Martin Advanced Environmental Systems (Lockheed).  In phase II,
contractors would compete for a contract to process the remaining
tank waste. 

DOE's experience during the initial part of phase I led to a change
in the contracting approach.  In May 1998, after reviewing the
preliminary designs and plans submitted by the two competing teams,
DOE decided to continue phase I with only one contractor--BNFL.  DOE
and outside expert reviewers found that the approach set forth by the
Lockheed team presented an unacceptably high technical risk in
attaining DOE's cleanup goals, requiring additional development and
using technologies that, in some cases, were largely unproven.  In
contrast, DOE concluded that BNFL's technical approach was sound,
using technologies for waste treatment and vitrification that were
well developed and had been used in other waste treatment situations. 
On August 24, 1998, DOE signed a fixed-price contract with BNFL to
continue with phase I by developing an approach that would process at
least 10 percent of Hanford's tank waste by 2017.  The contract cost
was estimated at about $6.9 billion (in constant 1997 dollars).  The
contract is considered a fixed-price contract because DOE and BNFL
agreed to a target fixed-price per unit of processed waste.  This
price may be adjusted in the year 2000 after technical aspects of the
project become more clearly known.  DOE estimated that its other
costs related to supporting BNFL's efforts would be about $2 billion,
bringing the project's total estimated cost to about $8.9 billion. 


--------------------
\2 Hanford's tanks contain highly radioactive waste.  When separated
into high-level and low-activity components, most of the waste will
be low-activity radioactive waste.  Low-activity waste has a wide
range of characteristics, but most of it contains small amounts of
radioactivity in large volumes of materials.  The tanks also contain
hazardous chemicals and heavy metals. 


   RESULTS IN BRIEF
------------------------------------------------------------ Letter :2

The project as currently envisioned is substantially different from
DOE's 1996 initial privatization strategy.  Although the project
award was made on the basis of a fixed-price contract, further
competition between contractors and short-term demonstration
facilities have been eliminated in favor of more permanent facilities
that could operate for 30 years or more and, therefore, would be
available to treat additional tank waste.  In addition, the design
phase as well as the date when DOE and BNFL are to reach agreement on
final contract price have been extended by 2 years to August 2000. 
BNFL's specific project financing arrangements which were to be
established in May 1998 have been deferred until August 2000.  In
addition, to ensure that BNFL can obtain affordable private
financing, DOE has agreed to repay much of the project debt if BNFL
defaults on its loans and DOE terminates the contract.  This is an
unusual feature of a fixed-price contract because the government
normally does not agree to pay a contractor's debt as an allowable
cost.  If this commitment were structured as a conventional loan
guarantee, DOE would have had to estimate the potential subsidy cost
over the term of the loans and have the budget authority to fund them
before making the guarantee. 

The revised approach extends the completion date for processing the
first portion of the waste from 2007 to 2017, and total costs rise
from $4.3 billion to $8.9 billion, including $2 billion in DOE's
support costs (in constant 1997 dollars).  The increased costs are
mainly the result of DOE's decision to build permanent facilities
that will take longer and cost more to design and build, and the
higher financing costs and contractor profits involved in operating
these facilities over a longer period of time.  However, DOE
estimated that this approach would save 26 to 36 percent over
contracting approaches it had used in the past, such as reimbursing a
contractor for its costs plus an agreed-upon profit amount.  Because
of questions about DOE's methodology for estimating savings,
considerable caution is needed in assuming how much the revised
approach will save. 

The revised approach represents a dramatic departure from DOE's
original privatization strategy of shifting most financial risk to
the contractor.  The contract now calls for DOE to pay BNFL for most
of the debt incurred in building and operating the facility if BNFL
should default on its loans.\3 Thus, DOE faces a financial risk not
initially contemplated on the project that could be in the billions
of dollars.  DOE agreed to assume this risk because it did not think
BNFL would be able to obtain affordable financing unless the
government provided some assurance that the loans would be repaid. 
Given that the project still has a number of technical uncertainties
such as using waste treatment technology that has yet to be
successfully tested at production levels on Hanford's complex and
unique wastes, and management challenges such as obtaining needed
contracting expertise, DOE's financial risks are significant. 

In an attempt to avoid repeating past mistakes in managing large
projects, DOE has identified additional expertise it needs and has
developed several management tools to strengthen its oversight of the
project.  For example, DOE plans to have about an 80-member team to
manage this effort, and it has taken a number of steps to plan for
better coordination among BNFL, the contractors providing support
services at Hanford, and its own staff.  The success of the project,
however, will depend heavily on how well DOE implements these plans. 
DOE has a history of not fully implementing its management and
oversight plans, and there are some early indications on this project
that DOE may be having difficulty ensuring that the proper expertise
is in place and fully funding project support activities. 


--------------------
\3 Under the terms of the contract, if the lenders declare BNFL in
default and accelerate the debt due, DOE will terminate the contract
for default or for the convenience of the government.  In this event,
DOE will pay BNFL, as an allowable cost, the outstanding principal
amount of the loans plus all accrued and unpaid interest, less
certain other adjustments. 


   DOE'S CURRENT APPROACH DIFFERS
   SIGNIFICANTLY FROM ORIGINAL
   PROJECT STRATEGY
------------------------------------------------------------ Letter :3

DOE's August 1998 contract with BNFL is a substantial departure from
DOE's original privatization strategy.  DOE said its contracting
approach evolved as it received feedback from private companies and
financial advisors, as well as input from the two contractor teams
that submitted proposals under the original approach.  According to
DOE, changes to its initial approach were made to optimize the
technical approach and to make the project financially feasible or to
reduce the likelihood of performance problems.  These changes fall
into four main areas:  competition, financial issues, facility
issues, and schedule revisions.  Table 1 summarizes these changes,
and the sections that follow discuss why DOE made them. 



                                Table 1
                
                   Summary of Major Changes to DOE's
                   Privatization Contracting Approach

Type of change                  Old approach        New approach
------------------------------  ------------------  ------------------
Competition                     Two contractors     One contractor
                                throughout phase I

Financial issues-risk           Contractors         Contractor
                                provide equity and  provides equity
                                obtain private      and obtains
                                financing           private financing
                                guaranteed by       but DOE commits to
                                corporate assets    repay contractor's
                                                    debt if contract
                                                    terminates

Financial issues-payment        Payments made upon  Up to $50 million
                                delivery of         interim payment
                                treated waste       made before waste
                                                    is processed;
                                                    remaining payments
                                                    made upon delivery
                                                    of treated waste

Financial issues-setting        Final price         Final price
prices                          agreement set by    agreement delayed
                                May 1998            2 years to August
                                                    2000

Facilities issues               Contractors build   Contractor builds
                                temporary           permanent
                                demonstration       production
                                facilities to       facilities to
                                process 6% of       process at least
                                Hanford waste by    10% of Hanford
                                2007                waste by 2017

Schedule revision               Construction to     Design period
                                begin December      extended;
                                1999                construction to
                                                    begin July 2001
----------------------------------------------------------------------

      COMPETITION
---------------------------------------------------------- Letter :3.1

Unlike DOE's original approach, the current approach is not
competitive.  Because DOE found that the approach set forth by the
Lockheed team was unacceptable, DOE authorized only BNFL to proceed
through the remainder of phase I.  The extent to which competition
will be present in phase II is unknown. 


      FINANCIAL ISSUES
---------------------------------------------------------- Letter :3.2

DOE's approach to financing the project has shifted.  Instead of
requiring the contractor to obtain all needed financing without
recovery from DOE if the project fails, DOE is now planning to repay
BNFL's debts above its equity, insurance, and other limited funds if
BNFL defaults on its loans and DOE terminates the contract.  DOE
officials said that the government's commitment to repay the
contractor's debt was needed, in large part, to make the project
financially feasible.  DOE concluded that requiring 100 percent
private financing without government backing of the private debt
could increase the financing costs so much that the project would not
be affordable.  DOE also concluded that an increased governmental
role in backing the debt would reduce the overall cost of the
project, but DOE had no specific estimate of the amount of the
reduced costs.  Government backing of the private debt is an unusual
feature for a fixed-price contract because the government normally
does not agree to pay a contractor's debt as an allowable cost. 

DOE has also shifted its financing strategy from providing payment to
the contractor only upon delivering the processed waste, to one where
BNFL may now receive a payment of up to $50 million before waste
processing begins.\4 DOE said that such interim payments will help
reduce the total project cost by reducing long-term financing costs
and providing an incentive for BNFL to reduce the actual costs of
waste processing. 

Finally, neither contractor was willing to commit to a fixed-unit
price and schedule by May 1998 without adding significant contingency
to the price.  The August 1998 contract identified a target price and
set August 2000 as the date at which the unit price will be fixed and
BNFL's funding commitments will be established.  DOE determined that
this delay would strengthen the feasibility and economics of the
project, although agency officials also indicated that information
gained during this delay could also lead to increased prices. 


--------------------
\4 The payment includes a base amount of $20 million for specific
deliverables and an incentive fee based on the extent that BNFL is
able to reduce the project's construction and operations costs or
identify contract changes acceptable to DOE that will reduce the
overall system cost. 


      FACILITY ISSUES
---------------------------------------------------------- Letter :3.3

DOE is contracting for waste processing services that will be much
different from what was initially envisioned.  DOE originally planned
that two contractors would build temporary waste processing
facilities during phase I that would test each contractor's
technological approach.  These facilities were estimated to have a
useful life of approximately 10 years and were considered
"throw-away" buildings.  According to DOE, both BNFL and Lockheed
concluded that shorter-term, throw-away facilities were not feasible
and that longer-life facilities were needed to provide the required
levels of safety, operability, and maintainability.  The contract now
requires the waste treatment facilities to be designed to operate for
a minimum of 30 years and have the capability to increase capacity. 
DOE said that although this approach means much more expensive
facilities than originally anticipated and, therefore, an increase in
project costs for phase I, longer-life, expandable facilities allow
DOE more flexibility and options in how the waste cleanup is
completed. 


      SCHEDULE REVISION
---------------------------------------------------------- Letter :3.4

In addition to more permanent, costly facilities, the new contract
extends the design period and delays the start of construction about
19 months beyond what was originally planned.  DOE expected that
construction of the facilities would begin by December 1999. 
However, in their January 1998 proposals, both BNFL and Lockheed
indicated that additional time was needed to further develop project
design and plans for meeting regulatory and permitting requirements. 
The contractors believed that adhering to the original schedule would
carry too many uncertainties, and that they would be unable to obtain
needed project financing unless a more realistic schedule could be
negotiated.  DOE believes that the change will allow further design
development before construction begins, thereby reducing the risks
associated with design uncertainties. 


   DIFFERENT APPROACH TO THE
   PROJECT EXTENDED SCHEDULE AND
   INCREASED COSTS
------------------------------------------------------------ Letter :4

Current schedule and cost estimates for the project are substantially
greater than DOE's original estimates.  In 1996, DOE estimated that
in the first phase of the project, two contractors would process 6
percent of the waste by 2007 and up to 13 percent of the waste by
2011.  DOE is now estimating that phase I will last until at least
2017, an extension of up to 10 years.  Several interim steps in phase
I also have revised completion dates.\5 (See table 2.)



                                Table 2
                
                   Changes in Estimated Schedule for
                   Hanford's Waste Treatment Project

                              Original      Revised          Extension
                              schedule      schedule          (months)
----------------------------  ------------  ------------  ------------
Contract modification signed  May 1998      Aug. 1998                3
                                            (actual)

Fixed-unit price established  May 1998      Aug. 2000               27

Facility construction         Dec. 1999     July 2001               19
started

Waste vitrification started   Dec. 2002     Feb. 2007               50

Waste processing completed    2007\a        2017\b                 120
----------------------------------------------------------------------
\a DOE's original schedule required the contractors to process 6
percent of the waste by 2007.  DOE had the option to extend waste
processing an additional 4 years to 2011 to process a total of 13
percent of the waste, depending on the contractors' performance. 

\b DOE's funding schedule shows that the project will be funded only
through 2017.  Additional time will be required to deactivate the
facilities. 

One change in the project schedule was an extension allowing BNFL
more time to design and construct permanent, more durable waste
pretreatment and treatment plants.  BNFL proposed extending the
design phase by 24 months.  BNFL reasoned that the additional design
time would help reduce the risks associated with a fast-track
design/construct project and was needed because more extensive
structures were being proposed.  BNFL's proposal to build more
durable facilities with a longer useful life also required a longer
construction time.  BNFL proposed this approach, and DOE agreed to
it, because (1) BNFL said that nuclear and worker safety requirements
could not be efficiently incorporated into the demonstration
facilities initially proposed and (2) permanent facilities provided
advantages in processing the tank waste that would remain after phase
I.  The lengthened construction added about 4 years to the original
schedule.  Also, it will take BNFL about 10 years to process the
waste, or about 5 years longer than if two contractors were doing so. 
In addition, BNFL included additional schedule contingency to deal
with possible start-up and production problems. 

Estimated costs for the project have also increased significantly. 
The total project costs for phase I, including DOE's support costs,
increased from $4.3 billion in the original estimate to process 6
percent of the waste to $8.9 billion in the current estimate to
process 10 percent of the waste, as measured in constant fiscal year
1997 dollars.  (See table 3.)



                                Table 3
                
                Changes in Cost Estimates for Hanford's
                        Waste Treatment Project

                 (Constant fiscal year 1997 dollars in
                               millions)

                                   Initial       Revised       Percent
Cost element                      estimate      estimate        change
----------------------------  ------------  ------------  ------------
Contract price                    $3,200\a      $6,960\b          +118
DOE support costs\c                 $1,050        $1,970           +88
Total costs                         $4,250      $8,930\d          +110
Cost per metric ton of waste         $0.76          $1.5           +97
 processed
----------------------------------------------------------------------
\a Includes initial project costs. 

\b Includes contract target price of $6.93 billion, initial project
costs to date, and a $20 million fee for successfully completing
project requirements and funding arrangements to begin facility
construction. 

\c Support costs are understated because they do not include the cost
of the regulatory unit or federal personnel costs.  These costs were
not readily available. 

\d This estimate does not include the potential cost to the
government of a possible BNFL default on its loans with the private
sector. 

The waste processing facilities now being designed will cost nearly
$1 billion more to build than the demonstration facilities DOE
originally proposed.  Because of the longer period during which
investors will expect a return on investments, equity and debt
financing costs are expected to increase from about $1 billion to
more than $3 billion. 


--------------------
\5 Some of these dates are later than the dates DOE has agreed to
with its regulators, the Environmental Protection Agency, and the
Washington State Department of Ecology.  DOE will need to renegotiate
these dates with its regulators.  One of the milestones, the date to
start high-level waste processing, is earlier than the date agreed to
with the regulators. 


      COST SAVINGS ESTIMATE MUST
      BE VIEWED WITH CAUTION
---------------------------------------------------------- Letter :4.1

As part of its revision of the project's cost and schedule, DOE
analyzed whether the BNFL fixed-price approach was likely to save
money when compared with two alternatives:  a management and
operations (M&O) contract or a cost-reimbursement contract with
performance-based incentives.  In July 1998, DOE estimated the range
of savings under its revised approach for phase I at 26 to 36 percent
when compared with these two alternatives.  The savings estimate of
36 percent was based on comparing the proposed BNFL fixed-price
approach with an M&O approach based on past Hanford management and
operating contractor cost data; the estimate of 26 percent was based
on a comparison with the estimated cost for BNFL to perform the work
under a cost-reimbursement contract.  Our review of DOE's most recent
estimates indicate that the savings amounts should be viewed with
considerable caution.  For example: 

M&O contracting approach should not be used as a comparison because
an M&O contracting strategy is no longer an approach that DOE would
seriously consider using.  Until recently, DOE has generally used M&O
contractors to manage its facilities.  DOE's M&O approach involved
paying an on-site contractor for management and operating services
regardless of what was accomplished.  However, DOE is shifting to
management and integration contracts involving performance measures
and incentive-based contracts.  DOE officials at Hanford told us that
they used an M&O contract estimate because they had historical
experience with M&O contracts and had an idea what the costs would
be.  However, we believe DOE's cost savings analysis could be more
meaningful if it included a range of contracting options that DOE
would actually consider using, such as various combinations of
government and private financing.  Because of the high cost of
private financing on this project and DOE's agreement to assume the
risk associated with the debt, direct federal funding of part or all
of the project may actually lower total project costs without
significantly increasing the government's level of risk.\6 DOE
discussed government versus private financing in its report to the
Congress on this project but did not present information on the
differences in total project costs.  According to DOE officials,
during the next 2 years they will further evaluate the trade-offs
between using government and private debt to determine the best
overall mix of equity, debt, and government financing for the
project. 

Rough estimates are presented as precise numbers.  Cost projections
for two of the contract alternatives DOE considered in its analysis
are based on what are called "rough order of magnitude" estimates. 
The margin of error for these estimates is plus or minus 40 percent,
meaning that the actual cost could be up to 40 percent less than or
greater than the estimate presented.  Because the order of magnitude
estimates are subject to so much variability, it is difficult to
assign much credence to an overall savings estimate based on these
numbers. 

Cost growth estimates are not used consistently.  For the comparison
between a fixed-price contract and a cost-reimbursement contract with
performance incentives, DOE assumed that cost growth would be 68
percent for the cost-reimbursement contract, and the fixed-price
contract would have no cost growth.  However, other evidence
indicates that fixed-price contracts may have greater cost growth
than cost-reimbursement contracts.\7

Analysis does not reflect full range of cost savings estimates. 
Although the potential cost savings DOE reported to the Congress show
a range of 26 to 36 percent, DOE documents supporting the analysis
show a range of 10 to 36 percent.  According to DOE's Director of the
Office of Project and Fixed Asset Management, DOE did not disclose
the lower number in its report to the Congress because it did not
believe the lower number was adequately supported. 

This is not the first time DOE has based cost savings on questionable
analytical practices.  In January 1997, we reported on the accuracy
of the information DOE provided to the Congress to support the
Department's fiscal year 1997 request for privatization funds.  DOE
claimed that six privatization projects had saved $1.1 billion. 
However, we found that DOE had understated some project costs, used
incorrect cost data, and made cost comparisons using projects of
different scopes.\8 In addition, our 1996 report on Hanford's tank
waste project also discussed the problems associated with DOE's
practice of presenting rough order of magnitude numbers as point
estimates.\9


--------------------
\6 For a discussion of alternative financing approaches and related
risks, see Department of Energy:  Alternative Financing and
Contracting Strategies for Cleanup Projects (GAO/RCED-98-169, May 29,
1998). 

\7 DOE based the 68-percent figure on a study that found a 68-percent
difference between two cost estimates--an estimate developed in the
study assuming an M&O contracting approach and a government cost
estimate for a privatized approach to the project.  However, the
study assumes that the M&O costs will be greater and uses judgment to
estimate the amount of the increase in costs, rather than estimating
the differences using actual historical data.  (See DOE M&O
Contractor Cost Estimate, Burns and Roe, June 1998.) Other
DOE-commissioned studies have reached different conclusions about
cost growth on DOE contracts by studying actual cost growth on DOE
contracts.  For example, a 1993 study of DOE's environmental
management contracting and project management practices found that
for a representative sample of contracts, the actual cost of
fixed-price contracts exceeded the estimated costs by almost 75
percent.  The primary cause of the cost increase was that fixed-price
contracts were used on poorly defined projects, which led to changes
during construction contributing to increased costs and schedule
delays.  In contrast, the study found that cost-reimbursement
contracts exceeded the estimated costs by 35 percent.  See Department
of Energy, Office of Environmental Restoration and Waste Management,
Project Performance Study, Independent Project Analysis, Inc. 
Reston, Va.  (Nov.  30, 1993). 

\8 See Nuclear Waste:  DOE's Estimates of Potential Savings From
Privatizing Cleanup Projects (GAO/RCED-97-49R, Jan.  31, 1997). 

\9 See Hanford Waste Privatization (GAO/RCED-96-213R, Aug.  2, 1996). 


   REVISED APPROACH SHIFTS
   SIGNIFICANT FINANCIAL RISK TO
   THE GOVERNMENT
------------------------------------------------------------ Letter :5

Under the revised contract approach, DOE faces a substantial
financial risk that could be in the billions of dollars.  This risk
comes mainly in the form of an agreement to pay BNFL for much of the
debt incurred in constructing and operating the waste treatment
facilities if BNFL defaults on its loan payments and DOE terminates
the contract.  This agreement has the same practical effect as a loan
guarantee and is a dramatic departure from the original privatization
strategy.\10 If DOE had provided a guarantee for BNFL's loans from a
private lender, the Federal Credit Reform Act would have required DOE
to estimate the net present value of the subsidy cost of the loan
guarantee over the term of the loan and to have budget authority
available for this full cost before the guarantee could be provided. 
DOE officials told us they agreed to back BNFL's loans because
lenders told DOE that BNFL would not be able to obtain affordable
financing without it.  The officials said the increased governmental
role would likely reduce the contract's overall cost by allowing BNFL
to borrow at lower rates. 

So far, the amount of DOE's potential liability is unknown, because
the amount of borrowing that will be covered under the commitment
will likely not be determined until the contract price is established
and financial closure occurs in August 2000.  However, BNFL's vice
president and project manager told us that DOE's potential liability
could be as much as $3 billion.  He said that in the case of a
default, $3 billion is about the maximum debt that would be
outstanding after BNFL's equity and contingency funds were
applied.\11

Apart from this financial risk, DOE also faces other financial risks
that are not as significant as a default by BNFL, but could increase
the project's overall price.  For example, DOE is subject to making
idle facility payments to BNFL if DOE is unable to supply waste from
the tanks.  Also, the contract contains provisions for renegotiating
the agreed-upon price if certain changes occur that could affect cost
or schedule, such as DOE's failure to provide required support
services, changes in environmental law, or other events that are
beyond the control of the contractor.  Some of these risks would also
exist under a more traditional contracting arrangement. 

DOE's financial risks hinge on a number of factors that could
potentially affect the project.  We identified six main types of
factors, which we believe merit continued attention as the project
proceeds. 


--------------------
\10 DOE's agreement to pay BNFL its outstanding debt as an allowable
cost if the contract is terminated is close to, but not the same as,
a federal loan guarantee.  DOE's agreement is a commitment DOE has
with BNFL, not with BNFL's lenders, and therefore it does not meet
the definition of a loan guarantee.  A federal loan guarantee is
provided directly to a lender, not to the borrower.  Agencies need
legislative authority to provide a loan guarantee. 

\11 Debt for financing the project can be of two types:  debt that is
secured by BNFL's assets (called "recourse" debt) and debt that is
secured only by the revenues BNFL expects to receive from treating
the waste (called "nonrecourse" debt).  The agreement between DOE and
BNFL applies only to recourse debt.  However, to this point, lenders
appear reluctant to provide a significant amount of nonrecourse
funding because of the project's numerous technical and operating
risks.  Thus, borrowing against assets is likely to be the main
source of capital for the project.  DOE's risk is made even more
substantial because BNFL is a separate corporation from its parent
company and, therefore, lenders may not be able to pursue BNFL's
parent company in the event of a contract termination. 


      UNVERIFIED TECHNOLOGY
---------------------------------------------------------- Letter :5.1

BNFL officials acknowledge that although the technology they plan to
use has been successfully applied in other settings, it has been
tested only on small amounts of Hanford waste in laboratories, and
has not been used at production facilities to vitrify the unique
types of waste at Hanford.  Under DOE's original approach, the
success of the selected technologies was to be demonstrated in
temporary plants; in DOE's revised approach, permanent plants will be
built.  However, BNFL has developed various other approaches to deal
with the need to ensure that the technology will work.  These include
conducting tests on certain aspects of the technology at existing
facilities at other DOE sites and in the United Kingdom and
constructing a prototype melter\12 for the low-activity waste
vitrification process.  These efforts are expected to continue as the
vitrification facilities are being designed and built.  BNFL has
assured DOE that its technology will be fully tested and demonstrated
before beginning operations of its full-scale, high-level waste
treatment plant in February 2007 and its low-activity waste treatment
plant in January 2008.  To ensure that technologies are fully
demonstrated, DOE expects to hire experts to review BNFL's
demonstration plans and testing results.  While DOE Hanford officials
also expect DOE headquarters to commission an independent review of
BNFL's testing results, no plans for this review have been developed. 

Under its revised approach, DOE retains a significant part of the
risk for the success of this technology.  In the worst case, if
demonstration activities fail or if they prove inadequate to ensure
the success of full-scale operations, the overall project may fail,
and DOE will be liable for paying off a significant portion of BNFL's
debt after BNFL's resources are exhausted.  If demonstration
activities show that the technology is usable but flawed, treatment
facilities may require expensive retrofitting to make them viable. 
This could raise the cost of the fixed-price contract that DOE will
negotiate with BNFL. 


--------------------
\12 The melter is a large furnace that heats the waste to a high
temperature and combines it with other materials to produce a
glass-like product. 


      RAPID PLANT CONSTRUCTION
---------------------------------------------------------- Letter :5.2

Although the revised approach gives BNFL additional time to design
the waste treatment and vitrification facilities, the schedule still
poses some potential risk.  To give BNFL more time to design the
facilities, DOE set back the start of construction by about 2 years. 
However, even with this change, construction will begin well before
all of the design work is completed.  BNFL officials estimate that
overall design work will be less than 50 percent complete at the
start of construction. 

DOE and BNFL officials expressed confidence in the time line of the
revised approach.  The officials said that the schedule is comparable
to other nuclear facilities BNFL has successfully built and operated. 
However, BNFL officials also acknowledged that conducting
simultaneous design, construction, and technology testing carries
some risk.  To reduce this risk, BNFL performs a periodic risk
assessment to ensure that design and technology testing concerns will
be addressed as quickly as possible in the next 24 months.  Because
of our experience in analyzing similar schedules that have
contributed to problems on other DOE projects, we believe that the
construction schedule is a potential factor affecting DOE's risk. 
Specifically, DOE projects such as the Defense Waste Processing
Facility at Savannah River, the Pit-9 cleanup project at the Idaho
National Engineering and Environmental Laboratory, and the Spent Fuel
Storage Project at Hanford\13 experienced cost overruns and schedule
changes that suggest designs should be developed enough to mitigate
such results. 

There do not appear to be agency or industry guidelines on the extent
to which facility designs for complex, one-of-a-kind nuclear
processing facilities like vitrification plants should be complete
before construction begins.  In an analysis of an earlier DOE
proposal to build a waste vitrification plant at Hanford, we raised
similar concerns about concurrent design and construction and pointed
out that for an advanced light water reactor, the Electric Power
Research Institute recommended that construction not begin until the
detailed design is 90 percent complete.\14 A manager of the
Institute's Nuclear Power Group told us that advanced light water
reactors are similar in complexity to a vitrification plant. 


--------------------
\13 See Nuclear Waste:  Defense Waste Processing Facility--Cost,
Schedule, and Technical Issues (GAO/RCED-92-183, June 17, 1992),
Nuclear Waste:  Department of Energy's Project to Clean Up Pit 9 at
Idaho Falls Is Experiencing Problems (GAO/RCED-97-180, July 28,
1997), and Nuclear Waste:  Management Problems at the Department of
Energy's Hanford Spent Fuel Storage Project (GAO/T-RCED-98-119, May
12, 1998). 

\14 Nuclear Waste:  Hanford Tank Waste Program Needs Cost, Schedule,
and Management Changes (GAO/RCED-93-99, Mar.  8, 1993). 


      SAFETY AND REGULATORY ISSUES
---------------------------------------------------------- Letter :5.3

Another factor potentially affecting the success of the project--and
therefore DOE's financial risk--is whether the safety and other
regulatory requirements can be successfully met.  For example, DOE's
Regulatory Unit raised 90 issues with safety documents that BNFL
submitted in January 1998.  DOE's manager of the Regulatory Unit
described the quality of the BNFL safety documents as "poor" and said
that the next set of safety documents, submitted in August 1998, was
also poorly done.  These latter documents were subsequently withdrawn
by BNFL.  The DOE manager said that problems with safety documents
could affect the project schedule and cost and that BNFL needed to
make immediate improvements in its approach to safety.  Several
additional safety documents are required before BNFL can begin
construction of the facilities in the year 2001.  Unless the required
safety documentation is approved, BNFL will be unable to start
construction on schedule.  DOE's Regulatory Unit is working to
prevent problems we found on two recent major projects--the Pit 9
project in Idaho and the Spent Fuel Storage Project at Hanford--where
problems with the safety basis of the work delayed project schedules
and caused additional rework. 

The BNFL project manager attributed the safety documentation problems
primarily to the early level of project design and said that BNFL
will greatly increase the staff addressing safety-related issues
during the rest of phase I.  BNFL also has recently hired an
experienced nuclear facilities licensing manager to lead this effort. 
DOE has also taken steps to help ensure that BNFL is addressing
safety issues.  For example, DOE has negotiated into the contract
provisions which (1) require periodic meetings between its regulatory
staff and BNFL to discuss safety issues and (2) provide for DOE
attendance at BNFL's safety committee facility design review
meetings. 

The project also presents another regulatory challenge.  DOE planned
to have the Occupational Safety and Health Administration (OSHA)
regulate worker safety at the plant.  However, in May 1998, OSHA
declined to assume responsibility, citing a need first for statutory
and regulatory changes to be in place, as well as a full complement
of the resources required.  If OSHA does not regulate worker safety,
then DOE must do so.  The manager of DOE's Regulatory Unit said that
if this issue is not resolved by January 2000, his unit will assume
responsibility for regulating worker safety so that construction can
begin on schedule. 


      DOE'S SUPPORT ACTIVITIES
---------------------------------------------------------- Letter :5.4

DOE is responsible for the following major support activities: 
sampling and analyzing tank waste (characterization); providing
infrastructure, which includes roads, water, electricity, and
wastewater treatment; retrieving waste, which requires DOE to
retrieve waste from the tanks and deliver it to BNFL while keeping
the chemical makeup of the waste within specified ranges; and storing
and disposing of waste after processing, which requires DOE to
temporarily store the high-level waste and permanently store
low-activity waste.  DOE estimates that support activities will cost
about $2 billion including about $600 million for waste retrieval,
$40 million for characterization, and about $370 million for waste
storage and disposal.  Although support activities are essential to
project success, many of them are still in the planning stages, and
potential problems are not yet apparent.  At this time, the areas
that appear to be most prone to problems are waste retrieval and
waste storage and disposal. 

DOE's ability to successfully retrieve waste for processing depends,
among other things, on the availability of double-shell tank space
for storage and transfer activities.  According to DOE and contractor
officials, double-shell tank space is a major uncertainty because
double-shell tanks are also being used for other cleanup activities
at the site.  If the capacity needed for waste storage exceeds the
space available in the 28 double-shell tanks, DOE's ability to supply
waste to the private contractor for processing will be affected. 

The storage of immobilized high-level waste in Hanford's canister
storage building poses another risk to DOE's ability to successfully
support the project.  The risk is that the installation of equipment
and subsystems needed to store immobilized high-level waste in the
canister storage building could conflict with a schedule for another
DOE project--storing spent nuclear fuel removed from its current
deteriorating storage facilities. 

DOE's site support contractor concluded that these two problems have
a high risk of adversely affecting the project.  As a result, DOE
could have to make idle facility payments.  In response, the site
support contractor identified a set of mitigating actions that it
believes will reduce the risk that the problems will adversely affect
the project. 


      DOE'S FUNDING STREAM FOR THE
      PROJECT
---------------------------------------------------------- Letter :5.5

DOE's ability to fund the project within its own budget is an
important factor in ensuring that lack of funding does not lead to
project termination.  DOE estimates that it will need more than $10
billion in actual year dollars from fiscal year 1999 through 2017 to
fund the $6.9 billion project cost--an average of $537 million
annually.  During 7 of those years, payments to BNFL are expected to
exceed more than $1 billion per year. 

This funding represents a substantially increased need for funding at
the Hanford Site, where current annual budgets for all on-site
cleanup activities total about $1 billion.  If DOE is unable to
provide funding for the privatization project when needed, the
contract would likely be terminated, triggering DOE's liability to
pay BNFL for the amounts borrowed against the company's assets. 

DOE officials said that they did not yet have a detailed funding plan
for how they would find the additional funding within their budget. 
However, assuming no significant increase from the Congress, DOE
indicated that a major source of funds would likely be funding made
available when other DOE sites, such as Rocky Flats and Fernald, are
cleaned up and closed.  Given DOE's track record in completing
environmental cleanup projects, however, we are concerned that the
funds may not be available when they are needed. 

Another issue that could potentially affect DOE's ability to ensure
that sufficient funding is available for the project relates to how
the new contracting approach is classified in the budget.  Because of
budget limitations contained in the Budget Enforcement Act, cost
estimates are prepared for programs, including projects in DOE's
privatization program, to ensure that the limitations are not
exceeded.  Federal agencies such as DOE may acquire long-term assets
in several ways, and each acquisition strategy may be scored
differently.  For example, if the federal agency offered a federal
government guarantee to a private lender for a contractor's debt
financing, the agency would have to estimate the subsidy cost of the
loan guarantee.  This is a complex process and is based on the risk
of a default or nonpayment of the loans and other factors.  The
agency would then need the budget authority for the full net present
value of the subsidy cost before it could make the guarantee. 

Although the tank waste project is not structured as an explicit loan
guarantee, there is an increase in the government's potential
liability associated with making BNFL's loans an allowable contract
cost in the case of a default.  Neither DOE nor the Office of
Management and Budget has estimated this potential additional cost. 
This scoring is of consequence because it affects how much funding
DOE will have to have on hand for the project, and when. 


      INCONSISTENCIES WITH
      GUIDELINES FOR FIXED-PRICE
      CONTRACTS
---------------------------------------------------------- Letter :5.6

The remediation of Hanford's tank waste is a very costly, complex,
and risky effort regardless of the contracting approach DOE selects. 
In an effort to balance risks and realize cost savings, DOE selected
a fixed-price approach.  Federal acquisition regulation guidelines
note that fixed-price contracting works best when the possibility is
low for changes with cost and schedule implications.  However, the
BNFL contract cites at least 15 events, such as regulatory changes or
failure to provide waste on a timely basis, that could cause cost or
schedule increases.  The consequence of such changes is that they
would constitute a potential basis for adjusting the fixed price or
paying agreed-upon additional amounts. 

Federal guidelines state that another factor contributing to the
successful use of fixed-price contracting is competition, which helps
determine a price that minimizes the cost to the government while
providing a fair profit to the contractor.  DOE's revised approach
removes competition as a check on price.  Without competition, DOE
may not have the same assurance of obtaining the best value for the
negotiated price.  To compensate for the lack of competition, DOE has
required BNFL to provide certified cost or pricing information for
evaluating BNFL's basis for its proposed fixed-unit prices.  Our
reviews of DOE contracts have demonstrated that entering into a
fixed-price contract is no guarantee that DOE will be successful in
minimizing the cost to the government. 


   EFFECTIVE DOE OVERSIGHT IS
   CRITICAL TO PROJECT SUCCESS
------------------------------------------------------------ Letter :6

Managing this large, complex project presents a significant challenge
to DOE.  The agency's continuing challenge will be to translate the
plans it has made into sustainable oversight efforts that are capable
of overcoming problems that have plagued many past waste cleanup
projects. 

DOE has had difficulty managing other large projects.  Our past
reviews have shown a consistent pattern of poor management and
oversight by DOE.  For example, in our 1996 report on DOE's major
system acquisition projects (generally projects costing $100 million
or more), we reported that at least half of the ongoing projects and
most of the completed ones had cost overruns and/or schedule
slippage.\15 Some of the reasons for cost overruns and schedule
slippage included inadequate project oversight and insufficient
attention to technical, institutional, and management issues.  In
addition, our reviews of individual DOE cleanup projects such as the
Defense Waste Processing Facility at Savannah River, the Pit 9
cleanup at Idaho Falls, and the Spent Fuel Storage Project at Hanford
all identified problems with DOE's oversight activities as factors
contributing to project difficulties. 

At least in part to respond to these past difficulties, DOE has
developed several systems and processes to manage the tank waste
project at Hanford and has subjected its plans to outside review. 
Despite these efforts, however, outstanding issues concerning
technical staff, site support activities, and project administration
may keep DOE from being fully prepared to oversee the project. 

Technical staff:  DOE has established a team eventually expected to
number about 80 technical and managerial staff to oversee the
project.  This team, the Waste Disposal Division at Hanford, will
have authority both for managing the BNFL contract and for overseeing
the support activities of the contractors that provide day-to-day
management of the site.  As of August 31, 1998, the Division had
about 30 vacancies, including key staff such as the Deputy Project
Manager and five of nine DOE staff in the contract management group. 
DOE's Director of Contract Reform and Privatization said that the
Hanford unit does not have all of the technical skills necessary to
ensure success in overseeing the project.  He was especially
concerned about the shortage of contract expertise related to
administering fixed-price contracts.  According to DOE's contracting
officer at Hanford, none of the current DOE staff are experts in
fixed-priced contracting, although he hopes to be able to hire staff
with these skills soon.  Staff with these and other skills are needed
very soon because DOE will be negotiating critical details of the
contract, including the fixed price, during the extended design phase
of the project, which is currently under way.  DOE officials at
Hanford plan to fill the vacancies during fiscal year 1999. 

Site support activities:  Also critical to project success will be
the support that site contractors must provide in preparing
infrastructure improvements, retrieving waste, and removing and
storing the containers of vitrified material.  DOE must ensure that
Fluor Daniel, the main contractor managing the Hanford site, is able
to provide the support necessary for the project.  In August 1997,
DOE directed Fluor Daniel to conduct an extensive study (called a
"readiness to proceed") to determine what was needed to support the
vitrification effort and to identify potential problems.  At the
conclusion of the study, outside reviewers commissioned by DOE and
Fluor Daniel concluded that the support could be provided if adequate
funding was forthcoming.  However, DOE and tank farm officials said
that the project is funded at about $23 million less than needed for
fiscal year 1999.  DOE has requested full funding for fiscal year
2000, but the budget has not yet been finalized.  According to the
Director of the Waste Disposal Division, failure to fully fund
support activities in the next couple of years could delay the
project. 

Project administration:  Carefully administering the contract may
also be critical to ensuring that DOE and the contractor work
together effectively.  Our review of another large cleanup project,
DOE's Pit 9 project in Idaho, disclosed that substantially different
views about the degree of oversight, involvement, and interaction
that was appropriate for the contract were held by DOE, the
contractor managing the contract for DOE, and the subcontractor
responsible for carrying out the project.  Due in part to this
problem, the Pit 9 project has failed to meet cost and schedule
targets and the parties are now involved in lawsuits to settle their
claims. 

In part because of the Pit 9 failure, DOE paid considerable attention
to developing an approach to overseeing BNFL's operations.  For
example, to resolve procedural and other issues that may come up, DOE
required BNFL to establish four teams specifically covering project
management; contract and finance matters; interfaces; and
environment, safety, and health-related matters.  To help with the
complicated interrelationships between DOE and its contractors, DOE
has also followed a systems engineering process that involves using
"interface control documents" for those areas where DOE or the site
contractor have interrelationships with the BNFL contract.  Overall,
the project has 23 such documents covering areas such as
infrastructure, emergency response, and permitting. 

The contract also ensures DOE's access to key information.  For
example, BNFL will be conducting numerous tests to ensure that its
treatment processes will work.  The contract stipulates that BNFL
must deliver completed test reports to DOE for numerous activities,
such as validation of chemical processes, qualification of proposed
products, and effectiveness of a nonradioactive pilot melter. 

Finally, DOE has subjected its entire management process to both
internal and external review.  As a first step, in January 1998, DOE
issued a self-assessment report on the Division's readiness to
proceed.  The report concluded that the management systems were
behind schedule and identified six key actions to be taken, including
revising the project management plan and developing a staffing plan. 
DOE then convened independent assessments to look at various aspects
of the authorization to proceed.  As of September 15, 1998, 20 of 91
recommendations made by the reviewers were still open, including the
need to establish interfaces between the division and headquarters
organizations and requiring training in administering fixed-price
contracts. 

The potential problem here is not with DOE's efforts to date but with
its willingness to fully implement the oversight plans it has
developed for the project.  Our work over several years and on a
variety of DOE activities has disclosed a consistent pattern of DOE's
failure to fully implement the plans that it develops.  In a number
of instances, we have tied project or program difficulties directly
back to these failures to implement key management and oversight
components.  For example, in a 1993 testimony on the environmental
restoration management contract approach,\16 we found several DOE
management and oversight weaknesses, including not fully implementing
project management plans.  In a 1997 report,\17 we found that two
projects at the Fernald, Ohio, site had weaknesses, including
insufficient DOE oversight of the contractor, inadequate testing of
the technology, and delays in completing planning documents.  These
problems contributed to a $65 million cost overrun and almost 6 years
of schedule slippage.  More recently, in a review of DOE's management
of contaminated soils above the groundwater at Hanford,\18 we found
that although DOE drafted a management plan by 1994, it never
implemented the plan.  Four years later, after admitting that the
tank waste has leaked to the groundwater, DOE has still not
implemented a comprehensive management strategy. 


--------------------
\15 See Department of Energy:  Opportunity to Improve Management of
Major System Acquisitions (GAO/RCED-97-17, Nov.  26, 1996). 

\16 DOE Management:  Implementing the Environmental Restoration
Management Contractor Concept (GAO/T-RCED-94-86, Dec.  1, 1993). 

\17 Department of Energy:  Management and Oversight of Cleanup
Activities at Fernald (GAO/RCED-97-63, Mar.  14, 1997). 

\18 Nuclear Waste:  Understanding of Waste Migration at Hanford is
Inadequate for Key Decisions (GAO/RCED-98-80, Mar.  13, 1998). 


   CONCLUSIONS
------------------------------------------------------------ Letter :7

Remediating Hanford's radioactive tank waste will be difficult and
very costly.  In addition, given the nature of the tank waste and the
challenges associated with converting it to a more stable form for
long-term storage, the project involves substantial risk of
encountering problems that could result in further increases in
schedule and cost.  Because of contract clauses that provide for
adjustments in contract price, these risks are largely shouldered by
DOE.  Furthermore, in order to make private financing of the project
feasible, DOE has also decided to pay, as an allowable cost, any debt
costs that BNFL is unable to pay in case it defaults on its loans and
DOE terminates the contract.  As a result, DOE's potential liability
could amount to several billion dollars.  Given these circumstances,
it is important that DOE have in place a skilled project oversight
team fully prepared to face the challenges ahead.  However, problems
already exist in implementing the staffing plan and securing adequate
funding for the contract's support services.  We are concerned that
if these problems are not addressed quickly, they may adversely
affect the project.  Although a contract has been signed, the $6.9
billion is an estimated cost because design and financing issues have
not been finalized and the final unit price has not been set.  Since
this important information will be developed over the next 21 months,
the end of the design phase in August 2000 is another critical point
to assess the project before the most significant project costs are
incurred. 


   RECOMMENDATION
------------------------------------------------------------ Letter :8

We recommend that the Secretary of Energy take immediate action to
fully implement the Department's management and oversight plan for
the Hanford tank waste project, including ensuring that (1) the
oversight team is fully staffed with the expertise required and (2)
adequate funding is available to provide the site support services
called for in the contract. 


   MATTERS FOR CONGRESSIONAL
   CONSIDERATION
------------------------------------------------------------ Letter :9

Because key aspects of the project and the associated contract are
still being developed, significant changes in project cost, schedule,
and overall approach could occur before the price is set.  The end of
the extended design phase in August 2000 provides another decision
point at which these and other aspects of the project could be
reviewed before the most significant project costs are incurred. 
Therefore, the Congress may wish to require DOE to include in its
annual status report on privatization contracts (1) the results of
BNFL's technology demonstration and testing using Hanford's waste,
(2) a reassessment of the cost-effectiveness of the proposed approach
including the results of DOE's analysis of different financing
alternatives, and (3) DOE's overall preparedness to effectively
oversee the project. 


   AGENCY COMMENTS
----------------------------------------------------------- Letter :10

We provided a draft of this report to DOE for review and comment. 
DOE generally agreed with the report's conclusions and
recommendations and also provided several comments and technical
clarifications to the report.  The full text of DOE's comments and
our evaluation of them are included in Appendix I. 


   SCOPE AND METHODOLOGY
----------------------------------------------------------- Letter :11

To determine how DOE's current project strategy has changed from the
original proposal, we reviewed the contract between DOE and BNFL and
DOE's report issued to the Congress in July 1998.  We also reviewed
DOE, BNFL, and other contractor documents related to the tank waste
project.  In addition, we interviewed DOE and contractor officials at
Hanford and in Washington, D.C., to discuss the changes in the
contract. 

To determine how the project's schedule, cost, and savings estimates
have changed from the original proposal, we reviewed DOE's
documentation providing original schedule, cost, and savings
estimates and compared those to the current contract as outlined in
the report to the Congress.  We also interviewed DOE and BNFL
officials and outside financial experts to understand the changes
that have occurred to the project's schedule, cost, and savings
estimates. 

To determine the risks that DOE is now assuming with the change in
contract approach, we analyzed the contract, DOE's report to the
Congress, and DOE and contractor documents describing the risks to
the project.  We also interviewed those officials, in addition to
officials at the Office of Management and Budget and the
Congressional Budget Office to discuss the risks and budget issues
associated with the change in contract approach. 

To determine the steps that DOE is taking to carry out its
responsibilities for overseeing the project, we reviewed DOE's report
to the Congress describing how the Department is organized to manage
the project, as well as Department and contractor plans for managing
the project.  In addition, we reviewed internal and outside
assessments of DOE's efforts to oversee the project.  Finally, we
interviewed contractor and DOE officials, including the Director of
DOE's Office of Contract Reform and Privatization. 

Our review was performed from June through September 1998 in
accordance with generally accepted government auditing standards. 


--------------------------------------------------------- Letter :11.1

We are sending copies of this report to the Secretary of Energy.  We
will also make copies available to others on request. 

Please call me at (202) 512-8021 if you or your staff have any
further questions.  Major contributors to this report were Chris
Abraham, John Cass, Dwayne Curry, Doreen Feldman, Susan Irwin, Nancy
Kintner-Meyer, Tom Perry, Tim Schindler, Stan Stenersen, and William
Swick. 

(Ms.) Gary L.  Jones
Associate Director
Energy, Resources, and
 Science Issues




(See figure in printed edition.)Appendix I
COMMENTS FROM THE DEPARTMENT OF
ENERGY AND GAO'S EVALUATION
============================================================== Letter 



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)


The following are GAO's comments on DOE's letter. 

Regarding our comments on the lack of competition for the remainder
of phase I, DOE emphasized that competition played an important role
in reducing costs early in the project and that DOE had instituted
other controls on the BNFL contract to help control and reduce costs
such as requiring certified cost or pricing data.  DOE also
emphasized that its decision to back BNFL's debt was essential for
BNFL to obtain private financing for the project.  We believe our
report adequately states these views. 

Concerning the payment of up to $50 million that DOE will make during
the extended design phase, DOE stressed that the payment will be made
only for specific deliverables under the contract and that DOE does
not consider this to be a progress payment.  We clarified this in our
report. 

DOE said that it was important to mention that the revised project
schedule actually accelerated by 2 to 3 years the milestone for
beginning to treat high-level waste as set forth in the Tri-Party
Agreement.  We modified the report to clarify that some milestones
would be delayed while one milestone, the date to start high-level
waste processing, would be advanced. 

Regarding the changes in cost estimates, DOE said that the two
approaches involved processing plants of substantially different
useful lives and, therefore, capacity, which would ultimately bring
down the unit cost of operation.  We presented processing quantities
and costs only for phase I because the current contract is only for
phase I and BNFL intends to recover all of its costs during phase I
operations.  Costs related to phase II operations are not known.  DOE
also said the cost estimates for phase I should be compared on a
present-value basis because the two time frames are quite different. 
Our analysis does compare the two costs on a present-value basis, in
constant fiscal year 1997 dollars. 

DOE disagreed with our statement that direct federal funding of the
project may not significantly increase the government's level of
risk.  DOE said that direct federal funding would be accompanied by a
significant increase in the government's level of risk.  However, it
is not clear how much of an increase in risk there is between private
debt which is in effect guaranteed by the government and direct
federal funding of the project.  DOE did not quantify this difference
in risk or compare it to the significantly lower financing costs that
could be achieved through direct federal funding of the project.  Our
point was that DOE needs to conduct such an analysis in order to
ensure that it has a cost-effective approach to funding the project. 
DOE said that this analysis will be accomplished during the next
phase of the project. 

DOE said that our reference to a 1993 study by Independent Project
Analysis, Inc., was incomplete because we did not mention the reason
why cost growth on fixed-price contracts averaged 75 percent.  The
study found that fixed-price contracts were used on poorly defined
projects, which led to changes during construction contributing to
increased costs and schedules.  DOE asserted that a similar condition
does not exist on the Hanford tank waste contract.  We added to our
report the causes for cost growth identified in the study.  However,
we disagree with DOE's assertion that similar conditions do not exist
on the Hanford project.  Instead, the project risks and uncertainties
we discuss in this report increase the chances that the Hanford tank
waste project could also experience cost and schedule increases. 

Finally, DOE felt that we inaccurately stated the risks associated
with two DOE support activities--waste retrieval and storage of
high-level waste.  DOE said that Phase I waste will come almost
entirely from double-shell tanks and, therefore, tank capacity for
storage and transfer of the waste should not be a problem.  However,
sufficient double-shell tank space must be available to blend and
stage the waste before delivering it to BNFL in batches.  In May
1998, the Hanford Site support contractor determined that there was a
high risk of having inadequate tank storage capacity to meet the
needs of this project and at the same time support other site
initiatives such as the project to pump liquids from single shell
tanks into the double-shell tanks and the need to continue to receive
waste into the tanks from other Hanford activities.  As part of the
multi-year planning process, the contractor is currently reevaluating
this issue. 

Regarding the on-site storage of high-level waste, DOE is correct in
saying that the risk of having a problem is reduced by the extended
schedule.  However, DOE's contractor still described it as a high
risk in May 1998 due to the modifications required to the storage
building which will have to be made during the same years that the
spent fuel program is moving its canisters into the building. 

While DOE may be correct that potential waste retrieval and storage
problems can be managed, they clearly represent a risk to the project
and illustrate the need for an aggressive risk management effort. 


*** End of document. ***