Index

National Ignition Facility: Management and Oversight Failures Caused
Major Cost Overruns and Schedule Delays (Letter Report, 08/08/2000,
GAO/RCED-00-141).

Pursuant to a congressional request, GAO provided information on the
Department of Energy's (DOE) National Ignition Facility (NIF), focusing
on the: (1) magnitude of NIF's cost and schedule overruns; (2) reasons
for these cost and schedule problems; (3) effects of NIF's cost and
schedule on other weapons and science programs; and (4) DOE's and
Lawrence Livermore's actions to correct these problems.

GAO noted that: (1) DOE and Lawrence Livermore now estimate that NIF
will eventually cost about $3.3 billion and will be completed in 2008;
(2) these new estimates mean NIF will cost over $1 billion more than
originally planned and take 6 years longer to complete; (3) however, on
the basis of analysis of figures from DOE and Lawrence Livermore, GAO
estimates that NIF's cost is closer to $4 billion because DOE's estimate
does not include all research and development costs from other program
areas that are needed to support NIF; (4) furthermore, since significant
research and development activities to support NIF remain to be
completed and technical uncertainties persist, the cost of NIF could
grow even higher and completion could take even longer; (5) NIF's cost
increases and schedule delays were caused by a combination of poor
Lawrence Livermore management and inadequate DOE oversight; (6) since
NIF's beginning, the absence of effective independent reviews has
enabled the project's costs and schedules to grow undetected by DOE
program officials at headquarters and at Lawrence Livermore; (7)
furthermore, none of these reviews has examined both the construction
project and its supporting research and development activities; (8)
paying for NIF's cost overruns has broad implications for DOE's nuclear
weapons program; (9) the Secretary of Energy has said he wants to
complete NIF but will not ask Congress for additional appropriations to
pay for NIF's cost increases; (10) instead, he announced that the
Department will pay for NIF's overruns by reallocating funds from within
DOE's existing nuclear weapons budget; (11) however, DOE has not fully
disclosed which programs will be cut to pay for NIF, nor when or how
this will be done; (12) nor has DOE evaluated how this reallocation will
affect components of the nuclear weapons program that might be
eliminated, reduced in scope, or extended in order to fund NIF; (13) DOE
tried but was unable to secure agreement among its three weapons
laboratories that will use NIF--Lawrence Livermore, Los Alamos, and
Sandia; (14) consequently, DOE's June "interim" cost and schedule plan
was not developed by taking into account the many potential impacts on
DOE's nuclear weapons program; and (15) in addition, because DOE has not
determined how it intends to pay for NIF's cost overruns, the potential
impacts on other science programs are unknown.

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

 REPORTNUM:  RCED-00-141
     TITLE:  National Ignition Facility: Management and Oversight
	     Failures Caused Major Cost Overruns and Schedule Delays
      DATE:  08/08/2000
   SUBJECT:  Financial management
	     Contract oversight
	     Cost overruns
	     Schedule slippages
	     Laboratories
	     Test facilities
	     Nuclear weapons testing
	     Future budget projections
IDENTIFIER:  DOE National Ignition Facility Program
	     DOE Stockpile Stewardship Program
	     DOE Inertial Confinement Fusion Program
	     DOE Oakland Operations Office

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO Testimony.                                               **
**                                                              **
** 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/RCED-00-141

Appendix I: Scope and Methodology

36

Appendix II: Comments From the Department of Energy

37

Appendix III: GAO Contacts and Staff Acknowledgements

45

Table 1: NIF's Estimated Cost and Schedule Changes 10

Figure 1: The Completed National Ignition Facility 8

Figure 2: Timeline of Laboratory's Disclosure of NIF Problems 22

DOE Department of Energy

ICF Inertial confinement fusion

NIF National Ignition Facility

Resources, Community, and
Economic Development Division

B-284392

August 8, 2000

The Honorable F. James Sensenbrenner, Jr.
Chairman
The Honorable Ralph M. Hall
Ranking Minority Member
Committee on Science
House of Representatives

The University of California, under contract to the Department of Energy
(DOE) to operate DOE's Lawrence Livermore National Laboratory in California,
is building the National Ignition Facility (NIF) at the Laboratory. In this
stadium-sized laser facility, DOE's goal is to produce intense pressures and
temperatures that may, for the first time, simulate in a laboratory the
thermonuclear conditions created in nuclear explosions. If successful, NIF
would allow scientists to evaluate the behavior of nuclear weapons without
explosive testing.

DOE considers NIF an essential component of its multibillion-dollar
Stockpile Stewardship Program, which is responsible for ensuring the safety
and reliability of nuclear weapons in the absence of nuclear testing.1
Stockpile Stewardship entails building and operating science facilities in
addition to NIF, replacing aging laboratory and production plant
infrastructure, and refurbishing weapons in the stockpile. DOE has
experimental facilities to support Stockpile Stewardship in all three of its
defense laboratories: Lawrence Livermore, Los Alamos National Laboratory in
New Mexico, and Sandia National Laboratories in New Mexico and California.
In addition, NIF will be used to address basic and applied science issues in
areas such as astrophysics, materials properties, and fusion research.2

NIF was originally expected to cost about $2.1 billion when completed in
2002. In 1997, DOE approved a $100 million increase and delayed completion
until 2004. In September 1999, DOE announced that NIF would cost up to $350
million more and be completed in 2006. In the wake of these reports, you
asked us to

 determine the magnitude of NIF's cost and schedule overruns,

 document the reasons for these cost and schedule problems,

 assess the effects of NIF's cost and schedule on other weapons and science
programs, and

 evaluate DOE's and Lawrence Livermore's actions to correct these problems.

To address these objectives, we (1) examined the reviews of NIF conducted by
DOE and its advisory board, the University of California, and by Lawrence
Livermore Laboratory's advisers and (2) analyzed NIF cost data from DOE,
Lawrence Livermore, and other DOE laboratories conducting research in
support of the project. We also conducted extensive interviews with Lawrence
Livermore and DOE officials responsible for NIF and with officials from
other DOE laboratories and scientific institutions whose programs are
affected by NIF. Appendix I describes our scope and methodology in more
detail.

DOE and Lawrence Livermore now estimate that NIF will eventually cost about
$3.3 billion and will be completed in 2008. These new estimates mean NIF
will cost over $1 billion more than originally planned and take 6 years
longer to complete. However, on the basis of our analysis of figures from
DOE and Lawrence Livermore, we estimate that NIF's cost is closer to $4
billion because DOE's estimate does not include all research and development
costs from other program areas that are needed to support NIF. Furthermore,
since significant research and development activities to support NIF remain
to be completed and technical uncertainties persist, the cost of NIF could
grow even higher and completion could take even longer.

Our estimates of NIF's costs are based on DOE's "interim" report to the
Congress made on June 1, 2000. Last September, a congressional
appropriations conference committee directed DOE to present a NIF cost and
schedule plan with supporting details by June 1, 2000, or prepare to
terminate the project. But DOE said that it was unable to present a "final"
plan until mid-September 2000. DOE's "interim" report did not contain
details on NIF's cost and schedules.

NIF's cost increases and schedule delays were caused by a combination of
poor Lawrence Livermore management and inadequate DOE oversight. Troubles
began when Lawrence Livermore officials planned, and DOE approved, a NIF
budget and a construction cost contingency that were inadequate, virtually
ensuring that NIF would exceed cost and schedule estimates. The project
manager put in charge of NIF had little experience directing large projects
and had no control over separately funded laser research and development
programs that were essential for NIF's success. This resulted in a poorly
integrated management team. Furthermore, the Laboratory's former laser
director, who oversaw NIF and all other laser activities, assured Laboratory
managers, DOE, the university, and the Congress that the NIF project was
adequately funded and staffed and was continuing on cost and schedule, even
while he was briefed on clear and growing evidence that NIF had serious
problems. DOE's inadequate oversight contributed to these problems by
failing to uncover the increasing costs and delays in schedule until more
than 6 months after they were first documented within the Laboratory. A
number of key DOE program officials in headquarters and at the Laboratory
told us that while they suspected NIF problems earlier, they did not take
aggressive action to force the Laboratory to address these problems sooner.
Since NIF's beginning, the absence of effective independent reviews has
enabled the project's costs and schedules to grow undetected by DOE program
officials at headquarters and at Lawrence Livermore. Furthermore, none of
these reviews has examined both the construction project and its supporting
research and development activities.

Paying for NIF's cost overruns has broad implications for DOE's nuclear
weapons program. The Secretary of Energy has said he wants to complete NIF
but will not ask the Congress for additional appropriations to pay for NIF's
cost increases. Instead, he announced that the Department will pay for NIF's
overruns by reallocating funds from within DOE's existing nuclear weapons
budget. However, DOE has not fully disclosed which programs will be cut to
pay for NIF, nor when or how this will be done. Nor has DOE evaluated how
this reallocation will affect components of the nuclear weapons program that
might be eliminated, reduced in scope, or extended in order to fund NIF. DOE
tried but was unable to secure agreement among its three weapons
laboratories that will use NIF--Lawrence Livermore, Los Alamos, and
Sandia--about how, when, or at what cost NIF should be completed.
Consequently, DOE's June "interim" cost and schedule plan was not developed
by taking into account the many potential impacts on DOE's nuclear weapons
program. In addition, because DOE has not determined how it intends to pay
for NIF's cost overruns, the potential impacts on other science programs are
unknown.

While Lawrence Livermore and DOE have made changes to improve NIF management
and strengthen oversight, concerns remain about NIF's cost and technical
uncertainties, the absence of an independent review, and the impact of
paying for NIF's cost overruns on the Department's nuclear weapons program.
To ensure that DOE has an effective independent assessment of NIF, we
recommend that the Secretary of Energy arrange for an outside scientific and
technical review of the technical challenges remaining for NIF that could
affect the project's cost and schedule risks. To ensure an appropriate
balance between NIF and the rest of the nuclear weapons program, we further
recommend that the Secretary of Energy not reallocate funds from the nuclear
weapons program to pay for NIF until DOE (1) evaluates the impact of the
current cost and schedule plan, as well as any other options for NIF, on the
overall nuclear weapons program and (2) certifies that the selected NIF cost
and schedule plan will not negatively affect the balance of the Stockpile
Stewardship Program.

NIF's primary mission is to support DOE's Stockpile Stewardship Program,
which is intended to maintain the nation's nuclear arsenal indefinitely
through scientific research and periodic weapon refurbishment. In 1990, the
National Academy of Sciences recommended that DOE build a laser capable of
simulating thermonuclear conditions, which would allow nuclear weapons
scientists to evaluate nuclear weapons' behavior without testing. In 1993,
DOE approved plans for the laser, now called the National Ignition Facility,
and in 1996, chose to build it at Lawrence Livermore National Laboratory in
Livermore, California. DOE has two other weapons laboratories that also
perform essential research in support of maintaining the safety and
reliability of the nuclear weapons stockpile: the Los Alamos National
Laboratory in New Mexico and the Sandia National Laboratories in New Mexico
and California.

The NIF project was originally approved in 1995 at a cost $1.1 billion,
excluding another $1 billion in related research and development costs and
program funding, for completion in 2002.3 Increases approved in 1997 brought
the project cost to $1.2 billion, with completion in 2004, which is the
current approved baseline. DOE now estimates that NIF project costs will be
about $2.2 billion, with completion delayed until 2008.

The stadium-sized NIF (85 by 200 meters) will focus 192 laser beams on a
tiny capsule of nuclear fuel, compressing it by pressure. If NIF achieves
its ultimate goal, the compressed nuclear fuel will "ignite" to release more
energy than was added and simulate conditions during a thermonuclear
explosion. Achieving ignition is an important goal in certain types of
weapons physics experiments and is important for understanding basic fusion
science. Achieving ignition is not expected until 2010, or about 2 years
after NIF begins full operations. Originally, Lawrence Livermore estimated
it would achieve ignition by 2003.4

NIF, when completed, will be nearly 60 times as powerful as any previous
laser. As shown in figure 1, NIF has many different components. It is
composed of a conventional facility, or building; the laser beam path
infrastructure, which includes supporting structures, vessels, cabling, and
utilities; the laser, which includes thousands of glass and optical
components; and a target chamber surrounded by lenses known as "final
optics."

Source: DOE.

NIF is managed by DOE's Inertial Confinement Fusion Program, which is part
of the agency's Office of Defense Programs.5 This office conducts essential
research for the NIF effort. According to DOE plans, about 85 percent of the
facility's experiments will be for nuclear weapons physics. The remaining
experiments will be for nuclear weapons effects and basic and applied
sciences. DOE's Oakland Operations Office is responsible for day-to-day
oversight of NIF and is expected to ensure that the Laboratory is meeting
its contracting responsibilities. The Oakland office also prepares the
Laboratory's annual performance appraisal. For the NIF project, Oakland
staff also review monthly and quarterly Laboratory progress reports and
transmit them to DOE headquarters for review. The University of California
manages the Laboratory for DOE and reviews Laboratory performance annually
as part of its oversight role.

Using preliminary projections by DOE and Lawrence Livermore, we estimate
that the total cost of building NIF is about $3.9 billion, when all
supporting research and related program activity costs are included.
Moreover, because significant research and development remains to be
completed, the cost of completing NIF could grow even higher. DOE has not
provided details on how much NIF will cost each year, how cost increases
will be paid for, or what existing programs and activities will be cut to
fund NIF in the future. DOE knew in June 1999 that a new NIF cost and
schedule was needed, and a congressional appropriations conference committee
directed DOE to present a new cost and schedule plan by June 1, 2000.
However, DOE provided only "interim" estimates on June 1, 2000, and
announced it will give the Congress a "final" cost and schedule plan by
mid-September 2000.

Delayed 6 Years

We estimate that, on the basis of DOE and Laboratory data, NIF's total cost
could reach at least $3.9 billion. Although DOE's June 2000 "interim" NIF
cost estimate is about $3.3 billion, we identified additional costs of over
$600 million that directly support the completion of NIF and should be
considered part of its total costs. As shown in table 1, DOE's latest
estimate includes $2.12 billion for total project costs (funds for
construction activities associated with the NIF building and its contents)
and $1.14 billion for research and development and related program funds
that directly support NIF (drawn from DOE's Inertial Confinement Fusion
Program).

Although DOE officials included most of the supporting research and
development and related program funds as part of the Department's June 2000
"interim" estimate, they did not include all supporting costs. Specifically,
they excluded costs for designing and fabricating a NIF target and funds
directly contributing to the NIF effort at other laboratories and
contractors.6 DOE officials told us that funds for developing a NIF target
are more of a weapons research concern and therefore should be considered
outside the program. Since these funds are necessary for NIF's success, we
believe they should be included in any discussion of total NIF costs. The
costs for designing and fabricating a NIF target ($491 million) and NIF
support costs from other laboratories and contractors ($136 million) should
be added to DOE's estimate, according to our analysis of Laboratory costs
and discussions with the individual laboratories contributing to the NIF
effort. As shown in table 1, this would bring the total estimated cost of
completing NIF to about $3.9 billion.

 Dollars in billions
                              Supporting research
 Baseline change     Project  and development and   Total   Fiscal year of
                     costs    related program       costs   completion
                              funds
 1995 Original
 baseline            $1.07    $1.00a                $2.07   2002
 1997 Baseline
 change              1.20     1.04a                 $2.24   2004b
 2000 Estimates--DOE
 (6/1/00)            2.12     1.14                  $3.26   2008
 2000 Estimates--GAO
 (6/1/00)            2.12c    1.77d                 $3.89   2008e

a1995 and 1997 costs are GAO estimates based on available Laboratory data.

b DOE defined completion in the 1997 baseline as occurring in fiscal year
2003, when half of NIF's beams become operational. This baseline assumed all
192 beams would be operational in fiscal year 2004.

cWe did not assess the accuracy of DOE's estimates of project costs ($2.12
billion), supporting research and development (R&D) and related program
funds ($1.14 billion).

dIncludes the following additions to DOE's 6/1/00 estimate: $491 million for
target physics and $136 million for direct support of NIF by other DOE
laboratories and inertial confinement fusion contractors.

eWe did not assess the accuracy of DOE's revised completion date.

Unresolved technical problems may further drive up the cost of NIF. A
November 1999 review by a technical subgroup to Lawrence Livermore's
advisory panel (the NIF Council7) highlighted several technical problems
that must be solved for NIF to operate as planned.8 The most challenging
technical issue raised by the subgroup is the need to develop optical
components that can withstand the laser's powerful beams as they reach the
spherical target chamber. When operating, the laser beams focus on a system
of glass lenses that surround the target chamber. These lenses must be able
to withstand high levels of intense ultraviolet energy. Several Laboratory
officials agree that lenses for the target chamber pose a major technical
challenge and that there is currently no solution to this problem. So
serious is this issue that the subgroup recommended that NIF scientists
should be thinking "out of the box" for design alternatives. A May 2, 2000,
technical review by another NIF Council subgroup also noted that the optical
lenses remain an unresolved issue. Unless the Laboratory can fabricate
lenses to withstand the enormous energy from the laser beams, the lenses
will have to be replaced more often, making laser operation at high power
levels much more expensive than planned.

Substantial technical and cost uncertainties also persist in the research
and development needed to design and build a target for NIF's laser beams,
according to the NIF Council and its subgroups. Still to be determined are
physics and engineering issues for designing and fabricating a NIF target.
The NIF Council's May subgroup report noted that fundamental work also
remains on this technical issue.

The NIF Council and its subgroups have all concluded that the difficult
technical issues facing NIF pose serious challenges but should not present
insurmountable obstacles to completing the project. Laboratory officials
said that they are confident that all these difficult technical challenges
can be solved and noted the Laboratory has met other NIF technical
challenges successfully.

In its January 2000 interim report on NIF, the Secretary of Energy Advisory
Board NIF Laser System Task Force also noted the technical issues cited by
the NIF Council and acknowledged that technical issues still remain that
will require significant attention. However, it stated that these issues
should not prevent the completion of NIF.9 Although both the NIF Council and
the Advisory Board's task force expressed confidence in the Laboratory's
ability to solve its remaining technical problems, neither group speculated
on the risks posed by the many technical uncertainties, nor estimated their
impact on future costs and schedules.

The details of DOE's June 1, 2000, NIF cost estimates are sketchy, even
though DOE senior officials learned in late June 1999 that NIF was over
budget and behind schedule and that a new cost and schedule plan was needed.
In its September 1999 report on energy and water development appropriations
for fiscal year 2000, the conference committee directed DOE to submit a new
cost and schedule by June 1, 2000, after declaring "disappointment" at
learning of NIF's cost overruns and schedule delays.10 The committee also
directed that if DOE could not prepare a new cost and schedule by this date,
it should develop a plan for terminating the NIF project. However, DOE's
June 1, 2000, report provided only "interim" figures for a new NIF cost and
schedule and no termination plan. DOE did not provide annual funding
profiles for its cost and schedule plan, stating only that it will seek "up
to" specific amounts in fiscal years 2002 to 2005 ($130 million to $150
million), and "additional funds as needed" to complete the project in later
years. DOE officials also did not disclose the programs from which NIF
increases will be funded or the impacts of these changes. DOE promised to
submit a "final" plan to the Congress by mid-September 2000. This final
plan, according to DOE, will include cost and schedule estimates with
supporting details on annual funding requirements, funding sources, and
deadlines, and an independent cost analysis. On June 27, 2000, DOE submitted
a revised budget request for fiscal year 2001 that provides more cost
information but does not give many details beyond fiscal year 2001.

Schedule Delays

NIF's cost increases and schedule delays were caused by poor Laboratory
management, which included weaknesses in planning, budgeting, and project
control. DOE's oversight weaknesses contributed to cost and schedule
problems, and coupled with no effective independent review of NIF, allowed
NIF problems to go undetected. Standards we have developed require federal
agencies to establish and maintain an effective system of internal controls
over their operations.11 Such a system is the first line of defense in
safeguarding assets and preventing and detecting errors. Under our
standards, managers should, among other things, ensure that their staff have
the required skills to meet organizational objectives, that the
organizational structure clearly defines key areas of authority and
responsibility, that progress be effectively measured, and that operations
be effectively monitored. These conditions were not present in the way NIF
was managed by the Laboratory and overseen by DOE.

Laboratory managers told us that NIF will cost more and take longer because
they (1) had not planned for the complexities associated with assembling and
installing the highly intricate and tightly packed network of laser beams in
a structure that requires high cleanliness standards and (2) underestimated
the cost of building major laser components. These results occurred,
according to Laboratory officials who examined the project, because the
original NIF managers lacked adequate project management capabilities and,
in particular, had no systems engineering focus within the NIF management
hierarchy. As a result, these managers did not properly identify the full
scope of the NIF effort and had greatly underestimated the project's
engineering complexities. On the basis of our discussions with current and
former DOE officials, Laboratory managers, and other experts, we believe
several additional reasons contribute to NIF's cost overruns and schedule
delays. For example:

 DOE and Laboratory officials initially developed a budget for the NIF
project that was clearly inadequate, given its technical risks, complexity,
and large size. DOE and Laboratory officials associated with NIF told us
that they recognized it would cost more than planned but that they accepted
this unrealistic budget in the belief that the Congress would not fund NIF
at a higher cost and that the value of NIF to the future of the Laboratory
overshadowed potential cost concerns. Current and former DOE and Laboratory
officials also told us that NIF would help the Laboratory maintain its
nuclear weapons and science capabilities and attract new scientists.

 DOE and Lawrence Livermore used a construction budget contingency of 15
percent, which is now widely recognized by DOE and the Laboratory as being
too low for such a complicated project. Contingencies are an essential
planning factor in any construction project, allowing for unanticipated cost
increases. The NIF Council and an independent engineering firm indicated
that a 25- to 35-percent contingency in the beginning of the project would
have been more realistic given the risks associated with NIF. DOE guidelines
suggest a 20- to 40-percent contingency, depending on project complexity.

 After the project was approved, Laboratory officials began construction in
tandem with essential research and development. While such a "fast-track"
strategy is successful on projects with well-defined costs and designs and
with little technical complexity, the NIF project does not meet these
conditions. Starting construction early can save costs initially but is more
costly if extensive changes are needed later. For NIF, the fast-track
approach may have caused the Laboratory to focus too much attention on
meeting construction goals at the expense of conducting and integrating
necessary research and development solutions. DOE and Laboratory officials
acknowledge that this approach affected basic construction decisions on the
size and design of the facility and eventually led to cost increases and
delays when changes had to be made to accommodate the results from
subsequent research and development.

 According to Laboratory officials, to cut costs early in the project, they
considered, but rejected, hiring outside industrial engineering firms to
manage major parts of NIF construction. Such firms could have provided the
experience to manage this large project, especially for assembling laser
components with exacting cleanliness standards, such as those in the
semiconductor assembly industry. The Laboratory is now contracting with such
firms to help complete NIF.

Unrealistic budgets, a low contingency, and the fast-track construction
strategy imposed substantial cost pressures that affected project management
decisions. The NIF Council criticized Laboratory officials for failing to
develop engineering prototypes that could have tested various NIF components
and technologies before planning to assemble them--a strategy rejected, in
part, because of cost pressures. For example, according to the Council, to
save money, Laboratory managers dismantled a laser scientific
prototype--called the Beamlet--that would have been useful for conducting
experiments on technical problems now affecting NIF.

After DOE approved NIF in 1993, the Laboratory did not appoint project
managers with the management and technical expertise needed for such a large
and complex undertaking and failed to install an effective reporting
structure that could have flagged emerging technical issues and project
risks. For example:

 The NIF Project Manager originally selected had no previous experience
managing large projects, and other managers had little or no expertise in
lasers, hindering their ability to understand the significance of some
problems as they arose. Furthermore, the project lacked a chief engineer,
whose responsibilities could have focused on integrating the construction
project with needed research and development efforts.

 The NIF Project Manager held a relatively low position in the Laboratory's
organization. Despite the NIF project's high profile and its critical
importance to the future of the Laboratory and to DOE's Stockpile
Stewardship Program, the NIF Project Manager reported to a Deputy Associate
Director who, in turn, reported to the Associate Director for Laser
Programs, who then reported to the Director of the Laboratory. A separate
manager was in charge of the inertial confinement fusion (ICF) research and
development programs, which were crucial to the success of NIF. The NIF
project manager and the ICF research manager each reported on different
milestones. No one coordinated the two efforts or estimated how one
program's missed milestones could affect the other program or the overall
NIF effort. Both the project and research and development managers filed
separate progress reports, and no one at the Laboratory or DOE assessed the
impacts of problems and risks on projected budgets and schedules.

 Senior officials from Lawrence Livermore, the University of California,
and DOE concluded that the NIF managers were overconfident about their own
abilities to solve project problems. These officials now acknowledge that
they should not have placed as much trust in the NIF managers and that the
Laboratory's "can-do" culture prompted the NIF managers to convince
themselves, then DOE and others, that they needed no outside help to
successfully build NIF. The Laboratory's former top laser manager embodied
this self-reliant attitude and was very persuasive in convincing others that
NIF would be built within cost and schedule. In its review of NIF, the
Secretary's Advisory Board task force reported that DOE's weapons
laboratories "have a tendency not to flag problems to the outside world,
where they actually might find counsel and help, but rather prefer to go it
alone for periods too long."

DOE acknowledges that its managers in headquarters and at the Laboratory
site office did not properly oversee NIF and, as a result, remained unaware
of major cost and schedule problems until several months after Laboratory
managers had first documented them. As a result, the Laboratory continued to
present--unchallenged by DOE--an inaccurate picture of NIF's progress.

Several reasons help to explain DOE's oversight weaknesses. At the core of
DOE's problems is its historical reliance on contractors to conduct the
Department's missions in the absence of an effective oversight process and
structure. For example, in response to the charge by outside reviewers of
DOE's "micro-management," the Department, in the mid-1990s, shifted from a
"compliance-based" to a "performance-based" management style. DOE's Oakland
Operations Office, which is directly responsible for NIF, explained to us in
1997 that it had changed its "approach to oversight." The old approach
"relied on DOE identifying problems and solutions" while the new approach
"relies on [the] contractor identifying problems and solutions."

In addition, the Department lacked staff with sufficient management and
technical skills to properly oversee NIF's managerial and technical
complexities. Neither DOE headquarters staff nor field managers were skilled
in managing large projects, nor did the field staff have technical
proficiency in laser operations. Also, these staffs' oversight roles were
diluted by other competing responsibilities--neither headquarters staff nor
field managers were full-time NIF overseers. Only four DOE on-site technical
staff worked on NIF oversight, along with one senior headquarters manager,
and each had other duties.

To gauge NIF's progress, DOE relied heavily on Laboratory-prepared monthly
and quarterly NIF reports. However, the reports did not reveal emerging cost
and schedule problems, even though they required an assessment of project
risk. These progress reports conveyed project status for senior DOE
officials, based on criteria developed jointly by DOE and the Laboratory.
The progress reports focused mainly on meeting construction and research
milestones, not on linking specific successes and failures to impacts on
future milestones, the budget, or the overall project schedule. The reports
also did not identify significant technical or managerial risks facing the
project, even when these were recognized by senior project managers.
Moreover, the Laboratory downplayed potential problems in its reports,
hoping that as problems emerged they would be resolved quickly.

There was not a clear chain of command between DOE's headquarters and its
on-site field office at Lawrence Livermore, thereby diffusing accountability
for NIF. DOE's Defense Programs funded NIF and assigned a project director
in headquarters, who provided Laboratory managers with programmatic
direction. Day-to-day supervision of NIF, however, was assigned to a project
manager from DOE's field office located on site at Lawrence Livermore.
However, this field office manager reported to another part of DOE--the
Office of Science--and not to the project director in Defense Programs.12
While a 1997 memorandum of understanding defined the field project manager's
responsibilities, it left unchanged the misdirected reporting to Science
rather than Defense officials in Washington, D.C., who were responsible for
the project. Laboratory officials said that they considered DOE's chain of
command confusing and really did not know to whom they reported on a
day-to-day basis. Because of this confusion, the DOE field manager had
difficulty in taking active control of a project that fell outside of his
chain of command. When suggesting revisions to the memorandum of
understanding in 1998, DOE's field manager noted that the memorandum
continues to serve a useful purpose in the "absence of consistently accepted
understanding of the roles of HQ and the Field."

While DOE managers were hampered by their own technical, managerial, and
organizational weaknesses, they missed opportunities to enforce Laboratory
accountability. For example:

 DOE officials did not aggressively act on their own suspicions. DOE field
and headquarters officials told us they started suspecting NIF cost and
schedule problems early in 1999 but then allowed the Laboratory's monthly
and quarterly reports, which were reviewed and approved by on-site DOE
staff, to continue reporting satisfactory progress until June 1999. These
DOE officials, including both headquarters and field staff, did not
aggressively pursue their suspicions by demanding more information in
progress reports or by alerting senior DOE officials to potential problems.
When the Secretary of Energy visited the Laboratory in early June to
dedicate the NIF target chamber--and declared NIF to be within budget and on
time--many DOE staff already knew that NIF had cost and schedule problems.

 DOE's Office of Field Management, a headquarters office that was
responsible for, among other things, preparing status reports for senior DOE
managers on large construction projects, raised management issues regarding
NIF that were largely ignored by NIF program managers in headquarters. In
June 1999, as a result of growing concern with NIF, this office proposed a
management action plan for NIF that recommended DOE track eight different
elements, including status measures, risk assessments, pending actions, and
completion dates. The office was abolished in June 1999 as part of a broader
DOE reorganization. Its functions were transferred to other parts of DOE,
but the recommendation was never acted upon.

 DOE negotiated a contract with the University of California that placed
little emphasis on NIF performance, even though the project dominated the
Laboratory's budget and mission. Also, DOE's evaluation of the Laboratory's
performance did not reflect NIF cost and schedule problems. As late as
August 1999, the Laboratory's self-evaluation claimed "outstanding"
performance for the Laser Directorate, which included NIF. This score
contributed to an overall DOE score for the Laboratory of "excellent." The
Laboratory's performance fee is based on these scores.

 DOE did not enforce the requirement that the Laboratory update its
estimates for the project's total cost and schedule in its monthly and
quarterly reports. Enforcing this requirement could have revealed growing
cost and schedule problems months earlier, giving DOE and Laboratory
officials more time to prepare a new baseline.

The absence of effective independent reviews to ensure that project risks
were recognized and addressed in a timely and effective manner was a major
flaw in how DOE and the Laboratory managed NIF. Attempts to provide
independent reviews all had limitations. For example:

 In 1996, the Laboratory established the NIF Council as an advisory group
to provide Livermore staff with technical advice; however, this group was
not independent. Members were chosen by the Laboratory's senior laser
manager, who also prepared its charter, and received and responded
personally to its reports. Some NIF Council members described themselves to
us as "friendly advisers" rather than as independent overseers. Although the
Council provided valuable advice on wide-ranging technical issues, its
members were unaware of the project's cost and schedule problems for many
months. Also, the NIF Council was not aggressive in providing managerial
advice, although this was one of its charter responsibilities.

 To fulfill a congressional request that DOE contract for independent
reviews of its projects, a NIF review was conducted in early 1999, but the
contractor performing the review did not report any significant cost,
schedule, or technical issues for NIF.13 At the time of this review, over
two dozen senior NIF managers knew the project faced growing problems that
threatened both its costs and schedules, but these concerns were not shared
with the reviewers. The contractor's report concluded that NIF was "by far
the best managed of any U.S. Government Project" that the contractor had
reviewed. DOE's actions may have compromised the contractor's independence.
For example, DOE imposed strict time limits on the contractor, which
prevented a detailed review. DOE also instructed the contractor to not
examine NIF supporting research and development, which was a vital part of
the NIF program. We discovered that a DOE manager and the Laboratory's NIF
Project Manager edited the contractor's draft report before it was submitted
to the Congress. For example, they suggested changes to the final report
that had the effect of downplaying the extent to which some research and
development tasks remained unresolved and posed cost and schedule risks.
Finally, DOE's on-site office at Livermore urged headquarters to discourage
the contractor from making a second visit to the Laboratory.

of California

The failure of senior Laboratory officials to disclose technical issues that
threatened the NIF project's costs and schedules was a serious error in
judgment, according to Laboratory and DOE officials we interviewed. Many
Laboratory managers were aware of serious NIF problems months before these
became public, yet they told us that they were not in a position of
responsibility to report the problems outside the Laboratory. When asked why
he did not disclose that a consensus of his senior managers believed NIF
could no longer be completed within current budgets and schedules, the
Laboratory's highest ranking laser manager responded that he first wanted to
"validate" the extent of the problems before reporting them to the
Laboratory Director, the University of California, or DOE. Even after these
problems were validated, the laser manager delayed disclosing NIF's cost and
schedule problems. As figure 2 shows, technical issues affecting NIF's cost
and schedule were first raised in the summer of 1998, were documented in
early December, and were more completely documented and presented to the
Laboratory's laser director in March 1999. Yet DOE was not formally told
about the significance of these issues until late June 1999. The Secretary
of Energy was not made aware of the cost and schedule problems until August
1999, 2 months after he and DOE's Chief Operating Officer had publicly
praised NIF's budget and schedule successes.

Programs

NIF's primary mission is to support DOE's nuclear weapons program--also
known as the Stockpile Stewardship Program. This program, which costs $4.5
billion annually, is intended to maintain the nation's nuclear arsenal
indefinitely through scientific research and periodic refurbishing of
weapons. Three different weapons laboratories--Livermore, Los Alamos, and
Sandia--conduct significant weapons-related scientific research. Paying for
NIF's cost overruns from within DOE's current nuclear weapons program
accounts--as the Energy Secretary has pledged--could significantly affect
portions of these laboratories' programs, as well as the Stockpile
Stewardship Program as a whole. In developing its new cost and schedule
estimate, DOE did not take into account which Stockpile Stewardship Program
components might be eliminated, reduced in scope, or extended in order to
fund NIF.

DOE's Stockpile Stewardship Program is intended to maintain the nation's
nuclear weapons in a safe and reliable state through a program of scientific
study and periodic refurbishing of the weapons. Currently, the scientific
study of nuclear weapons is organized around the concept of "campaigns." DOE
describes campaigns as focused scientific and technical efforts to develop
and maintain critical capabilities needed to certify that the nation's
nuclear weapons are safe and reliable. Campaigns have definitive milestones,
specific work plans, and specific end dates. All campaigns have milestones
for completing scientific studies by 2005, although some will be completed
later, up to 2010. NIF experiments would be useful for a number of
campaigns, including the campaigns to (1) certify the yield (explosive
energy) in a nuclear weapon's "primary," (2) determine what is needed to
produce a militarily effective nuclear weapon's "secondary," and (3) certify
that a nuclear weapon can withstand a hostile (i.e., wartime) environment
during use.14

The other key feature of the Stockpile Stewardship Program is the effort to
extend the lifetime of the nation's nuclear weapons through a series of
planned refurbishments. These refurbishments will involve disassembling
weapons and replacing selected components in order to extend each weapon's
life for several more decades. Refurbishments for two weapons in the
stockpile--the W76 and the W80--are scheduled to begin later this decade.
NIF could be useful to the refurbishment process by allowing weapons
scientists to study the potential effect of proposed changes on a weapon's
safety and reliability.

Stockpile stewardship is being carried out as an integrated program
involving the three nuclear weapons laboratories, four production plants,
and a test site. Integration can occur on several levels. For example, the
three laboratories are expected to share experimental facilities, like NIF,
in order to reduce the overall cost of the program. In addition to sharing
facilities, the DOE laboratories work together to develop an annual
certification to the President and the Congress that the nation's nuclear
weapons remain safe and reliable without the use of underground testing.

DOE and its three weapons laboratories believe that NIF is a major component
of the Stockpile Stewardship Program. Experiments conducted on NIF are
expected to assist in the work of the campaigns as well as in the
refurbishment of the stockpile. NIF is also expected to help attract and
train new weapons scientists to replace the program's currently aging
workforce.

In general, the delays NIF will experience will not affect meeting the
campaigns' 2005 milestones since NIF was not expected to be ready in time
for those. Some weapons scientists told us that, even before NIF's problems
became public, they had already planned alternative experiments in case NIF
was not ready on time. However, according to program documents, NIF's delays
will in turn delay the achievement of some of the experiments needed for the
2010 milestones by as much as 3 years. With respect to weapons
refurbishment, the expected delays in NIF will mean that it will not be
available to help analyze the impact of any proposed changes in the W76 or
W80 weapons.

Many important experiments can still be conducted on other facilities or on
a partially completed NIF. For example, studies of the high-energy density
properties of materials used in nuclear weapons can be conducted with only
48 to 96 beams installed in NIF (NIF is designed for 192 beams). This
information is vital in modeling the performance of nuclear weapons.
Livermore's Target Physics Review Committee, a NIF Council technical
subgroup, recommended in May 2000, that while NIF should be completed to its
full 192-beam configuration, a first cluster of 48 beams should be given a
high priority and brought up for experiments as soon as practical, both for
weapons physics experiments and to attract and retain the best scientists.
The former Associate Director for Defense and Nuclear Technologies at
Lawrence Livermore told us that while he believes not having NIF available
may introduce some risk into the stockpile, this increased risk does not
approach any level that would indicate that the stockpile is unsafe and
unreliable.

In a September 1999 statement responding to the problems NIF was
experiencing, the Secretary of Energy said,

"I expect all cost issues to be handled within our DOE defense programs and
Lawrence Livermore National Laboratory (LLNL) budget funding lines. We will
reprioritize our national security program to reallocate dollars, people,
and other resources--so the U.S. taxpayer does not foot additional bills
because of these problems."

Assuming the Secretary enforces his pledge to fund NIF's cost overruns from
the Stockpile Stewardship Program's current budget, significant portions of
the nuclear weapons program could be affected. Currently, several DOE
officials and veteran weapons experts believe the program is under
significant stress as numerous activities compete for its limited budget.
These activities include building and operating additional science
facilities beyond NIF, replacing aging laboratory and production plant
infrastructure, and refurbishing weapons in the stockpile. This stress is
expected to increase as the result of increased security requirements,
potential new problems with the stockpile, and the need to address the
program's aging scientific and production workforce.

DOE has not revealed how it intends to pay for NIF's cost overruns, and Los
Alamos and Sandia officials are concerned about how NIF's cost increases
will affect the Stockpile Stewardship Program at their laboratories. DOE
attempted to gain consensus among the weapons laboratories on a path forward
for NIF. An April 2000 "White Paper" signed by DOE and its three weapons
laboratories concluded that NIF was important to the Stockpile Stewardship
Program, but no agreement was reached on how many beams NIF should have or
how it should be funded. The paper's authors explicitly stated that the
White Paper

"does not address the appropriate revised scope for NIF, the schedule for
deployment, the proper cost baseline, or the impact on the balance within
the Stockpile Stewardship Program. Consequently, concurrence with the views
stated here should not be construed as support for any specific path forward
or for other yet to be resolved issues."

This lack of agreement regarding how NIF should be developed and funded was
echoed in comments made to us by senior weapons scientists at the three
laboratories. For example:

 The Associate Director for NIF Programs at Livermore believes that NIF is
critical to certifying the nuclear weapons stockpile and supports completing
a full 192-beam NIF. In discussions with us, he recognized that NIF is not
the only essential facility in the Stockpile Stewardship Program and agreed
that the rest of the program should not become unbalanced by funding NIF. He
believed Livermore might accept a "pause" in completing NIF of about 3 years
at the 120-beam level.

 The Deputy Director of the Nuclear Weapons Program at Los Alamos told us
that he believes NIF should be stopped at 48 beams. This level of NIF would
support about 80 percent of the experiments the program needs to perform. He
believes that funding a larger NIF would seriously unbalance the stewardship
program by taking funds away from needed research and production activities.
Other Los Alamos weapons scientists also expressed the view that NIF had
only a 50-percent chance of achieving ignition.

 The Senior Vice President for Nuclear Weapons Programs at Sandia National
Laboratory told us he believes there is no clear role for NIF that requires
it to be built to full power immediately. He believes that NIF should be
built to a smaller scale--such as 48 or 96 beams--and operated for some
period of time before additional investments are made. Taking this approach
would allow the NIF program to demonstrate with certainty how much it costs
to build and operate NIF.

Senior DOE officials told us that when such fundamental disagreements
existed in the past, they would hold a "summit" to reconcile the views.
However, according to our interviews with senior DOE officials, the
Secretary did not reconcile these views before he made his decision
regarding the "interim" cost estimate (rebaseline). The Secretary's
rebaseline decision was not made in the context of which stewardship program
components might be eliminated, reduced in scope, or extended in order to
fund NIF. Rather, the Secretary relied on overall funding options. None of
the various rebaselining options considered by the Secretary described what
the specific effects would be on the overall stewardship program for fiscal
year 2001 and in the future. DOE is still trying to determine how to adjust
the Stockpile Stewardship Program for fiscal year 2001 and beyond to pay for
NIF.

The Secretary's pledge notwithstanding, some scientists believe that science
programs are vulnerable to budget cuts. DOE sponsors a wide variety of
science activities in such areas as high-energy physics, advanced computing,
and fusion. Scientists at several universities told us that they believe
funding for civilian research may be cut as a result of NIF cost overruns.
Several scientists pointed out that each fusion energy experiment may take
several years to plan, design, and conduct, requiring a long-term commitment
for funding. Budget cuts would delay the development and implementation of
these experiments--the very research that in 1990 the National Academy of
Sciences recommended NIF be designed to conduct.

Scientists expressed mixed views about NIF's effect on science research
schedules. While some scientists told us that schedule problems may delay
needed experiments on NIF, others said that delays could actually be
beneficial. Delays could provide additional time to observe the initial
operations of NIF and to better design and modify planned experiments.

DOE, Laboratory, and University of California officials are taking several
actions designed to strengthen the management and oversight of NIF. Some
actions are complete, but most are still under way. Many of these actions
address specific deficiencies in existing processes and organizational
structures. These actions are encouraging, but implementation has only just
begun in many areas, so an assessment of their ultimate value is premature.

The Secretary of Energy has taken several actions to establish
accountability for NIF's problems and to improve DOE's oversight. Stating
that he was "deeply disturbed" to learn of NIF's cost and schedule problems
and "gravely disappointed" with both the Laboratory and the University of
California, he ordered DOE to withhold at least $2 million of the
University's 1999 program performance fee in recognition of the "significant
mission disruption caused by the project's problem." The Secretary also
ordered a report, due later in the year, that can recommend disciplinary
action against DOE officials responsible for NIF's mismanagement. Other
actions announced by the Secretary include the following:

 creating a new headquarters NIF Project Office whose manager is dedicated
to NIF,

 assigning six full-time employees to oversee NIF on site at Lawrence
Livermore, and

 placing NIF on the Department's Chief Operating Officer's list of projects
that require senior management attention (watch list), where it is subject
to more stringent reporting requirements.

NIF has also been designated a pilot project for the new Project Management
and Oversight function. DOE intends to hire an outside contractor who will
give DOE project management expertise on-site at the Laboratory. This
contractor will be paid from the $2 million fee that the University was
required to return to DOE. Although not yet fully in place, other key
features of this new function include the following:

 creating a headquarters Office of Engineering and Construction Management
to provide guidance and expertise on project management,

 developing a career track for project managers, long a weak link in DOE's
ability to attract and hire highly qualified project managers, and

 requiring and budgeting for more research and development before
committing to major construction funding.

This new project management system will also attempt to establish clearer
lines of responsibility from DOE headquarters to the field units that
perform day-to-day contractor oversight at major construction projects. This
is to be accomplished by requiring that DOE's on-site oversight managers
report to the headquarters program offices responsible for the projects, as
well as reporting through their field offices. DOE's headquarters chain of
command is now more direct for the NIF project. DOE's on-site field manager
for NIF will report directly to Defense Programs in headquarters, and DOE
has advised us that his performance rating will be the responsibility of
Defense Programs in headquarters, not the Oakland Operations Office.

The Secretary of Energy Advisory Board NIF Task Force also made
recommendations in its January 2000 interim report. A main recommendation
was to have DOE develop intensive internal review processes similar to those
used successfully by its Office of Science to track projects for budget and
schedule. DOE plans to start this new process in August 2000. The task force
also called on DOE to clarify roles and responsibilities between
headquarters and field offices and recommended that NIF be completed in
phases. DOE is still studying these recommendations.

While improved processes and organizational fixes are helpful, DOE has
difficulty implementing management reforms. For example, a 1991 DOE
initiative, called the "Financial and Project Management Improvement
Program," sought to give line managers early review authority over projects
and outlined a plan to hire more well-qualified project managers. DOE
acknowledged in 1999 that its project management suffers from "serious
systemic issues needing correction." In addition, past DOE reorganizations
have attempted to clarify chains of command and upgrade project management
skills, also with little success.

The Laboratory is taking a variety of corrective actions to improve its
management of NIF. The original NIF project manager was reassigned to
another laboratory, and the Laboratory's laser director resigned. In
addition, the University of California did not give its Lawrence Livermore
director a salary raise. Other Laboratory actions include the following:

 The Laboratory is reorganizing the NIF project to create clearer lines of
responsibility and to better focus on major risk areas. The new structure
promises to give the NIF project manager more authority and a more direct
line to top management. This new structure could more closely integrate the
NIF project with the laser research and development program that supports
it. The new project manager has experience with large projects and has
appointed a chief engineer.

 The reporting details and procedures for NIF progress reports are being
changed to better reflect project risks and their potential impacts on
future budgets and schedules. While reporting standards already exist, these
new requirements are intended to make the reporting process more accurately
reflect the NIF's project status.

 The Laboratory has replaced its NIF Council with a new NIF Programs Review
Committee that will report to the Laboratory director, not just the laser
director. In turn, the Laboratory director will forward each report to DOE.
Laboratory officials told us that they will "invite" DOE to participate in
selecting new committee members.

The University of California, as the contractor operating Lawrence Livermore
for DOE, maintains its own oversight structure for the Laboratory. In its
report on NIF's problems, the University called for clearer roles and
responsibilities within the Laboratory, involving the Laboratory director
more directly in NIF management, creating a new senior position of associate
director for NIF, and improving the Laboratory's communications
procedures.15 The report also called on DOE to demand accountability from
the Laboratory. The University has restructured its own advisory council to
better oversee all of its national laboratories and has created a new panel
on project management. The Laboratory is addressing these recommendations.

DOE, the Lawrence Livermore National Laboratory, and the University of
California all share responsibility for NIF's cost and schedule problems.
Each lacked the appropriate skills to manage and oversee the project and
exercised oversight powers poorly. In addition, inadequate independent
reviews enabled the Laboratory to withhold significant information from
external reviewers and the Congress for months. The result is a
billion-dollar overrun and multiyear schedule delay that could worsen
because substantial research and development is still incomplete.
Unfortunately, the Congress cannot know with assurance just how much NIF
will cost, where in DOE's budget the money will come from, what impact NIF
will have on the overall nuclear weapons program, or how long it will take
to complete, even though 1-1/2 years have passed since the Laboratory first
questioned whether it could meet its most recent cost and schedule.

These conditions have now spawned new reforms by the Laboratory and the
University to improve their management and organizational shortcomings. NIF
has also taught DOE to continue its efforts at streamlining chains of
command between headquarters program offices and field management units. But
the Department must work hard to hire--and retain--competent project
managers. Also, since construction began, the NIF project has not been
subject to a comprehensive and effective independent review that covers both
the project and its supporting research and development activities. DOE must
provide for effective independent review of its projects to ensure that it
has the best possible advice on technical and managerial issues and that it
is aware of the remaining risks to NIF's cost and schedule.

NIF was designed to be a national facility to support an integrated
Stockpile Stewardship Program for maintaining the safety and reliability of
the nation's nuclear arsenal. As such, we would have expected any decision
about future NIF funding to be made in the context of what stewardship
program components might be eliminated, reduced in scope, or extended in
order to fund NIF's cost increases, as well as what impact any decision
would have on the success of the overall program. The lack of agreement
among NIF's users--the nuclear weapons laboratories--about the appropriate
path forward indicates to us that considerably more work needs to be done
before a final decision on NIF is made. Without such an agreement, the
Secretary of Energy is unable to assure himself, and the Congress, that the
appropriate cost/benefit trade-off has been made between funds for finishing
NIF and funds for the other portions of the Stockpile Stewardship Program
needed to maintain a safe and reliable nuclear weapons arsenal.

To ensure that DOE has an effective independent review of NIF, we recommend
that the Secretary of Energy arrange for an outside scientific and technical
review of NIF's remaining technical challenges as they relate to the
project's cost and schedule risks.

To ensure an appropriate balance between NIF and the rest of the nuclear
weapons program, we recommend that the Secretary of Energy not reallocate
funds from the nuclear weapons program to NIF until DOE (1) evaluates the
impact of its cost and schedule plan, as well as any other options for NIF,
on the overall nuclear weapons program and (2) certifies that the selected
NIF cost and schedule plan will not negatively affect the balance of the
Stockpile Stewardship Program.

In commenting on a draft of this report, DOE raised a number of issues
concerning how NIF's costs should be counted and our characterization of the
technical uncertainties facing the project. DOE also said that it is already
implementing our recommendation for an external independent review and that
it has met the "intent" of our recommendation to evaluate the impact of
NIF's cost increases and schedule delays on the overall nuclear weapons
program.

With respect to the first issue, DOE said that a substantial portion of the
cost increases we cited is for developing a target for NIF and that this
amount should be not included as part of NIF's construction costs. We are
not suggesting that these costs be included as construction costs. But we
are suggesting that all NIF costs be identified. Both our figures and DOE's
show cost increases in related programs and research and development. To
give the Congress a more accurate picture of the total costs associated with
the NIF project, the cost of designing and building a target--without which
the NIF cannot operate--is an essential component of the entire NIF effort,
regardless of its budget account label. In addition, we included costs borne
by other laboratories that directly contribute to the NIF effort that are
not reflected in the NIF project figures that DOE has prepared.

Concerning our presentation of the project's technical uncertainties, DOE
disagreed with our characterization of final stage optics as an unresolved
technical risk, noting that problems associated with optics are only an
issue of "economics" and that NIF's technical uncertainties are well
understood. We disagree. As DOE's own expert reviewers have noted, final
optics is NIF's most challenging technical area and remains an unresolved
issue. As DOE noted in its letter, if optics cannot be made to withstand
NIF's high-intensity lasers, they will suffer damage and will have to be
replaced more often. This outcome has the effect not only of increasing
NIF's operating costs but also of limiting its performance because fewer
laser firings would result for a given operating budget. In addition, as we
noted in our report, substantial technical and cost uncertainties persist in
the research and development needed to design and build a target for NIF's
laser beams.

DOE said that it is implementing our recommendation on the need for an
independent scientific and technical review of NIF. DOE also cited a large
number of past and ongoing technical and cost reviews of NIF, noting that it
has been aggressive in obtaining expert reviews of NIF's technical
challenges. DOE also said that it is planning an internal review of NIF in
August 2000 that will include an independent cost estimate, and that this
review is designed to give the Department "a higher level of confidence"
that it can complete NIF as planned. We agree that the technical and cost
components of NIF have been frequently reviewed in the past. However, past
reviewers did not discover and report on the NIF's fundamental project and
engineering problems, bringing into question their comprehensiveness and
independence. Also, since cost and schedule problems were first identified
in late 1998, there have been no effective independent reviews of NIF that
have related technical challenges to its cost and schedule risks. The
internal review DOE is planning for August 2000 resembles the process DOE's
Office of Science uses to evaluate its major science projects and has
potential for providing valuable insights into technical and managerial
issues. However, we believe internal review processes supplement--but cannot
replace--the value of a separately conducted scientific and technical
evaluation managed by experts outside DOE and its laboratory system.

Finally, DOE also said it has already met the "intent" of our recommendation
to evaluate the impact of NIF's funding options on the Stockpile Stewardship
Program. DOE stated that its June 1, 2000, "interim" report to the Congress
and its fiscal year 2001 revised budget request of June 27, 2000, contain
"an option that considers" the impacts of NIF on the overall Stockpile
Stewardship Program. However, neither of these reports contains an analysis
of the impacts of NIF's cost increases on the balance of the Stockpile
Stewardship Program. DOE's June 1 report contained no details on how NIF
cost increases will be funded. Its June 27 revised budget request only
contains details for fiscal year 2001 on how the Department intends to
reallocate funds to pay for NIF's cost increases. These reallocations
included reductions in various Stockpile Stewardship Program components
managed by Lawrence Livermore and other national laboratories but did not
describe how its proposed reallocations and associated delays will affect
the Stockpile Stewardship Program's campaigns and their objectives and
milestones. In addition, the revised budget request did not describe how
paying for NIF affects Stockpile Stewardship Program activities beyond
fiscal year 2001. Because this is a long-term integrated program, we believe
that the Secretary of Energy needs to present in more detail how planned
cuts and delays caused by NIF affect the overall Stockpile Stewardship
Program's objectives and milestones. The Laboratory's own technical advisers
have recommended that better planning is needed to show how NIF is to be
used within the Stockpile Stewardship Program and that such a plan should
assess the impacts of delays on both NIF and other Stockpile Stewardship
Program facilities.16

Appendix II includes the full text of DOE's comments and our response.

As arranged with your offices, unless you publicly announce its contents
earlier, we plan no further distribution of this report until 30 days after
the date of this letter. At that time, we will send copies to the Honorable
Bill Richardson, Secretary of Energy, and the Honorable Jacob J. Lew,
Director, Office of Management and Budget. We will also make copies
available to others on request.

If you or your staff have any questions about this report, please call me at
(202) 512-3841. Key contributors to this report are listed in appendix III.

Jim Wells,
Director, Energy, Resources,
and Science Issues

Scope and Methodology

To determine the magnitude of the National Ignition Facility's (NIF) cost
and schedule overruns, we examined all Laboratory progress reports and
quarterly documentation and interviewed Laboratory and Department of Energy
(DOE) program officials. We worked closely with budget officials from each
of the three defense laboratories to determine the amount of their funds
that supported the NIF project. From these discussions, we constructed a
detailed table of NIF's costs.

To document the reasons for the cost and schedule problems, we examined past
reviews of NIF, including those conducted by the University of California
and the Secretary of Energy Advisory Board Task Force on NIF. We also held
extensive interviews with the authors of these reviews. We also interviewed
all former and current project managers and past senior managers at DOE and
the Office of Management and Budget. We also examined documents from the NIF
Council and interviewed its members on several different occasions. We also
interviewed senior budget officials from DOE and officials from the now
defunct Office of Field Management. These individuals provided us with
substantial documentation of past reviews of NIF's progress.

To assess the impacts from the problems, we interviewed a wide range of
scientists from Lawrence Livermore, Los Alamos, and Sandia national
laboratories. We also interviewed past and current officials and scientists
from universities and from private interest groups.

To evaluate DOE and Laboratory actions to correct these problems, we
examined the previously mentioned reviews and interviewed the authors of
these reviews. This included interviewing officials from the University of
California and Lawrence Livermore Laboratory. We also attended each of the
four Secretary of Energy Advisory Board meetings on NIF and interviewed
members of the board to ascertain their understanding of NIF's technical and
managerial problems. We also interviewed the authors of the engineering firm
that conducted a review of NIF.

Our review was performed from October 1999 through July 2000 in accordance
with generally accepted government auditing standards.

Comments From the Department of Energy

The following are GAO's comments on the Department of Energy's letter dated
July 28, 2000.

1. Our response is included in the body of the report.

2. Our response is included in the body of the report.

3. Our response is included in the body of the report. In addition, DOE's
listing of a GAO review of the NIF Project/ICF Program in 1998 is misleading
because we have not previously analyzed the NIF Project or ICF Program. We
did receive a one-day briefing in 1998 on NIF and ICF program issues that
was part of another assignment. Also, DOE stated that the Lockwood-Greene
review was conducted before the NIF cost and schedule problems were
identified. However, as we noted in our report, Lawrence Livermore's NIF
managers were aware of cost and schedule problems before Lockwood-Greene
conducted its visits in early 1999.

4. Our response is included in the body of the report.

GAO Contacts and Staff Acknowledgements

Jim Wells, (202) 512-3841

Gary Boss, (202) 512-3841

In addition to those named above, James Noel, Thomas Kingham, William
Lanouette, and Robert Sanchez made key contributions to this report.

(141393)

Table 1: NIF's Estimated Cost and Schedule Changes 10

Figure 1: The Completed National Ignition Facility 8

Figure 2: Timeline of Laboratory's Disclosure of NIF Problems 22

  

1. On July 3, 1993, President Clinton announced that the United States would
no longer conduct nuclear tests, as a result of the Cold War's end and the
desire to reduce the threat of nuclear weapons. As an alternative to such
testing, DOE created the Stockpile Stewardship Program to certify that
nuclear weapons are safe and reliable.

2. Fusion results when light atoms, such as isotopes of hydrogen, are
combined under very high temperatures and pressures to make heavier ones,
such as helium. Fission is the opposite: It starts with very heavy atoms,
such as uranium, and splits them into two or more pieces called fission
products. Both fusion and fission release large quantities of energy as heat
or radiation. Heat energy can be converted into steam to generate
electricity.

3. Completion occurs when the building, infrastructure, laser systems, and
optical components are installed. At that point, experiments can begin.

4. In commenting on a draft of this report, DOE said that 2004 would be the
earliest year for gaining ignition.

5. During our review, the National Nuclear Security Administration was
created and assumed responsibility for DOE's Defense Programs (including the
Stockpile Stewardship Program and NIF), effective March 1, 2000. The
stockpile stewardship activity was reorganized and renamed. Since our work
was performed before and after these administrative changes were occurring,
for clarity we refer throughout this report to the organization and
terminology in use before the changes took place.

6. Participating Internal Confinement Fusion contractors besides Lawrence
Livermore National Laboratory are Los Alamos National Laboratory, Sandia
National Laboratories, the University of Rochester, and General Atomics.

7. The NIF Council, which advises the Laboratory on NIF-related issues, is
composed of leading experts on lasers from academia, government, and
scientific institutions. It has recently been replaced by the NIF Programs
Review Committee.

8. NIF Technology Review, Technology Resource Group of the NIF Council (Nov.
4, 1999).

9. On October 6, 1999, Secretary Richardson requested that the Secretary of
Energy Advisory Board form a task force to review the engineering and
management aspects of the assembly and installation of the NIF laser system.
The task force is composed of board members and individuals with expertise
in systems engineering, laser science, and project management. It issued an
interim report on January 10, 2000.

10. H.R. Rep. No. 106-336, at 96-97 (1999).

11. Standards for Internal Control in the Federal Government
(GAO/AIMD-00-21.3.1, Nov. 1999).

12. We previously highlighted weaknesses in DOE's chain of command. See
Department of Energy: Views on DOE's Plan to Establish the National Nuclear
Security Administration (GAO/T-RCED-00-113 , Mar. 2, 2000).

13. External Independent Review of the Department of Energy National
Ignition Facility Project at the Lawrence Livermore National Laboratory
Site, Detailed Review and Assessment Report. Lockwood Greene Technologies
(Mar. 29, 1999).

14. The "primary" is the fission stage of a thermonuclear weapon. Detonated
first, the primary produces the extremely high temperatures and pressures
required to produce fusion in the weapon's "secondary." The "secondary," or
thermonuclear stage, of a nuclear weapon produces its energy through the
fusion of deuterium and tritium nuclei.

15. Report of the University of California President's Council National
Ignition Facility (NIF) Review Committee, University of California, (Nov.
18, 1999).

16. Report of the National Ignition Facility Target Physics Program Review
Committee, May 2, 2000.
*** End of document. ***