Nuclear Weapons: Improvements Needed to DOE's Nuclear Weapons Stockpile
Surveillance Program (Letter Report, 07/31/96, GAO/RCED-96-216).

Pursuant to a congressional request, GAO provided information on the
Department of Energy's (DOE) Nuclear Weapons Stockpile Surveillance
program, focusing on DOE efforts to get the testing program on schedule.

GAO found that: (1) DOE is behind schedule in conducting flight tests,
nonnuclear system laboratory tests, and nuclear and nonnuclear component
laboratory tests; (2) these schedule slippages are a result of
unapproved safety studies, suspended testing at certain nuclear
facilities, and inappropriate transfer of testing functions; (3) DOE has
reduced its plan for testing the Air Force's intercontinental ballistic
missiles from three tests per year to two tests per year; (4) flight
testing of W88 warheads is suspended until a safety study plan is
approved; (5) DOE has taken actions to increase the number of stockpile
surveillance tests, but DOE does not have formal contingency plans for
continuous stockpile testing; (6) one DOE facility is unable to conduct
surveillance tests due to procedural safety problems; (7) DOE uses the
Stockpile Management Preferred Alternative Report to determine
alternative locations for weapons testing depending on the nature of the
problem at the original testing facility, length and time of the outage,
and particular weapon involved; and (8) DOE prefers to develop a
specific plan of action after testing problems occur to confront the
wide range of problems and variables involved in the surveillance
testing process.

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

 REPORTNUM:  RCED-96-216
     TITLE:  Nuclear Weapons: Improvements Needed to DOE's Nuclear 
             Weapons Stockpile Surveillance Program
      DATE:  07/31/96
   SUBJECT:  Nuclear weapons testing
             Atomic energy defense activities
             Test facilities
             Missile warheads
             Intercontinental ballistic missiles
             Nuclear weapons plant safety
             Advanced weapons systems
             Safety regulation
             Safety standards
IDENTIFIER:  DOE Nuclear Weapons Stockpile Surveillance Program
             ICBM
             W62 Warhead
             Minuteman III Missile
             W78 Warhead
             W87 Warhead
             W88 Warhead
             Trident II Missile
             Peacekeeper Missile
             DOE Nuclear Explosive Safety Study
             DOE Stockpile Management Preferred Alternatives Report
             
******************************************************************
** 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 the Committee on Armed Services, U.S.  Senate

July 1996

NUCLEAR WEAPONS - IMPROVEMENTS
NEEDED TO DOE'S NUCLEAR WEAPONS
STOCKPILE SURVEILLANCE PROGRAM

GAO/RCED-96-216

Nuclear Weapons Stockpile

(302193)


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

  DOE -
  DOD -
  GAO -

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


B-272465

July 31, 1996

The Honorable Strom Thurmond
Chairman
The Honorable Sam Nunn
Ranking Minority Member
Committee on Armed Services
United States Senate

The Department of Energy (DOE) is responsible for managing the
nation's nuclear weapons stockpile, including surveillance of weapons
currently in the stockpile to identify reliability and safety
problems.  DOE conducts three types of stockpile surveillance
tests--flight tests, nonnuclear systems laboratory tests, and nuclear
and nonnuclear component tests--on nine types of nuclear weapons. 
Your offices asked us to provide information on the status of the
stockpile surveillance program.  If the program is not on schedule,
you asked us to determine why it is not and provide you with
information on the steps being taken to get the program back on
schedule. 


   RESULTS IN BRIEF
------------------------------------------------------------ Letter :1

DOE is behind schedule in conducting many of the stockpile
surveillance tests.  These include flight tests of three types of
weapons, nonnuclear systems laboratory tests of one type of weapon,
and laboratory tests of most key components.  In some instances, DOE
is several years behind schedule.  Only through testing can DOE
identify problems or defects that would warrant changing the
reliability level it assigns to a particular weapon.  Being behind
schedule in the testing program provides DOE with less information on
the weapons, thereby providing less confidence in the reliability
levels assigned. 

The stockpile surveillance program is behind schedule for a variety
of reasons.  At one facility, testing was suspended because the
facility lacked an approved safety study required to disassemble and
inspect one type of weapon.  Testing was suspended at another
facility because of concerns about safety procedures.  Testing delays
also occurred during the transfer of testing functions to new
facilities. 

DOE does not have written plans indicating in detail how it will get
the testing program on schedule.  For some of the tests, DOE
officials told us that it will be years before the tests are back on
schedule.  Furthermore, several factors, including the possible
expiration of required safety studies, future limitations on the
number of flight tests, and the lack of contingency plans in the
event a test facility is temporarily shut down, could prevent DOE
from being able to conduct tests and return to its schedule. 


   BACKGROUND
------------------------------------------------------------ Letter :2

The purpose of DOE's stockpile surveillance program is to ensure,
primarily through three types of tests, that the safety and
reliability of nuclear weapons are maintained.  Flight tests involve
the actual dropping or launching of a weapon, which has had the
nuclear components removed.  Nonnuclear systems laboratory tests
involve testing a weapon's nonnuclear systems to detect defects due
to handling, aging, manufacturing, or design.  The nuclear and
nonnuclear components laboratory tests involve destructive analysis
to identify defects or failures in individual weapon components. 

Weapons are randomly selected for flight and nonnuclear systems
laboratory tests from the stockpile each year.  Weapons chosen for
the nuclear and nonnuclear components laboratory tests are
judgmentally selected from the weapons that have been selected for
the other two tests.  For testing purposes, DOE considers the active
stockpile to consist of nine weapon types--three bombs and six
missile warheads, each with unique capabilities.  From 1958 to 1996,
DOE's stockpile surveillance program has tested about 14,000 weapons,
with about 2,400 findings documented.  Over 50 percent of these
findings were considered "significant." A significant finding is the
identification of a defect or failure in a weapon system.  A defect
is an observable anomaly, while a failure is a flaw or malfunction in
the weapon that would prevent the weapon from operating as
intended.\1

When a significant finding is disclosed, DOE may perform additional
tests to confirm the finding, determine the cause of the problem,
assess its impact on the stockpile, and recommend a corrective plan. 
Of the 2,400 findings, 370 were "actionable." DOE defines an
actionable finding as a finding that lowers the weapon's reliability
or for which some action is taken.  About 1 in 3 actionable findings
(118 findings) have resulted in retrofits and major design changes. 
The remainder required either process changes or no physical changes. 
When a weapon's reliability is lowered because of a finding, the
result is reported to the Department of Defense (DOD). 

DOE and the national nuclear laboratories\2 have determined that they
generally need to test about 40 to 44 weapons of each type in the
stockpile over a 4-year period.\3 According to DOE officials, over
that 4-year time frame, the tests should consist of 8 to 12 flight
tests per weapon type (an average of 2 or 3 tests per year) and 28 to
36 laboratory tests of nonnuclear systems (an average of 7 to 9 per
year).  Finally, from the weapons scheduled for testing each year,
DOE designates components from certain weapons for laboratory tests. 
DOE considers five components to be key--the pit, the secondary, the
detonator sets, the gas transfer system, and the high explosives.\4
On average, for each weapon type, DOE believes that one pit, one
secondary, two to five detonator sets, one or two gas transfer
systems, and one high-explosive system should be tested each year. 

According to DOE officials, when a significant number of tests are
cancelled or delayed, the Department lacks information on the
reliability of the weapon.  While lack of testing will not affect the
reliability level assigned to a weapon (only a test finding can alter
the reliability level), the lack of test information reduces DOE's
confidence in the assessed reliability of the weapon. 


--------------------
\1 About 1.3 percent of the tests have identified failures. 

\2 The Sandia National Laboratories, Los Alamos National
Laboratories, and Lawrence Livermore National Laboratories. 

\3 The number of weapons needed to be tested can vary somewhat
according to the history of the weapon type and the number of weapons
of that type in the stockpile at that time. 

\4 The pit is a metal casing that contains plutonium.  The pit is
where the fission explosion originates.  A nuclear weapon secondary
is an assembly in the weapon where a fusion explosion originates. 
The detonator sets off the high explosive, and the gas transfer
system includes a tritium reservoir and associated hardware. 


   STOCKPILE SURVEILLANCE TESTS
   ARE BEHIND SCHEDULE
------------------------------------------------------------ Letter :3

DOE is currently behind schedule in conducting some flight tests,
nonnuclear systems laboratory tests, and nuclear and nonnuclear
components laboratory tests.  For some tests, DOE is several years
behind schedule.  These schedule slippages are the result of a
variety of factors, including an unapproved safety study, suspension
of testing at some facilities, and the transfer of testing functions
to new facilities. 


      FLIGHT TESTS
---------------------------------------------------------- Letter :3.1

Flight tests involve the actual dropping or launching of a weapon
from which the nuclear components have been removed.  DOE uses
specially designed equipment--referred to as telemetry packages--to
test the integration and functioning of the weapon's electrical and
mechanical subsystems.  Until November 1992, DOE planned to conduct a
minimum of 3 flight tests per year--or 12 flight tests over a 4-year
period--for bombs, InterContinental Ballistic Missiles, and
Submarine-Launched Ballistic Missiles.  According to DOE officials,
in November 1992, DOE reduced its plan for testing Air Force
InterContinental Ballistic Missiles from three tests per year to
two--or eight tests over a 4-year period.  DOE officials informed us
that they made the reduction based on an evaluation of applicable
existing test data and in preparation for the Air Force's
implementation of the START I and START II treaties.\5 Under these
treaties, the Air Force will have to reduce the number of warheads
carried on missiles.  The plan for testing bombs and
Submarine-Launched Ballistic Missiles was not altered and remains at
3 per year, or 12 over a 4-year period. 

DOE officials told us that they believe the reduction in flight tests
from three to two per year for InterContinental Ballistic Missiles
represents an acceptable increase in the risk of having undetected
problems in weapons.  The officials explained that by flight testing
three weapons per year, there is a 90-percent chance of discovering a
"flight-unique" defect if the defect occurs in 18 percent of the
weapons.  By testing only two weapons per year, the risk increases. 
With two tests per year, the defect would have to occur in 22 percent
of the weapons to have a 90-percent chance of discovering it.  DOE
officials believe that conducting fewer than two tests per year (or
eight tests over a 4-year period) is a concern and a significantly
increased risk to the program. 

Three weapon types--the W62, W78, and W88 warheads--have had, on
average, fewer than two tests conducted per year over the past 4
years.  Table 1 shows the three weapon types, DOE's plan for testing,
and the number of tests conducted over the past 4 years (fiscal years
1992 through 1995). 



                                Table 1
                
                  Flight Tests Conducted During Fiscal
                  Years 1992 Through 1995 for Selected
                              Weapon Types

                                       4-year plan        Actual tests
                                     (no. of tests           completed
Type of weapon                          scheduled)           (4 years)
------------------------------  ------------------  ------------------
W62                                              8                   6
W78                                              8                   7
W88                                             12                   3
----------------------------------------------------------------------
The W62, a warhead used by the Air Force on the Minuteman III
missile, has been flight-tested six times (of the eight planned) over
the past 4 years.  Two planned tests were not conducted because DOE's
Pantex facility had trouble preparing warheads for flight testing and
could not deliver the test warheads to the Air Force in time for the
test flights.  The W78 warhead, also used on the Minuteman III, has
had seven flight tests (of the eight planned) over the past 4 years. 
DOE and the national laboratory officials told us that a flight test
with telemetry equipment was not conducted because the Department
decided to use the available warhead test slot on the test missile
for a nontelemetry DOE test of the W78.\6

The W88 is a warhead used by the Navy on the Trident II missile. 
Only 3 W88 stockpile flight tests (of the 12 planned) were conducted
during the 4-year period from fiscal year 1992 through 1995.  Flight
testing of W88 warheads taken from the stockpile was halted for more
than 1 year because an important safety study required for
disassembly and inspection of the warhead at DOE's Pantex plant
lacked approval.  A Nuclear Explosive Safety Study is required for
each weapon type before DOE's Pantex Plant can disassemble and
inspect a weapon selected for testing.  Without disassembly and
inspection capability, surveillance tests, including flight tests of
sample warheads from the stockpile (the nuclear components must be
removed and replaced by the telemetry equipment), cannot be
conducted.  DOE and national laboratory officials are not concerned
about the reliability of the W88 warhead because they have collected
considerable data over the past few years by testing W88 warheads
that had never been placed in the stockpile.  Because the W88 warhead
is a relatively new weapon, DOE officials believe that the
information from these "new material" tests provides good reliability
data. 


--------------------
\5 The START (Strategic Arms Reduction Treaty) I and START II
treaties are arms control agreements between the United States and
what was formerly the Soviet Union that mandate reductions in
strategic offensive nuclear weapons. 

\6 DOE has periodically conducted these nontelemetry tests on the W78
and W87 over the past 3 years because they provide for flight
dynamics that are more typical of a real warhead than the telemetry
warheads.  The nontelemetry tests do not, however, provide discrete
information on the operation of the weapon's components. 


      NONNUCLEAR SYSTEMS
      LABORATORY TESTS
---------------------------------------------------------- Letter :3.2

Of the nine weapon types, only the W88 warhead is considered by DOE
to be of concern in relation to nonnuclear systems laboratory tests. 
These tests involve testing the nonnuclear systems--such as the radar
systems and fuzes--in the weapon to detect defects due to handling,
aging, manufacturing, or design.  DOE officials said the Department
should have conducted about 28 laboratory tests, but over the past 4
years, only 15 (or 54 percent) tests were performed.  According to
DOE and national laboratory officials, the tests were not conducted
because of the absence of an approved safety study at Pantex.  DOE
officials said that in this case, the lack of testing reduces their
confidence in the weapon's reported reliability.  DOE officials told
us that they could not quantify the decrease in confidence. 


      LABORATORY TESTS OF NUCLEAR
      AND NONNUCLEAR COMPONENTS
---------------------------------------------------------- Letter :3.3

From the weapons selected for testing each year, one weapon of each
type is selected to have individual nuclear and nonnuclear components
destructively tested.  Although many other components are tested
(such as cables and electrical components), according to DOE
officials, the five key components tested are the pit, the secondary,
the detonator assembly, gas transfer system, and high explosive. 
Testing of four of these key components has been behind schedule in
recent years.  Only the high explosives tests have been conducted on
schedule.\7

The pit is a part of the nuclear package that, until 1989, was
manufactured and tested at DOE's Rocky Flats facility in Colorado. 
According to DOE officials, the Department ideally tests one pit per
year per weapon type.  In December 1989, the Rocky Flats facility
ceased production operations.  At first, DOE believed that Rocky
Flats would reopen; however, in 1992 DOE decided to move pit tests to
the Los Alamos National Laboratory.  This lapse created a backlog of
up to 4 to 5 years, but testing is currently nearly back on schedule. 

The secondary is tested at DOE's Y12 facility in Oak Ridge,
Tennessee.  Ideally, one secondary should be tested per weapon type
per year.  Few have been tested since September 1994, when Y12 was
placed in a "stand-down" mode because of problems related to safety
procedures that had been noted by the Defense Nuclear Facilities
Safety Board.\8 Most of these problems did not involve unsafe
conditions, but were related to not following approved procedures. 
According to DOE officials, a 1-year backlog of secondaries to be
tested currently exists. 

DOE's Mound facility in Ohio tested detonator sets through 1994.  At
that time, responsibility for testing detonator sets was moved to
DOE's Los Alamos and Lawrence Livermore Laboratories.  Ideally, DOE
tests two to five detonator sets per year per weapon type.  Los
Alamos began testing in June 1996, and Lawrence Livermore is
scheduled to begin testing later this year.  In the meantime, a
1-1/2-year backlog of detonator sets to be tested exists. 

DOE's Mound facility also tested gas transfer systems through 1994. 
Ideally, one or two gas transfer systems are tested per weapon type
per year.  Responsibility for testing gas transfer systems was moved
to DOE's Savannah River facility in South Carolina.  Savannah River
began testing some gas transfer systems earlier this year, but a
1-1/2-year backlog currently exists. 

According to DOE officials, the lack of nuclear component testing
decreases DOE's confidence in the reliability assessments of the
weapons in the nuclear stockpile.  DOE officials said that they could
not estimate the degree to which confidence in the reliability
assessments of the weapons had decreased because of the backlogs in
nuclear components laboratory tests.  However, the officials said
that the confidence had not diminished to a point of concern.  The
officials explained that pits, secondaries, detonator assemblies, and
gas transfer systems are long-lived items, and generally, testing
could be suspended for 3 years without confidence diminishing to a
point of concern. 


--------------------
\7 According to DOE officials, all other component tests are on
schedule. 

\8 The Defense Nuclear Facilities Safety Board was created in 1988 to
investigate situations at DOE's defense nuclear facilities that could
adversely affect public health and safety. 


   DOE'S ABILITY TO CONDUCT SOME
   FUTURE TESTS IS UNCERTAIN
------------------------------------------------------------ Letter :4

DOE has taken actions to increase the number of stockpile
surveillance tests but has not prepared detailed plans for returning
the stockpile surveillance program to its schedule.  Without such
plans, it is difficult for us to assess the likelihood that stockpile
surveillance tests will return to the schedule.  Furthermore, we
believe that issues and factors such as the availability of test
missile launches, expiration of approved safety studies, or cessation
of operations at test facilities could have an adverse effect on
DOE's future ability to remain on schedule. 


      FLIGHT TESTS AND NONNUCLEAR
      SYSTEMS LABORATORY TESTS
---------------------------------------------------------- Letter :4.1

For most weapon types, DOE has taken actions that may return flight
tests and nonnuclear systems laboratory tests to the schedule in the
short term.  However, in the longer term, implementation of the START
I and START II treaties, the availability of telemetry packages used
in flight testing, and the expiration of safety studies could cause
these testing programs to fall behind schedule. 

Based on the Air Force's agreement to provide for sufficient flight
tests on test missiles, DOE estimates that W78 warhead flight tests
will be back on schedule by the end of fiscal year 1996.  The W62
warhead is behind schedule for flight testing because DOE could not
deliver the test warheads to the Air Force in time for the tests. 
DOE officials told us that this should not recur, but, as discussed
later, DOE may not be able to maintain the W62 warhead flight test
schedule in the long term because of limited inventories of testing
equipment. 

The safety study that caused delays in the W88 warhead testing has
been approved, and both flight testing and nonnuclear laboratory
systems tests have been resumed.  To get flight tests back on
schedule, DOE plans to conduct six flight tests in fiscal year 1996
(as of July 1996, DOE had conducted three telemetry and one
nontelemetry test during fiscal year 1996), four in 1997, three in
fiscal year 1998, and three in fiscal year 1999.  To get nonnuclear
laboratory systems tests on schedule, DOE plans to consolidate 3
years of testing into 2 years.  DOE estimates that flight tests will
be back on schedule sometime during fiscal year 1999 and nonnuclear
systems laboratory tests will be back on schedule in fiscal year
1998. 

In the longer term, tests of the W78 warhead--as well as the W62 and
W87 warheads--could be a problem.\9 DOE officials told us that when
the START I and START II treaties are fully implemented, the Air
Force may be limited in its ability to conduct flight tests.  Air
Force officials confirmed that providing for future InterContinental
Ballistic Missile flight tests may be difficult because of
limitations imposed by the START treaties.  These treaties require a
transition from Multiple Independent Reentry Vehicles to Single
Reentry Vehicles.  Until recently, multiple flight tests were
routinely conducted on one missile firing.  After the treaties are
fully implemented, only one test warhead per missile will be allowed. 
A reduction in the number of warhead tests per flight reduces the
overall number of tests that can be conducted because the number of
missiles available for testing purposes is limited. 

Future flight tests of the W62 warhead could also be limited by a
lack of telemetry packages.  Initially, DOE had enough telemetry
packages to test this warhead during its projected life.  However,
retirement of this warhead has been delayed, and DOE is running out
of telemetry packages.  Also, the company that produced the package
has gone out of business.  DOE is studying the possibility of using
parts from W68 warhead telemetry packages (the W68 has been retired,
but telemetry parts remain that may be recertified for use in the
W62) to increase the number of telemetry packages available.  If this
is done, DOE could test W62 warheads for 4 years at the rate of two
per year.  DOE officials told us that a decision will be required in
1998 to determine if this warhead will remain in the stockpile long
enough to make the redesign and purchase of new telemetry packages
worthwhile. 

Finally, while the W88 warhead safety study has been approved, the
expiration of other approved safety studies at Pantex could affect
DOE's future ability to conduct stockpile surveillance tests in the
future.  To conduct any of the three major types of stockpile
surveillance tests, Pantex must be able to inspect the weapon,
disassemble the weapon, reassemble the weapon, and replace the
nuclear package with telemetry for the flight test.  Without a valid
safety study for each weapon type, Pantex cannot conduct any of these
operations.  The safety studies are valid for 5 years, and an
extension can be granted for an additional 5 years.  The safety study
for the W78 warhead expired in April 1995 but has since been
revalidated.  Safety studies for the W87 and B83 warheads will expire
within the next year.  DOE does not anticipate a problem as
revalidation of the studies is scheduled to occur before the old
studies expire. 


--------------------
\9 The W87 is the warhead used on the Air Force's Peacekeeper
missile.  Testing of this weapon is currently not behind schedule. 


      LABORATORY TESTS OF KEY
      COMPONENTS
---------------------------------------------------------- Letter :4.2

Although DOE has no formal written plans specifically for returning
laboratory tests of key components to the schedule, DOE officials
told us that activities have been undertaken and progress is being
made toward eliminating the backlog of tests.  Table 2 shows the type
of component, the number of tests normally conducted for each
component, and the approximate number of components in the backlog as
of July 1996. 



                                Table 2
                
                  Backlog of Laboratory Tests for Key
                               Components

                                 Planned tests per
Type of component                             year    Backlog of tests
------------------------------  ------------------  ------------------
Pit                                              9                   2
Secondary                                        9                   7
Detonator sets                            18 to 45                 114
Gas transfer systems                       9 to 18                  22
----------------------------------------------------------------------
Pit testing began at Los Alamos in fiscal year 1993.  By conducting
19 tests per year, DOE officials said that the 4- to 5-year backlog
that once existed will be eliminated by the end of this fiscal year. 
DOE officials also told us that about 10 pit tests were "written
off." This means that DOE determined that it was not necessary to
conduct the tests because sufficient past data existed or because
testing one or two pits out of a backlog of three or four for a
specific weapon would, in its opinion, provide sufficient data. 

Regarding the secondary tests, Y12 is still in a stand-down mode. 
Tests of seven secondaries are currently being conducted under
"special operations." Special operations are defined as discrete
activities or operations that can be performed before resuming normal
activities within a nuclear facility.  Completion of these seven
tests is scheduled before the end of fiscal year 1996.  DOE also is
in the process of testing three secondaries at Los Alamos.  (Los
Alamos has the capability to test secondaries in very limited
numbers.) However, about a 1-1/2-year backlog still exists.  DOE
plans to conduct a readiness assessment for restarting normal
operations by October 1, 1996.  Currently, DOE is considering
conducting 15 tests of secondaries (at least one of each type of
weapon in the active stockpile) during fiscal year 1997.  This would
put secondary testing back on schedule. 

Detonator set testing began at Los Alamos in June 1996 and will begin
at the Lawrence Livermore Laboratory later this year.  DOE plans to
eliminate the backlog by the end of fiscal year 1997.  This should
not involve overtime or reallocation of resources.  DOE officials
explained that once the laboratories are set up to test the detonator
sets, doing additional tests will require very little extra time. 

Gas transfer system testing began at Savannah River in 1996. 
Savannah River will conduct a phased approach to eliminate the
backlog of tests one weapon type at a time.  As a result, some weapon
types will be back on schedule within a year while others will fall
further behind.  DOE officials believe that the Department will
eliminate all backlogs sometime during fiscal year 2000. 

DOE does not have formal written plans describing how it will return
component laboratory tests to the schedule.  DOE officials at the
Albuquerque Operations Office informed us that in the case of gas
transfer systems and detonators, the Activity Transfer Plan prepared
when testing responsibility was transferred from the Mound facility
establishes the testing capability at the new locations.  Beyond the
plan, however, planning for reducing the backlogs and returning to
the testing schedule is done informally.  Officials representing all
organizations involved in the testing meet periodically to resolve
problems affecting the testing program.  In this manner, DOE
officials said that they reach agreement on what to do and how to do
it.  However, without formal documents detailing testing plans,
costs, and schedules, it is difficult--if not impossible--to review
the plans and assess their adequacy, determine the cost-effectiveness
of the plans, or measure progress the test facilities are making. 


   DOE DOES NOT HAVE CONTINGENCY
   PLANS FOR STOCKPILE TESTING
------------------------------------------------------------ Letter :5

In the past, DOE had more facilities and more alternatives for
shifting functions and operations.  However, in DOE's current nuclear
complex, if a particular facility cannot perform testing for an
extended period, there is little redundant capability for stockpile
surveillance testing.  Without redundancy, planning for continued
testing operations in the event of problems at one or more of the
existing facilities takes on added importance.  However, DOE does not
have formal contingency plans for continuation of stockpile
surveillance tests in the event that one or more of the testing
facilities experienced serious operational problems and could not
perform testing for an extended period of time. 

In the past, several facilities have been unable to conduct testing
for extended periods of time.  Most recently, as discussed
previously, Y12 was unable to conduct surveillance tests because of
procedural safety problems.  When the stand-down occurred, DOE did
not have a plan that established how or where surveillance tests
should or could be resumed.  As a result, secondary testing was
halted until special operations began earlier this year at Y12, and
DOE decided to test several secondaries at Los Alamos.  In the
meantime, a backlog of secondaries accumulated.  Perhaps the most
drastic example occurred when operations at Rocky Flats ceased in
1989.  No contingency plan for testing existed, and in the time it
took to make a decision on where testing should be conducted and
complete the transfer arrangements, a 4- to 5-year backlog of pits
waiting to be tested accumulated. 

DOE has a draft report that discusses alternate locations for
conducting weapons-related activities.  For example, DOE's draft
Stockpile Management Preferred Alternatives Report shows that for
detonator-related functions, Los Alamos would be the alternative. 
DOE officials indicated, however, that this does not mean that these
locations have surveillance testing capability available, although
the facility or operations at the facility could possibly be modified
to perform the function. 

In the event of a disruption of operations at a facility that would
preclude testing, DOE officials said that they would use the
Stockpile Management Preferred Alternatives Report to devise a
specific plan.  Depending on the nature of the problem at the
original facility, the length and nature of the outage, and the
specific weapon(s) involved, DOE would determine the best course of
action.  That course of action could be (1) to wait for the problem
to be fixed at the site and resume normal operations at the original
facility, (2) conduct operations at the original facility under
special operations, or (3) alter an existing facility to assume
surveillance operations.  DOE officials said that they believe that
developing a specific plan after the problems occur is the best
course of action because of the wide range of problems that could
occur and the variables related to outage length and potential
remedies. 


   CONCLUSIONS
------------------------------------------------------------ Letter :6

Confidence that the nation's nuclear weapons are reliable is taking
on added importance because these weapons are aging, and no new
weapons are being produced to replace the existing weapons.  As a
result, the stockpile surveillance program's role in assessing
weapons' reliability and ensuring confidence in the reliability takes
on increased importance.  DOE's confidence in the reliability levels
assigned to some nuclear weapons has been diminished because some
needed tests have not been carried out.  To ensure nuclear weapons'
reliability, it is important that DOE's stockpile surveillance be
maintained on schedule.  However, without formal written plans
detailing how DOE will increase the number of surveillance tests in
order to return the program to its schedule, it is difficult to
determine if DOE's estimates on getting the surveillance testing back
on schedule are reasonable and cost-effective.  Furthermore, without
contingency plans, DOE's ability to respond to possible future major
disruptions in its testing operations is uncertain. 


   RECOMMENDATIONS
------------------------------------------------------------ Letter :7

We recommend that the Secretary of Energy direct that the Assistant
Secretary for Defense Programs develop

  -- detailed, written plans to restore stockpile surveillance tests
     to the schedule and

  -- contingency plans for testing facilities to provide for
     continued testing operations in the event that a testing
     facility is shut down for an extended period of time. 


   AGENCY COMMENTS AND OUR
   EVALUATION
------------------------------------------------------------ Letter :8

We provided a draft of this report to DOE for its review and comment. 
We met with officials from DOE's Office of Nuclear Weapons Management
and its Albuquerque Operations Office, Weapons Quality Division,
including the Director of the Office of Nuclear Weapons Management,
who agreed that the report was accurate and agreed with our
conclusions and recommendations. 

During our discussions, the DOE officials stressed that they are
making every effort to get the stockpile surveillance program on
schedule and, over the past year, have made much progress toward that
goal.  In addition, DOE officials stressed that reliability of the
nuclear weapons in the stockpile have not been adversely affected by
a lack of testing. 


   SCOPE AND METHODOLOGY
------------------------------------------------------------ Letter :9

Our objectives in this review were to (1) provide information on the
status of DOE's stockpile surveillance program; (2) if the program is
not on schedule, determine why it is not; and (3) provide information
on the steps being taken to return the program to the schedule. 

To determine if DOE's stockpile surveillance program is on schedule,
we obtained statistics from DOE and DOD and compared those statistics
with DOE's test schedules.  For weapon types or components that were
behind schedule, we discussed with DOE and laboratory officials the
reasons why they were behind schedule and the efforts being made to
return to the schedule.  We also discussed with DOE and DOD officials
the prospects for problems in keeping future tests on schedule.  We
reviewed the safety study expiration and approval schedule for each
weapon type and discussed with DOE officials the contingencies in the
event a testing facility could not operate.  We verified DOE's
statistical analysis of confidence levels and defect discovery
probabilities.  We conducted our review between April and July 1996
in accordance with generally accepted government auditing standards. 


---------------------------------------------------------- Letter :9.1

As arranged with your offices, unless you publicly announce its
contents earlier, we plan no further distribution of this report
until 10 days after the date of this report.  At that time, we will
send copies of the report to the Secretary of Energy; the Secretary
of Defense; and the 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.  Major contributors to this report include
Bernice Steinhardt, Associate Director; William F.  Fenzel, Assistant
Director; Kenneth E.  Lightner Jr., Evaluator; William M.  Seay,
Evaluator; and John D.  Gentry, Evaluator. 

Victor S.  Rezendes
Director, Energy, Resources,
 and Science Issues


*** End of document. ***