General Accounting Office Reports

Access to Space: Issues Associated With DOD's Evolved Expendable Launch Vehicle Program (Letter Report, 06/24/97, GAO/NSIAD-97-130).

Pursuant to a congressional request, GAO reviewed the Department of
Defense's (DOD) progress in acquiring the Evolved Expendable Launch
Vehicle (EELV), focusing on: (1) factors associated with program cost,
schedule, and performance; and (2) selected aspects of EELV's
relationship to the commercial launch vehicle market.

GAO noted that: (1) reducing the cost of launching satellites into orbit
is the paramount objective of the EELV program; (2) however, DOD faces
many program risks in making the transition to EELV that could increase
costs, cause schedule delays, and possibly jeopardize some satellite
schedules and missions; (3) EELV development is less than 25 percent
complete, and DOD has about 1 year to address these risks before
proceeding into engineering and manufacturing development which is
scheduled for June 1998; (4) with several billion dollars at stake, risk
mitigation efforts are essential; (5) cost risk is inherent in the
current acquisition plan because EELV production could be initiated from
1 to 2 years before the first system development test flight; (6)
pursuing such a strategy could result in costly modifications to the
production vehicles because historically, most launch systems have had
several failures during their early flight period; (7) also, existing
satellite programs expect to incur at least $117 million in added costs
as a result of the transition to EELV, and these costs are not included
in the Office of Secretary of Defense or Air Force cost estimates for
the EELV program; (8) there are schedule risks that could seriously
affect the EELV program; (9) as currently planned, DOD will purchase the
last of its existing expendable launch vehicles before the first EELV
system development test flight is scheduled to occur; (10) an
unsuccessful test flight, coupled with the expiration of existing
vehicle contracts, could create a void in the government's launch
capability; (11) DOD has not developed contingency plans to address this
potential risk to national security and civil satellite schedules and
missions; (12) however, it did indicate that commercial launch vehicles
could be used for an emergency procurement in the event of an EELV
failure or schedule delay; (13) the Air Force has identified the meeting
of launch facility preparation schedules as a significant program risk;
(14) in addition, there are technical issues that raise concerns about
potential EELV performance; (15) the Air Force has identified vehicle
propulsion, systems integration, and software as technical risk areas;
(16) the commercial application of EELV poses a unique situation for the
government; (17) the space industry expects a large international market
for commercial satellites, particularly communication satellites, and t*

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

 REPORTNUM:  NSIAD-97-130
     TITLE:  Access to Space: Issues Associated With DOD's Evolved 
             Expendable Launch Vehicle Program
      DATE:  06/24/97
   SUBJECT:  Satellites
             Defense cost control
             Military procurement
             Concurrency
             Aerospace industry
             Defense contracts
             Product performance evaluation
             Defense conversion
             Testing
IDENTIFIER:  DOD Evolved Expendable Launch Vehicle Program
             Defense Satellite Communications System
             Defense Meteorological Satellite Program
             Defense Support Program
             DOD Space-Based Infrared Satellite Program
             
**************************************************************************
* 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.                                          *
*                                                                        *
* A printed copy of this report may be obtained from the GAO Document    *
* Distribution Facility by calling (202) 512-6000, by faxing your        *
* request to (301) 258-4066, or by writing to P.O. Box 6015,             *
* Gaithersburg, MD 20884-6015. We are unable to accept electronic orders *
* for printed documents at this time.                                    *
**************************************************************************


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


Report to the Chairman, Subcommittee on National Security, Committee
on Appropriations, House of Representatives

June 1997

ACCESS TO SPACE - ISSUES
ASSOCIATED WITH DOD'S EVOLVED
EXPENDABLE LAUNCH VEHICLE PROGRAM

GAO/NSIAD-97-130

Access to Space

(707131)


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

  DMSP - Defense Meteorological Satellite Program
  DOD - Department of Defense
  DSCS - Defense Satellite Communications System
  DSP - Defense Support Program
  EELV - Evolved Expendable Launch Vehicle
  EMD - Engineering and Manufacturing Development
  GPS - Global Positioning System
  OSD - Office of the Secretary of Defense
  SBIRS - Space-Based Infrared System

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


B-272655

June 24, 1997

The Honorable C.W.  Bill Young
Chairman, Subcommittee on National Security
Committee on Appropriations
House of Representatives

Dear Mr.  Chairman: 

The Department of Defense (DOD) is acquiring a multi-billion dollar
Evolved Expendable Launch Vehicle (EELV) system to replace the
existing fleet of expendable launch vehicles in the 21st century. 
For an initial investment of about $2 billion, the EELV program goal
is to reduce the costs of launching satellites into space by at least
25 percent, compared to using existing vehicles. 

As you requested, we reviewed DOD's progress in acquiring the EELV
system.  We specifically reviewed factors associated with program
cost, schedule, and performance and examined selected aspects of
EELV's relationship to the commercial launch vehicle market.  This
report discusses program issues and risks that need to be addressed
before the program proceeds into the engineering and manufacturing
development (EMD) phase of the acquisition process, which is
currently scheduled for June 1998. 


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

The federal government currently uses a fleet of expendable launch
vehicles--Delta, Atlas, and Titan--to transport a variety of national
security and civil satellites into space.  According to DOD, these
vehicles (which are acquired by DOD), currently operate at or near
their maximum performance capability.  Also, DOD and congressional
sources consider these vehicles to be very costly to produce and
launch.  Since 1987, the government has made several attempts to
develop a new launch vehicle system, but these attempts were canceled
either because of funding issues, changing requirements, or
controversy regarding the best solution. 

In 1994, by congressional direction, DOD developed a space launch
modernization plan that led to the initiation of the EELV program. 
Currently, two contractors--Lockheed Martin Astronautics and
McDonnell Douglas Aerospace--are competing in a pre-EMD phase, and
one is to be chosen for the EMD phase.  Of the total planned $2
billion investment, the EMD phase is expected to cost about $1.6
billion and take approximately 6 years.  Concurrent with the EMD
decision, DOD plans to authorize the start of EELV production.  An
initial quantity of 29 launch vehicles is estimated to cost about
$1.5 billion.  In addition, toward the end of the EMD phase, a
decision is to be made on whether to produce a significantly larger
quantity that would cost several billion dollars. 


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

Reducing the cost of launching satellites into orbit is the paramount
objective of the Evolved Expendable Launch Vehicle program.  However,
DOD faces many risks in making the transition to the vehicle program
that could increase costs, cause schedule delays, and possibly
jeopardize some satellite schedules and missions.  Vehicle
development is less than
25 percent complete, and DOD has about 1 year to address these risks
before proceeding into engineering and manufacturing development,
which is scheduled for June 1998.  With several billion dollars at
stake, risk mitigation efforts are essential. 

Cost risk is inherent in the vehicle acquisition plan because
production could be initiated from 1 to 2 years before the first
system development test flight.  Pursuing such a strategy could
result in costly modifications to the production vehicles because
historically, most launch systems have had several failures during
their early flight period.  In addition, there is program cost
uncertainty as evidenced by significant estimating differences
between the Office of the Secretary of Defense and the Air Force. 
Also, existing satellite programs expect to incur at least $117
million in added costs as a result of launch vehicle transition, and
these costs are not included in the Office of the Secretary of
Defense or Air Force cost estimates for the vehicle program. 

There are schedule risks that could seriously affect the program. 
First, as currently planned, DOD will purchase the last of its
existing expendable launch vehicles before the first system
development test flight is scheduled to occur.  An unsuccessful test
flight, coupled with the expiration of existing vehicle contracts,
could create a void in the government's launch capability.  DOD has
not developed contingency plans to address this potential risk to
national security and civil satellite schedules and missions. 
However, it did indicate that commercial launch vehicles could be
used for an emergency procurement in the event of an Evolved
Expendable Launch Vehicle failure or schedule delay.  Second, the Air
Force has identified the meeting of launch facility preparation

schedules as a significant program risk.  At the Cape Canaveral and
Vandenberg launch ranges, there are conflicts between the planned use
of certain facilities for the Evolved Expendable Launch Vehicle
program and the current use of these same facilities by other
programs.  Also, the environmental regulatory process that is
required before facility construction can begin could cause an
8-month program delay. 

In addition, there are technical issues that raise concerns about
potential system performance.  The Air Force has identified vehicle
propulsion, systems integration, and software as technical risk
areas.  Propulsion systems are expected to require significant
development.  Integrating all design, engineering, testing,
manufacturing, and launch functions and the software information
system are expected to be challenging tasks.  Although risk
mitigation plans have been developed, problems could arise in these
areas, adversely affecting program cost and schedule goals. 

The commercial application of the Evolved Expendable Launch Vehicle
poses a unique situation for the government.  The space industry
expects a large international market for commercial satellites,
particularly communication satellites, and therefore, for launch
vehicles.  As a result, the winning contractor will enjoy an enhanced
competitive position in the international launch vehicle market from
DOD's investment in the program.  Although the competing contractors
have indicated that they intend to make private investments in
program development, they are not contractually obligated by the
government to do so under the existing pre-engineering and
manufacturing development contracts.  Given this situation, the
question arises as to how the government should be compensated for
its major investment by the winning engineering and manufacturing
development contractor who stands to benefit substantially in the
commercial marketplace from that investment.  Alternatives could be
for DOD to employ a cost-sharing contract for the engineering and
manufacturing development phase and/or arrange for the government to
recoup part of its investment based on commercial launch vehicle
sales. 


   PROGRAM COST UNCERTAINTY
------------------------------------------------------------ Letter :3

DOD has emphasized reducing the costs of space launches as the
paramount EELV program objective.  OSD established a $2-billion
development cost objective.  The Air Force established a production
and launch cost-reduction goal of 25 to 50 percent, compared with the
cost of using existing launch vehicles.  However, considering the
uncertainty in program cost, as evidenced by risk in the acquisition
plans and the differences in the cost estimates done by OSD and the
Air Force, the potential exists for program cost increases. 


      ACQUISITION PLANS CONTAIN
      COST AND MISSION RISKS
---------------------------------------------------------- Letter :3.1

Cost risk is inherent in the EELV acquisition plan because production
could be initiated from 1 to 2 years before the first system
development (medium-lift) test flight.  Initial procurement funding
is planned for fiscal year 2000, and the system test flight is
scheduled for as early as June 2001, but not later than December
2001.  This test flight could have a relatively high risk of failure
because, as indicated in DOD's space launch modernization plan, (1)
historically, most launch systems have had several failures during
their early flight period and (2) generally, failure rates increase
subsequent to a major design or operational change.  In the event
that the EELV test flight does not perform as required, the result
could be costly modifications to production vehicles. 

DOD usually considers the initial production strategy and authorizes
the initial production quantities concurrently with the EMD decision. 
Our experience has shown, however, that once the initial production
decision is made, the options available to decisionmakers, when a
system is found to be deficient, are significantly limited.\1 DOD has
the latitude to modify its usual concurrent approach and schedule a
separate initial production milestone authorization at a later point
when more program risk assessment information is available.  The
opportunity to do this would extend up to 15 months from June
1998--the planned EMD decision date--to October 1999--the earliest
that fiscal year 2000 procurement funds could be obligated. 

EELV acquisition plans show that all 29 of the initial production
vehicles--20 for DOD and 9 for other U.S.  government
organizations--will be used for launching operational satellites, and
none will be used solely for operational test and evaluation
purposes.  Generally, DOD's major programs include separate
production-representative articles for operational testing.  However,
according to the acquisition plans, cost dictates that there not be
any EELV operational "test article" per se.  Instead, assessments are
to be performed on the operational flights.  Although this strategy
may be economically sound, there is increased mission risk to costly
national security and civil satellites because of not having
assurance that a production-representative vehicle will perform as
intended. 


--------------------
\1 Weapons Acquisition:  Low-Rate Initial Production Used to Buy
Weapon Systems Prematurely (GAO/NSIAD-95-18, Nov.  21, 1994). 


      PROGRAM COST ESTIMATING
      DIFFERENCES ARE A MAJOR
      ISSUE
---------------------------------------------------------- Letter :3.2

EELV program cost uncertainty became evident during the December 1996
Defense Acquisition Board review process when estimates were prepared
for system development and production and launch costs.  The
uncertainty was manifested in significantly different cost estimates
done by OSD and the Air Force, which used different assumptions and
methodologies.  Both OSD and Air Force estimates were higher than the
$2-billion development cost objective, and the calculations differed
by several hundred million dollars.  Regarding production and launch
costs, both OSD and Air Force primary calculations showed cost
reductions that exceeded the 25- percent cost-reduction goal. 
However, differences between the two organizations ranged from over
$1 billion to about $2 billion.  In addition, a separate OSD
analysis, using a different assumption, showed that the minimum
25-percent cost-reduction goal would not be met.  Overall, OSD's
position was that the program would likely cost more than the Air
Force estimated. 

A complicating cost-estimating factor was fluctuations in the
national mission model\2 --the primary concern being a decrease in
heavy-lift vehicle requirements.  The model was used to prepare a
launch cost baseline, assuming the use of existing launch vehicles,
to compare with the estimated EELV production and launch costs for
the purpose of assessing the achievement of the program
cost-reduction goal.  In addition to the heavy-lift requirements
issue, the baseline was questionable because of long-term predictions
of U.S.  satellite launches that extended 25 years to fiscal year
2020.  OSD cost analysts characterized the production and launch cost
estimates as highly uncertain. 

DOD recognizes that EELV cost estimates need more attention.  The
Under Secretary of Defense for Acquisition and Technology directed
that every effort should be made to further understand the cost
differences that exist between the Air Force and OSD for both EMD and
production. 


--------------------
\2 The national mission model is a long-range mission requirements
plan, prepared periodically by the Air Force Space Command that lists
planned U.S.  space launches. 


      ADDITIONAL SATELLITE COSTS
      ARE ALREADY EVIDENT
---------------------------------------------------------- Letter :3.3

As discussed in DOD's 1994 space launch modernization plan,
redesigning satellites to fly on new launch vehicles is extremely
costly.  Therefore, a key consideration in establishing the EELV
program schedule was to minimize the cost of satellite redesign and
integration with the vehicle.\3 This was to be achieved by making the
transition from the existing Delta, Atlas, and Titan vehicles to EELV
at planned satellite design change or satellite constellation
replenishment points. 

We identified at least $117 million in additional costs, separate
from the EELV program, that satellite programs expect to incur as a
result of the transition to EELV.  These satellite systems--the
Defense Satellite Communication System (DSCS), Defense Meteorological
Satellite Program (DMSP), Defense Support Program (DSP), and Global
Positioning System (GPS)--were specifically designed to be launched
on the Delta, Atlas, or Titan vehicles.  Each system must now have
payload interface adapters, or associated equipment, designed and
developed to allow them to be launched on EELV.  A new satellite
system currently under development--the Space-Based Infrared System
(SBIRS)--is to be made compatible with both the Atlas and EELV in the
eventuality that EELV may not be available when needed. 


--------------------
\3 Integration in this context means mating the satellite and vehicle
to each other using compatible mechanical and electrical interfaces. 
To aid in reducing costs, EELV requires a single standard satellite
interface design for each vehicle class in the EELV family. 


   SCHEDULE ISSUES TO ADDRESS
------------------------------------------------------------ Letter :4

The interrelationship between the EELV program schedule and other
space-related activities involves some significant risks that DOD
needs to address.  The activities affected are existing launch
vehicles, future satellite launches, and launch facilities.  Ensuring
that the EELV program and these activities are effectively
coordinated is essential to preclude schedule disruptions, cost
increases, and adverse effects on operational satellite schedules and
missions. 


      VEHICLE TRANSITION PLANS
      CONTAIN FUTURE SATELLITE
      LAUNCH SCHEDULE AND MISSION
      RISKS
---------------------------------------------------------- Letter :4.1

DOD either has purchased or will purchase the last of its Delta,
Atlas, and Titan launch vehicles before the first EELV system
development test flight in fiscal year 2001.  An unsuccessful system
test flight, coupled with the planned expiration of existing vehicle
contracts, would (1) delay EELV's availability, (2) create a void in
the government's in-house medium-lift launch capability, and (3)
place national security and civil satellite launch schedules and
missions at risk. 

Replacing the existing medium- and heavy-lift launch vehicle fleet
with EELV requires effective planning to ensure that the continuity
of scheduled satellite launches is maintained while minimizing the
cost of retaining duplicate launch capabilities.  However, in
planning this transition, DOD representatives told us that they had
not yet assessed the need for or feasibility of either extending
existing vehicle contracts or using U.S.  commercial launch vehicles
(such as Delta III or Atlas IIAR) as an alternative to ensure the
continuity of planned satellite launches. 


      LAUNCH FACILITY TRANSITION
      REQUIREMENTS COULD DELAY
      PROGRAM
---------------------------------------------------------- Letter :4.2

The Air Force EELV program office identified the meeting of launch
facility preparation schedules at the Cape Canaveral and Vandenberg
launch ranges as a significant program risk.  The Air Force Space
Command characterized the required lead times for facility projects
as the greatest risk to meeting EELV operational milestones. 
Effective transition planning at the ranges is critical to ensure
that existing space launch facilities are available for satellite
launches and that EELV facilities are available on schedule. 

There are two specific areas of concern--(1) conflicts between the
planned use of facilities for the EELV program and the existing use
of these same facilities by other government and commercial launch
programs and (2) the length of time normally required to complete the
environmental regulatory process associated with modifying existing
facilities or building new ones for EELV. 

The most critical goal is to have the facilities available for the
first EELV system development test flight in June 2001.  However,
some significant launch facility conflicts could cause major
disruptions to ongoing programs, requiring DOD to address facility
priorities.  In addition, based on nominal time frames associated
with the environmental regulatory and construction process, the EMD
contract award, and possibly the system development test flight,
could be delayed by 8 months. 


   POTENTIAL PERFORMANCE ISSUES
------------------------------------------------------------ Letter :5

The Air Force identified vehicle propulsion, systems integration, and
software as technical risk areas.  Major changes in vehicle
propulsion systems are expected that could require significant
development, and propulsion represents a significant portion of EELV
estimated costs.  Systems integration, which involves combining all
design, engineering, testing, manufacturing, and launch functions and
the software information system necessary to complete a project, is a
common risk in most programs.  Although the Air Force has required
the contractors to submit risk mitigation plans, these risk areas
could still pose significant system development challenges and
adversely affect EELV program cost and schedule goals. 


   COMMERCIAL APPLICATION FOR EELV
------------------------------------------------------------ Letter :6

DOD's 1994 space launch modernization plan stated that although the
four national space community sectors--defense, intelligence, civil,
and commercial--have distinct space missions with their own unique
cultures and practices, they have one thing in common--the
requirement for space launch.  In this regard, the plan referred to a
natural synergy that could be created with the commercial sector.  It
cited the emergence of the commercial satellite market during the
past several years--particularly communication satellites--as a
significant driver for launch vehicles and an opportunity for
potential private sector investment in space launch. 

DOD has an interest in seeing that EELV is used for commercial
purposes in order to lower EELV costs.  For example, the EELV
acquisition plan states that the government is interested in the
competing contractors' ability to develop a successful commercial
EELV system, which should result in achieving recurring cost
reductions by virtue of a significantly larger customer base
(government and commercial) for the EELV contractor. 

In addition, both contractors have indicated that they intend to make
private investments in EELV development, and they have an incentive
to do so because of the potential to enhance their positions in the
international commercial markets.  In December 1996, the Air Force
informed the Defense Acquisition Board of DOD's potential to benefit
from contractor commercial (private) investment in EELV.  However,
the contractors are not obligated by the government to make such
investments under the existing pre-EMD contracts.  According to
program officials, the option of contractually binding the winning
EMD contractor to such an investment is available, if it is
considered prudent. 

Considering the commercial benefit to the winning EMD contractor from
using the EELV design, a mechanism to ensure some reduction to the
government's estimated $2 billion investment would be reasonable. 
From a government perspective, the question is how the contractor,
who stands to benefit substantially in the commercial market place
from the government's investment, should compensate the government
for that investment.  Alternatives could be for DOD to employ a
cost-sharing contract for the EMD phase and/or arrange for the
government to recoup part of its investment based on commercial
launch vehicle sales. 


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

Considering the cost and schedule issues associated with the EELV
program, we recommend that the Secretary of Defense: 

  Either (1) revise the program strategy, by decoupling the planned
     concurrent engineering and manufacturing development decision
     and initial production authorization, to take advantage of the
     most current program risk assessment information available prior
     to obligating procurement funds planned for fiscal year 2000 or
     (2) review the initial production authorization prior to
     obligating any procurement funds, if that authorization is made
     concurrently with the engineering and manufacturing development
     decision. 

  Develop contingency plans to (1) meet national security and civil
     satellite launch schedules when the existing launch vehicle
     production contracts expire and (2) address the potential for
     delay in the availability of launch facilities. 

In view of the expected compensating benefits to the winning EELV
contractor to enhance its competitive position in the international
commercial launch vehicle market, we recommend that the Secretary of
Defense devise a mechanism, such as a cost-sharing approach and/or a
recoupment arrangement for commercial launch vehicle sales, to help
reduce the government's investment in EELV and see that the mechanism
is included in the Air Force's request for proposal for the EMD
acquisition phase of the EELV program. 


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

In commenting on a draft of this report, DOD generally agreed with
two of our recommendations and disagreed with one.  We modified that
recommendation in an attempt to address DOD's concerns while still
retaining the thrust of what we believe needs to be done. 

  DOD stated that it had initially implemented a launch vehicle
     transition strategy and would continue to refine it as the EELV
     program matures.  DOD indicated that the commercial demand for
     Delta and Atlas vehicles would keep the production lines
     operational during the transition.  DOD also indicated that the
     Air Force was tracking the potential delay in the availability
     of EELV launch facilities and developing risk mitigation plans. 

  DOD stated that several cost-sharing approaches will be evaluated
     during the next 12 months; the actual arrangement for any cost
     sharing between the government and the winning EELV contractor
     would be included, as appropriate, in the source selection
     process; and the approach that provides the best value to the
     government would be incorporated into the EMD contract. 

  DOD disagreed with the recommendation in our draft report on
     refraining from authorizing EELV production concurrently with
     the EMD decision and holding a separate production decision
     meeting subsequent to the planned EMD decision to take advantage
     of the most current program risk assessment information
     available prior to obligating procurement funds.  Although DOD
     agreed that the decision to begin full production (milestone
     III) should be based on as much program risk data as possible,
     it viewed the implementation of our recommendation as limiting
     the Defense Acquisition Executive's flexibility in the milestone
     II acquisition review process. 

Concerning the last item, we believe that the extent of the risks in
the EELV program makes a concurrent EMD decision and initial
production authorization unwise.  We also believe that the Defense
Acquisition Executive's flexibility should not be restricted in
making the best milestone II decision.  Accordingly, we modified our
recommendation to call upon the Secretary to choose between (1)
revising the EELV program strategy by decoupling the planned
concurrent EMD decision and initial production authorization or (2)
reviewing the initial production authorization prior to obligating
any procurement funds, if that authorization is made concurrently
with the EMD decision. 

DOD's comments are reprinted in appendix II. 


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

We evaluated the Air Force's plans and progress in developing EELV. 
We reviewed the interrelationships among EELV and existing launch
vehicles, satellite programs, and launch facilities.  We also
reviewed program cost, schedule, and performance information; program
risk areas; transition plans; and national mission model data.  We
specifically examined acquisition planning documents, budget
information, launch requirements, contractor proposals, launch
facility plans, and space policies and studies. 

We performed our work primarily at the Air Force Space and Missile
Systems Center in El Segundo, California.  In addition, we held
discussions with representatives from the Office of the Deputy Under
Secretary of Defense for Space, the Office of the Secretary of
Defense's Program Analysis and Evaluation Directorate, the Air
Force's Office of the Assistant Secretary for Acquisition, and the
Air Force Cost Analysis Agency, Washington, D.C.  We also held
discussions with representatives from the Air Force Space Command,
Colorado Springs, Colorado; the Air Force's 30th and 45th Space
Wings, Vandenberg Air Force Base, California, and Patrick Air Force
Base, Florida, respectively; the Air Force's Phillips Laboratory
Propulsion Directorate, Edwards Air Force Base, California; the Air
Force Operational Test and Evaluation Center, Albuquerque, New
Mexico; and the Aerospace Corporation, El Segundo, California. 

Competition sensitive information associated with the ongoing EELV
program acquisition is not disclosed in this report. 

Appendix I provides detailed information on the EELV program. 

We performed our review between January 1996 and January 1997 in
accordance with generally accepted government auditing standards. 


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

We are sending copies of this report to the Ranking Minority Member,
Subcommittee on National Security, House Committee on Appropriations
and to the Chairmen and Ranking Minority Members of the House
Committee on National Security; the Senate Committee on Armed
Services; and the Senate Subcommittee on Defense, Committee on
Appropriations.  We are also sending copies to the Secretary of
Defense and the Director, Office of Management and Budget.  We will
make copies available to others upon request. 

This report was prepared under the direction of Thomas J.  Brew,
Associate Director, Defense Acquisitions Issues, who may be reached
on (202) 512-4841 if you or your staff have any questions.  Major
contributors to this report are listed in appendix III. 

Sincerely yours,

Louis J.  Rodrigues
Director, Defense Acquisitions Issues


EVOLVED EXPENDABLE LAUNCH VEHICLE
PROGRAM PLANS AND ISSUES
=========================================================== Appendix I

The federal government uses expendable launch vehicles to provide
transportation for communication, navigation, ballistic missile
warning, meteorological, environmental, intelligence, and scientific
satellites into space.  U.S.  policy asserts that access to and use
of space is central to preserving peace and protecting national
security as well as civil and commercial interests. 

Since 1987, the United States has been attempting to develop a new
launch vehicle.  The Advanced Launch System program during 1987 to
1990, the National Launch System program from 1991 to 1992, and the
Spacelifter program in 1993 were each subsequently canceled because
of funding issues, changing requirements, or controversy over the
best way to address national launch needs.  These development efforts
resulted from a combination of factors, including (1) policy
decisions in the late 1970s regarding exclusive reliance on the space
shuttle for space transportation, and the corollary action to
terminate investments in expendable launch vehicles; (2) the 1986
space shuttle Challenger accident; (3) restoration of expendable
launch vehicle programs that were based on 1960's and 1970's
technology, which did not take advantage of newer technology and
manufacturing concepts; and (4) a concern about reduced
competitiveness in the international launch market.\1

The existing launch vehicles--Delta II, Atlas II, Titan II, and Titan
IV--were derived, to one degree or another, from ballistic missile
systems, and currently operate at or near their maximum performance
capability.  In addition, these vehicles lack standardization, even
among different configurations of the same vehicle.  In its October
1993 Bottom-Up Review report, the Department of Defense (DOD)
characterized U.S.  military space launch capabilities as very
costly, with serious operational limitations.  The report stated that
the existing expendable launch vehicles (1) were able to meet their
performance requirement of delivering satellites to a specific orbit,
but with less than desired reliability and (2) fell short of the
operational flexibility requirement--meaning the capability to
perform rapid payload integration, servicing, substitution, and
launch. 

In November 1993, the Congress directed the Secretary of Defense to
develop a space launch modernization plan with clearly defined
priorities, goals, and milestones regarding modernization of space
launch capabilities for DOD or, if appropriate, the government as a
whole.\2

In May 1994, DOD's space launch modernization plan (known as the
Moorman study) discussed the increasing hardware costs associated
with DOD's medium- and heavy-lift launch vehicles, with particular
emphasis on the heavy-lift Titan IV and its inefficient production
rates.  In addition, it discussed the manpower intensive aspects of
launch system manufacturing and operations, also with particular
emphasis on Titan IV, and the multiple launch complexes at Cape
Canaveral and Vandenberg--the Air Force's two space launch ranges. 
The plan provided four options to alleviate these conditions--(1)
sustain existing systems, including austere upgrades; (2) evolve
existing systems; (3) develop a new expendable system; and (4)
develop a new reusable system.  DOD chose to pursue the second option
as a cost-saving measure and to accommodate schedule opportunities
when several satellite systems were to undergo design changes.  In
September 1994, the Congress provided the initial funds to develop a
new family of medium- and heavy-lift expendable launch vehicles
evolved from existing technologies.\3

In November 1994, DOD developed an Evolved Expendable Launch Vehicle
(EELV) implementation plan, stating that the program objective was to
reduce total cost for medium- and heavy-lift vehicle space launch. 
The plan summarized DOD's launch assets as including 11 launch pads,
5 launch teams, 3 launch vehicle production and processing
industries, 2 launch ranges, and various support resources.  It
discussed an EELV program strategy to incorporate industrial
competition, resulting in a single production contract that would (1)
maximize common systems and components to reduce procurement costs
and enhance production rates and (2) decrease the number of launch
complexes, launch crews, and support requirements to reduce operation
costs. 


--------------------
\1 Final Report to the President on the U.S.  Space Program, from the
Vice President, Chairman of the National Space Council, Jan.  7,
1993. 

\2 National Defense Authorization Act for Fiscal Year 1994 (P.L. 
103-160, Nov.  30, 1993). 

\3 Department of Defense Appropriations Act, 1995, (P.L.  103-335,
Sept.  30, 1994). 


   ACQUISITION STRATEGY AND STATUS
--------------------------------------------------------- Appendix I:1

EELV is intended to be the federal government's only medium- and
heavy-lift expendable space transportation capability for several
years after the beginning of the 21st century.  This planned family
of vehicles is intended to launch the government's portion of the
national mission model, which is currently being launched by the
existing fleet of Delta, Atlas, and Titan vehicles.  For EELV, this
model consists of 193 government launches for fiscal years 2002
through 2020--177 for defense and intelligence purposes and 16 for
the National Aeronautics and Space Administration. 

In May 1995, the Office of the Secretary of Defense (OSD) established
an EELV development cost objective of $2 billion (in then-year
dollars).  The Air Force's acquisition plan included a threshold
amount not to exceed $2.3 billion.  In addition, the Air Force Space
Command established a goal of reducing EELV production and launch
costs by a minimum of 25 percent from the estimated costs of using
existing launch vehicles, with an objective of 50 percent.  In August
1995, the Air Force awarded four competitive $30 million contracts
for a low-cost concept validation phase.  The purpose was to provide
system specifications, cost estimates, trade-off analyses, risk
mitigation results, environmental analysis reports, and manufacturing
plans. 

In December 1996, DOD decided to proceed into a 17-month
pre-engineering and manufacturing development (pre-EMD) phase, and
the Air Force awarded competitive $60 million development contracts
to both Lockheed Martin Astronautics and McDonnell Douglas Aerospace. 
The purpose of this pre-EMD phase is to refine system specifications,
update cost estimates, complete risk-reduction efforts, and support
the government in the environmental regulatory process. 

In June 1998, DOD plans to decide whether to proceed into a 6-year
EMD phase, and the Air Force anticipates issuing one
cost-plus-award-fee development contract for an anticipated $1.6
billion.  During this phase, the contractor is to provide detailed
system specifications, perform two system test flights (one
medium-lift and one heavy-lift), validate manufacturing processes,
and activate the launch sites.  Also, at this EMD decision point, DOD
plans to authorize the initial production strategy and quantities. 
Air Force documents show that 29 initial production vehicles are
planned--20 for DOD and 9 for other U.S.  government
organizations--which we estimated to cost about $1.5 billion, based
on Air Force budget information.  A decision on whether to produce
larger quantities that would cost several billion dollars is also to
be made during the EMD phase.  In its fiscal year 1998 research,
development, test, and evaluation budget, the Air Force is requesting
$91.6 million for EELV.  Of this amount, $28.4 million is to initiate
EMD. 


   ACQUISITION PLANS CONTAIN COST
   AND MISSION RISKS
--------------------------------------------------------- Appendix I:2

Air Force EELV acquisition plans show that the 29 initial production
vehicles would be procured over a 4-year period (fiscal years
2000-2003) during the EMD phase.  Assuming appropriations are
provided, missile procurement funds for the first six initial
production vehicles would be obligated in fiscal year 2000--at least
1 year, and possibly
2 years, before the first system development (medium-lift) test
flight.  The preferred date for this test flight is June 2001, and
the threshold date is not later than December 2001.  Missile
procurement funds for a second set of six vehicles are planned to be
requested for fiscal year 2001.  If these funds are appropriated at
the outset of the fiscal year, they could also be obligated before
this test flight actually occurs. 

Although the Air Force acquisition plan assumes a 2-year lead time
from initiating production to delivering and launching a vehicle, it
states that the industry has historically required over 2 years. 
According to the plan, if more than 2 years are needed, missile
procurement funds could be requested as early as fiscal year 1999--at
least 2 years, and possibly 3 years, before the test flight.  
Figure I.1 shows the current funding and launch plans for the 29 
initial production vehicles and the associated concurrency of 
these plans relative to the EMD schedule. 


Source:  Air Force EELV acquisition plan. 

We have reported on numerous occasions about the risks associated
with program concurrency and initiating production without adequate
testing.  For example, in 1990, we concluded that although
concurrency can expedite the development and production of weapon
systems, entering production before critical tests are successfully
completed has resulted in the purchase of systems that do not perform
as intended.  In 1994, we reported that programs are often permitted
to begin production with little or no scrutiny, and the consequences
have included procuring substantial inventories of unsatisfactory
weapons requiring costly modification to achieve satisfactory
performance.  Once production is started, the options available to
decisionmakers, when a system is found to be deficient, are
significantly limited.\4 DOD usually considers the initial production
strategy and approves the quantities concurrently with the EMD
decision.  Generally, a favorable decision authorizes the program
manager to initiate production. 

The EELV acquisition plan states that there will be no initial
production vehicles dedicated solely for operational testing because
of high vehicle cost.  Instead, the Air Force plans to use all of
these vehicles for operational purposes by launching navigational,
missile warning, communications, meteorological, scientific, and
classified satellites.  Although this strategy may be economically
sound, there is increased mission risk to costly national security
and civil satellites because of not having assurance that a
production-representative vehicle will perform as intended.  The
planned involvement by the Air Force Operational Test and Evaluation
Center in the EELV program is to acquire and evaluate data from the
first system development test flight and the initial six operational
flights to support a decision to continue long-term production in
fiscal year 2003. 


--------------------
\4 Weapon Systems:  Concurrency in the Acquisition Process
(GAO/T-NSIAD-90-43, May 17, 1990) and Weapons Acquisition:  Low-Rate
Initial Production Used to Buy Weapon Systems Prematurely
(GAO/NSIAD-95-18, Nov.  21, 1994). 


   PROGRAM COST-ESTIMATING
   DIFFERENCES ARE A MAJOR ISSUE
--------------------------------------------------------- Appendix I:3

EELV program cost-estimating differences between OSD and the Air
Force became evident during the December 1996 Defense Acquisition
Board review process.  The primary reasons given were the use of
different assumptions and methodologies. 


      DEVELOPMENT COST OBJECTIVE
------------------------------------------------------- Appendix I:3.1

DOD's development cost objective of $2 billion was for all three
development phases--low-cost concept validation, pre-EMD, and EMD. 
This objective was based on the 1994 space launch modernization plan,
which estimated that nonrecurring costs for evolving a family of
medium- and heavy-lift launch vehicles were in the range of $1
billion to $2.5 billion.  The study's wide cost range was largely due
to the lack of detailed engineering and program estimates for this
particular evolved expendable launch vehicle approach. 

Of the $2-billion objective, $1.6 billion was planned for EMD, which
is to be performed by one contractor.  OSD's development cost
assessment for the EMD phase (in constant 1995 dollars) was several
hundred million dollars higher than the Air Force estimate, and both
calculations exceeded the objective.  (Details regarding cost
estimates are considered competition sensitive and therefore not
disclosed.)


      PRODUCTION AND LAUNCH
      COST-REDUCTION GOAL
------------------------------------------------------- Appendix I:3.2

The Air Force Space Command's 25 to 50 percent production and launch
cost-reduction goal meant that an evolved family of vehicles should
cost less than if existing vehicles were used.  This goal was to be
measured by first establishing an estimated recurring cost--called
the launch cost baseline--of producing and launching Delta, Atlas,
and Titan vehicles to satisfy the government launch needs during
fiscal years 2002 through 2020.  This calculation was then to be
compared with the EELV contractors' proposed cost estimates for
satisfying these launch needs during the same period.  The Air
Force's latest baseline was estimated at $20.6 billion.  It was
prepared by the existing vehicle program offices based on the Air
Force Space Command's January 1996 national mission model for 193
predicted launches. 

OSD's assessment of EELV production and launch costs for fiscal years
2002 through 2020 (in constant 1995 dollars) ranged from over $1
billion to about $2 billion higher than the Air Force's estimate. 
OSD calculations showed cost reductions that exceeded the 25-percent
goal, and the Air Force showed reductions that exceeded the
50-percent goal.  However, there was considerable uncertainty
regarding these recurring costs.  For example, OSD cost analysts
believe that some of the estimated component costs were too low, and
in the process of performing a net present value analysis, determined
that the minimum 25-percent reduction would not be met.  (Details
regarding cost estimates are considered competition sensitive and
therefore not disclosed.)

In addition, the validity of the national mission model, which was
used to prepare the launch cost baseline, was questionable because
(1) of the uncertainty in predicting government space launches 25
years into the future; (2) the model had fluctuated from 171 to 193
launches since the EELV program was established in 1995; and (3) the
model is expected to continue changing, probably downward, because of
recent OSD analyses regarding decreased heavy-lift vehicle
requirements.  According to DOD cost analysts, such long-term
predictions and fluctuations made credible assessments of production
and launch costs and the comparative baseline more complicated. 


      DOD RECOGNIZES THAT COST
      ESTIMATES NEED MORE
      ATTENTION
------------------------------------------------------- Appendix I:3.3

As a result of a Defense Acquisition Board review, the Under
Secretary of Defense for Acquisition and Technology authorized the
Air Force to proceed with the EELV program into the pre-EMD phase of
the acquisition process.  However, considering the differences in
estimated program costs, the Under Secretary emphasized that every
effort should be made to understand the cost differences between the
Air Force and OSD estimates for both EMD and production.  He
indicated that the variability of these cost estimates, which stem
from both increases and decreases in the national mission model, as
well as the effect of varying requirements for heavy-lift
capabilities, should be fully explored. 

OSD established criteria for the program to exit the pre-EMD phase
and enter the EMD phase, which is scheduled for June 1998.  These
criteria included (1) preparing an updated life-cycle cost estimate
with a detailed cost risk analysis and (2) performing an
independently reviewed economic investment analysis that would
identify projected recurring cost savings and investment payback.  An
independent program cost estimate is a statutory requirement under 10
U.S.C.  2434 for entry into EMD. 


   ADDITIONAL SATELLITE COSTS ARE
   ALREADY EVIDENT
--------------------------------------------------------- Appendix I:4

Five Defense Satellite Communications System (DSCS) satellites, which
have been built, remain to be launched during fiscal years 1998
through 2003 as the transition is made to EELV.  Although these
satellites were designed to be launched on Atlas II vehicles, the
last two are now scheduled to be launched on EELV.  According to
program representatives, an additional $25 million has been budgeted
for a payload interface adapter design and modification that will be
needed for these last two satellites. 

Six Defense Meteorological Satellite Program (DMSP) satellites, which
have been built and are currently in storage, remain to be launched
during fiscal years 1998 through 2007 as the transition is made to
EELV.  Although these satellites were designed to be launched on
Titan II vehicles, the last four are now scheduled to be launched on
EELV.  According to program representatives, an additional $28 to $30
million has been estimated for a satellite payload interface adapter
design and modification effort that will take about 5 years.  They
believe that other costs could be incurred because there is concern
that the satellites may need to be modified to withstand the stress
anticipated from an EELV launch.  Whether this will be necessary will
not be known until the EELV launch tolerance parameters are
demonstrated during the first system development flight test in
fiscal year 2001. 

According to program officials, five Defense Support Program (DSP)
satellites are either built and in storage or being fabricated and
are scheduled to be launched during fiscal years 1998 through 2003 as
the transition is made to EELV.  Although these satellites were
designed to be launched on either Titan IV vehicles or the Space
Shuttle, the last satellite is now planned to be launched on the EELV
heavy-lift test vehicle in fiscal year 2003.  According to program
representatives, an additional $29 million is needed for hardware and
cable associated with satellite and vehicle integration. 

The Global Positioning System (GPS) is currently launched on Delta II
vehicles, and the satellite system is undergoing a redesign, referred
to as a block change, from the IIR to the IIF version.  Although most
of the IIF satellites are planned to be launched on EELV, the first
satellite must be launched on Delta II.  This is because the
scheduled launch date is in fiscal year 2001 and the first
operational EELV flight is not scheduled until fiscal year 2002.  In
addition, according to a program official, all IIF satellites need to
be compatible with both launch vehicles because there is some
uncertainty regarding the stress anticipated with an EELV launch. 
GPS program representatives informed us that about $35 million would
be needed to develop and build a payload interface adapter for EELV. 

The new Space-Based Infrared System (SBIRS) satellite program
includes a geosynchronous earth orbit component and a low-earth orbit
component.  This program is intended to replace the DSP satellite
system.  SBIRS program officials initially intended to use EELV for
the first geosynchronous launch scheduled for fiscal year 2002. 
However, because they want to mitigate the risk of delay in EELV
availability, SBIRS is to be compatible with both EELV and Atlas. 
The deployment schedule for the low-earth orbit component is yet to
be finalized.  DOD originally scheduled the first launch for fiscal
year 2006 using EELV; the Congress directed the first launch to be in
fiscal year 2002; and now DOD is revising its schedule for launch in
fiscal year 2004. 


   VEHICLE TRANSITION PLANS
   CONTAIN FUTURE SATELLITE LAUNCH
   SCHEDULE AND MISSION RISKS
--------------------------------------------------------- Appendix I:5

The Delta, Atlas, and Titan programs are managed under separate Air
Force production contracts and each have a different expiration date. 
Considering when the last procurement action is scheduled under these
vehicle contracts relative to the schedule for the first EELV system
development test flight, DOD appears to be taking some risk regarding
future satellite launch schedules and missions.  Figure I.2 shows the
last scheduled procurement actions and the last planned launches for
the government's existing launch vehicles, relative to key EELV
scheduled events. 



Source:  Air Force Program acquisition plans. 

The last of 21 Delta II launch vehicles for the government is to be
procured in fiscal year 2000--1 year before the first EELV system
development test flight.  The last government Delta launch is planned
for fiscal year 2002--the same year that medium-lift EELV operational
flights are to begin. 

Regarding Atlas II, a contract option is to be executed for the last
of six government launch vehicles in fiscal year 1998--3 years before
the first EELV system development test flight.  The last government
Atlas launch is planned for fiscal year 2002--the same year that
medium-lift EELV operational flights are to begin. 

The Titan II space launch vehicles were converted from deactivated
intercontinental ballistic missiles, thus there was no production
activity.  According to program officials, after the last Titan II
launch in fiscal year 1999, the Air Force plans to deactivate the
launch pad. 

Since 1991, Titan IV production rates have declined from 10 to 2
vehicles per year because of reduced requirements, with the resulting
effect of increasing unit costs.  The last Titan IV purchase was in
fiscal year 1995 for 2 of 41 vehicles.  Any follow-on procurement is
in doubt because of uncertainties regarding future DOD heavy-lift
requirements.  The last Titan IV launch is planned to occur 2 years
after the scheduled heavy-lift test flight. 

In commenting on a draft of this report, DOD indicated that although
these last procurement actions are scheduled, the vehicle contracts
will still be in place, giving DOD the opportunity to procure
additional vehicles, if required.  However, such extensions usually
require contract renegotiations, and the result is usually price
increases.  DOD also indicated that commercial launch vehicles may be
available to mitigate some of the risk. 


   LAUNCH FACILITY TRANSITION
   REQUIREMENTS COULD DELAY
   PROGRAM
--------------------------------------------------------- Appendix I:6

There are three factors affecting a smooth launch facility transition
at the Cape Canaveral and Vandenberg launch ranges--existing facility
conflicts, environmental regulatory requirements, and the amount of
time needed for facility construction. 


      EXISTING FACILITY CONFLICTS
------------------------------------------------------- Appendix I:6.1

In April 1996, the Air Force Space Command prepared an EELV plan that
identified the need to coordinate efforts at the Cape Canaveral and
Vandenberg launch ranges for making the transition from existing
launch vehicles to EELV.  At these ranges, the competing contractors
were expected to use a combination of existing facilities, modified
facilities, and possibly new facilities for the various launch
support functions, such as material receiving and storage, vehicle
and upper stage processing, payload integration, final assembly, and
launch services and operations. 

In August 1996, the Command completed a launch site facility baseline
study to provide the EELV program office with facility information
necessary for risk assessment, management decisions, and answering
contractors' questions.  The study identified several existing
facilities at Cape Canaveral and Vandenberg that were of interest to
the EELV contractors but that were being used by other government or
commercial programs.  Some of these facilities were forecasted to be
used through mid-to-late fiscal year 2000 or beyond and would require
negotiations with the existing user regarding availability for EELV. 
Although resolution of several conflicts is the responsibility of Air
Force launch range officials, there were indications that significant
issues associated with certain facility priorities may have to be
addressed at the DOD level.  (Specific conflicts are considered
competition sensitive and therefore not disclosed.)


      ENVIRONMENTAL REGULATORY
      REQUIREMENTS
------------------------------------------------------- Appendix I:6.2

Before construction of EELV facilities can begin at the Cape
Canaveral and Vandenberg launch ranges, the Air Force must complete
an environmental impact statement.  According to the EELV acquisition
plan, the environmental process, which is based on National
Environmental Policy Act requirements, is to be completed during the
pre-EMD phase, prior to committing EMD funds.  However, the plan also
states that it takes about
24 months to obtain approval.  A representative from the Air Force
Center for Environmental Excellence confirmed this estimate. 

EELV program officials informed us that some preliminary
environmental work was started during the low-cost concept validation
phase.  However, the actual contract for the environmental effort was
not awarded until February 1997.  This means that to initiate EMD on
schedule in June 1998, only 16 months are available to complete the
nominal 24-month environmental effort.  Thus, the EMD contract award
could be delayed up to 8 months, until February 1999. 


      FACILITY MODIFICATION AND
      CONSTRUCTION REQUIREMENTS
------------------------------------------------------- Appendix I:6.3

Existing facilities cannot be modified and new facilities cannot be
constructed at the launch sites until (1) all environmental
requirements are completed and (2) DOD decides to proceed into the
EMD phase.  In its transition plan, the Air Force Space Command urged
that sufficient time be set aside to allow for facility planning,
programming, design, construction, and acceptance.  It used
traditional planning factors in estimating the duration of facility
projects, such as 2 to 3 years for modifications to existing
facilities and 4 years for new construction. 

According to EELV program officials, the nominal facility
construction time is 36 months.  Given the requirement for a
sequential environmental and construction process, and the possible
8-month delay in initiating EMD because of environmental regulatory
requirements, the first system development test flight could be
delayed by 8 months from the preferred date of June 2001 to February
2002. 


   POTENTIAL PERFORMANCE ISSUES
--------------------------------------------------------- Appendix I:7

The Air Force identified vehicle propulsion, systems integration, and
software as technical risk items.  EELV propulsion could be a primary
area of risk because of the significant anticipated development. 
Although systems integration, including development and reuse of
software, is a common risk area, it is expected to be a major system
development challenge.  Problems in these areas could result in an
adverse effect on EELV program cost and schedule goals. 


      PROPULSION COULD BE A
      PRIMARY RISK
------------------------------------------------------- Appendix I:7.1

From inception, the EELV system design was intended to be evolved
from the existing medium- and heavy-lift vehicles into one family of
vehicles.  Despite this evolutionary concept, major changes in
vehicle propulsion systems are expected that could require
significant development.  Changing propulsion systems appears to be
necessary considering that the 1994 space launch modernization plan
stated that (1) existing propulsion systems (both solid and liquid
propellent variants) were the cause for 25 to 50 percent of the
launch vehicle failures in previous years and (2) there was general
consensus that propulsion technology was the most serious area of
deficiency in the existing U.S.  launch vehicle fleet.  In addition
to the expected technical risk, propulsion represents a significant
portion of the estimated launch vehicle costs.  Thus, development,
testing, and integration of propulsion components could pose special
challenges to the EELV program. 


      SYSTEMS INTEGRATION AND
      SOFTWARE ARE USUAL RISKS
------------------------------------------------------- Appendix I:7.2

Systems integration involves combining all design, engineering,
testing, manufacturing, and launch functions, as well as the software
information system, that are essential to complete the intended
project.  Systems integration problems can occur, even though the
various components and subsystems performed successfully on previous
systems.  As an example, similar to EELV, the Air Force's C-17
aircraft program intended to use current, available, and proven
technology to minimize development costs and structure a low
technical risk effort.  The integration of sophisticated technologies
into a workable aircraft design was a major engineering and
management task that eventually contributed to significant cost
increases and schedule delays.\5 EELV system integration could be
similarly challenging. 

Air Force officials informed us that new computer software would be
developed and existing software (from other programs) would be reused
for EELV.  They stated that software issues are particularly
challenging and that program cost and schedule could be affected. 
Because of this, they have performed an assessment of the
contractors' ability to develop software and intend to perform
another one to help mitigate the risk. 

Reusing software can be cost-effective, but it can present
significant problems.  For example, in June 1996, the initial flight
of the European Ariane 5 launch vehicle failed because of inadequate
software testing.  The software was used successfully on previous
Ariane 4 launch vehicles and then reused on Ariane 5.  According to
the report of an inquiry board established to perform an
investigation, the failure was caused by the complete loss of
guidance and attitude information resulting from specification and
design errors in the software of the inertial reference system.  The
report stated that (1) there was inadequate analysis and testing of
systems that included the reused software and (2) if testing had been
performed, the potential failure could have been detected. 





--------------------
\5 Military Airlift:  Cost and Complexity of the C-17 Aircraft
Research and Development Program (GAO/NSIAD-91-5, Mar.  19, 1991). 

Appendix II
COMMENTS FROM THE DEPARTMENT OF
DEFENSE
=========================================================== Appendix II






MAJOR CONTRIBUTORS TO THIS REPORT
========================================================= Appendix III

NATIONAL SECURITY AND
INTERNATIONAL AFFAIRS DIVISION,
WASHINGTON, D.C. 

Homer H.  Thomson

LOS ANGELES OFFICE

Samuel S.  Van Wagner
Allen D.  Westheimer
Larry J.  Bridges
Lorene S.  Sarne


*** End of document. ***