CONGRESS OF THE UNITED STATES

CONGRESSIONAL BUDGET OFFICE

CBO
STUDY
March 2006

Alternatives for
Long-Range
Ground-Attack
Systems

CBO
S T U D Y

Alternatives for Long-Range

Ground-Attack Systems
March 2006

The Congress of the United States O Congressional Budget Office

Note
The cover shows the three types of long-range bombers in service with the United States Air
Force. From the top: the B-2A Spirit stealth bomber (USAF photo by Master Sgt. Val
Gempis); the B-1B Lancer (USAF photo by Master Sgt. Robert W. Valenca); and the B-52H
Stratofortress (USAF photo by Master Sgt. Michael A. Kaplan).

Preface

he recent air campaigns in Afghanistan and Iraq highlighted the utility of long-range
ground-attack systems. The Air Forces fleet of B-52, B-1, and B-2 heavy bombers helped coalition forces overcome the limited availability of local air bases by operating from more distant
bases and provided responsive air support to ground forces by orbiting over the battlefield for
long periods of time. Recognizing those contributions, the Department of Defense (DoD) is
in the process of developing new concepts for the role of long-range systems in future conflicts
and is also beginning to examine new systems that could be used to attack targets anywhere in
the world.
This Congressional Budget Office (CBO) studyprepared at the request of the Subcommittee on Strategic Forces of the Senate Committee on Armed Serviceslooks at the capabilities
and costs associated with alternative long-range strike systems that DoD might develop and
procure to improve its ability to conduct ground-attack operations. The study compares the
advantages, disadvantages, and costs of eight alternative systemsfive aircraft-based systems
and three missile-based systems. In keeping with CBOs mandate to provide objective, impartial analysis, this study makes no recommendations.
Robie Samanta Roy and David Arthur of CBOs National Security Division prepared the
study under the supervision of J. Michael Gilmore. (Robie Samanta Roy has since left CBO.)
David Newman, Raymond Hall, and Matthew Schmit of CBOs Budget Analysis Division
prepared the cost estimates and wrote the appendix under the supervision of Jo Ann Vines.
Kevin Perese and Adebayo Adedeji of CBO and Dr. Mitch Nikolich of CACI-NSR, Inc., provided thoughtful comments. (The assistance of an external participant implies no responsibility for the final product, which rests solely with CBO.)
Janey Cohen edited the study, and John Skeen proofread it. Christian Spoor edited the figures, and Leah Mazade edited the tables. Cynthia Cleveland produced drafts of the study and
formatted the tables. Maureen Costantino designed the cover and prepared the study for publication, and Lenny Skutnik printed the initial copies.

Donald B. Marron
Acting Director
March 2006

Contents
Summary ix

The United States Long-Range Strike Capabilities 1

The Current Long-Range Strike Force 2
Capabilities for Future Long-Range Strike Systems 4

Alternative Designs for Long-Range Strike Systems 7

Technical Considerations for the Design of Long-Range
Strike Platforms 7
The Specific Long-Range Strike Alternatives CBO Examined 11

Comparing the Capabilities of the Long-Range Strike

Alternatives CBO Examined 15
Ability to Reach Targets Anywhere on Earth 16
Responsiveness 18
Sustained Firepower 23
Payload Flexibility 25
Survivability 25

Cost and Force-Structure Implications 27

Aircraft-Based SystemsAlternatives 1 Through 5 27
Common Aero Vehicle SystemsAlternatives 6 Through 8 33
The Relative Difficulty of Developing the Long-Range
Strike Alternatives Considered by CBO 35

Appendix: The Methodology Behind the Cost Analysis 37

VI

ALTERNATIVES FOR LONG-RANGE GROUND-ATTACK SYSTEMS

Tables
S-1.

Long-Range Strike Alternatives Examined by CBO

S-2.

Estimated Costs of Long-Range Strike Alternatives Examined by CBO

1-1.

Characteristics of Current U.S. Long-Range Bombers

1-2.

Maximum Number of Weapons Carried on Current U.S. Long-Range Bombers

2-1.

Long-Range Strike Alternatives Examined by CBO

3-1.

Flight Time Needed to Reach a Target in Afghanistan

20

3-2.

Examples of Target Types for Air-to-Ground Missions

25

4-1.

Estimated Costs of Long-Range Strike Alternatives Examined by CBO

28

A-1.

Characteristics of Supersonic Missiles for the Arsenal Aircraft (Alternative 1)

38

A-2.

Acquisition Costs of Supersonic Missiles for the Arsenal Aircraft

Alternative

39

Summary of CBOs Cost Estimating Methods for Supersonic Missiles in

Alternative 1

40

A-4.

Characteristics of Bomber Aircraft

43

A-5.

Acquisition Costs of Aircraft in Alternatives 2, 3, 4, and 5

44

A-6.

Summary of CBOs Cost Estimating Methods for Aircraft in

Alternatives 2, 3, 4, and 5

45

A-7.

Characteristics of Common Aero Vehicles in Alternatives 6, 7, and 8

49

A-8.

Acquisition Costs of Common Aero Vehicles in Alternatives 6, 7, and 8

50

A-9.

Summary of CBOs Cost Estimating Methods for Common Aero Vehicles in

Alternatives 6, 7, and 8

52

S-1.

Response Times of Alternative Strike Systems for Preplanned Missions

xiii

S-2.

Number of Aircraft Needed for 24-Hour Coverage of 25 Percent of

Afghanistan with a 10-Minute Response Time

xiii

S-3.

Bomb Delivery Rates for Alternative Strike Aircraft

xiv

1-1.

Potential Firepower of Current Air Force Strike Aircraft, Without Aerial Refueling

2-1.

Response Times of Long-Range Common Aero Vehicles

A-3.

x
xv

Figures

6
12

CONTENTS

Figures (Continued)
3-1.

Distribution of Worst-Case Penetration Distances for Countries of the

World, Measured by Geographic or Operational Limits

16

3-2.

Ability of Alternative Strike Systems to Attack Targets Deep in Enemy Territory

17

3-3.

Response Times of Alternative Strike Systems for Preplanned Missions

19

3-4.

Area Covered from Loitering Orbit for Different Response Times

22

3-5.

Number of Aircraft Needed for 24-Hour Coverage of 25 Percent of

Afghanistan with a 10-Minute Response Time

23

3-6.

Bomb Delivery Rates for Alternative Strike Aircraft

24

4-1.

Potential Firepower of Alternative Air Force Strike Forces, Without

Aerial Refueling

30

4-2.

Total Potential Firepower of Alternative Bomber Forces, by Mission Distance

31

4-3.

Cumulative Development and Acquisition Costs for Weapons and Their

Delivery Systems, by Number of Weapons Used

34

VII

Summary

he United States maintains a considerable capability to attack ground targets with conventional weapons
anywhere in the world. Air Force, Navy, and Marine
Corps tactical aircraft deployed to forward locations can
be used for ground attack, or strike missions, in the region where their bases or aircraft carriers are located, and
a similar regional capability is offered by Tomahawk
cruise missiles launched from submarines or surface
ships. If air bases or aircraft carriers are not available in
the region, long-range strike capability can be provided
by the Air Forces smaller fleet of long-range bombers,
which can conduct missions from more-distant bases, including ones in the United States.
The performance of the bomber force in Afghanistan
(Operation Enduring Freedom) and Iraq (Operation
Iraqi Freedom) confirmed the value of long-range strike
systems that are less dependent on having access to air
bases close to the conflict. In Afghanistan, strike aircraft
were forced to fly very long missions with extensive airborne tanker supportfighters typically had to refuel
many times during a missionbecause of basing and airspace restrictions in neighboring countries. In Operation
Iraqi Freedom, the availability of air bases was limited,
and the air bases in Kuwait were vulnerable to attack by
Iraqi cruise and ballistic missiles. Although both air campaigns were successful, if air operations of greater intensity or length had been needed, those adverse circumstances might have posed problems. In contrast, longrange bombers contributed to the campaigns over great
distances from secure bases by operating from places such
as Diego Garcia, an island in the Indian Ocean.
Although the Air Force continues to upgrade the existing
bomber fleet to deliver most types of conventional weapons and to participate more effectively in tactical groundattack operations, until recently there have not been de-

finitive plans for expanding long-range strike capabilities.

Numerous studies of which capabilities might be desired
and several plans for potential long-range systems had
been proposed, but none had resulted in decisions on a
way to move forward.1 Now, however, the Department of
Defense (DoD) has begun to define initial plans for developing new long-range strike systems. For example, the
2006 Quadrennial Defense Review (QDR) report states
that DoD intends to develop a new land-based longrange strike capability and to deploy an initial capability
to deliver precision-guided conventional warheads using
long-range ballistic missiles. However, because specific requirements for system performance and force levels have
yet to be defined, considerable uncertainty remains as to
which capabilities DoD will require of new long-range
strike systems, how well different types of systems might
provide those capabilities, and what it might cost to develop and deploy such systems.
This Congressional Budget Office (CBO) study, which
was prepared before the release of the 2006 QDR, examines those questions. It compares how well eight longrange strike systems might perform in several areas that
DoD studies have identified as important for future operations. Those systems reflect general classes of long-range
weapons that have been proposed within the defense
community and include aircraft, long-range missiles, and
space-based weapons. Although all of the systems CBO
examined would have the common characteristic of a
range no less than about 1,500 nautical miles (nm)
greater than that of current or planned strike fighters carrying typical weapon loadstheir performance in other
1. See, for example, Office of the Secretary of Defense, Report to
Congress on: Prompt Global Strike Plan (June 2005); and Department of Defense, Defense Planning Guidance: Long Range Global
Precision Engagement Study (April 2003).

ALTERNATIVES FOR LONG-RANGE GROUND-ATTACK SYSTEMS

Summary Table 1.

Long-Range Strike Alternatives Examined by CBO

Unrefueled Range
a
with Full Payload
(Nautical miles)

Payload
(Pounds)

Speed
(Mach)

Arsenal Aircraft
C-17
Supersonic missile

1,500
500

134,000
n.a.

0.76
3

Medium-Range Subsonic Bomber

1,500

20,000

0.85

Medium-Range Supersonic
Dash Bomber

1,500

10,000

0.85 (Sustained)
1.5 (Dash)

Long-Range Subsonic Bomber

2,500

40,000

0.85

Long-Range Supersonic Cruise Bomber

2,500

40,000

2.4

Medium-Range Surface-Based CAV

3,200

2,000

14

Long-Range Surface-Based CAV

Global

4,000

20

Space-Based CAV

2,000

20

Nearly Global

Source: Congressional Budget Office.

Note: CAV = common aero vehicle; n.a. = not applicable.
a. The ranges shown are the maximum distance from an air base or launcher location to the target. For the aircraft alternatives, the total
distance flown on an unrefueled mission would be double the values shown.
b. Aircraft payloads represent combat loads; CAV payloads are per missile.
c. From an equatorial orbit, space-based CAVs could reach any point on Earth between the latitudes of 60 degrees north and 60 degrees
south.

areas would differ substantially, as would the estimated

costs to develop and procure them.2
To compare the military utility of different approaches
for striking at long range, CBO used several specific measures to quantify the types of capabilities that have been
identified for future long-range strike systems:
B

Reachthe ability to attack targets regardless of

location.

Responsivenessthe ability to attack targets quickly.

2. Unless it is specified otherwise, an aircrafts range in this study

refers to its combat radius with a full load. The combat radius is
defined here as the distance that an aircraft could fly from its base
to attack a target and still have enough fuel to return without
aerial refueling. The total distance the aircraft would fly on such a
mission would be twice that range. Missile ranges are simply the
maximum distance from the launch location to the target.

Firepowerthe ability to sustain attacks over time.

Survivabilitythe ability to avoid or defeat air

defenses.

CBO did not assess how much of the particular capabilities might be desired or needed but rather compared how
well its alternatives (as well as todays forces) could provide them.

Alternatives for Improving Long-Range

Strike Capabilities
The eight alternatives examined by CBO represent very
diverse approaches to improving long-range strike capabilities (see Summary Table 1). Each offers advantages
and disadvantages in how it contributes to the ability to
strike at long range.

SUMMARY

Alternative 1 would provide for the delivery of fast missiles from a large cargo aircraft, an approach that has been
dubbed an arsenal aircraft in past studies. The system
CBO examined would consist of a new supersonic missile
capable of flying 500 nm at Mach 3 (three times the
speed of sound) that would be launched by C-17 cargo
aircraft with internal rack systems from which the missiles could be extracted through the aircrafts rear door.
Aircraft from the planned fleet of 180 air mobility C-17s
could carry the racks, or additional C-17s could be purchased and dedicated to the strike mission. Because cargo
aircraft are not designed to elude air defenses, the arsenal
aircraft itself could not penetrate hostile airspace. They
would have to launch their missiles from secure airspace.
Alternative 2 and Alternative 3 would develop stealthy
medium-range bombers capable of penetrating air defenses. Those aircraft would have ranges and payloads between those of todays strike fighters and long-range
bombers. (The now-retired F-111 is a recent example of a
U.S. medium-range bomber.) The aircraft in Alternative
2 would have a higher payload than that of Alternative 3
but would be limited to subsonic speeds. The aircraft in
Alternative 3, a concept similar to proposals for a socalled FB-22, would be capable of dash speeds up to
Mach 1.5 for limited distances. Those aircraft could be
designed as either manned or unmanned systems. (CBOs
cost estimates assume they would be manned.)
Alternative 4 and Alternative 5 would develop long-range
bombers also capable of penetrating air defenses. Those
aircraft would have ranges and payloads similar to those
of todays heavy B-52, B-1, and B-2 bombers. Alternative
4s aircraft would be similar in concept (although not
necessarily in specific design) to the stealthy, subsonic B2. Alternative 5s aircraft would be an advanced bomber
capable of maintaining speeds greater than Mach 2 over
most of its mission. Achieving higher speed (with a similar range and payload) than that of Alternative 4 would
require about a 40 percent larger and heavier aircraft to
accommodate more fuel and more powerful engines. Although probably less stealthy than the other aircraft
alternativesits large size and other design characteristics
for sustained supersonic flight are not as amenable to
stealththe high speed of the supersonic cruise bomber
would contribute to its survivability. Those aircraft could
also be designed as either manned or unmanned systems.
(CBOs cost estimates assume they would be manned.)

Alternatives 6 through 8 would develop maneuvering

warheads called common aero vehicles (CAVs) similar in
concept to hypersonic systems that have been explored by
the Defense Advanced Research Projects Agency and the
Air Force under the FALCON (Force Application and
Launch from the Continental United States) program.
CAVs are missile- or spacecraft-launched unmanned vehicles capable of flying through space on suborbital trajectories. CAVs are shaped to generate sufficient lift so
that, after reentering the atmosphere, they can glide
many thousands of miles to their targets at hypersonic
speeds with a combination of thrusters and flaps providing maneuvering control. CBO examined CAV systems
that could be launched at their targets using a groundbased intercontinental ballistic missile (ICBM) or a
smaller ground- or ship-based medium-range missile, as
well as CAVs that would be placed in equatorial lowEarth orbits and de-orbited when needed.
The next section provides quantitative comparisons of
the eight specific systems CBO analyzed. However, that
analysis pointed to several general observations that can
be made about each of the classes of systems CBO examined, independent of detailed design specifications:
Arsenal aircraft armed with supersonic missiles offer the potential to provide significant firepower and responsiveness
at costs substantially lower than those of new penetrating
bombers. However, their vulnerability to enemy air defenses would limit their reach into defended airspace to
the range of the missile, a much shorter distance than
those of the other alternatives CBO examined.
Stealthy manned or unmanned medium-range bombers
would offer reach and firepower improvements over current long-range strike fighters but would not offer the
global reach or long loitering capability of long-range
bombers. (Loitering in the target area enables aircraft to
respond very quickly to fleeting targets.) A greater number of medium-range bombers could be fielded for a
given investment, although the net firepower would not
necessarily be higher than that of a smaller number of
larger-payload long-range bombers.3
Stealthy manned or unmanned long-range bombers offer
global reach and substantial sustained firepower. Subsonic
bombers would offer global response times on the order
3. See Chapter 4 for a comparison of the effects of purchasing different quantities of the penetrating bombers.

XI

XII

ALTERNATIVES FOR LONG-RANGE GROUND-ATTACK SYSTEMS

of 15 hours with long loitering endurance to provide fire

against fleeting targets with response times on the order
of several minutes. Supersonic bombers would offer
shorter global response times but would have higher cost
and might have a limited ability to loiter.
Hypersonic CAVseither space-based or launched by ballistic missileswould offer responsiveness on the order of
one hour against targets anywhere on the globe and
would be the most difficult systems for enemy defenses to
intercept. However, their high unit cost implies that they
probably could not be purchased in sufficient numbers to
provide the sustained firepower offered by aircraft forces.

Comparison of Long-Range Strike

Alternatives Considered by CBO
The long-range strike systems examined by CBO would
provide diverse capabilities, with each alternative offering
advantages and disadvantages for different types of missions. In addition to differences in capability, the alternatives would have unique implications for the future force
structure. Fielding CAVs, for example, would provide a
new rapid-strike capability but would not address the issue of the aging bomber force. Conversely, a supersonic
bomber could replace todays bombers but would offer
less responsiveness than CAVs would. Because of such
distinctions, the alternatives CBO examined should not
necessarily be viewed as independent alternatives. Depending on the specific requirements that DoD eventually establishes for its long-range strike systems, the preferred solution might include more than one of the
systems CBO examined.

Capabilities
In comparing the capabilities of the alternative longrange strike systems, CBO examined how far each system
could reach, how responsive it would be in several settings, how much firepower it could provide, and how
safely it could operate in the face of enemy air defenses.
Reach. The reach of a long-range strike system can be important for two reasons. First, long range allows missions
to be conducted from greater distances, either before local
bases can be established or when they are not available.
Aerial refueling, however, means that even short-range
aircraft can fly missions much longer than their unrefueled range would allow. Consequently, all of CBOs alternatives would offer the potential for global reach.

Nonetheless, the two long-range CAVs and the two longrange bombers would be best suited for missions requiring global reach. The medium-range bomber alternatives
could, in principle, be used for intercontinental missions,
but the need for more frequent refueling would complicate operations, and crew endurance in the probably
cramped confines of a small cockpit might present problems. The C-17 arsenal aircraft could carry relief crews to
help remedy that problem, but it would still require more
frequent refueling than would the long-range bomber alternatives. Although lacking inherent global reach, the
medium-range CAV alternative could cover most of the
globe from just a few forward land bases (for example,
Guam in the western Pacific Ocean and Diego Garcia, a
territory of Great Britain in the Indian Ocean) or ships.
A second facet to the military value of long range is the
capability it offers to reach targets deep in hostile airspace, where support from airborne tankers would not be
available. That can be important against larger countries
or in situations where tanker operations are otherwise
constrained. All of the alternatives CBO examined except
the arsenal aircraft provide very good capability to reach
any point within all or most countries in the world. The
long-range bombers and the long-range surface-based
CAV could do so for all countries. The medium-range
bombers could do so for 95 percent of all countries under
conservative assumptions about access to airspace adjacent to the target country. Easing those assumptions
slightly would enable full coverage. The medium-range
CAV could fully cover all countries given suitable launch
locations. The space-based CAV could fully cover almost
all countries (about 97 percent of them) from its equatorial orbit. The exceptions are those countries with territory at North or South latitudes greater than about 60 degrees. The arsenal aircraft, carrying a supersonic missile
with a range of 500 nm after launch, would provide the
least coverage. If the C-17 delivering the missile must
stand off outside hostile airspace, it could fully cover only
about 75 percent of the worlds countries.4
4. This geography-based analysis does not distinguish among nations
that are more or less likely to be considered potential threats. Such
judgments are subjective and change over time. For three nations
commonly mentioned as potential adversariesNorth Korea,
Iran, and Chinathe CAV and long-range bomber alternatives
would provide total coverage, the medium-range bombers could
not reach parts of China, and the supersonic air-launched missile
could not reach parts of Iran and China.

SUMMARY

Summary Figure 1.

Response Times of Alternative Strike

Systems for Preplanned Missions
(Hours to strike target)
25

20

Arsenal Aircraft
Medium-Range
Bombers and
Long-Range
Subsonic Bomber

15

Long-Range
Supersonic
a
Cruise Bomber

10

Common Aero Vehiclesb

2,000

4,000

6,000

8,000

Against fleeting targets or in a ground-support role, none

of the long-range strike alternatives would provide, from
a standing start, response times on the order of a few
minutes. Even a space-based CAV would need at least 15
minutes if the launcher was in the right orbital location
and no atmospheric maneuvering was necessary. Very
short response times require maintaining systems close to
the locations where targets are expected to appear, a tactic
that could not be accomplished with the CAV alternatives. For the other alternatives, the area that can be covered by an aircraft orbiting over a particular location for a
given response time and the number of aircraft needed to
maintain such orbits become the critical factors. (See
Chapter 3 for a more detailed discussion.) From that perspective, the arsenal aircraft would provide the shortest
response times with the fewest aircraft because the Mach
3 missile it would employ can dash much farther in a
Summary Figure 2.

because it could not maintain supersonic speed during

the entire transit to its target.

10,000

Nautical Miles to Target

Source: Congressional Budget Office.

a. The steps in the line for the supersonic cruise bomber result
from its need to slow for aerial refueling.
b. Response times for the common aero vehicle alternatives will
vary for a given distance to the target, depending on the specific
flight profile needed. The medium-range common aero vehicle
has a maximum distance to target of 3,200 nautical miles.

Responsiveness. CBO considered responsiveness in two

contexts: preplanned missions that would require responsiveness on the order of hours, and fleeting-target or
ground-support missions that would require response
times on the order of a few minutes. Assuming a similar
planning process, a systems speed will be the primary determinant of its responsiveness in a preplanned mission.
The greater the distance to be traveled, the greater the cumulative advantage of higher speed (see Summary
Figure 1). The hypersonic CAV alternatives, consequently, would offer by far the shortest response times
among CBOs alternatives. The long-range supersonic
cruise bomber (Alternative 5) would have a response time
between that of the CAVs and the other aircraft alternatives. The medium-range supersonic bomber would not
be significantly more responsive than the subsonic aircraft

Number of Aircraft Needed for

24-Hour Coverage of 25 Percent of
Afghanistan with a 10-Minute
Response Time
100

Medium-Range
Supersonic
Dash Bomber

80

Long-Range
Subsonic
Bomber

60

40

Medium-Range
Subsonic
Bomber

20

Arsenal
Aircraft

0
0

500

1,000

1,500

2,000

2,500

Nautical Miles Between Base and Orbit Location

Source: Congressional Budget Office.

Note: This figure assumes that an aircraft expends all of its munitions before returning to base. Endurance would be lower if
an aircraft returned with unused munitions.

XIII

XIV

ALTERNATIVES FOR LONG-RANGE GROUND-ATTACK SYSTEMS

Summary Figure 3.

Bomb Delivery Rates for Alternative

Strike Aircraft
(Number of 2,000-pound JDAM-equivalents per aircraft
per day)
80

Arsenal Aircraft
60

Long-Range
Subsonic
Bomber

Long-Range
Supersonic
Cruise Bomber

40

Medium-Range
Subsonic Bomber

20

Medium-Range
Supersonic
Dash Bomber
0
0

500

1,000

1,500

2,000

2,500

Nautical Miles to Target

Source: Congressional Budget Office.

Note: Data are shown out to the unrefueled radius of each type of
aircraft when carrying a full bomb load. All alternatives could
achieve greater ranges with aerial refueling or reduced
bomb loads.
JDAM = Joint Direct Attack Munition.

given amount of time than can the other aircraft alternatives (see Summary Figure 2). The long-range subsonic
bomber (Alternative 4) would offer good performance
because of its long endurance, especially for orbits far
from base or far from aerial refueling support. Performance of the medium-range bomber alternatives suffers
from the lack of the supersonic missiles high speed (Alternative 1) and the long-range subsonic bombers endurance (Alternative 4), although the supersonic dash capability of Alternative 3 gives it good performance for orbits
close to base. The supersonic cruise bomber could not
meet the responsiveness criteria in Summary Figure 2 because CBO assumed its design would not be suitable for
maintaining sustained low-speed orbits.

Firepower. Another requirement for long-range strike

systems will probably be support of operations requiring
the delivery of a high volume of munitions such as those
seen in large conflicts. Long-range systems could be especially important early in those situations, before significant numbers of tactical aircraft have been deployed to
the theater. Because of their large payloads, the arsenal
aircraft and the long-range subsonic bomber (Alternatives
1 and 4) would offer the highest per-platform weapon delivery rates for mission distances of less than 2,400 nm
(see Summary Figure 3). For mission distances of longer
than 2,400 nm, the supersonic cruise bombers speed
would provide it an advantage in firepower despite the
longer time needed to prepare it for each mission.5
The number of systems purchased also affects the achievable firepower. Although Alternative 1 can physically
achieve a high weapon delivery rate, the relatively high
cost of each supersonic missileabout $1.4 million per
round as compared with about $31,000 for a satelliteguided bomb such as the Joint Direct Attack Munition
(JDAM)may ultimately constrain its use. CBO did not
include the CAV alternatives in the firepower comparison
because their much higher unit costs would almost certainly make them unsuitable for such sustained operations. Similarly, the number of aircraft purchased under a
given alternative would affect the forcewide weapon delivery capability. For example, an additional 25 mediumrange subsonic bombers (for a total of 300) would provide similar firepower (albeit over shorter distances) as the
150 long-range bombers under Alternative 4 but for
about $19 billion less. In contrast, 600 medium-range supersonic bombers would be needed to provide an equivalent firepower at a cost of about $38 billion more than
that for 150 long-range subsonic bombers.
Survivability. The ability to reach the targetand, in the
case of aircraft, return safelyin the face of air defenses is
important for long-range strike systems because they will
most likely be tasked in circumstances where the timely
suppression of enemy air defenses will not be possible.
The CAV alternatives examined by CBO would have the
greatest survivability. After launch, their hypersonic
speeds and ability to maneuver unpredictably could only
be countered by a very sophisticated missile defense sys5. CBO assumed that the greater complexity of a supersonic cruising
aircraft and the greater rigors of sustained supersonic flight would
result in servicing times between missions of about twice the
length of the long-range subsonic-bomber alternative.

SUMMARY

Summary Table 2.

Estimated Costs of Long-Range Strike Alternatives Examined by CBO

Quantity

Arsenal Aircraft

Medium-Range Subsonic
Bomber

2.8/6.1

275

19

52

72

188

Medium-Range Supersonic
Dash Bomber

275

23

61

85

220

Long-Range Subsonic Bomber

150

31

61

93

409

Long-Range Supersonic
Cruise Bomber

150

69

137

207

912

Medium-Range SurfaceBased CAV

48

2.4

1.2

3.7

26

Long-Range Surface-Based CAV e

24

2.5

0.9

4.0

36

4.0

7.7

11.7

55

6
7
8

Space-Based CAV

128

4.3/7.6

Average Unit Procurement Cost

(Millions of 2006 dollars)

1.5

2,000

Costsa
(Billions of 2006 dollars)
RDT&E
Procurement
Totalb

1.4

Source: Congressional Budget Office.

Note: RDT&E = research, development, test, and evaluation; CAV = common aero vehicle.
a. The costs for Alternatives 2 to 5 exclude munitions.
b. Includes additional military construction costs of about $1 billion for the aircraft alternatives (2 to 5) and $600 million for Alternative 7.
c. The quantity shown is the number of supersonic missiles purchased. The lower of the two costs assumes that those missiles would be carried by C-17 aircraft in the current fleet. The higher of the two costs assumes that 15 additional C-17s would be purchased to support the
strike mission.
d. Average unit procurement cost is for supersonic missiles only.
e. Alternative 7 assumes that 24 Peacekeeper missiles would be converted to carry two CAVs per missile. If more missiles were desired, as
many as 60 Peacekeepers might be available for conversion. The cost of additional missiles would be much higher than the cost shown
here because new boosters would be needed.
f.

Enough satellites would be purchased to maintain the constellation for 30 years. Only 40 space-based CAVs would be available for use at
any one time.

tem. Attacking CAVs before they were launched would

be difficult as well: survivable long-range systems would
be needed to hit the surface-based CAV systems, and an
antisatellite capability would be needed against orbiting
CAVs. In the case of the arsenal aircraft, although the supersonic air-launched missile would also be a challenge
for air defenses, its C-17 delivery platform could be vulnerable if an adversary was able to send fighters out to intercept it. The stealth designs for the subsonic bombers,
Alternative 2 and Alternative 4, should give them good
survivability against ground-based air defenses and
against fighters at night. They could be vulnerable to
fighters during the day, however, when the bombers

could be detected visually. The long-range supersonic

cruise bomber has a combination of limited stealth and
high speed that should enable it to avoid engagement by
many surface defenses and to outrun fighters sent to intercept it. Similarly, the medium-range bomber in Alternative 3 could use stealth to survive surface defenses and
its supersonic dash speed to elude fighters, although with
less certainty because it would have less of a speed advantage than the supersonic cruise bomber would.

Cost and Force-Structure Implications

Comparing the long-range strike alternatives is complicated by the significant differences in the estimated costs

XV

XVI

ALTERNATIVES FOR LONG-RANGE GROUND-ATTACK SYSTEMS

to develop and field them (see Summary Table 2). CBO

estimates that research, development, test, and evaluation
(RDT&E) costs could vary by more than an order of
magnitude among the alternatives, with the arsenal aircraft and CAV alternatives costing the least and the advanced supersonic cruise bomber the most. Similarly, estimates of production costs also vary over a broad range,
although the differences among alternatives are in part
the result of assumptions about how many of each type of
system would be purchased. The cost estimates are presented in 2006 dollars. Significant uncertainty exists
about the costs associated with developing, purchasing,
and operating weapon systems envisioned in the alternatives because those programs are either conceptual in nature or in the early stages of development. Consequently,
they entail a greater risk of cost and schedule overruns
than do programs that are better defined and based on
proven technologies. CBOs cost estimates represent one
possible outcome, calculated under specific assumptions.
Without established DoD requirements as a guide, CBO
based its estimates of procurement quantities for each alternative on the current force structure and forcestructure plans wherever possible. Procurement quantities
were not adjusted to try to achieve equivalent capabilities.
(For example, the 40 operational CAVs purchased under
each CAV alternative would be equivalent to the payloads
of only about two long-range bombers.) Details of CBOs
methods for developing its cost estimates are described in
the appendix.
In general, the alternatives that involve penetrating
bombersAlternatives 2, 3, 4, and 5are much more
expensive to develop and procure, but they offer the advantage of being able to repeatedly deliver relatively inexpensive munitions. The very short response times of the
CAV alternatives come with the disadvantage of high unit
costs, which could limit the number procured.
Arsenal Aircraft. The arsenal-aircraft alternative would be
less costly than the penetrating-bomber alternatives and
less costly than the CAVs on a per-missile basis, although
as noted earlier, the arsenal aircraft would have several
drawbacks. It would, however, represent a new type of
strike system not comparable with any of todays systems.
The lower cost shown in Summary Table 2 assumes that
aircraft could be drawn from the Air Forces planned inventory of 180 C-17s purchased for strategic airlift. If
purchases of more C-17s dedicated only to strike missions were necessary, the cost of this alternative would be

higher. For example, 15 new aircraft would cost an additional $3.3 billion, CBO estimates. Such additional aircraft would be able to augment the airlift fleet when not
needed as strike aircraft.
Medium-Range Bombers. The costs for the mediumrange bombers in Alternatives 2 and 3 are based on purchasing 275 of those aircraft, a number similar to the current inventory of F-117 and F-15E strike fighters that a
medium-range bomber might augment or replace.6 Replacing those strike fighters with medium-range bombers
would represent an improvement in strike capability because the new aircraft would have longer range, and all of
them would be stealthy. (Only the 55 F-117s are stealthy
today.) Although the current F-117s and F-15Es are aging, that force is not as old as todays long-range bombers.
Because the aircraft in Alternatives 2 and 3 would have
limited ability to conduct global-range strikes, they
would leave unaddressed the issue of replacing todays
long-range bombers. Thus, DoD might have to consider
developing a new long-range bomber at some other time,
or it could consider ceding the global-range mission to
the lower-cost CAV alternatives.
Long-Range Bombers. The costs for the long-range
bombers associated with Alternatives 4 and 5 are based
on purchasing 150 of those aircraft, a number similar to
the current long-range bomber force, and they could be
used to either augment or replace those aircraft. Those
new bombers, when coupled with the improved range expected for the Joint Strike Fighter over the F-16s that aircraft is expected to replace, could reduce the need for a
direct replacement for the F-117 and F-15E forces. Alternatives 4 and 5 would maintain a manned global-strike
capability, although CAVs might still be desired because
of their much shorter response times.
Hypersonic CAVs. The alternatives associated with hypersonic CAVs that were examined by CBO would have significantly lower costs than those for the penetratingbomber alternatives. CAVs have the disadvantages, however, of being less flexible than aircraft and very costly to
purchase in more than limited numbers. CBO estimates,
for example, that it would cost over $200 billion to purchase enough space-based CAVs to provide the same
number of weapons as one days delivery of 2,000-pound
6. The F-117 and F-15E are the longer-range tactical strike aircraft
in the Air Force. The bulk of the force is made up of shorter-range
F-16 multirole fighters.

SUMMARY

JDAMs by 100 supersonic cruise bombers flying missions

against targets 7,000 nm from their base. CBO sized its
CAV alternatives on the basis of the availability of Peacekeeper ICBM boosters to launch them. Of course, larger
numbers could be purchased, although at a higher unit
cost because new boosters would have to be designed and

built. Because the procurement quantities of space-based

CAVs or ballistic missiles armed with CAVs would probably be limited by their high unit costs, those missiles
would be unable to fully replace aircraft in the role of sustained long-range strike operations.

XVII

CHAPTER

1
The United States Long-Range
Strike Capabilities
Chapter 1: The United States Long-Range Strike Capabilities

he U.S. military possesses considerable capability

to attack targets anywhere in the world with conventional
weapons. Shorter-range Air Force and Navy tactical aircraft, forward deployed and supported by airborne tankers, can attack targets in the region where their bases or
aircraft carriers are located, and Tomahawk cruise missiles
launched from submarines or surface ships offer a similar
regional capability. If nearby air bases or aircraft carriers
are not available, global reach can be achieved with the
Air Forces smaller fleet of long-range bombers supported
by airborne tankers.1 Tactical ground-attack aircraft were
used very effectively in Operation Desert Storm in 1991,
where approximately 900 U.S. strike fighters based in the
Persian Gulf region or on aircraft carriers nearby flew
more than 40,000 ground-attack sorties against targets in
Iraq during the 43-day air campaign. Bomber contributions, although important, were limited because only
B-52 aircraft were available to deliver conventional munitions at that time. (Of the other two bomber types in todays inventory, the B-1 was not used, and the B-2 was
not yet operational.)
Despite the effectiveness of the Desert Storm air campaign, planners saw a potential vulnerability in its execution. The heavy reliance on shorter-range tactical aircraft
suggested that U.S. airpower could be severely limited if
local air bases were unavailable. That scenario could occur if nations in a region were reluctant to host U.S. military forces, if operations had to take place in an undeveloped area that lacked such bases (or proximity to the sea
for aircraft carriers), or if an adversary was able to attack
bases hosting U.S. aircraft with ballistic or cruise missiles.
Some planners argued that enlarging long-range strike
forces would mitigate that vulnerability because such sys1. Intercontinental ballistic missiles also offer global reach but currently only for nuclear warheads.

tems could be based farther away without an overreliance

on aerial refueling. Having a greater number of longrange strike aircraft also would increase the number of
airfields where strike aircraft could be based, and those
bases could be beyond the range of an adversarys missiles.
A decade later, the experiences in Operation Enduring
Freedom (Afghanistan) and Iraqi Freedom confirmed the
value of long-range strike systems that are less dependent
on having access to air bases close to the conflict. In the
case of Afghanistan, basing and airspace restrictions
forced strike aircraft to fly very long missions with extensive airborne tanker support. (Fighters typically had to refuel several times during a mission.) Against Iraq, strike
fighters operating from air bases in Kuwait were within
range of Iraqi cruise and ballistic missiles. Although both
air campaigns were ultimately successful, those adverse
circumstances might have posed serious problems, especially if air operations of greater size or length had been
needed. In contrast, long-range bombers operating from
Diego Garcia, an island in the Indian Ocean, contributed
to those campaigns at long ranges from a secure base and
with less reliance on airborne tanker support.2 Navy aircraft operating from aircraft carriers also contributed to a
greater relative extent than they did in Desert Storm.
Although the Department of Defense (DoD) is pursuing
numerous programs to improve its strike capabilities,
most are focused on modernizing the shorter-range strike
fighter force or developing more capable munitions.
Notable among those efforts are the F/A-18E/F and Joint
Strike Fighter aircraft, which are scheduled to replace
most Air Force, Navy, and Marine Corps strike fighters
over the next 20 years or so. Although those fighter pro2. Bombers were usually refueled in flight as well but not as frequently.

ALTERNATIVES FOR LONG-RANGE GROUND-ATTACK SYSTEMS

grams have been in existence for more than a decade, attention is now being given to improving long-range strike
capabilities. Beginning in the 1990s, the existing bomber
fleet was progressively upgraded so that it could deliver
most types of conventional weapons, and the bombers
were also equipped with the communications gear necessary to take part in tactical ground-attack operations.
Those types of upgrades continue today. Until recently,
however, there were no firm plans to expand the bomber
force or to develop alternative means for conducting conventional strike missions over extended ranges. The Air
Force has recently begun to examine a variety of strike alternatives with longer range than tactical fighters have,
including a new medium-range bomber and intercontinental ballistic missiles (ICBMs) armed with conventional warheads. Additionally, the 2006 Quadrennial Defense Review (QDR) report states that DoD intends to
develop a new long-range strike capability based on land
and to deploy an initial capability to deliver conventional
warheads with ballistic missiles fired from Trident submarines.
This Congressional Budget Office (CBO) study looks at
the potential operational effectiveness and costs of alternatives for improving strike capabilities at combat ranges
longer than about 1,500 nautical miles (nm) without refuelingranges that are greater than those of current or
planned strike fighters carrying typical weapon loads.3
The study considers a spectrum of systems including aircraft, air- and surface-launched missiles, and space-based
weapons. It compares the capabilities offered by such systems with one another as well as to existing means of providing the same military capability. The alternatives CBO
examined reflect general classes of weapon systems, not
specific systems proposed by DoD or by industry. They
are not intended to identify a preferred solution but
rather to offer a comparison of the capabilities and costs
that can be expected from different types of long-range
strike systems.
3. Unless it is specified otherwise, an aircrafts range in this study
refers to its combat radius with a full payload. The combat radius
is defined here as the distance that an aircraft could fly from its
base to attack a target and still have enough fuel to return without
aerial refueling. The total distance the aircraft would fly on such a
mission would be twice that range. Missile ranges are simply the
maximum distance from the launch location to the target.

The Current Long-Range Strike Force

Todays conventional long-range strike capability is provided by the Air Forces fleet of long-range bombers. The
origins of the current bomber force go back to World
War II, when large fleets of heavy and medium bombers
were built to attack the industrial infrastructure of Germany and Japan and interdict surface forces and supplies
moving to the combat zones. Shorter-range fighters were
usually used to counter enemy fighters or to provide close
air support to ground units.4 Later, as part of the U.S.
nuclear deterrent during the Cold War, bombers were designed with the long ranges and heavy payloads needed to
carry large nuclear weapons from bases in the continental
United States to targets deep within the Soviet Union. Although many strike fighters could also carry nuclear
bombs, they were primarily designed for conventional
warfare. Fighter-like aircraft were preferred for conventional missions because accurately delivering unguided
conventional munitions against heavily defended tactical
targets required low-altitude operations and fighter-like
speed and agility. In that case, long range was less important because the fighters were expected to operate from
bases in Western Europe, close to any fighting in a conventional war between the North Atlantic Treaty Organization and the Warsaw Pact.
At the height of the Cold War, the size of the U.S.
bomber fleet was significantly larger than it is today. In
1963, there were 709 B-52s as well as more than 1,000
other bombers such as the B-47 and B-58. In addition to
their long-range bombers, the Air Force and Navy had
intermediate-range bomberswith payloads and range
less than those of the heavy bombersbut by the early
1990s, the Air Force had retired its F-111s and the Navy
had retired its A-6 Intruders.5 Today, the bomber force
numbers 182 aircraft, of which 96 are combat-ready (see
Table 1-1 for a description of the quantities and capabili4. Although optimized for their given missions, heavy bombers,
medium bombers, and fighter bombers could be used interchangeably if circumstances required and conditions permitted.
For example, Navy fighter bombers were occasionally used to
attack industrial targets on Japans main islands, and the U.S.
Eighth Air Force used heavy bombers to support the D-Day
landings.
5. In terms of distance from the United States, short-range aircraftcarrier-based Navy strike fighters could be thought of as providing
a long-range capability. Targets would still need to be relatively
close to the sea, however.

CHAPTER ONE

THE UNITED STATES LONG-RANGE STRIKE CAPABILITIES

Table 1-1.

Characteristics of Current U.S. Long-Range Bombers

B-1B

B-2A

B-52H

Bomber Inventory
Total active
Combat-ready

67
36

21
16

94
44

Average Age (Years)

17

11

43

1,800

2,000

3,000

190,000
477,000

154,000
337,000

185,000
488,000

54,000

34,000

Usually 50,000

24

16

12

0.78

0.8

Yes

No

Combat Radiusa (Nautical miles)

Weight (Pounds)
Empty
Maximum takeoff
Maximum Dropped Payload (Pounds)
Number of 2,000-Pound JDAMs
Crew
Employment Speed (Mach)
Stealth Features e

4
0.85

Some

Source: Congressional Budget Office based on Air Force briefings and data from the Congressional Research Service.
Note: JDAMs = Joint Direct Attack Munitions.
a. Measured on an unrefueled basis with a full combat payload.
b. The B-52 can deliver up to 65,000 pounds of payload when armed with conventional air-launched cruise missiles.
c. The number of JDAMs that the B-52 can carry is limited by the number of weapon stations that are able to provide target coordinates to
the weapon before it is dropped.
d. The B-1 is capable of dashing at speeds up to about Mach 1.2 for limited distances.
e. Features designed to reduce the ability of defensive systems to detect or track the aircraft.

ties of aircraft in the current inventory).6 The three types

of bombers have significantly different characteristics that
reflect the evolution of mission tactics and aeronautical
technologies spanning the nearly four decades over which
they were developed and fielded.

B-52H Stratofortress
The B-52 is the Air Forces oldest bomber. Between 1952
and 1962, the Boeing Company built 744 B-52s of various models. The H model in the current fleet first flew
in 1961 and has more powerful engines, greater payload,
and longer range than do earlier models. Todays fleet of
6. Aircraft might not be combat-ready for several reasons. For example, aircraft undergoing scheduled maintenance at a depot are not
considered combat-ready. Similarly, the level of funding allocated
for operating a type of aircraft also determines the number that
can be kept combat-ready.

94 aircraft includes 44 that are combat-ready for both nuclear and conventional missions. B-52s are based at
Barksdale Air Force Base in Louisiana and Minot Air
Force Base in North Dakota.
Previous Air Force plans projected that B-52s would remain in the Air Force inventory until 2037, although the
recently released 2006 QDR report states DoDs intention to reduce the B-52 inventory to 56 aircraft. Plans remain in place for maintaining and upgrading the smaller
force. Planned upgrades to the B-52 include improvements to navigation systems, onboard computers, electronic countermeasures, and communications links. The
Air Force has also considered replacing the engines with
ones that have lower operating costs, although there are
no current plans to do so. Continuing to operate the B52s for another two to three decades could present problems with airframe life. Concerns have been voiced about

ALTERNATIVES FOR LONG-RANGE GROUND-ATTACK SYSTEMS

corrosion in the airframe and the fatigue life of the upper

wing surface. As of May 2003, the average age of B-52s in
the force measured in flight hours was around 15,860
hours, with the oldest having about 20,700 hours. DoDs
Future Years Defense Program (FYDP) for fiscal years
2006 to 2011 includes about $1 billion for B-52 upgrades over that period. (That total does not include
funds included to develop electronic warfare pods to enable using B-52s as standoff jammers against enemy air
defenses. Those plans were dropped in the fiscal year
2007 budget request.)

The B-1 carries all its munitions in three internal weapons bays. The three-bay configuration gives the bomber
the flexibility to carry a different type of munition in
each. While the B-52 has a larger maximum payload in
terms of weight, the B-1 can carry a greater number of
most munition types (see Table 1-2). Although not considered to be a stealthy aircraft, the B-1 design incorporates features that give it a considerably lower radar signature than that of the B-52. In addition, the B-1 is the
fastest of todays bombers and, with its afterburning engines, can dash at supersonic speeds for limited distances.

The B-52 can carry more types of weapons than the B-1
or B-2 canmore than 20 different types of conventional and nuclear bombs and missiles from the current
Air Force inventory. Weapons can be carried in the internal weapons bay or externally under the wings. Planned
payload improvements include the installation of a laser
targeting pod and the ability to carry new precision weapons as they are introduced into service. Although the airframe design does not incorporate stealth characteristics,
the B-52 can attack targets with long-range cruise missiles, such as the Conventional Air-Launched Cruise Missile (CALCM), that enable it to remain beyond the reach
of air defenses.

B-2A Spirit

B-1B Lancer
After a controversial development history that included
the cancellation of the A model by the Carter Administration, the first B-1B bomber entered service in 1986.
Rockwell International (now Boeing) produced a total of
100 aircraft by 1988. The current fleet size is 67 aircraft,
with 36 that are combat-ready and configured for nonnuclear missions. The B-1 no longer has a nuclear mission. B-1s are currently based at Dyess Air Force Base in
Texas and Ellsworth Air Force Base in South Dakota.
The Air Forces plans for the B-1, like those for the B-52,
call for maintaining and upgrading the force until at least
2025. Upgrades include installation of improved defensive electronic countermeasures for better survivability
and an improved computer system that would allow the
aircraft to carry more types of precision weapons. In addition, the Air Force is improving communications links on
the B-1B with the B-One Next Enhancement program,
which seeks to integrate the Link-16 communications
suite for jam-resistant communications as well as other
equipment for beyond line-of-sight satellite communications. The 2006 FYDP included about $1.1 billion for
B-1B upgrades over the 2006-2011 period.

The B-2 is the newest aircraft in the bomber fleet. Built

by Northrop Grumman, the B-2 flew for the first time in
1989 and was declared operational in 1993. Only 21 aircraft were produced. The current fleet of 21 aircraft includes 16 that are combat-ready and are configured for
both nuclear and conventional missions. The entire fleet
is based at Whiteman Air Force Base in Missouri.
DoD also plans to maintain and upgrade the B-2 fleet
until at least 2025. Planned improvements include
changes to the radar, integration of new munitions, and
changes to some of the stealth features to make the aircraft easier to maintain. DoD included about $1.9 billion
for B-2 upgrades in its 2006-2011 FYDP.
Although the B-1 has some characteristics that reduce the
ability of radars to detect and track it, the B-2 was designed around stealth to enable missions that could penetrate heavy Soviet air defenses. The primary focus on
stealth came at the expense of other design attributes such
as speed, payload, and maintainability. The B-2 is the
smallest and slowest of todays Air Force bombers, and
the need to maintain its radar-evading characteristics
makes it difficult to deploy to forward bases unless special
shelters are available. During Operation Enduring Freedom, B-2s flew missions to Afghanistan from the United
States. Those missions required several aerial refuelings
plus a brief stop at Diego Garcia to change crews. During
Operation Iraqi Freedom, B-2s were based in special shelters on Diego Garcia to allow much-reduced mission
times for the aircraft and their crews.

Capabilities for Future Long-Range

Strike Systems
Although they are now used in conventional roles, todays
bombers were originally designed to carry nuclear weap-

CHAPTER ONE

THE UNITED STATES LONG-RANGE STRIKE CAPABILITIES

Table 1-2.

Maximum Number of Weapons Carried on Current U.S. Long-Range Bombers

Weapon

B-1B

B-2A

B-52H

Mk 82 Unguided 500-Pound Bomb

84

80

45

Mk 84 Unguided 2,000-Pound Bomb

24

16

18

CBU-87/97 Unguided Cluster Bomb

30

34

24

Wind-Corrected Munitions Dispenser

30

16

2,000-Pound Joint Direct Attack Munition

24

16

12

Joint Air-to-Surface Standoff Weapon

24

16

12

EGBU-28 Bunker Buster

Source: Congressional Budget Office based on data from the Air Force.

ons from bases in the United States to targets deep in the

Soviet Union. They had to be large to achieve the long
ranges and carry the large payloads necessary for that mission. For example, the distance from the B-52 base at
Minot, North Dakota, to Moscow is more than 4,000
nm, a much longer distance than what would be anticipated for regional conflicts, and a cruise missile armed
with a nuclear warhead weighs about 3,500 pounds, significantly more than the 2,000 pounds of a GBU-31
Joint Direct Attack Munition (JDAM), one of the heavier
conventional munitions that is used in large numbers.
The nuclear mission was also characterized by deliberate
planning, operations from a well-equipped home base,
and the need for only a few (or perhaps just one) nuclear
attack sorties per aircraft.

bination of weapons and external fuel tanks carried, typical combat radii for U.S. strike fighters range from less
than 350 nm to about 900 nm. Consequently, the bulk
of U.S. strike capability is in that range. The figure shows
that the delivery capacity of bombers at long ranges is
about a quarter of the total delivery capacity of the Air
Force at strike-fighter ranges.7 Although the long range of
heavy bombers is necessary for global missions such as
those flown from the continental United States, a strike
system with an intermediate rangegreater than fighters
but less than heavy bombersmight provide a less costly
yet still very capable force. Other approaches might instead develop long-range missiles that could be launched
from the ground or from ships or submarines.

For a new conventional long-range strike system, DoD

could select an aircraft design with range and payload
similar to those of todays bombers but with additional
characteristicssuch as ease of deployment to forward
bases and ease of refueling and rearming for repeated missionsthat are desirable in a conventional strike system.
Large aircraft, however, have the disadvantage of high development, procurement, and operations costs.

Prior to the QDR, the Air Force had proposed a threephase approach for improving long-range strike capabilities. For the near term, roughly from 2005 to 2015, that
proposal called for continuing bomber upgrades such as
those described above. For the mid-term, roughly from
2015 to 2020, the proposal called for fielding an interim
system that, to the extent possible, used current technologies and avoided the risks of substantial development of
advanced technology. The third phase called for fielding a
highly advanced capability in about 2035 and beyond.

Todays nonnuclear strike capability, measured in terms of

JDAM-equivalent payload delivered per day without
aerial refueling (which captures range, payload, and the
number of sorties achievable per day), is heavily weighted
toward fighters with much shorter ranges than those of
heavy bombers (see Figure 1-1). Depending on the com-

7. Aerial refueling increases the useful range of both fighters and

bombers, although delivery capacity within a given period of time
still drops with increasing mission distances because each mission
takes more time.

ALTERNATIVES FOR LONG-RANGE GROUND-ATTACK SYSTEMS

Figure 1-1.

Potential Firepower of Current

Air Force Strike Aircraft, Without
Aerial Refueling

Responsiveness: the ability to rapidly attack targets at

any time under any conditions. Desired response
times range from hours to minutes.

Flexibility: the ability to carry a wide variety of munition types and to easily incorporate new munitions or
other new technologies.

Survivability: the ability to evade or defeat defenses in

daylight as well as at night.

Situational Awareness: the ability to use onboard sensors plus connectivity to external information sources
to improve combat effectiveness.

(Number of 2,000-pound JDAM-equivalents per day)

20,000

Fighters and
Bombers

18,000

Bombers Only

16,000
14,000
12,000
10,000

The Air Force has not identified specific systems to satisfy

those objectives, although several systems such as a
medium-range-bomber version of the F-22 fighter have
been informally proposed.

8,000
6,000
4,000
2,000
0
0

500

1,000

1,500

2,000

2,500

3,000

Nautical Miles to Target

Source: Congressional Budget Office.

Notes: This figure is based on the Air Forces total inventory of aircraft. Numbers for operational aircraft would be lower.
JDAM = Joint Direct Attack Munition.

The QDR confirmed the need for continuing upgrades

to existing bombers (Phase 1 of the Air Force plan) and
called for fielding a new land-based long-range strike capability by 2018 (Phase 2 of the Air Force Plan) but did
not describe that new capability as an interim one, leaving unaddressed the status of the third phase of the Air
Forces proposal.
The general capabilities that the Air Force has proposed
for future long-range strike systems include:
B

Long Range: the ability to attack targets anywhere on

Earth.

Different performance characteristics can be incorporated

into a strike system to achieve a given capability. For example, speed can be used to achieve survivability by reducing the amount of time the strike aircraft is within
range of defenses. Alternatively, reducing the radar cross
section could increase survivability by decreasing the
range at which the aircraft would be detected. This study
compares the capabilities of eight alternatives for future
long-range strike systems and assesses how basic performance characteristics such as range, payload size and flexibility, signature, and speed contribute to a variety of desired capabilities similar to those identified by the Air
Force. CBO used those basic performance characteristics
to calculate design characteristics such as size, fuel load,
airframe materials, and engine power. In turn, the development and procurement cost of each alternative was
based on those design characteristics. Chapter 2 provides
a detailed description of each CBO alternative and an estimate of its cost. Chapter 3 compares how well the alternatives provide the capabilities desired in a new longrange strike system. Details of the cost estimates are in the
appendix.

CHAPTER

2
Alternative Designs for Long-Range Strike Systems
Chapter 2: Alternative Designs for Long-Range Strike Systems

he Congressional Budget Office considered eight

alternative systems for conducting long-range strikes that
span a range of performance and costs. The types of systems CBO analyzed would have performance characteristics (for example, range, speed, and payload) similar to
systems that have been considered in recent studies prepared for the Department of Defense.1 The next section
describes the technical analyses that CBO used to estimate the general physical characteristics (for example, size
and weight) of a system, given a set of desired performance characteristics. The subsequent section describes
the physical characteristics for the specific systems CBO
examined. Those physical characteristics, in turn, are necessary to make estimates of the costs to develop and purchase each system. The methods CBO used to develop
cost estimates are described in the appendix.
The long-range strike alternatives CBO examined included five aircraft-based systems and three missile-based
systems. The missile-based systems, which would use
common aero vehicles to carry a warhead to the target,
would be launched either from the ground on ballistic
missiles or from satellites placed in advance in low-Earth
orbit. After reentering the Earths atmosphere, the CAVs
would be able to glide to their target at hypersonic speed.

Technical Considerations for the

Design of Long-Range Strike Platforms
The purpose of CBOs analysis is to demonstrate how
military effectiveness and cost are related to key characteristics of long-range strike systems, particularly their
range, payload, and speed. Considering both aircraft and
CAVs, those key characteristics vary widely for the alternatives CBO analyzed. The speed of the alternatives var1. See, for example, Department of Defense, Defense Planning Guidance: Long Range Global Precision Engagement Study Final Report
(April 2003).

ies from subsonic (for example, a C-17 arsenal aircraft

that can cruise at about 76 percent of the speed of
soundMach 0.76) to hypersonic (the long-range
CAVs, which reenter the atmosphere at speeds exceeding
Mach 20). The payloads carried by the long-range strike
alternatives vary from 2,000 pounds for an individual
CAV missile to about 134,000 pounds for the arsenal aircraft. The maximum range at which targets can be attacked varies from global (that is, the 10,000-nauticalmileor morerange of the space-based CAVs) to
1,500 nm for the medium-range bombers and arsenal aircraft.2 In addition, the takeoff and empty weights of the
aircraft alternativeswhich are directly related to their
range, payload, and speed and are key determinants of
costvary by about a factor of 5 (see Table 2-1).
Whether an aircraft flies at subsonic or supersonic speeds,
the greater its range and payload, the heavier its weight at
takeoff will be (to accommodate both the payload and
needed fuel) and the higher its cost. The speed at which
an aircraft flies also affects key aspects of its design and
cost. For example, engines providing sufficient thrust for
flight at supersonic speeds are generally less fuel-efficient
and more costly than those designed to propel subsonic
aircraft. The aerodynamic characteristics of supersonic
aircraft are also different from subsonic aircraft, which is
reflected in their differing shapes. Supersonic aircraft tend
to be longer and narrower than subsonic aircraft in order
to have an acceptable ratio of lift to drag; a narrow fuselage can, however, constrain the payload an aircraft can
carry. Supersonic aircraft must also be constructed of materials that can withstand the increased friction-generated
heat associated with supersonic flight, which also tends to
2. Ranges of 3,000 to 4,000 nm are frequently quoted for the C-17
flying airlift missions. The 1,500-nm mission radius used here
would correspond to a total unrefueled flight distance of 3,000
nm.

ALTERNATIVES FOR LONG-RANGE GROUND-ATTACK SYSTEMS

Table 2-1.

Long-Range Strike Alternatives Examined by CBO

Unrefueled
Range with
Full Payload
(Nautical miles) a

Speed
(Mach)

Time Required
to Strike
a Target
at Maximum
Range (Hours)

Full
Expended
Payload
(Pounds)

3.4

134,000

277,000

569,509

3.1

20,000

60,000

120,000

3.1

10,000

59,000

126,000

Empty
Gross Takeoff
Weight
Weight of Aircraft or
of Aircraft Launch Weight of CAV
(Pounds)
(Pounds)

1 Arsenal Aircraft (C-17 with

supersonic missiles)

1,500

0.76/3.0

2 Medium-Range Subsonic
Bomber

1,500

0.85

3 Medium-Range Supersonic
Dash Bomber

1,500

0.85/1.5

4 Long-Range Subsonic Bomber

2,500

0.85

5.1

40,000

128,000

283,165

5 Long-Range Supersonic
Cruise Bomber

2,500

2.4

1.8

40,000

165,000

439,990

6 Medium-Range SurfaceBased CAV

3,240

14

0.5

2,000

n.a.

48,000

Global

20

1.1

4,000

n.a.

193,000

Nearly Global

20

1.1

16,000

n.a.

1,584,000

7 Long-Range SurfaceBased CAV

8

Space-Based CAVd

Source: Congressional Budget Office.

Note: CAV = common aero vehicle; n.a. = not applicable.
a. The ranges shown are the maximum distance from an air base or launcher location to the target. For the aircraft alternatives, the total
distance flown on an unrefueled mission would be double the values shown.
b. The aircrafts speed is Mach 0.76; the missiles speed is Mach 3.0.
c. The bombers cruise speed to target is Mach 0.85; it can dash (with a reduction in range) at speeds up to Mach 1.5.
d. The weights shown for space-based CAVs are for an eight-missile satellite. The launch weight includes the space-launch rocket used to
put the satellite into orbit.

increase their cost. On the other hand, several of the

long-range subsonic strike aircraft CBO considered were
assumed to be stealthy (in particular, to be difficult to detect with radar), implying the use of radar-absorbing materials and construction techniques that also increase cost.
(An aircrafts shapefor example, a tailless designcan
also affect its stealth. Because of the associated heat environment, CBO assumed that for supersonic cruise aircraft, shape would be the primary means used to reduce
radar signature.)

ture designed to be able to maneuver and withstand the

deceleration and heat generated by atmospheric friction
as the CAV reenters the atmosphere at high speed and
glides to its target. Generally, longer-range CAVs must be
launched on trajectories that have them fly higher above
the atmosphere, reenter at higher speeds, and experience
higher deceleration than do shorter-range CAVs.3 Providing the needed thermal protection and structure to withstand the forces associated with the deceleration affects
cost. Additionally, costs are greater for the larger booster

Range and payload also affect the costs of the CAV systems considered by CBO. A CAV contains a warhead
made of conventional high explosives encased in a struc-

3. That general rule applies to the CAVs considered in CBOs analysis. Conceptual designs exist, however, for long-range CAVs that
would fly on lower-altitude trajectories.

CHAPTER TWO

rockets needed to launch both longer-range CAVs and

CAV systems with greater payloads (that is, simultaneous
launch of a larger number of individual CAVs). The
heaviest and most expensive launchers are those used to
place multiple CAVs in low-Earth orbit.
The remainder of this chapter discusses in greater detail
the design considerations incorporated in CBOs analysis
of long-range strike alternatives.

Aircraft
CBOs analysis considers aircraft that could replace existing long-range heavy bombers (the B-52, B-1, and B-2)
in terms of their range and payload, as well as aircraft that
could provide improved responsiveness, measured by the
time required to strike a target at maximum range. Because the Air Force has been discussing the possibility of
pursuing a medium-range or regional bomber potentially derived from the F-22 fighter, CBO also considered
three medium-range alternatives spanning a wide range
of potential payloads. Needed range, payload, speed, and
stealth determine the trade-offs that can be made among
an aircrafts aerodynamics, propulsion, and structure.
Propulsion. All of the new aircraft examined use airbreathing turbine engines for propulsion. Performance of
the propulsion system is typically measured by specific
fuel consumption (SFC), which is the pounds of fuel
consumed per hour per pound of thrust generated. Typical SFC values for jet engines range from 0.5 for an efficient turbofan to 1.0 or greater for high performance, but
less efficient, supersonic aircraft engines. The lower the
SFC, the less fuel consumed and the greater the aircrafts
potential range or payload. According to CBOs calculations, aircraft that dash at supersonic speeds can have fuel
consumption rates up to six or seven times higher than
rates during subsonic cruise.
Aerodynamics. The shape of the wings and the fuselage of
an aircraft determine how well it can maneuver through
the air, as well as how much engine thrust is required to
overcome the drag caused by the friction of air. One measure of aerodynamic performance is the ratio of the lift
generated (which must be equal to the aircrafts weight in
cruise) to the drag. The higher the lift-to-drag (L/D) ratio, the longer the range of the aircraft for a given speed
and load of fuel. In addition, low-drag aircraft can cruise
at higher speeds. Typical values of the L/D ratio for aircraft cruising at optimum speed (the speed that maximizes their range) vary from 10 to more than 20 for sub-

ALTERNATIVE DESIGNS FOR LONG-RANGE STRIKE SYSTEMS

sonic aircraft; ratios of less than 10 are common for

supersonic aircraft.4
Structures. The ratio of the empty weight of an aircraft
that is, its weight without fuel, crew, or payloadto its
maximum takeoff weight indicates its capacity to carry
fuel and payload. That ratio is called the empty weight
fraction, which varies from 40 percent to 45 percent for
bombers and cargo aircraft. The use of modern composite materials in aircraft construction can reduce the
weight of an aircrafts structure while preserving needed
strength. In general, for a given set of engines, the lighter
the structure of an aircraft, the higher its cruise speed
will be.
The effect that propulsion, aerodynamics, and structures
have on an aircrafts range, speed, and payload can be
quantified by the Breguet range equation, as follows: 5
R = (Vc / SFC)(L/D)ln(Wo / W).
In the equation, R is the aircrafts range, Vc is the aircrafts
cruise speed, SFC is the specific fuel consumption of the
engines used on the aircraft, L/D is the aircrafts lift-todrag ratio, Wo is the aircrafts gross takeoff weight, and W
is the weight of the aircraft minus the fuel burned to
reach the range R. Thus, the aircrafts range is determined
by the performance of its propulsion systems (Vc / SFC),
its aerodynamics (L/D), and its structure (Wo / W).6
Starting with a desired payload and assumptions regarding an aircrafts aerodynamics, propulsion, and structure,
CBO used a model based on the Breguet range equation
to estimate the gross takeoff and empty weights for the alternative aircraft displayed in Table 2-1. For the range estimates shown in the table, CBO assumed that the longrange strike aircraft would not be refueled during flight
4.

Work has been done, however, on designs for supersonic aircraft

incorporating features that could yield cruise lift-to-drag ratios
approaching 10.

5. See, for example, Daniel P. Raymer, Aircraft Design: A Conceptual

Approach, 3rd ed., American Institute of Aeronautics and Astronautics Educational Series (Reston, Va.: American Institute of
Aeronautics and Astronautics, Inc., 1999).
6. For this formulation of the Breguet equation to be used, an aircrafts flight is divided into a number of phasessuch as climb to
altitude, cruise, and landingduring which SFC, velocity, and L/
D are assumed not to vary (although they can be different for each
phase). The formulation can also be adjusted to account for payload expended (that is, bombs dropped) during an aircrafts flight.

10

ALTERNATIVES FOR LONG-RANGE GROUND-ATTACK SYSTEMS

and would cruise until it reached the maximum range

such that its remaining fuel would be sufficient to enable
a safe return to its base. At that maximum range, the aircraft would drop its full payload of weapons and return to
base, preserving about 10 percent of its full fuel load as a
reserve in case of an emergency. Actual missions would be
extendable if the aircraft was capable of aerial refueling.
The total payload carried by an aircraft consists of the
weapons it expends, any racks or other equipment needed
to install the weapons in the aircraft, and the aircrafts
crew and their equipment.
Other Design Considerations. CBO assumed that its alternatives for long-range strike aircraft would be manned.
Unmanned versions of such aircraft have been proposed.
The empty weights and costs of such unmanned aircraft
might be less than those of the aircraft considered by
CBO because unmanned aircraft do not need to carry
crew support equipment. On the other hand, unmanned
aircraft could require the use of high-bandwidth satellite
communications for command and control and would
require the development and testing of complex software
automating their operation. Those factors would tend to
increase the cost of unmanned aircraft relative to manned
aircraft.

Common Aero Vehicles

For the foreseeable future, achieving very fast response
times (that is, on the order of an hour or less) at global
ranges will require systems such as CAVs that have trajectories through space. The CAVs CBO examined are reentry vehicles shaped to generate sufficient lift to glide
many thousands of miles at hypersonic speed as they reenter the Earths atmosphere. The CAVs would use a
combination of thrusters and flaps to maneuver as they
glide. CBO considered CAVs that could be launched at
their targets using a surface-based intercontinental ballistic missile (ICBM) or a smaller surface-based mediumrange missile, as well as CAVs that would be placed in
equatorial low-Earth orbits (at an altitude of 270 nautical
miles) and de-orbited when needed. Because CAVs have
the ability to glide and maneuver in the atmosphere, they
would be more costly and technically challenging than
simpler reentry vehicles that might only be able to make
minor course corrections as they reach the target. However, the CAVs ability to shape trajectories means they

could avoid flying over diplomatically sensitive territory

and could be launched on initial trajectories that would
preclude spent booster rockets from falling on populated
areas.7
The Defense Advanced Research Projects Agency was
conducting a program known as FALCON (Force Application and Launch from the Continental United States)
to develop the technology to enable CAVs to be built and
made operational.8 The CAV design that the FALCON
program was considering would weigh about 2,000
pounds and carry a payload of 1,000 poundseither a
single conventional (that is, nonnuclear) warhead or multiple smaller munitions. The CAVs would be shaped to
generate sufficient lift at hypersonic speeds to enable
them to glide long distances downrange (that is, along the
projection of their ballistic flight path) as well as maneuver to strike targets up to 2,500 to 3,000 nm crossrange
(the distance perpendicular to their ballistic flight path).
Their ability to glide and maneuver makes CAVs more
technologically challenging than the reentry vehicles currently used to carry nuclear warheads, such as the Mark
12 launched on the Minuteman III ICBM. For example,
to generate lift, the surface area of a CAV would need to
be several times as large as that of an existing reentry vehiclean area that must be protected from the heat generated during the prolonged flight of the CAV at hypersonic speeds. (A CAV would be in flight at hypersonic
speeds for up to 30 times longer than the times for existing ballistic reentry vehicles. Some observers consider the
development and testing of materials that could withstand the heat generated for that period of time to be the
aspect of a CAV program incorporating the greatest technical risk.) The CAVs speed could give it an advantage
7. The Navy is exploring simpler maneuvering reentry vehicles with
its Enhanced Effectiveness Initiative for Trident II ballistic missiles. See Congressional Research Service, Conventional Warheads
for Long-Range Ballistic Missiles: Background and Issues for Congress
(January 2006).
8. Although the FALCON program still exists, its focus has shifted
from developing an operational weapon to more generic research
into hypersonic flight. In this study, CBO used common aero
vehicle as a generic term for such weapons because it is the most
commonly known. The CAV is, however, a specific concept of the
Air Forces and the Defense Advanced Research Projects Agencys.
The Hypersonic Glide Vehicle is a related concept of the Armys.

CHAPTER TWO

over aircraft-delivered conventional munitions if it was

used against hardened or deeply buried targets.9
The period spanning the time between the launch of a
surface-based CAV and its arrival at a target comprises a
number of phases. Those phases include the period during which the CAV is riding on its rocket launcher under
thrust, the ballistic suborbital flight through space that
occurs after the CAV separates from its launcher and lasts
until the CAV reenters the atmosphere, and the period of
hypersonic glide and maneuver through the atmosphere
to a target. For long-range CAVs, the boost phase would
last about three to four minutes, the suborbital phase
would last about 30 minutes, and the glide phase would
last from a few minutes (if little crossrange flight or additional downrange flight was required) up to 30 minutes.
For medium-range surface-based CAVs, the boost phase
would last somewhat more than two minutes, the suborbital phase would last several minutes, and the glide phase
would be comparable with that of a long-range CAV.
The use of a space-based CAV would involve a set of
phases having some commonality with surface-based
CAVs but also with key elements that are different. In
particular, a space-based CAV could not, in general, be
fired immediately following a decision to attack a target.
Because there would probably be a limited number of
CAVs in orbit, a certain amount of time would elapse before some CAVs orbit carried it to a position in space
within range of the target to be attacked. Once a CAV
satellite was in position, the orbiting CAV would have to
fire its attached retro-rocket to slow itself to suborbital
speed and reenter the atmosphere. Finally, after reentry,
the space-based CAV would glide and maneuver to its
target in the same way as a surface-launched CAV would.
CBO determined the number of different orbital locations needed for the space-based CAVs and the size of
their attached retro-rockets by imposing the requirement
that the responsiveness of space-based and surfacelaunched long-range CAVs be equal (see Figure 2-1).
CBO chose a retro-rocket providing a braking speed of
9. However, if the impact speed of a warhead is too great, its structure will be transformed from solid to fluid and it will fail to penetrate the target. Studies have shown that for steel impacts on
granite and concrete, maximum penetration depths will occur for
an impact speed of about 4,000 feet per second (or about Mach
4). See United States Air Force Scientific Advisory Board, Why
and Whither Hypersonics Research in the U.S. Air Force (December
2000).

ALTERNATIVE DESIGNS FOR LONG-RANGE STRIKE SYSTEMS

0.5 kilometers per second, corresponding to a de-orbit

time of about 15 minutes from a 270-nautical-mile circular orbit.10 Given a 15-minute de-orbit time, five orbital
locations at the equator would provide the same responsiveness as would a long-range surface-launched CAV, assuming each CAV had the ability to glide and maneuver
for up to 30 minutes within the atmosphere. (A constellation with more satellites in equatorial low-Earth orbit
would provide only marginally shorter response times at
greater cost.) The 2,500-to-3,000-nm crossrange flight
that the CAVs could achieve would enable them to strike
targets located in any of the worlds major populated areas
from an equatorial orbit.
Other choices for retro-rocket braking speeds are possible, with different braking speeds requiring a different
number of satellites to achieve the chosen response time.
However, there is no optimum choice of braking speed
and number of satellites that would minimize the total
mass (and, therefore, the launch costs) of a CAV satellite
constellation for a fixed number of CAVs in orbit. A design with lower braking speeds would have lighter satellites because each retro-rocket would be smaller. However, to compensate for the increase in de-orbit time that
would result from lower braking speeds, a greater number
of those lighter satellites would be needed in orbit.

The Specific Long-Range Strike

Alternatives CBO Examined
CBO analyzed eight alternative long-range strike systems
that reflect several types of systems that have been proposed within the defense community:

Alternative 1Arsenal Aircraft: Supersonic

Missiles on C-17 Aircraft
A number of previous studies on long-range strike systems have examined the concept of an arsenal aircraft
that could be a military cargo plane or commercial transport converted to carry large quantities of munitions.11
That long-range strike aircraft would not be stealthy; so
to provide it the capability to attack targets in the pres10. CBO chose a 270-nm altitude for the space-based CAVs because
it is the lowest orbit that is high enough to prevent friction with
the residual atmosphere that would cause the CAVs orbits to
decay substantially over their expected lifetime in orbit.
11. See, for example, Department of Defense, Defense Planning Guidance: Long Range Global Precision Engagement Study Final Report
(April 2003).

11

12

ALTERNATIVES FOR LONG-RANGE GROUND-ATTACK SYSTEMS

Figure 2-1.

Response Times of Long-Range Common Aero Vehicles

(Minutes to strike target)
70

60

50

30

Atmospheric Gliding and Maneuvering

30

Ballistic Flight

40

De-orbit
Orbital Positioning

30

15
Boost

20

33

21

10

Surface Based

Space Based

Source: Congressional Budget Office.

Notes: Response time for the space-based common aero vehicle assumes five satellites in equatorial orbit at an altitude of 270 nautical miles
and with a 0.5 kilometer per second de-orbiting rocket.
Atmospheric gliding and maneuvering time will vary depending on the targets location.

ence of sophisticated air defenses, CBO assumed that the

aircrafts payload would consist of high-speed cruise missiles with a speed of Mach 3 and a range of 500 nm.
Those high-speed missiles would be able to reach their
targets in about 20 minutes or less after launch, although
times on the order of several hours might be needed for
the aircraft to fly to the missile-launch location. CBO estimated that, consistent with the C-17s capabilities, the
arsenal aircraft would cruise at Mach 0.76 and carry a
payload of 167,000 pounds, about 80 percent of which
would be expendable munitions.12 CBOs analysis indicates that the arsenal aircraft would have a combat range
of 1,500 nautical miles, which would enable it to attack
targets at a range of 2,000 nm with the addition of the
supersonic missiles 500-nm range.
12. The C-17s total payload is about 167,000 pounds. CBO assumed
that racks and dispenser mechanisms for the high-speed cruise
missiles carried by the aircraft would compose about 20 percent of
that total.

Alternative 2Medium-Range Subsonic

Stealth Bomber
As mentioned earlier, the Air Force has recently discussed
the potential need for a new medium-range bomber. In
this analysis, CBO examined a subsonic design having a
combat range of about 1,500 nautical miles without refueling, a payload of 20,000 pounds, and a cruising speed
of Mach 0.85. That design assumed a crew of two and
propulsion from two engines derived from existing designs. CBOs analysis indicates that the aircrafts empty
weight would be about 60,000 pounds, with a maximum
gross takeoff weight of 120,000 pounds. The design was
assumed to be highly stealthy, providing the capability to
conduct operations in the presence of sophisticated air
defenses.

Alternative 3Medium-Range Stealth Bomber with

Supersonic Dash
CBO also examined a medium-range bomber that would
cruise at subsonic speeds but that would have the capability to dash at supersonic speeds, hereafter referred to as

CHAPTER TWO

the medium-range supersonic bomber. That aircraft

would be similar in concept to a medium-range bomber
derived from the F-22 supersonic stealth fighter, which is
being considered by the Air Force. The alternative examined by CBO would have a cruising speed of Mach 0.85
and could dash at speeds up to Mach 1.5 while carrying
10,000 pounds. If it cruised exclusively at subsonic speed,
the aircrafts combat range would be 1,500 nautical miles.
However, if the aircraft dashed to its target at Mach 1.5
and then returned at Mach 0.85, its combat range would
be about 800 nautical miles; flying exclusively at Mach
1.5, its range would be somewhat less than 600 nautical
miles. The aircraft would have a crew of two and be powered by two high-thrust engines, probably derivatives of
the F119-100 engines used on the F-22. CBOs analysis
indicates that the aircrafts empty weight would be about
59,000 pounds, and maximum gross takeoff weight
would be about 126,000 pounds. The design was assumed to be highly stealthy.

Alternative 4Long-Range Subsonic

Stealth Bomber
When examining a design for a new long-range bomber,
CBO considered what could be viewed conceptually as
an improved version of the existing B-2. CBO assumed
that such an aircraft would have a subsonic cruise speed
of Mach 0.85, could carry 40,000 pounds of payload,
and would have a combat range, without refueling, of
2,500 nautical miles (compared with about 2,000 nautical miles for a fully loaded B-2). CBO assumed that, as
with the current B-2, this long-range strike aircraft would
have a crew of two but would use two high-performance
engines derived from current designs instead of the four
engines used on the B-2. Assuming that the aircrafts design could incorporate improvements in both aerodynamic performance and engine fuel efficiency relative to
the B-2, CBOs analysis indicates that the aircrafts empty
weight could be about 128,000 pounds (compared with
the B-2s 154,000 pounds) and that its gross takeoff
weight would be about 283,000 pounds. The design
would be highly stealthy.

Alternative 5Long-Range Supersonic

Cruise Bomber
CBOs analysis includes a bomber that could cruise at
Mach 2.4 and carry a payload of 40,000 pounds. Assuming a maximum combat range of 2,500 nautical miles,
CBOs analysis indicates that such an aircraft would have
an empty weight of about 165,000 pounds and a gross

ALTERNATIVE DESIGNS FOR LONG-RANGE STRIKE SYSTEMS

takeoff weight of about 440,000 pounds.13 A supersonic

bomber that could cruise to its target at Mach 2.4 would
have responsiveness about 65 percent better than that of
the long-range subsonic bomber considered by CBO (1.8
hours to reach a target 2,500 nautical miles away versus
5.1 hours). The aircraft would have two crew members
and four (newly developed) engines. CBO assumed that
the aircraft would be optimized for high-speed flight and
would therefore have limited ability to loiter at subsonic
speeds.14 For this design, the aircrafts shape would be the
primary means of achieving stealthimplying that the
aircraft would be less stealthy than those in Alternatives
2, 3, and 4because it is unclear to what extent the special materials used to help further reduce aircraft signatures would be robust enough to withstand the high temperatures of sustained supersonic flight.

Alternative 6Medium-Range Surface-Based

Common Aero Vehicle
On the basis of information provided by the Army, CBO
assumed that a 2,000-pound CAV could be launched on
a solid-fuel booster rocket small enough to be housed on
a mobile launcher/erector. The booster could launch a
single CAV on a trajectory with an apogee of about 49
nautical miles. Under CBOs assumptions, the CAV
would be able to glide up to 3,200 nautical miles downrange and would have a maximum crossrange maneuvering capability of somewhat less than 950 nautical miles.
CBO assumed that a medium-range CAV battery would
include two missiles (each carrying a single CAV) placed
13. CBOs analysis assumed that technology available in the relatively
near term would be used to build the Mach 2.4 long-range strike
aircraft. Over the longer term, technology that might emerge from
continuation of the Defense Advanced Research Project Agencys
Quiet Supersonic Program (or related efforts) could make lighterweight designs for long-range supersonic aircraft possible. If
improvements in engine SFC and aerodynamics consistent with
that programs goals were achieved, CBOs analysis indicates that a
Mach 2.4 bomber with a 20,000-pound payload and a range of
2,500 nautical miles would have an empty weight of about 56,000
pounds. CBO did not consider alternatives for long-range strike
aircraft that had cruising speeds well above Mach 3 (so-called
hypersonic aircraft) because it assumed that the required advancements in technology could not be incorporated into an operational aircraft for more than 20 years.
14. There are proposals for variable-cycle supersonic engines with
improved fuel efficiency at subsonic speeds. The designs for such
engines introduce complexities in their operation and maintenance, however, and CBO did not assume that they would be
developed for and used on the supersonic bomber considered in
this analysis.

13

14

ALTERNATIVES FOR LONG-RANGE GROUND-ATTACK SYSTEMS

in two canisters carried on a single mobile launcher. Vehicles carrying launch control systems, command-and-control systems, and other support equipment would also be
included in the CAV battery, which would be transportable by air. CBO assumed that 24 batteries would be purchased, 20 of which would be for operational use and
four for spares.

Alternative 7Long-Range Surface-Based Common

Aero Vehicle
CBO assumed that long-range surface-launched CAV capability could be provided by 2,000-pound CAVs
launched on Peacekeeper ICBMs being retired from their
strategic nuclear mission. Two CAVs could be launched
on a single Peacekeeper. CBO assumed that two sets of 10
Peacekeepers would be modified for operational use, with
each set capable of launching 20 CAVs. One set would be
placed in silos constructed at Cape Canaveral Air Force
Station on the East Coast of the United States, and the
second set would be placed at Vandenberg Air Force Base
on the West Coast. That basing arrangement would enable targets to be attacked worldwide while precluding
the impact of a spent Peacekeeper booster within a populated region (if used, the Peacekeepers would burn out
over the Atlantic or Pacific Oceans). Four additional systems would be built as spares.

Alternative 8Space-Based Common Aero Vehicle

CBO also considered an alternative that would base the
2,000-pound CAV in space. CBO assumed that five sets
of eight CAVs would be placed in low-Earth orbit, with
each set housed in a protective satellite providing power,
thermal control, station-keeping (the ability to maintain
proper orbital position and orientation), and communications for command and control. (The earlier discussion
of general design considerations for CAVs provides the
rationale for CBOs choices of orbit and number of satellites.) With eight CAVs housed in each of five orbiting
protective satellites, this alternative would provide 40 operational CAVs, the same number assumed in Alternatives 6 and 7.15 CBO assumed that the CAV satellites
would have a service life of about 10 years in orbit. Maintaining a space-based CAV capability for 30 years, a period comparable with the typical service life of military

aircraft such as those considered in Alternatives 1 through

5, would require that enough CAV satellites be purchased
to replace each one twice. With the addition of eight
CAVs (and one protective satellite) as spares, a total of
128 CAVs would be purchased under Alternative 8.16
As in its analysis of potential designs for space-based interceptors used for missile defense, CBO examined two
potential weights for the protective satellites: 50 percent
of the weight of the CAVs to be housed and 20 percent of
their weight.17 CBO estimates that the total weight of
each space-based CAV would be about 2,730 pounds: the
CAV would weigh 2,000 pounds, and the solid fuel deorbit retro-rocket would weigh about 730 pounds.
Therefore, eight CAVs housed in a heavier protective satellite would weigh about 32,760 pounds, and eight CAVs
housed in a lighter-weight protective satellite would
weigh about 26,200 pounds. CBO assumed that eight
CAVs housed in a heavier protective satellite could be
launched from Cape Canaveral into a 270-nautical-mile
circular equatorial orbit using a heavy-payload version of
the Delta IV Evolved Expendable Launch Vehicle
(EELV).18 A Delta IV medium-plus launcher should be
capable of placing eight CAVs housed in a lighter-weight
protective satellite in orbit.19 CBOs cost estimates assume the heavier protective satellite and launch on a
heavy EELV.
15. As mentioned previously, the crossrange maneuver capability of
2,500-to-3,000 nautical miles assumed by CBO would enable a
CAV in an equatorial orbit to reach targets in North Korea or at
higher latitudes. If development of CAVs with that crossrange
capability proved infeasible, multiple sets of CAVs in multiple
inclined orbits would be required to provide the ability to attack
targets worldwide within an hour, substantially increasing costs.
16. The spare satellite would be needed if a launch vehicle failed during launch or if a satellite malfunctioned while in orbit.
17. That analysis is described in Congressional Budget Office, Alternatives for Boost-Phase Missile Defense (July 2004), available at
www.cbo.gov.
18. See Boeing Corporation, Delta IV Payload Planners Guide (October 2000 and later supplements), available at www.boeing.com/
defense-space/space/delta/docs/DELTA_IV_PPG_2000.PDF.
19. Ibid.

CHAPTER

3
Comparing the Capabilities of the Long-Range Strike
Alternatives CBO Examined
Chapter 3: Comparing the Capabilities of the Long-Range Strike Alternatives CBO Examined

wo likely roles for future long-range strike systems

will be to provide a new prompt, nonnuclear global strike
capability and to contribute to the strike capabilities
available to commanders in regional conflicts, especially
those that are conducted over long distances because of
local geography or basing restrictions. The Department
of Defense, however, has not established future requirements for long-range strike systems with any detail beyond identifying general types of capabilities such as
those described in the Secretarys Prompt Global Strike
Plan and the Air Forces nascent three-phase plan for such
systems (see Chapter 3).1 The Long Range Global Precision
Engagement Study prepared by the Air Force as directed
by the Fiscal Year 2004 Defense Planning Guidance also
established some initial structure to guide thinking about
future long-range strike capabilities. The study defined
four capability focus areas that describe what DoD expects future long-range strike systems to provide:
B

Prompt Global Strikethe ability to strike anywhere

(from very long ranges if necessary) within about 12
hours, with an emphasis on global access, survivability,
and speed at the expense of volume of fire.
Prompt Theater Strikesurvivable strike capability
over theater distances (up to 2,000 nautical miles)
with a greater emphasis on volume of fire.
Persistent Area Strikestrike capability for higher volume of fire from shorter ranges, and attacks against
time-sensitive targets (10 to 15 minutes from an execute order) provided by high speed or long loiter
times.

1. Office of the Secretary of Defense, Report to Congress on: Prompt

Global Strike Plan (June 2005).

Battle Management, Command, Control, Communications, Computers, and Intelligence, Surveillance and Reconnaissance (BMC4ISR)a combination of onboard
sensors and connectivity to external sensors and communications supporting the other three focus areas.

To make quantitative comparisons of the military utility

of different approaches for long-range strike systems, the
Congressional Budget Office identified several specific
capabilities that would enable a future system to contribute substantively to both prompt global strike missions
and to air campaigns in regional conflicts. Those capabilities include:
B

Reachthe ability to attack targets regardless of

location.

Responsivenessthe ability to attack targets quickly.

Firepowerthe ability to sustain attacks over time.

Payload Flexibilitythe ability to deliver different

types of munitions.

Survivabilitythe ability to avoid or defeat air

defenses.

Those capabilities are related to, although not the same

as, the Long Range Global Precision Engagement Studys capability focus areas. CBO did not assess how much of a
particular capability might be desired or needed but
rather compared how well the postulated alternatives (as
well as todays forces) could provide them. CBO assumed
its alternatives would be designed to incorporate other attributes that have been identified as necessary for longrange strike, such as the ability to strike day or night and
in poor weather, as well as connectivity to the networked

16

ALTERNATIVES FOR LONG-RANGE GROUND-ATTACK SYSTEMS

Figure 3-1.

Distribution of Worst-Case Penetration Distances for Countries of the World,

Measured by Geographic or Operational Limits
(Cumulative percentage of countries)
100

Geographic Limitsa
Operational Limitsb

80

Approximate
Radius of a
B-2 Bomber
60

Approximate Radius
of a B-1 Bomber

40

Approximate Radius
of an F-16 Fighter
20

0
Less
Than
250

500

750

1,000

1,250

1,500

Worst-Case Penetration Distance (Nautical miles)

1,750

2,000

More
Than
2,000

Source: Congressional Budget Office based on data from the Central Intelligence Agency, World Fact Book (2003), and the Air Force.
Note: Radius of a B-52 bomber is approximately 3,000 nautical miles.
a. Geographic limits are based on a countrys long axis or maximum distance from open ocean.
b. Operational limits are based on geographic limits plus the need for a tanker to stand off 100 nautical miles and a 30 percent range penalty
to account for threat avoidance.

force (for example, the ability to send and receive targeting information).

Ability to Reach Targets Anywhere on

Earth
With the exception of Alternative 6the medium-range
common aero vehicleall of CBOs long-range strike alternatives offer sufficient range or endurance to fly from
any base or launch location to almost any other point on
Earth. Long-range CAVs (Alternative 7) and space-based
CAVs (Alternative 8) could do so unassisted, and the aircraft alternatives could do so with aerial refueling support. Operational restrictions, however, can prevent a system from being able to attack a target that it would
otherwise be physically able to reach. For example, because vulnerability to air defenses usually prevents air-

borne tanker operations inside denied or hostile airspace,

aerial refueling may not be available in all cases where it
might be needed to allow strike aircraft to attack targets
deep within a large countrys borders.2
In general, the longer the unrefueled range of a system is,
the greater its flexibility to attack targets wherever
needed. To quantify the flexibility offered by increasing
unrefueled range, CBO estimated what fraction of the
2. Todays tanker fleet comprises converted commercial airliners that
are easily detected by radar, and the task of aerial refueling requires
the tanker and the receiver to fly a predictable (and easy-to-target)
profile while fuel is being transferred. Consequently, most refueling is done in secure airspace, although tankers can be sent into
hostile airspace if the need warrants it. There have been proposals
to develop stealthy tankers, but the feasibility of achieving a
stealthy refueling capability is undetermined.

CHAPTER THREE

COMPARING THE CAPABILITIES OF THE LONG-RANGE STRIKE ALTERNATIVES CBO EXAMINED

Figure 3-2.

Ability of Alternative Strike Systems to

Attack Targets Deep in Enemy Territory
(Percentage of countries in the world covered)
100

80

60

40

20

0
1

Alternatives

Source: Congressional Budget Office.

Note: Alternative 1 = Arsenal aircraft
Alternative 2 = Medium-range subsonic bomber
Alternative 3 = Medium-range supersonic dash bomber
Alternative 4 = Long-range subsonic bomber
Alternative 5 = Long-range supersonic cruise bomber
Alternative 6 = Medium-range surface-based common aero
vehicle
Alternative 7 = Long-range surface-based common
aero vehicle
Alternative 8 = Space-based common aero vehicle
a. The results for the medium-range common aero vehicle in Alternative 6 assume that the launcher can be located near the target
country.

worlds countries would be entirely accessible to each

long-range strike alternative. That entailed comparing an
alternatives unrefueled range (in the case of the penetrating-aircraft alternatives) or its missile range (in the case of
the arsenal-aircraft and the CAV alternatives) with the
geographic dimensions of each country. To provide full
coverage, CBO assumed that the necessary penetration
distance for each country would be the smaller of the
longest distance between two points within a countrys
borders or the farthest distance from an ocean (because
strike aircraft could approach from the sea). By that measure, even shorter-range tactical fighters could reach any
point in most of the worlds countries (see Figure 3-1).

For example, an F-16 carrying two 2,000-pound Joint

Direct Attack Munitions could reach any point within
about 80 percent of the worlds countries if aerial refueling was available at the target countrys border (and along
the route to the border, if necessary) and if its route between the tanker and the target was a straight line.3 The
F-16s coverage would drop to about 65 percent of all
countries if additional operational considerationsin
this example, the need for a tanker to stand off 100 nm
outside of the adversarys border and a 30 percent range
penalty to account for indirect routes to avoid air defensesare included. (This discussion does not address
whether the strike system could successfully avoid air defenses. Survivability is examined in a later section.) Todays heavy bombers offer full coverage, even with the examples operational constraints (as shown in Figure 3-1).
Most of CBOs long-range strike alternatives offer complete or nearly complete coverage as well (see Figure 3-2).
The long-range surface-based CAV and the two longrange-bomber alternatives would fully cover all countries,
and the two medium-range-bomber alternatives would
fully cover about 95 percent of all countries. The
medium-range-bomber alternatives would achieve full
coverage if they were allowed to enter enemy airspace
from points on the border closer to the target. The spacebased-CAV alternatives crossrange maneuverabilitythe
ability to fly north or south from its equatorial orbit
would limit its coverage to North or South latitudes less
than about 60 degrees but allow it to still fully cover
about 97 percent of the worlds countries. The mediumrange-CAV alternative would offer full coverage if its
launcher could be positioned close enough to the target
countryno farther than about 500 nm for worst-case
geography and the largest countries but as far as 2,000
nm in most cases. Consequently, medium-range CAVs
carried on board surface ships or submarines would offer
good coverage. (CBO did not estimate the costs of putting CAVs aboard ships or submarines.)
The arsenal aircraft would offer the least coverage among
CBOs alternatives. Its 500-nm missile range and the inability of C-17s to penetrate air defenses would limit its
coverage to around 75 percent of the worlds countries.
(CBO assumed that the C-17 would operate 100 nm
from the border but that the missile would fly directly to
3. The percentages cited are based on a list of countries found in
Central Intelligence Agency, World Fact Book (2003). They
include all countries or autonomous areas listed.

17

18

ALTERNATIVES FOR LONG-RANGE GROUND-ATTACK SYSTEMS

its target.) That percentage would improve if the C-17

could launch its missiles from a more favorable location
on the perimeter of the target country or if air defenses
were suppressed enough to allow the C-17 itself to penetrate the target countrys borders.
This geography-based approach to examining the penetration capability of CBOs alternatives does not distinguish among nations that are more or less likely to be
considered potential threats. Those judgments are subjective and change over time. For three nations commonly
mentioned today as current or possible future adversariesNorth Korea, Iran, and Chinathe CAV and longrange-bomber alternatives would provide total coverage.
The medium-range bombers could not reach parts of
China, and the supersonic air-launched missile could not
reach parts of Iran and China.

Responsiveness
Responsiveness in a long-range strike system is the ability
to quickly attack a target after the order to do so is given.
Systems that can execute their attacks in a shorter period
of time are said to be more responsive than those needing
more time.
In its analysis, CBO considered two operational contexts
for assessing responsiveness:
B

Preplanned Missionsthose for which the identity of

the target is known before the mission begins. Preplanned missions typically achieve responsiveness on a
timescale of hours or days.

Fleeting Target/Ground Support Missionsthose for

which the specific target is not known before the mission begins, such as missions involving close air support targets identified by ground controllers when the
aircraft are already overhead. Such missions typically
aim for responsiveness on the order of a few minutes.

vanced several hours in an attempt to hit leadership targets that intelligence sources were thought to have
located. Responsiveness to fleeting targets was illustrated
later when a B-1B bomber over Iraq was diverted to attack a building where Saddam Hussein was thought to be
located. The bomber struck the target less than 15 minutes after receiving the order to attack.
The responsiveness needed against a target depends on
what the target is and what the target is doing. For example, high responsiveness would usually be desired against
an adversarys mobile ballistic missile launchers because
they might launch their missiles at any time and because
they are difficult to track for long periods of time. The responsiveness desired against an armored unit, on the
other hand, would depend on what that unit was doing.
If it was far from friendly forces, responsiveness on the order of hours or even days might be sufficient because the
unit would pose no immediate threat and would be easy
to track. (It might not even be deemed worth an attack.)
If the same unit was about to overrun friendly ground
forces, however, responsiveness on the order of minutes
would be needed.
Several steps contribute to the time that elapses between
when a target is detected and when it is attacked. Those
steps include time for transmission and processing of data
from the sensor that detects the target, time for analyzing
the data and deciding to attack, time for the decision to
reach the unit that will conduct the attack, and time to
prepare and execute the mission. Total times can vary
widely depending on factors such as the political or military situation, the type of target, the volume of data being
processed, and the type of strike system tasked to perform
the attack. This analysis focuses on the time that would
elapse between when a strike order is received by the attacking unit and when weapons hit the target.

Responsiveness in Preplanned Attacks

Those two types of responsiveness are examined separately because each would require that strike systems be
employed differently.

Responsiveness in preplanned attacks is primarily a function of the speed of the strike system and the distance to
the target. Thus, higher speeds and the ability to be based
close to potential targets contribute significantly to a
strike systems responsiveness.4 The benefit of higher

Many strike missions are preplanned and can be prepared

long in advance of the actual attack or rapidly as a conflict progresses. Examples of preplanned missions that
were prepared rapidly were seen at the beginning of Operation Iraqi Freedom when the initial air strikes were ad-

4. Other contributing factors include the ability to be quickly

launched after an attack order is received and the ability to rapidly
plan the mission (or to conduct detailed planning during transit
to the target). CBO assumed that its alternatives would be
designed to incorporate those characteristics.

CHAPTER THREE

COMPARING THE CAPABILITIES OF THE LONG-RANGE STRIKE ALTERNATIVES CBO EXAMINED

Figure 3-3.

Response Times of Alternative Strike Systems for Preplanned Missions

(Hours to strike target)
25

20

Arsenal Aircraft
(Alternative 1)
15

Medium-Range
Bombers and
Long-Range
Subsonic Bomber
(Alternatives 2, 3, and 4)

10

Long-Range
Supersonic
Cruise Bomber
a
(Alternative 5)

Common Aero Vehicles (Alternatives 6, 7, and 8)b

0

2,000

4,000

6,000

8,000

10,000

Source: Congressional Budget Office.

a. The steps in the line for the supersonic cruise bomber result from its need to slow for aerial refueling.
b. Response times for the common aero vehicle alternatives will vary for a given distance to the target, depending on the specific flight
profile needed. The medium-range common aero vehicle has a maximum distance to target of 3,200 nautical miles.

speed, however, depends on the distance to be traveled,

because over longer distances, a faster system has more
time to pull ahead of a slower system. To capture that dependence, CBO compared the preplanned mission responsiveness of its long-range strike alternatives over different mission distances (see Figure 3-3).
Over global distances, the CAV alternatives offer by far
the best response times because they travel at speeds
many times faster than aircraft do. The response times for
CAVs are shown as a band because those times do not depend only on the distance between launch point and target. The specific flight times are dependent on the trajectory that must be flown and are therefore influenced by
other factors such as the amount of glide trajectory relative to ballistic trajectory needed and, in the case of Alternative 8, the locations of the space-based CAVs in orbit at
the time a launch order is received.5 Shots requiring a
crossrange trajectory would need the longer times because
the difference in range off the ballistic trajectory would be
achieved by CAV maneuvers in the atmosphere. For the

space-based CAV, those times could vary from about 15

minutes to one hour depending on the delay needed for
an orbiting CAV satellite to move into position and the
amount of maneuvering the CAV would require after it
reentered the atmosphere. The upper bound of responsiveness for the space-based CAVs could be reduced by
placing more CAVs in orbit because that would reduce
the wait for a CAV to reach the correct orbital position.
Response times could vary from about 30 minutes to one
hour for the long-range surface-based CAV. The
medium-range CAV could offer faster response times at
shorter ranges because of its shorter ballistic phase of
flight. Short response times for the medium-range CAV
assume that the launcher is within range of the target.
The response time could be many hours or days if
5. Although ranges for the CAV options are referred to in terms of
ground distance between the launcher and the target, CAVs would
actually fly a much longer distance along their suborbital trajectories. Similarly, the actual distance flown by the space-based CAVs
would include the vertical distance from orbit.

19

20

ALTERNATIVES FOR LONG-RANGE GROUND-ATTACK SYSTEMS

Table 3-1.

Flight Time Needed to Reach a Target in Afghanistan

Time to Target (Hours)
a
b
Flying at Mach 0.85
Flying at Mach 2.4

Launch Location

Type of Base

United States

U.S. territory

15.0

6.0

Guam

U.S. territory

10.0

4.0

Diego Garcia

Territory of an ally

6.0

2.2

Indian Ocean

Aircraft carrier

2.5

Source: Congressional Budget Office.

Notes: Diego Garcia is a British-held island in the Indian Ocean.
Aerial refueling would be necessary for the notional missions presented in this table.
a. A speed of Mach 0.85 is representative of the aircraft in Alternatives 2, 3, and 4 and todays bombers or fighters.
b. A speed of Mach 2.4 is representative of the aircraft in Alternative 5.
c. With foreseeable technology, a Mach 2.4 cruise bomber would probably be too large to operate from an aircraft carrier.

launchers must first be moved to a location within range

of the target.

ternative 5 would be about eight hours shorter than the

time for the subsonic-aircraft alternatives.

The response times for the aircraft alternatives are a

strong function of speed. The three subsonic aircraft alternatives (1, 2, and 4) have similar response times over a
wide band of ranges because their speeds are about the
same over most of their flight to the target. The arsenal
aircraft is slightly less responsive because the C-17 aircraft
is slower than the penetrating bombers, at Mach 0.76
versus Mach 0.85, although the supersonic missile would
fly the final 500 nm of its mission at Mach 3. The response time of Alternative 3 would be similar to that of
the subsonic-only aircraft because it could dash at speeds
up to Mach 1.5 only for limited distances. Its range
would be less than 1,000 nm if it cruised at supersonic
speeds. The supersonic cruise bomber (Alternative 5) offers a considerably quicker response, although it would
have to slow for refueling during missions over distances
longer than its unrefueled radius (a consideration that results in the stair-step shape of the line in Figure 3-3 on
page 19).6 The response-time advantage for Alternative 5
increases with mission distance. Over long regional conflictsdistances of about 1,000 nm to 1,500 nmthe
response time of the supersonic cruise bomber would be
one to two hours shorter than the time for the subsonic
alternatives. For global missions, the response time for Al-

None of todays systems could match the responsiveness

of the CAV alternatives over long ranges. Only airborne
aircraft loitering in the area (or, possibly, aircraft on alert
at a base very close to the target) could do so. Todays subsonic bombers would provide ranges and response times
similar to those of CBOs subsonic-aircraft alternatives if
they were flying over similar distances. Navy aircraft have
the potential to offer better response times than those of
the aircraft alternatives examined by CBO if an aircraft
carrier was in the area (see Table 3-1). For example, a
strike aircraft from a carrier in the Indian Ocean could
have similar responsiveness against a target in Afghanistan
to that of the Mach 2.4 bomber in Alternative 5 based on
the island of Diego Garcia. However, arranging tanker
support for the shorter-range Navy aircraft might not be
possible on such short timelines. A carrier-based aircraft
with a range like that of the medium-range-bomber alternatives would not need such support, however. Operating
from a carrier would have the advantage of obviating permission to use foreign bases but would still require that a
carrier be in the area.

6. The subsonic bombers might also slow to refuel, but the difference in speed would be much smaller.

Responsiveness Against Fleeting Targets

Although the preplanned response times estimated above
are adequate against many types of targets, they are too
long for others. Highly time-sensitive targets include
those that pose an immediate danger to friendly forces or

CHAPTER THREE

COMPARING THE CAPABILITIES OF THE LONG-RANGE STRIKE ALTERNATIVES CBO EXAMINED

civilians, those that are exposed for only brief periods of

time, and those on the move if they cannot be reliably
tracked. Inadequate responsiveness can lead to unsuccessful attacks because:
B

The target is able to complete its hostile action before

the strike is conducted (for example, an enemy unit
destroys a friendly unit before close air support arrives);

Unbeknownst to intelligence, surveillance, or reconnaissance (ISR) systems, the target moves before the
strike is conducted (for example, a terrorist leader
leaves a safe house before a strike delivers its attack on
a now-vacated building); or

The target moves and the strike system cannot locate

it to complete the attack (for example, a mobile missile launcher is several miles from its anticipated location before the strike arrives).

For time-sensitive targets, the probability of successfully

attacking decreases as response time increases. In the first
two examples above, as time passes, the probability that
the target will either be relevant or present decreases with
time. In the third example, the target may still be in the
general area, but the strike system might have to search
for it because the targets movement creates uncertainty as
to where it will be relative to its last known location.
When there is great uncertainty about the targets location, the strike system must be able to search the area
with its own sensors, a potentially time-consuming process that could expose the attacker to air defenses.
The very short response times needed against fleeting targetsthe Long Range Global Precision Strike Study set a
goal of five to 10 minutes from an execute ordercan be
achieved only with very high speeds or with systems that
are close to the target from the outset.7 Aircraft would almost certainly have to be airborne before receiving the
mission because the time needed to taxi and take off
could consume much of the available time. For example,
during Operation Desert Storm, the Air Force maintained combat air patrols of strike aircraft, usually F15Es, over Iraq in order to rapidly attack mobile SCUD
7. Improvements in ISR systems and communications links to strike
platforms could also help responsiveness. For that reason, good
connectivity to the ISR and communications networks is stressed
as an important attribute for future long-range strike systems.

missile launchers. Those aircraft could be minutes away if

a launch or launcher was detectedclose enough to find
and hit it before it could return to hiding. By comparison, aircraft waiting on the ground would have taken at
least an hour to reach the areanot soon enough to catch
the launchers that returned to hiding shortly after firing
their missiles.
In contrast to preplanned missions originating at air
bases, aircraft-based strike systems can have shorter response times than those for the CAV alternatives when
they operate from airborne orbits. If air defenses have
been suppressed, aircraft could be placed in orbits close to
or over areas where targets might be expected to appear. If
air defenses have not been suppressed, orbits could be located in the nearest secure airspace. Because the precise
locations of targets are not known in advanceif they
were, they would be preplanned targetsplanners must
provide enough orbits so that a strike aircraft is always
close enough to provide a sufficiently short response time.
Consequently, a primary disadvantage to loitering orbits
is the large number of aircraft that might be needed to
maintain them: aircraft that could be used elsewhere end
up tasked to perform orbits in the hope that targets will
reveal themselves.
In general, more effective loitering strike systems will require fewer aircraft to cover a given target area with a desired response time. Both speed and endurance can help
reduce the number of strike systems needed to provide
short response times over a target area. Higher speed enables greater area coverage from an orbit, and longer endurance means fewer systems are needed to maintain an
orbit around the clock.
The arsenal aircraft (Alternative 1) would have a significantly greater area of coverage from a single orbit than
would the other aircraft alternatives because the supersonic missile could fly up to 500 nm to the target at
Mach 3, faster than the other aircraft alternatives (see
Figure 3-4). Area coverage would increase slowly for targets that were more than 500 nm from the arsenal aircraft
because the C-17 must fly toward the target at Mach 0.76
before firing the Mach 3 missile. (That causes the sudden
decrease in the slope of the curve for Alternative 1 in the
figure.)
The subsonic penetrating-bomber alternatives (2 and 4)
would have the smallest area of coverage because of their
low speed, although that coverage would increase if, as

21

22

ALTERNATIVES FOR LONG-RANGE GROUND-ATTACK SYSTEMS

Figure 3-4.

Area Covered from Loitering Orbit for

Different Response Times
(Thousands of square nautical miles)
1,250

Arsenal Aircraft
(Alternative 1)

1,000

750

Medium-Range
Supersonic
Dash Bomber
(Alternative 3)

Medium- and
Long-Range
Subsonic Bombers
(Alternatives 2 and 4)

500

250

0
0

10

15

20

25

30

Response Time (Minutes to strike target)

Source: Congressional Budget Office.

with the arsenal aircraft, they could be armed with highspeed missiles. The medium-range supersonic bomber
would perform better than the subsonic alternatives, although it would lack the fuel for a Mach 1.5 dash to
ranges much greater than the supersonic missile could
cover at Mach 3. The higher-speed long-range supersonic
bomber (Alternative 5)not shown in the figure
would probably not be employed in this type of mission.
Its design would probably not be suitable for the long periods of efficient low-speed flight needed for a loiter mission. (Its endurance would be similar to that of the aircraft in Alternative 4 in terms of distance flown but much
shorter in terms of time aloft.)
The picture changes somewhat when endurance is factored into the number of aircraft required to continuously cover a given area (see Figure 3-5). The figure
shows the number of strike aircraft needed to continuously cover 45,000 square nautical miles (about 25 percent of Afghanistan) with a response time of not longer
than 10 minutes. Despite having less endurance than
some of the other alternatives, the large area covered by
the supersonic missiles enables the arsenal aircraft to provide coverage with fewer orbit locations than the other al-

ternatives couldand thus fewer total aircraftfor orbit

locations up to 1,000 nm from base.8 The number of aircraft needed to fill an orbit increases rapidly as the transit
distance to the orbit approaches the range of the aircraft
because the time on-station for each aircraft is rapidly
shortened. Beyond a 1,000-nm transit distance, only the
long-range subsonic bomber is well short of its maximum
range (2,500 nm versus 1,500 nm for the other alternatives). Consequently, if orbit locations are farther than
1,000 nm from a base (or last aerial refueling), long-range
subsonic bombers can provide 10-minute response times
with the fewest number of aircraft dedicated to filling the
orbits. The medium-range bombers require the largest
number of aircraft because they have neither the highest
speed nor the longest endurance. The faster mediumrange bomber, in Alternative 3, can provide coverage with
fewer aircraft than Alternative 2 can because its higher
dash speed allows it to cover a greater area from each
orbit.9
The preceding analysis assumes that a bombers payload is
fully expended before it returns to base. That would be
the case if targets were to arise at a rate that would use up
the bombers munitions during its time on-station or if
the bombers were tasked to strike preplanned targets at
the end of that time. If it was necessary to bring back unexpended weapons or if weapons were expended before
the available fuel, loiter times would be shorter.
Although the CAV alternatives could provide full and
continuous coverage from a single locationa country or
ship in the region for Alternative 6, the United States for
Alternative 7, or low-Earth orbit for Alternative 8their
responsiveness would be limited by the time required to
fly their trajectories. At best, that time would be on the
order of 10 minutes for a medium-range CAV if the warhead required minimal glide and maneuver time in the
atmospherethat is, if the launcher was relatively close
to the target. The space-based CAV would need at least
8. Distances shown in the figure are from a ground base. Coverage
could also be measured relative to the distance from an aerial refueling orbit. Relative results would be similar.
9. The figure shows Alternative 3 requiring more aircraft than Alternative 2 for orbits greater than about 1,000 nm from a base. That
occurs because at long ranges, the supersonic dash that gives the
bomber in Alternative 3 its extra reach would also seriously limit
orbit time because of the extra fuel consumed during the dash.
Because orbiting tactics would probably require air superiority,
tankers might be on hand to mitigate that limitation.

CHAPTER THREE

COMPARING THE CAPABILITIES OF THE LONG-RANGE STRIKE ALTERNATIVES CBO EXAMINED

Figure 3-5.

Number of Aircraft Needed for 24-Hour Coverage of 25 Percent of

Afghanistan with a 10-Minute Response Time
180

Medium-Range
Supersonic
Dash Bomber
(Alternative 3)

160
140
120
100
80

Long-Range
Subsonic Bomber
(Alternative 4)

60

Medium-Range
Subsonic Bomber
(Alternative 2)

40

Arsenal Aircraft
(Alternative 1)

20
0
0

500

1,000

1,500

2,000

2,500

Nautical Miles Between Base and Orbit Location

Source: Congressional Budget Office.

Note: This figure assumes that an aircraft expends all of its munitions before returning to base. Endurance would be lower (that is, more aircraft would be needed) if an aircraft returned with unused munitions. Endurance would be higher (fewer aircraft would be needed) if
distance was measured from an airborne tanker rather than from a ground base.

15 minutes to de-orbit, and the surfaced-based CAV

would need at least 30 minutes to fly from the United
States. If significant glide time in the atmosphere was
needed, response times could be as high as half an hour
for the medium-range CAV alternative or about an hour
for the other CAV alternatives.

The sortie rate for an aircraft depends on the duration of

the missions to be flown and on turn time, the time
needed between missions for servicing and rearming.
CBO assumed that all of the aircraft would be designed
for quick turn times but that larger and faster aircraft
would need longer turn times because they carry a larger
number of munitions and larger fuel loads.

Sustained Firepower

Under the alternatives considered here, CAVs would be

incapable of providing sustained firepower because they
would probably be purchased in limited numbers and because each CAV could be used only once. Among the
other alternatives, the two medium-range bombers offer
the lowest bomb delivery rate because they have smaller
payloads than the longer-range alternatives and fly at subsonic speeds (see Figure 3-6). The aircraft in Alternative 3
can dash for short distances at Mach 1.5 but would probably not fly entire missions at its dash speed. The arsenal
aircraft and long-range penetrating bombers have payloads higher than those of the medium-range bombers

The previous discussion focused on achieving very rapid

attacks against high-value targets that are likely to be
present in limited numbers. Long-range strike systems
can also be important contributors to large-volume attack
operations such as those seen in major theater wars. They
can be especially important if the availability of bases near
the conflict is limited.
A systems sustained firepower is determined by the
amount of munitions it can deliver per mission and its
sortie ratethe number of missions it can fly per day.

23

24

ALTERNATIVES FOR LONG-RANGE GROUND-ATTACK SYSTEMS

Figure 3-6.

Bomb Delivery Rates for Alternative Strike Aircraft

(Number of 2,000-pound JDAM-equivalents per aircraft per day)
80

Long-Range
Supersonic Cruise Bomber with
Short Turnaround Time

Arsenal Aircraft
(Alternative 1)

60

Long-Range
Subsonic Bomber
(Alternative 4)

Long-Range
Supersonic Cruise Bomber
(Alternative 5)

40

Medium-Range
Subsonic Bomber
(Alternative 2)

Medium-Range
Supersonic Dash Bomber
(Alternative 3)

20

0
0

500

1,000

1,500

2,000

2,500

Nautical Miles to Target

Source: Congressional Budget Office.

Notes: Data are shown out to the unrefueled radius of each type of aircraft when carrying a full bomb load. All alternatives could achieve
greater ranges with aerial refueling or reduced bomb loads.
JDAM = Joint Direct Attack Munition.

but similar to one anothers when measured in terms of

equivalent warhead weight.10 As a result, the relative performance of those alternatives depends on mission distance because speed and turn times are the differentiating
factors. At short ranges, the arsenal aircraft (Alternative 1)
and the long-range subsonic bomber (Alternative 4) offer
the highest bomb delivery rates because they have shorter
turn times than those of the long-range supersonic cruise
bomber (Alternative 5). The arsenal aircraft is the highest
at the shortest ranges because a substantial portion of the
mission is flown by the missile at Mach 3. At ranges
above about 1,000 nm, the higher payload and aircraft
10. With Alternative 1, the supersonic missile, the total payload
weight of the C-17 is much higher than that of the long-rangebomber alternative, but most of that weight is associated with the
launch racks, missile canisters, and the missile motor, not with the
warheads.

speed allow the long-range subsonic bomber to overtake

the arsenal aircraft. The long-range supersonic cruise
bomber has a lower bomb delivery rate than that of the
other two large-payload aircraft because CBO assumed a
significantly longer turn time (about double) for that
more advanced aircraft. Longer turn times would be
likely because Alternative 5 would have four highly advanced engines versus two engines that would be derivatives of proven designs and because the greater rigors of
sustained supersonic flight would require more careful
servicing of the aircraft during its time on the ground.
The higher speed of the bomber in Alternative 5 does result in shorter missions, however, allowing that alternative to overcome the disadvantage of longer ground times
for longer-range missions. If the turn time for Alternative
5 were the same as for Alternatives 1 and 4, its delivery
rate would be the highest at all mission distances
slightly higher than the arsenal aircraft at 500 nm, with a

CHAPTER THREE

COMPARING THE CAPABILITIES OF THE LONG-RANGE STRIKE ALTERNATIVES CBO EXAMINED

Table 3-2.

Examples of Target Types for

Air-to-Ground Missions
Type

Examples

Preferred Weapons

Hard Point

Deeply buried bunker

Unitary penetrators

Hardened aircraft
shelter
Hard Area

Dispersed armored
vehicles

Penetrating submunitions
with target seekers
Guided shaped charge
weapons

Soft Point

Buildings

Unitary blast/
fragmentation

Soft Area

Supply dumps

Unitary blast/
fragmentation

Dispersed trucks

Bomblet submunitions

Source: Congressional Budget Office.

Note: The terms hard and soft characterize the extent to
which a target is vulnerable to a weapons effects such as
its blast wave or its metal fragmentation pattern. The terms
point and area characterize the spatial extent of a
target.

growing advantage as mission distance increased. (That

case is shown as the dashed line in Figure 3-6.)

Payload Flexibility
Payload flexibility is a strike systems ability to deliver munitions with the most appropriate warhead for a given
mission. The desired warhead is usually dictated by target
characteristics. In particular, a targets hardness (its level
of armor or other protection) and size (how easy it is to
hit) combine to influence the choice of warhead. At the
highest level, targets are grouped in four ways that match
up against four types of warhead (see Table 3-2). The target types shown in the table represent generally preferred
target/weapon combinations. In many cases, a warhead
other than the preferred one could suffice. For example,
unitary blast/fragmentation warheads can be effective
against trucks if the trucks are arrayed close enough together. In other cases, only one type of warhead would be
effective. For example, only ground-penetrating weapons
could destroy hardened, deeply buried targets. (Other
warhead types might be able to damage such facilities by

attacking exposed features such as doors or ventilation

entrances.)
The penetrating-aircraft alternatives (2 through 5) would
offer the greatest payload flexibility among CBOs longrange strike alternatives because they could be designed to
carry all (or nearly all) existing or planned air-delivered
munitions. Different types of warheads could be developed for the supersonic air-launched missile and the
CAVs, but they would not be easily interchangeable and
would require assumptions about what distribution of
warhead types would be needed in the future. After being
launched into orbit, the space-based CAV warheads could
not be changed. The development of different versions of
each missile would also increase the costs of the alternatives. CBOs cost estimates assume that the CAVs and the
supersonic air-launched missile would be armed with
simple unitary warheads. More complicated warheads
such as submunitions that require dispensing mechanisms or warheads with seekers for greater accuracy
would be more costly to develop and integrate into the
missile systems.

Survivability
Survivability, the ability to complete a mission in the face
of enemy air defenses andexcept in the case of expendable systems such as the CAVsreturn safely, is an important attribute for long-range strike systems because
they would most likely be employed on missions lacking
substantial air-defense suppression support. Those types
of missions might include short-notice precision strikes
that must be executed before enemy air defenses can be
suppressed or missions conducted early in regional conflicts when suppression support might be available but
when a potentially dangerous air-defense threat could still
remain. As with responsiveness and firepower, substantially different survivability characteristics exist for CBOs
long-range strike alternativesdifferences that would influence the circumstances under which they could be employed.
The CAV alternatives would have the highest survivability against current air-defense systems and any new systems that might be expected in the near future. After a
CAV is launched, its combination of hypersonic speed,
small size, and maneuverabilitywhich means it does
not travel a predictable ballistic trajectorymakes it a
difficult target that only sophisticated ballistic missile defenses, which do not exist today, could defeat. A CAV

25

26

ALTERNATIVES FOR LONG-RANGE GROUND-ATTACK SYSTEMS

would be a challenging target even in cases requiring it to

expend much of its speed to reach targets at the limits of
its crossrange capability. Directed energy weapons, such
as lasers, might eventually be able to defeat a CAV, although such technologies are in their infancy, and the
CAVs thermal shielding against atmospheric heating
would make it inherently resistant to such defensive
measures.
Attacking a CAV system before it launches a missile
would also be difficult. The surface-based CAVs would
be located in secure areas far from an adversarys countryin the United States for the long-range CAV and
perhaps at sea for the medium-range CAV. Consequently,
attack from the ground would be difficult, and an adversary would need a long-range strike system to attack by
air. Attacking the orbiting CAVs would require an advanced antisatellite system capable of hitting a target in
equatorial low-Earth orbit. Although there is concern in
DoD about the vulnerability of space systems in the future, the existence of antisatellite systems, other than
those explored by the United States and the Soviet Union
during the Cold War, has not been confirmed.11
The survivability of the arsenal aircraft would be limited
by the C-17 launch platform. The missile itself would be
difficult for air defenses to engage because of its high
speed, small size, and its potential to be designed to incorporate some stealth characteristics. However, the systems survivability would be high only if the C-17 could
remain beyond the reach of enemy air defenses. The C17 is not stealthy and could be detected and tracked by
radar. Modern surface-to-air missile (SAM) systems, if located on a countrys border, could force the C-17 to remain well outside the enemys border, seriously limiting
the effective reach of the 500-nm supersonic missile. For
example, the Russian-built SA-20 system is reported to
have a maximum range of slightly over 100 nm.12 Enemy
fighters could also be a serious threat because they could
11. The antisatellite mission against a CAV is made more difficult
because, except for countries on the equator, the CAV satellite
does not pass directly over its target. Consequently, more powerful
booster rockets would be needed to reach a CAV orbit, and a
ground-based antisatellite laser would, in most cases, not have a
direct line of sight to the satellites.
12. Duncan Lennox, ed., Janes Strategic Weapons Systems, Issue 43
(July 2005).

choose to cross their borders and intercept the C-17s even

farther out. That threat could only be countered by U.S.
fighter support if air-superiority fighters were already
present in the region.
The survivability picture is more complicated for the alternatives that use penetrating stealthy bombers. The very
low observable characteristics of the medium-range
bombers and the long-range subsonic bomber (Alternatives 2 through 4) would offer good survivability against
surface air defenses for operations in darkness and probably also in daylight, although advanced optical systems
might be able to detect a high-altitude aircraft in daylight
and cue a SAM intercept. Enemy interceptors at altitudes
similar to those of the bombers could be a serious daytime threat, however. They could more easily detect the
bombers, either visually by their crews or with long-range
cameras, and such detection could occur at ranges short
enough for air-to-air missiles or cannons to be effective.
To help counter that threat, the bombers could be armed
with air-to-air missiles of their own, but their slower
speed and lack of maneuverability relative to fighters
would still put them at a disadvantage. Alternative 3
could have good survivability against that threat because
the bomber could turn away and try to escape at supersonic speed. Although interceptors might be faster, if the
bomber began its escape early enough, its greater endurance could allow it to remain out of range long enough
for the interceptors to exhaust their fuel. The fighters
would not have destroyed the bomber, but they would
have prevented it from completing its mission. As with
the arsenal aircraft, the air-to-air threat would be lower if
friendly fighters were available to support the bombers.
Although less stealthy than the other penetrating bombers, the long-range supersonic cruise bomber (Alternative
5) would have the advantage of even higher speed. That
speed would shrink the effective range of SAM systems by
reducing the amount of time available for an engagement. Although the bomber might be detected at a
greater distance, it could pass beyond the range of the
SAM before it was overtaken by the missile. Similarly, the
ability to sustain a speed of Mach 2.4 would make it difficult for airborne interceptors to catch the bomber. That
could give the aircraft in Alternative 5 better daytime survivability than the slower aircraft alternatives would have.

CHAPTER

4
Cost and Force-Structure Implications
Chapter 4: Cost and Force-Structure Implications

he long-range strike alternatives the Congressional

Budget Office examined would offer diverse capabilities.
No single alternative provides superior effectiveness
across all of the focus areas for long-range strike capability. In addition to simple differences in capability, each alternative would fit into the existing force structure in different ways. For example, fielding common aero vehicles
would provide a new rapid-strike capability but could not
replace the firepower offered by todays bombers when
those aircraft are retired from service. A supersonic
bomber, on the other hand, could replace the aged bombers but would offer significantly less improvement in responsiveness than CAVs would. Because of such distinctions, the alternatives CBO examined should not
necessarily be viewed as mutually exclusive. Depending
on how the Department of Defense eventually defines its
requirements for long-range strike capability, a mix of different systems might be needed.
Further complicating comparison of the alternatives are
the significant differences in the estimated costs to develop and field them (see Table 4-1). Research, development, test, and evaluation (RDT&E) costs could vary by
over an order of magnitude, CBO estimates, with the
missile alternatives at the lower end and the advanced supersonic bomber at the upper end. CBOs estimates of
production costs vary over a similarly broad range, although the variation results in part from assumptions
about how many of each type of system would be procured. In the absence of established DoD requirements,
CBO chose quantities consistent with current forcestructure plans and the procurement of past systems. Details of the methods CBO used to develop its cost estimates (with the exception of assumed procurement quantities) are described in the appendix. This chapter
describes the assumptions underlying the procurement
quantities used in the cost estimates and discusses the effect that cost and force-structure considerations, in conjunction with the capabilities described in Chapter 3,

could have on decisions about which future long-range

strike systems might be pursued.

Aircraft-Based SystemsAlternatives 1
Through 5
The alternatives that include new aircraft designs would
be significantly more costly to develop and procure than
the alternatives that only develop new missiles, CBO estimates.1 Aircraft development costs tend to be higher than
missile development costs because of an aircrafts much
greater complexity and size, as well as having to be designed for many years of operation rather than a single,
one-way mission. Consequently, it is unlikely that DoD
would simultaneously pursue more than one alternative
for long-range aircraft. However, combinations of
aircraft-based and missile-based alternatives could provide capabilities that DoD might judge to be costeffective.

Alternative 1: Arsenal Aircraft with

Supersonic Missiles
Along with the surface-based CAVs, the arsenal-aircraft
alternative, at just over $4 billion, would be among the
least expensive of the alternatives CBO examined if DoD
chose to not purchase additional C-17s as dedicated
launch platforms. DoD has nearly completed purchasing
180 C-17s for strategic and tactical airlift missions, and
some of those aircraft could be used to deliver missiles.
The 2,000 missiles slated for purchase under Alternative
1 would be a quantity on a par with the number in other
tactical missile systems such as the Navys Joint Standoff
1. The difference between the aircraft-based and missile-based alternatives would be even greater if operation and support costs were
included. CBO focused on acquisition costs for those comparisons because they are sufficient for assessing top-level distinctions
among the various alternatives.

28

ALTERNATIVES FOR LONG-RANGE GROUND-ATTACK SYSTEMS

Table 4-1.

Estimated Costs of Long-Range Strike Alternatives Examined by CBO

Quantity

Arsenal Aircraft

Medium-Range Subsonic
Bomber

2.8/6.1

275

19

52

72

188

Medium-Range Supersonic
Dash Bomber

275

23

61

85

220

Long-Range Subsonic Bomber

150

31

61

93

409

Long-Range Supersonic
Cruise Bomber

150

69

137

207

912

48

2.4

1.2

3.7

26

24

2.5

0.9

4.0

36

4.0

7.7

11.7

55

Medium-Range SurfaceBased CAV

e

Long-Range Surface-Based CAV

Space-Based CAV

128

4.3/7.6

Average Unit Procurement Cost

(Millions of 2006 dollars)

1.5

2,000

Costsa
(Billions of 2006 dollars)
b
RDT&E
Procurement
Total

1.4

Source: Congressional Budget Office.

Note: RDT&E = research, development, test, and evaluation; CAV = common aero vehicle.
a. The costs for Alternatives 2 to 5 exclude munitions.
b. Includes additional military construction costs of about $1 billion for the aircraft alternatives (2 to 5) and $600 million for Alternative 7.
c. The quantity shown is the number of supersonic missiles purchased. The lower of the two costs assumes that those missiles would be carried by C-17 aircraft in the current fleet. The higher of the two costs assumes that 15 additional C-17s would be purchased to support the
strike mission.
d. Average unit procurement cost is for supersonic missiles only.
e. Alternative 7 assumes that 24 Peacekeeper missiles would be converted to carry two CAVs per missile. If more missiles were desired, as
many as 60 Peacekeepers might be available for conversion. The cost of additional missiles would be much higher than the cost shown
here because new boosters would be needed.
f.

Enough satellites would be purchased to maintain the constellation for 30 years. Only 40 space-based CAVs would be available for use at
any one time.

Weapon (JSOW) and the Air Forces Joint Air-to-Surface

Standoff Missile (JASSM).
If DoD decided that aircraft dedicated to the strike mission were needed, procuring 15 additional C-17 aircraft
would cost about $3.3 billion, CBO estimates. That
number of aircraft could maintain continuous coverage
of about 45,000 square nautical miles with a 10-minute
response time. The estimated cost for those additional
aircraft would be considerably higher, however, if the decision to purchase them was not made before the C-17
production line closed. Because the 2006 Quadrennial
Defense Review did not identify plans to purchase more

than 180 C-17s, currently planned production will end

by 2010. Purchasing additional aircraft would have the
advantage of being able to augment the airlift fleet when
they are not needed as strike aircraft. Alternatively, C-17s
from the air-mobility fleet could be pressed into service as
strikers if the circumstances required, although that
would decrease the number of C-17s available to move
cargo.
In many respects, the arsenal aircraft would offer performance similar to that of the long-range subsonic penetrating bomber (Alternative 4), albeit for shorter distances. For example, the two alternatives would have

CHAPTER FOUR

similar firepower up to about 2,000 nautical miles and

would require a similar number of aircraft to maintain responsive strike orbits up to 1,500 nm from a base or
aerial refueling point. A major disadvantage of Alternative 1, however, would be the aircrafts lack of stealth,
which would make it unable to enter hostile airspace until after defenses had been suppressed. For strike missions
without defense suppression support, Alternative 1 would
have a penetration depth equal to only its missiles range,
about 500 nma distance only on the order of that for
short-range fighters. Penetration of hostile airspace would
be further limited if enemy fighters were able to fly out
beyond their borders to intercept the C-17s. Increasing
the range of the missile could alleviate that shortcoming
but at higher cost and a smaller carriage capability on the
aircraft.
To overcome the disadvantages of delivery from a nonstealthy C-17, Alternative 1s missile could be developed
in conjunction with a stealthy bomber. The size and
weight of the notional missile are similar to those of
GBU-28 bunker-buster bombs that can be delivered by
B-2 bombers. Any of the aircraft alternatives could be designed to accommodate the supersonic missile, although
it would be most difficult to fit such bulky weapons into
the smaller medium-range supersonic dash bomber in Alternative 3. If a long-range bomber (or one of todays B2s) was to carry it, the supersonic missile would enjoy the
advantages of stealth, global delivery range, and rapid responsiveness (from an orbiting aircraft) against fleeting
targets. Integrating the missile onto more than one aircraft would require additional development costs because
the missile would probably be launched in different ways:
from a container in an arsenal aircraft, from a rotary
launcher in a penetrating bombers internal weapons bay,
or from an external weapons station on an F-15E. Responsiveness at global ranges would still lag far behind
that of the CAV alternatives, however.

Alternatives 2 and 3: Medium-Range Bombers

The medium-range stealth bombers would improve strike
capabilities over regional distances but would not be as
well suited for missions with a global range, although
such missions might be possible with extensive tanker
support. As a result, medium-range bombers would probably not be a suitable replacement for some of the capabilities offered by todays long-range heavy bombers.
CBO assumed instead that DoD would purchase about
275 of the aircraft as replacements for F-117 stealth fighters and F-15E Strike Eagles.2 For that number of aircraft,

COST AND FORCE-STRUCTURE IMPLICATIONS

a medium-range subsonic bomber force (Alternative 2)

would cost about $72 billion, and a medium-range supersonic dash bomber force (Alternative 3) would cost about
$85 billion, CBO estimates. The result would be to provide substantial firepower capability at ranges beyond
those of todays strike fighters but no improvement at
heavy-bomber ranges (see Figure 4-1). That result is illustrated by the additional delivery capability beyond 700
nmabout a 70 percent increase for Alternative 2 and
about a 35 percent increase for Alternative 3 for ranges
up to 1,500 nm. For ranges greater than 1,500 nm, the
improvement offered by the medium-range bombers
drops quickly because the aircraft must carry reduced
bomb loads to fly those distances. In addition to having
greater range, that entire portion of the force would be
stealthy. Today, there are only 55 stealthy F-117s that are
operational compared with 217 conventional F-15Es.
Depending on the alternative selected, however, the new
force could lack the air-to-air capability of the dual-role
F-15E.
The purchase of 275 aircraft under Alternatives 2 and 3
would increase delivery capability at short ranges relative
to that of todays force but by a smaller proportion (as
shown in Figure 4-1). As an alternative, those notional
forces could be sized to match todays short-range capability, although that would result in a lesser improvement
between 900 nm and 1,500 nmabout 22 percent. Such
a force would need either 85 medium-range subsonic
bombers or 170 medium-range supersonic dash bombers.
For those purchase quantities, the costs of Alternatives 2
and 3 would drop to about $42 billion and $66 billion,
respectively, CBO estimates.
Although medium-range bombers alone could not replace all of the capabilities of heavy bombers, a mediumrange bomber force in concert with one of the CAV alternatives would provide a capability that some might argue
could reduce or eliminate the need for heavy bombers.
The CAVs would provide prompt strike at global ranges,
and the medium-range bombers would provide substantial firepower for regional conflicts. Replacing heavy
2. This approach assumes that the fraction of the inventory that is
maintained in a combat-ready status is similar for both the current
aircraft and the aircraft in CBOs alternatives. Although surface
CAV missiles would require periodic maintenance, their combatready status could be nearly 100 percent. The space-based CAVs
would have a 100 percent combat-capable rate unless there were
irreparable failures while in orbit.

29

30

ALTERNATIVES FOR LONG-RANGE GROUND-ATTACK SYSTEMS

Figure 4-1.

Potential Firepower of Alternative Air Force Strike Forces, Without

Aerial Refueling
(Number of 2,000-pound JDAM-equivalents per day)
25,000

20,000

Force with Medium-Range

Subsonic Bombers
(Alternative 2)

15,000

Force with Medium-Range

Supersonic Dash Bombers
(Alternative 3)

10,000

Current Force
5,000

0
0

500

1,000

1,500

2,000

2,500

3,000

Nautical Miles to Target

Source: Congressional Budget Office.

Notes: Future forces are assumed to retain current bombers and short-range fighters but to replace todays F-15Es and F-117As with 275
medium-range subsonic or supersonic bombers.
This figure is based on the Air Forces total inventory of aircraft. Numbers for combat-ready aircraft would be lower.
JDAM = Joint Direct Attack Munition.

bombers with CAVs for global missions would have the

disadvantage of reducing firepower over global ranges. Although missions of long durations over global ranges
limit the sortie rates of todays bombers to well below one
mission per aircraft per day, large long-range bomber
forces could still deliver a considerable number of bombs
per day.3 For example, a B-2 with a payload of 16 Joint
Direct Attack Munitions could generate an average delivery rate of about three to four JDAM per aircraft per day.
When multiplied by many bombers in a large force, that
delivery rate would exceed one that would be practical
with long-range CAVs. Historically, however, globalrange conventional strikes such as the B-2 missions from
Missouri to Bosnia have involved only small numbers of
3. Bombers are often deployed to forward locations during sustained
operations to overcome the limitation of that sortie rate.

bombers. That may be because target sets have been small

or because only stealth aircraftand hence only a small
fraction of the bomber forcewere suitable for those
missions.
Alternative 3 is similar to Alternative 2 in terms of range
and cruise speed. The additional ability to dash at Mach
1.5 comes with the penalty of having half the payload of
the medium-range subsonic bomber and an estimated
cost about 20 percent higher$85 billion versus $72 billionfor the same production quantity of 275 aircraft.
As with Alternative 2, this alternative would provide significantly improved strike capability relative to the F-117
and F-15E fighters that would be replaced.
Alternative 3 would have a sizable advantageabout a
factor of two for orbits less than 1,000 nm from the last
refueling base or tankerover Alternative 2 with respect

CHAPTER FOUR

COST AND FORCE-STRUCTURE IMPLICATIONS

Figure 4-2.

Total Potential Firepower of Alternative Bomber Forces, by Mission Distance

(Number of 2,000-pound JDAM-equivalents per day)
14,000

Medium-Range
Subsonic
Bomber
(Alternative 2)

Current Bomber
Force

12,000

10,000

8,000

Long-Range
Supersonic
Cruise Bomber
(Alternative 5)

6,000

4,000

Medium-Range
Supersonic
Dash Bomber
(Alternative 3)

2,000

Long-Range
Subsonic
Bomber
(Alternative 4)

0
0

500

1,000

1,500

2,000

2,500

Nautical Miles to Target

Source: Congressional Budget Office.

Notes: Data are shown out to the unrefueled radius of each type of aircraft when carrying a full bomb load. All alternatives could achieve
greater ranges with aerial refueling or reduced bomb loads.
This figure is based on the total number of aircraft purchased. Numbers for combat-ready aircraft would be lower.
The arsenal aircraft (Alternative 1) is not shown because that alternative does not assume a specific quantity of aircraft.
Unlike the display in Figure 4-1, the current bomber force firepower is shown here without daily sortie constraints.

to the number of aircraft needed to maintain very short

response times against time-critical targets because the
bombers supersonic dash speed would allow it to attack
targets at greater distances within a given response time.
However, the lower payload for Alternative 3 could negate that advantage somewhat if the aircraft had to attack
targets at a rate that would exhaust its munitions well before its fuel. The lower payload would result in significantly lower firepower than that of the subsonic bomber,
which could deliver munitions at about twice the rate of
the supersonic dash bomber during sustained operations.
Proponents of an aircraft similar to the supersonic
bomber in Alternative 3 argue that its superior ability to
elude enemy fighter aircraft could be an important advantage, especially for daytime strike missions. As was

discussed in Chapter 3, however, those circumstances

might be rather narrow, and, if the bomber was confronted by enemy fighters on the way to the target, the
act of eluding the fighters might prevent it from completing its mission.

Alternatives 4 and 5: Long-Range Bombers

Long-range stealth bombers would improve strike capabilities over global distances. The subsonic bomber (Alternative 4) would require tanker support comparable
with that of todays bombers. The supersonic cruise
bomber (Alternative 5) would require greater tanker support because of its high fuel consumption. CBO assumed
that DoD would purchase about 150 of the aircraft as replacements for the B-1, B-2, and B-52. Despite the similarity in numbers to the current bomber force, that re-

31

32

ALTERNATIVES FOR LONG-RANGE GROUND-ATTACK SYSTEMS

placement would result in some changes in firepower

capability (see Figure 4-2). For the long-range subsonic
bomber in Alternative 4, a smaller inventory (150 under
CBOs notional alternative versus 182 bombers currently
in place) plus a lower payload relative to that of the B-1
and B-52 would result in less total delivery capability
than that of the current force. In practice, the difference
might be smaller than shown in the figure if the Air Force
could maintain a larger fraction of the new bombers in
combat-ready status. The actual number of new bombers
DoD might decide it needs would depend on that consideration as well as assessments of future requirements.
The lower inventory and lower payload of the long-range
supersonic cruise bomber in Alternative 5 relative to
those of the B-1 and B-52 would also result in lower firepower at shorter ranges, but the faster bombers shorter
mission time would reduce that difference for longer mission distances. As with the medium-range-bomber alternatives, Alternatives 4 and 5 would result in an all-stealth
bomber force.4 Today, there are only 21 stealthy B-2s
compared with 161 conventional B-1s and B-52s.
CBOs estimate of Alternative 4s cost$93 billionis
somewhat higher than the cost for the medium-rangebomber alternatives$72 billion for Alternative 2 and
$85 billion for Alternative 3but only about half the
number of aircraft would be purchased. Despite its lower
numbers, Alternative 4 would still outperform the alternatives for medium-range bombers in terms of firepower.
However, only 25 additional medium-range subsonic
bombersfor a total of 300would be needed for Alternative 2 to match the firepower of Alternative 4 at ranges
up to 1,500 nm. CBO estimates that adding those aircraft would increase the cost of Alternative 2 to about
$74 billion, a total lower than the total cost for 150 longrange subsonic bombers. In contrast, because of the lower
payload of the medium-range supersonic dash bombers
in Alternative 3, about 600 of those would be needed to
match the firepower of the long-range subsonic bomber
in Alternative 4. That large force would cost about $131
billion, CBO estimates.
The long-range supersonic cruise bomber in Alternative 5
is the most advanced concept examined in this study.
4. As was noted in Chapter 3, the supersonic cruise bomber would
be less stealthy than the other alternatives because it would rely
only on its shape to achieve its low-observable characteristics. Its
very high speed, however, would augment its survivability.

Whereas the other aircraft could use many technologies

already developed for existing systems, Alternative 5
would represent an all-new design. Consequently, it
would come with a much higher price tagabout $207
billion to develop and field 150 bombers, CBO estimates. That is more than double the estimated costs for
Alternatives 3 and 4 and nearly triple the cost of Alternative 2. For its higher price, the long-range supersonic
cruise bomber would offer capabilities between those of
subsonic bombers and CAVs. At short mission distances,
the supersonic bomber in Alternative 5 would offer modest improvements in responsiveness relative to subsonic
aircraft but with lower sustained munitions delivery rates
(firepower). At very long ranges, Alternative 5 would offer the greatest firepower but would be far less responsive
than would the CAV alternatives. Depending on the level
of stealth that could be incorporated into its high-speed
design, the bomber in Alternative 5 might have better
survivability than that of the other aircraft alternatives because its high speed would give both air and ground
threats little time to engage it. However, CAVs would
have a better chance of reaching their targets.
The performance and cost of the bomber in Alternative 5
suggest it would be a preferred system only under a narrow set of requirementsspecifically, if large rates of munitions delivery are desired over global distances where
the bombers speed and range combine to give it a substantial firepower advantage over the other alternatives.
Such situations, however, would probably occur rarely or
not at all. The rapid delivery of large numbers of munitions is typically associated with conflicts that include
substantial forces on the ground, which implicitly suggests that some regional access for strike aircraft would be
available. Aircraft carriers could provide strike capability
far from the United States against those targets that can
be reached from the sea until regional bases could be established. Time-sensitive targets, those for which there is
not time to move a carrier into position or to fly to from
the United States at subsonic speed, typically do not require a great volume of firepower but rather a few wellplaced bombs. In those cases, the CAV alternatives could
offer adequate firepower and much better responsiveness
than the bomber in Alternative 5. Consequently, it could
be argued that an approach that develops a CAV option
and a new subsonic bomber would provide more broadly
useful capabilities at less cost than would a long-range
supersonic cruise bomber.

CHAPTER FOUR

Common Aero Vehicle Systems

Alternatives 6 Through 8
The common aero vehicle alternatives would offer much
shorter response times than those of the aircraft-based alternatives for missions over long distances because of
their hypersonic speeds. For missions where the launch
platform can loiter close to target areas, the CAVs do not
perform as well because their launchers must be located at
a considerable distance from enemy territoryat a secure
base or on a ship in the region for the medium-range
CAV (Alternative 6), in the United States for the longrange CAV (Alternative 7), and several hundred miles
away in low-Earth orbit for the space-based CAV (Alternative 8). The CAV alternatives would be much less expensive than developing new penetrating bombers, CBO
estimates, but somewhat more expensive than the airlaunched missile in the arsenal-aircraft alternative (for the
quantities considered). As was mentioned earlier, the
quantity purchased under each CAV alternative was set to
give the alternatives as similar a capability as possible. A
constellation of five satellites in equatorial low-Earth orbit would provide the space-based CAV with a response
time at least as good as that of the long-range surface
CAV. In CBOs alternative, arming each launcher with
eight CAVs provided a launch mass well suited to available space-launch rockets. The resulting quantity totaled
40 operational CAVs in orbit. However, satellites in lowEarth orbit typically have about a 10-year life in the harsh
environment of space. Consequently, CBO assumed that
two additional constellations worth of satellites would be
needed to maintain that capability over about 30 years, a
service life similar to what is often expected for new military aircraft. The surface-based alternatives had the same
number of test and operational missiles as a single spacebased constellation did to provide similar total firepower.5 Additional CAVs could be purchased under any
of the alternatives, although the unit cost for the missiles
under Alternative 7 would increase dramatically when the
inventory of retired Peacekeeper missiles was exhausted
and new intercontinental ballistic missile-sized rockets
had to be purchased as well. A new rocket the size of a
Peacekeeper would cost about $100 million and require
new development, CBO estimates, more than tripling the
5. Alternative 7 included half as many missiles, but each missile
could deliver two CAVs. Although that arrangement would be less
flexible than a one CAV/one missile configuration, a target of high
enough value to warrant a CAV shot might warrant two warheads
to improve the odds that the target would be destroyed.

COST AND FORCE-STRUCTURE IMPLICATIONS

cost of the round. Alternatively, it might be possible to

purchase submarine-launched Trident D-5 boosters that
are currently in production at about $40 million each and
convert them for surface launch.6 The CAV missiles are
much more expensive on a per-round basis than are Alternative 1s air-launched missiles, but they would provide much higher speed and much harder hitting power.
Because each medium-range CAV would have a lower
unit cost, total costs associated with a medium-range
CAV force would increase more slowly as purchase quantity increased.
High unit costs would probably make CAVs unsuitable
for replacing aircraft in roles that require attacking large
numbers of targets. For example, with aerial refueling
support, 100 long-range supersonic cruise bombers (Alternative 5)the fraction of a 150-aircraft fleet that
might be committed to a conflictcould deliver a payload equivalent to about 2,100 JDAMs per day to a range
of 7,000 nm.7 The procurement cost for a similar number of space-based CAVs would total over $200 billion
(see Figure 4-3). The costs for that number of weapons
would be about $30 billion for the medium-range surface
CAV and $80 billion for the long-range surface-launched
CAV.8 The aircraft alternatives costs change little with
increasing numbers of weapons delivered because the
costs of those weapons are small compared with the costs
of the aircraft. (Alternative 1 has the lowest cost, but as
noted earlier, it is considerably less capable than the other
alternatives.) Aside from the cost involved in achieving
high delivery rates with CAVs, other practical difficulties
exist. For example, forward deployment of several hundred medium-range CAVs would be a significant transportation burden, and substantial additional launch facilities would be needed to provide high rates of fire with
6. The 2006 Quadrennial Defense Review includes plans to rapidly
develop a system that would deliver conventional warheads on
Trident missiles. Those missiles would probably be launched from
Trident submarines instead of converted for launch from the
ground or from surface ships.
7. CBO used 100 aircraft, not the entire 150-aircraft force, to reflect
aircraft that might not be mission-capable at any given time. That
distinction was less important for aircraft-to-aircraft comparisons.
CBO used entire inventories under the assumption that comparable fractions of each type would be operational.
8. The unit cost of the long-range surface-launched CAVs that
require new production of boosters is more than double that of
the medium-range version but provides two CAVs per shot.

33

34

ALTERNATIVES FOR LONG-RANGE GROUND-ATTACK SYSTEMS

Figure 4-3.

Cumulative Development and Acquisition Costs for Weapons and Their Delivery
Systems, by Number of Weapons Used
(Billions of 2006 dollars)
400

Approximate number of 2,000-pound

JDAMs deliverable per day by

350

100 Alternative 5 bombers based

7,000 nm from their targets

300

Alternative 8
250

Alternative 5
200

150

Alternative 7

Alternative 3
Alternative 2
Alternative 4

100

50

Alternative 6
Alternative 1

0
0

500

1,000

1,500

2,000

2,500

3,000

3,500

4,000

Number of Weapons Used

Source: Congressional Budget Office.

Notes: Assumes aircraft are delivering 2,000-pound Joint Direct Attack Munitions (JDAMs).
nm = nautical miles.
Alternative 1 = Arsenal aircraft
Alternative 2 = Medium-range subsonic bomber
Alternative 3 = Medium-range supersonic dash bomber
Alternative 4 = Long-range subsonic bomber
Alternative 5 = Long-range supersonic cruise bomber
Alternative 6 = Medium-range surface-based common aero vehicle
Alternative 7 = Long-range surface-based common aero vehicle
Alternative 8 = Space-based common aero vehicle

long-range surface CAVs. Similarly, large numbers of

space-based CAVs would require a proportionally larger
number of Evolved Expendable Launch Vehicles to put
them in orbit and hence might require an expansion of
the nations space-launch infrastructure. Consequently,
CAVs most likely would be cost-effective only against
high-value targets that typically exist in small numbers.
However, as was discussed earlier, a combination of longrange CAVs and medium-range bombers could conceivably replace long-range bombers if DoD decided that sufficient CAVs could be purchased to handle the number of
targets that would require attack from global distances.

It might also prove cost-effective to purchase a combination of the missiles in Alternatives 6 and 7. Much of the
research and development cost, particularly that of the reentry vehicle, could be the same for the two systems. The
medium-range CAVs, which could be purchased in larger
quantities for a given level of investment, could be stationed at forward locations such as Guam or on ships so
as to be in a position to provide short response times in
areas of heightened tension, and the longer-range missiles
could be saved for circumstances when the mediumrange systems were not located close enough to targets
that might arise.

CHAPTER FOUR

COST AND FORCE-STRUCTURE IMPLICATIONS

The force-structure comparisons described above are in

the context of fully operational forces that might eventually be fielded. Those comparisons do not consider the
different lengths of time for development and procurement that might be needed to realize such forces, a factor
that could be important as threats evolve and as existing
systems age and are retired. The time required to develop
a weapon system is influenced by many factors, some of
which are inherent to the system (for example, the maturity of component technologies) and others that are external to the system (for example, funding levels and institutional support). This section briefly discusses how factors
that are inherent to developing new systems could influence how soon the alternatives examined by CBO might
be fielded.

Although the United States has test-flown maneuvering

reentry vehicles, it must still develop the technology to
field an operational CAV. In particular, materials would
have to be developed that could withstand for up to 30
minutes the heat generated by the CAVs hypersonic
flight through the atmosphere as it maneuvered to strike
its intended target. In addition, although components of
existing boosters might be used for the medium-range
surface-based CAV, the integration of those components
could prove problematic, as has been the case in an
ongoing program to develop and deploy ballistic missile
defenses. And the protective satellite that would house
the space-based CAVs would have to be designed and
tested. Therefore, the time and effort needed to develop
either surface-based or space-based CAVs (Alternatives 6,
7, and 8) would most likely be greater than that to develop a supersonic cruise missile delivered by an arsenal
aircraft.

Determining a schedule on which the development of a

long-range strike alternative might proceed requires detailed information regarding the work and testing to be
accomplished. CBO lacks that information for the alternatives considered in this study, and even if such information were available, DoDs experience indicates that predicting schedules for any of its major programs before
they are well under way is problematic. Although CBO
cannot provide definitive schedules for developing longrange strike alternatives, it can provide observations
indicating the relative difficulty of implementing those
alternatives.

The medium-range bombers (subsonic and with supersonic dash, Alternatives 2 and 3) and long-range subsonic
bomber (Alternative 4) would all be based largely on existing technology, much of which has already been implemented in operational aircraft. Nonetheless, their development would require new aircraft designs (not merely
modifications to existing designs), as well as the integration of a large number of subsystems, including complex
software supporting the aircrafts flight and operation of
its sensor systems. Therefore, those alternatives would
probably take longer to develop than would the supersonic cruise missile or the common aero vehicle.

The United States does not currently have an operational

supersonic cruise missile, but other nations, such as Russia and India, do. The supersonic missiles that currently
exist, however, lack the range and payload postulated for
the missile used in the arsenal-aircraft alternative considered by CBO. Nonetheless, the United States has investigated the technologies needed for such a missile for a
number of years. For example, in April 2002, the Defense
Threat Reduction Agency solicited information on technology available for executing an advanced concept technology demonstration program for a supersonic cruise
missile. Because supersonic cruise missiles exist and the
technology needed to build them is relatively well understood, the arsenal aircraft (Alternative 1) would probably
require less time to develop than the other alternatives for
long-range strike considered by CBO.

Of all the alternatives considered by CBO, the long-range

supersonic cruise bomber (Alternative 5) would most
likely require the greatest effort and time to develop. Although the United States has operated and test-flown supersonic bombers and reconnaissance aircraft in the past,
many years have passed since a U.S. manufacturer designed such an aircraft in detail. Moreover, incorporating
some amount of stealth (or radar cross-section reduction)
in a supersonic cruise bomber would probably require the
use of advanced materials not incorporated in past designs. As with the subsonic bombers, developing a supersonic bomber would require integrating a large number
of subsystems and complex software. Achieving the
engine fuel efficiencies postulated by CBO for this alternative would also require additional development of
technology.

The Relative Difficulty of Developing

the Long-Range Strike Alternatives
Considered by CBO

35

Appendix:
The Methodology Behind the Cost Analysis

he Congressional Budget Office (CBO) developed

several alternatives for maintaining and improving the
long-range strike capabilities of the U.S. military into the
next decade. Those alternatives include upgrading tactical
missile capabilities, improving the fleet of bomber aircraft, and building maneuverable reentry vehicles that are
launched from ballistic missiles on the ground or from
satellites in space.
For each alternative, CBO estimated the costs to develop
and procure the weapon system, as well as the costs to improve or build the facilities and infrastructure needed to
support the system.
Expenses in the development phase include the cost to
design and build the components of the long-range strike
system, the cost to test those components to ensure that
they meet performance requirements, and the cost to integrate the system into the militarys infrastructure and
support systems. Expenses in the procurement phase include the cost of special tools and equipment to manufacture system components; the cost of hardware, material,
and fabrication such as finished components and raw materials; and the cost to assemble the final product. Hardware, material, and fabrication are also important for determining estimates because they are inputs to statistical
relationships used to estimate costs for nonrecurring design and other support costs.
All of the cost estimates in this appendix are presented in
2006 dollars. Significant uncertainty exists about the capabilities, technologies, and costs associated with developing, purchasing, and operating weapon systems envisioned in the alternatives. Because the programs are either
conceptual in nature or in the early stages of development, they entail a greater risk of cost and schedule overruns than do programs that are better defined and based

on proven technologies. CBOs cost estimates represent

one possible outcome, calculated under specific assumptions regarding technologies and capabilities. Although
CBO attempted to account for that risk, those estimates
will change, perhaps significantly, as the designs of the
systems in the alternatives become more fully defined.

The Cost of Supersonic Cruise

Missiles for the Arsenal Aircraft
In Alternative 1, the Air Force would develop and procure a new supersonic cruise missile that would be
launched from an arsenal aircraft. Because the arsenal aircraft would be a modified version of an existing airframe
(the C-17), for which the costs are generally known, this
discussion focuses primarily on the methodology used to
develop the cost of developing and procuring the missile.
The need to purchase additional C-17 aircraft would depend in large part on future airlift requirements. The projected fleet of 180 C-17 aircraft might be sufficient to
perform the long-range strike mission without critically
hampering concurrent airlift efforts; thus, there would be
little additional costs beyond that of the missiles. However, if the existing fleet of aircraft was not sufficient to
meet that new requirement, additional C-17s would need
to be purchased. That could substantially increase the total cost of the arsenal aircraft alternative. On the basis of
recent acquisition cost data from the ongoing procurement program, CBO estimated, for example, that acquiring 15 additional C-17s for that purpose would cost
about $3.3 billion. That estimate assumes those aircraft
would be purchased at the end of current planned production, before the assembly line was shut down. If the
decision to purchase additional aircraft was delayed beyond 2008, that assembly line would have to be reestab-

38

ALTERNATIVES FOR LONG-RANGE GROUND-ATTACK SYSTEMS

Table A-1.

Characteristics of Supersonic Missiles

for the Arsenal Aircraft (Alternative 1)
Variable

Launch Weight (Pounds)

Maximum Speed
Payload
Range (Nautical miles)
Engine Design
Number of Engines

4,000
Mach 3
1,000-pound warhead
500
Ramjet

is also affected by the type of warhead chosen for the system. CBO assumed that the missile would employ a lessexpensive unitary warhead rather than precision submunitions that would require more advanced guidance and
control systems to permit complex maneuvers by the missile in the moments before it reached its target. CBO estimated that the cost to develop and procure 2,000 such
supersonic cruise missiles would total $4.3 billion, including the cost to integrate the weapon with a delivery
aircraft such as the C-17 cargo plane. On average, it
would cost $2.2 million to acquire those missiles (see
Table A-2 for a summary of acquisition costs for this alternative).

Source: Congressional Budget Office.

Note: Mach is the ratio of the speed of an object to the speed of
sound in air.

lished. In that case, it would cost more to acquire additional aircraft.

Several variables for the performance parameters and design of a supersonic cruise missile drive the cost of such a
weapon, although the exact values of those variables are
not known at this time. For this study, CBO assumed
that the missile would be capable of traveling 500 nautical miles and would carry a 1,000-pound warhead (see
Table A-1 for a listing of the missiles characteristics).
The most important variable is the required speed of the
weapon, which would affect the designand thus the
costof the propulsion system. The missile considered
in this study would use a ramjet propulsion system with a
top speed of Mach 3.1, 2 The cost of a supersonic missile
1. Mach is a ratio of the speed of an object to the speed of sound in
the air or fluid in which it is traveling. At standard sea-level conditions, Mach 1 is 1,225 kilometers/hour (766 miles per hour) in
the atmosphere.
2. In a typical turbojet engine, pressure is maintained in the combustion chamber through the use of a compressor fan. In contrast, a
ramjet propulsion system uses the forward speed of the vehicle to
ram external air into the combustion chamber to create the high
pressure required for combustion. However, air entering a ramjet
engine must be slowed down, as combustion can only take place
at subsonic speeds, which limits a ramjets top speed to about
Mach 5. New supersonic combustion ramjet (scramjet) technology is attempting to overcome that limitation by allowing combustion to occur at supersonic speeds.

CBO used several sources to estimate the development

and procurement costs of a new supersonic cruise missile,
such as methods and data contained in a 1998 cost estimate from the Naval Center for Cost Analysis for a ramjet missile, as well as information on hardware costs from
existing ramjet/scramjet efforts.3 CBO also looked at data
from prior tactical missile programs and other cost studies (see Table A-3 for a summary of the methods used to
estimate development and procurement costs for this alternative).
CBOs estimates for development and procurement are
broken down into costs for:
B

Hardware, material, and fabrication;

Nonrecurring design, testing, and tooling; and

Other support such as program management, initial

maintenance plans, and contractor fees.

Hardware, Material, and Fabrication

Hardware, material, and fabrication includes the cost of
purchasing finished components and raw materials from
subcontractors, as well as the cost of recurring labor associated with assembling the final product. Those costs are
historically the most significant part of the cost estimate
for missile production and also constitute a small portion
of costs in the development phase. Hardware, material,
3. Jeff Cherwonik, William Stranges, and Jeff Wolfe, Generic Supersonic Cruise Missile Life Cycle Cost Estimate, prepared by the Naval
Center for Cost Analysis in support of the Hard and Deeply Buried Target Defeat Capability Weapon Analysis of Alternatives
(Arlington, Va.: Naval Center for Cost Analysis, October 1998).

APPENDIX: THE METHODOLOGY BEHIND THE COST ANALYSIS

Table A-2.

Acquisition Costs of Supersonic

Missiles for the Arsenal Aircraft
(Alternative 1)
(Billions of dollars)
Cost

Research and Development

Hardware, material, and fabricationa
Nonrecurring design, testing, and toolingb
Other support c
Subtotal, research and development
Procurement (2,000 missiles)
Hardware, material, and fabrication
Nonrecurring design, testing, and tooling
Other support
Subtotal, procurement
Total Estimated Acquisition Costs

0.1
0.8
0.6
___
1.5
1.6
0.2
1.0
___
2.8
4.3

Source: Congressional Budget Office.

a. Includes the cost of purchasing finished components and raw
materials from subcontractors and the cost of recurring labor
associated with assembling the final product.
b. Includes one-time costs that do not vary with the total number
of missiles produced. Those costs include initial design and
engineering efforts, software development, systems testing and
evaluation, flight-testing, and platform integration.
c. Includes costs for contractor and government program management, data collection, contractor fees, and initial maintenance agreements.

and fabrication are also important for cost-estimating

purposes because they are inputs to statistical relationships CBO used to estimate costs for nonrecurring design
and other support costs.
Tactical missiles typically contain four main hardware
components: the engine, the airframe, the guidance and
control system, and the warhead. In addition, CBO assumed that each missile would require a booster rocket to
accelerate the weapon to a speed sufficient to operate the
ramjet.4 CBO estimated that hardware, material, and
fabrication for the supersonic cruise missile would cost
$140 million for the development phase and about $1.6
billion for the procurement phase.
CBO approximated the cost for the missile engine on the
basis of the material and fabrication costs of propulsion

hardware in the X-43A hypersonic research aircraft, the

GQM-163 missile (a supersonic sea-skimming target
used to test shipboard defensive systems), and the Scramjet Engine Demonstration (SED) program.5 The airframe cost estimate is based on the cost of other missile
airframes, such as the Joint Standoff Weapon (JSOW),
with adjustments for weight and material to account for
the higher temperatures that a supersonic cruise missile
would endure during flight.6 The cost estimates for the
guidance and control and warhead subsystems are based
on analogies to other missile programs such as the Tomahawk cruise missile and the JSOW.
CBO assumed that the missile payload would consist of a
unitary warhead, although it is possible that the Department of Defense (DoD) might add precision submunitions to a supersonic strike missile. However, such proposals would create additional technical challenges and
costs that are not included in CBOs estimate. CBO used
the cost of the Mk 72 booster from the Navys Standard
missile as an analogy to estimate the potential cost of the
booster rocket because the Mk 72 propels a load that
would be similar in weight to the proposed supersonic
cruise missile. (CBO assumed that the missile and the
booster considered in this alternative would weigh about
4,000 pounds.)
4. Ramjet engines use external air speed, generated by the velocity of
the vehicle itself, to create the pressure in the engine necessary for
combustion, but the air speed must first be provided by some
other means before the ramjet engine can function. While there
are several ways to generate that speed in a tactical missile system,
CBO assumed this missile would use a solid rocket booster that
would be jettisoned once the missile reached the necessary speed.
5. The costs can only be approximated because the programs mentioned above are not direct analogies to a missile propulsion system of the type that would be used in a new supersonic cruise
missile. The scramjets used in the National Aeronautics and Space
Administrations X-43A and the Air Force and the Defense
Advanced Research Projects Agencys SED programs are only
meant to be demonstration articles for use in testing and proving
concepts. The GQM-163 is a missile used by the U.S. Navy for
training and testing purposes. The propulsion system is a solidfuel ducted rocket ramjet and would not be powerful enough to
propel the strike weapon discussed in this study.
6. For information on the level of effort and cost involved with using
different types of airframe materials, see Obaid Younossi, Michael
Kennedy, and John C. Graser, Military Airframe Costs: The Effects
of Advanced Materials and Manufacturing Processes, MR-1370-AF,
(Arlington, Va.: RAND Project Air Force, 2001).

39

40

ALTERNATIVES FOR LONG-RANGE GROUND-ATTACK SYSTEMS

Table A-3.

Summary of CBOs Cost Estimating Methods for Supersonic Missiles in

Alternative 1
Research and Development

Procurement

Statistical cost estimating relationships

were used that relate nonrecurring costs
to the average cost of the first 1,000
production units and the time required for
development. Also, test results were
extrapolated from the JASSM program and
increased by 50 percent.

Statistical cost estimating relationships

were used that relate nonrecurring costs
to the average cost of the first 1,000
production units.

Costs were based on information from

Hyper-X, SED, and GQM-163 Coyote
propulsion hardware.

A step-down factor was estimated on the

basis of the cost of this hardware in the
research and development phase.

Airframe

A step-up factor was estimated on the

basis of the cost of this hardware in the
procurement phase.

Costs were based on those for the JSOW

system (dollars per pound) and adjusted
for weight and materials.

Guidance and control

A step-up factor was estimated on the

basis of the cost of this hardware in the
procurement phase.

Costs were based on those for the JSOW,

SLAM-ER, and Tomahawk systems.

Warhead

A step-up factor was estimated on the

basis of the cost of this hardware in the
procurement phase.

Costs were based on those for the

Tomahawk and other historical missile
systems.

Booster rocket

A step-up factor was estimated on the

basis of the cost of this hardware in the
procurement phase.

The Mk 72 booster was used as an

analogy, but costs were adjusted to
account for supersonic speeds.

Spares and support equipment

A percentage was added to the cost of all

hardware.

A percentage was added to the cost of all

hardware.

Statistical cost estimating relationships

were used that relate support costs to
hardware, material, and fabrication costs.
Contractor fees were included, which
added 15 percent to contractor costs.

Statistical cost estimating relationships

were used that relate support costs to
hardware, material, and fabrication costs.
Contractor fees were included, which
added 15 percent to contractor costs.

Nonrecurring Hardware Design, Testing,

and Tooling

Hardware, Material, and Fabrication

Engine

Other Support

Source: Congressional Budget Office.

Note: JASSM = Joint Air-to-Surface Standoff Missile; JSOW = Joint Standoff Weapon; SED = Scramjet Engine Demonstration.

APPENDIX: THE METHODOLOGY BEHIND THE COST ANALYSIS

CBO relied on data from the production programs of the

weapon systems just mentioned to estimate procurement
costs for the airframe, the guidance and control system,
the warhead, and the rocket booster. CBO adjusted those
data to calculate the cost of the first production unit for
each of the major components and then applied learning
curves to estimate the hardware costs for 2,000 units.7
For the estimate for the supersonic strike missile, CBO
selected a learning curve of 90 percent, consistent with
that in five other missile and munitions programs.
In contrast, CBO derived the cost of the engine by estimating the first set of hardware built during the research
and development phase (because data on production engines were not available, CBO used engine cost data from
development prototypes) and then used a step-down
equation to estimate the cost of the first production unit.
A step-down equation is based on the premise that fabrication in the development phase is usually more expensive because the units are being assembled at lower quantities and do not always have the benefit of special
manufacturing tools and other efficiencies available in the
production phase. For this estimate, CBO used an equation based on historical missile costs that relates the cost
of the first unit in the development phase to the cost of
the first unit in the production phase.8 After estimating
7. Learning curves are based on the theory that each production unit
will be easier to produce than the previous unit because of the
increased experience of workers and other efficiencies. The curve
is generally expressed as a logarithmic relationship; however, it is
frequently described as a percentage decrease in cost or labor hours
each time the number of units produced doubles. For this estimate, CBO used a learning curve of 90 percent, which means the
hardware for the second unit would cost 90 percent of the cost of
the first unit, and the fourth unit would cost 90 percent of the
cost of the second unit, and so on. A learning curve equation can
also incorporate a coefficient to account for efficiencies gained by
increasing the production rate. For this estimate, CBO used a rate
coefficient of 95 percent and assumed that the 2,000 missiles
would be purchased in lots of approximately 175 each year.
8. For this cost estimate, CBO used step-up/step-down factors and
equations contained in Paul L. Hardin and Daniel Nussbaum,
Analysis of the Relationship Between Development and Production
Costs and Comparisons with Other Related Step-up/Step-down Studies (Arlington, Va.: Naval Center for Cost Analysis, January 1994).

the cost of the first production engine, CBO applied a

production learning curve to estimate the cost to procure
2,000 engines.
In addition to the 2,000 production units, CBO assumed
60 missiles would be built in the development phase. The
Air Force would use those missiles to test and validate the
missile design. Except for their engines, the cost of hardware, material, and fabrication for missiles built in the development phase was estimated using step-up equations, which are the inverse of step-down equations.
CBO then estimated the cost of the 60 development missiles using a learning curve similar to that used for the
production missiles.
To complete the cost estimates for hardware, material,
and fabrication, CBO added costs for final integration
and system checkout, initial spares, and support equipment (such as containers to transport the missiles). Those
estimates were based on factors derived from other missile
programs that were applied to the recurring hardware
costs.

Nonrecurring Design, Testing, and Tooling

This category includes all of the one-time costs that are
not necessarily dependent on the total number of missiles
that would be purchased. It includes initial design and
engineering efforts, systems testing and evaluation (including flight-testing), tools to build missiles, software
development, and efforts to integrate the missile with a
delivery aircraft. CBO estimated that the nonrecurring
costs for the missile hardware in the research and development phase would total about $800 million.
About 85 percent of the nonrecurring costs in the research and development phase are for initial design and
engineering as well as systems testing and evaluation. Design and engineering estimates are based on a statistical
relationship that relates the cost of those efforts to the
length of the development phase and the average cost of
the first 1,000 production units. For that estimate, CBO
assumed that the development effort would last about
eight years on the basis of the development time of other

41

42

ALTERNATIVES FOR LONG-RANGE GROUND-ATTACK SYSTEMS

missile systems, with additional allowance for the design

and testing of new technologies.9

maintenance plans, and government program management expenses.

CBO used several methods to estimate costs for systems

testing and evaluation, including a statistical relationship
that uses missile weight and development time as independent variables, as well as actual costs from the Air
Forces Joint Air-to-Surface Standoff Missile (JASSM)
program. Because the United States does not currently
use ramjets in its tactical missiles, data from historical
programs may not accurately estimate the cost of testing
that new technology. To account for that uncertainty,
CBO increased by 50 percent the estimates for systems
testing and evaluation produced using the historical data.

Program management is the largest component of that

category. For the research and development phase, CBO
estimated those costs using a statistical cost estimating relationship that incorporates the development time and
nonrecurring design costs as independent variables. For
the procurement phase, program management is generally a percentage of the recurring hardware costs adjusted
for the lot number because program management costs
are generally higher in earlier procurement lots.

Although most nonrecurring costs are incurred in the research and development phase, CBO estimated that there
would be about $170 million in nonrecurring costs in the
procurement phase for the special tooling and facilities
needed for full-rate production of the missile. CBO estimated those costs using a statistical relationship that relates such tooling and facilities costs to missile hardware
costs.

Other Support
CBO estimated that other support costs would total
$590 million in the research and development phase and
$960 million in the procurement phase. Those costs
comprise expenses for data collection, contractor fees and
program management expenses, warranties and initial
9. For this estimate, CBO included only those development costs
that would be incurred once the initiation of a formal acquisition
program had begun. Generally, the estimates include all costs associated with the system development and demonstration phase of
the updated DoD Directive 5000 (DoD Directive 5000.1, The
Defense Acquisition System, October 23, 2000). However, the
estimates include some costs in the concept and technology demonstration phase, because those activities were included in program definition and risk reduction activities prior to the update
of DoD 5000 in October 2000. The estimates do not include current and planned expenditures needed to bring the necessary technologies to maturity. For instance, the Air Force, Navy, and
Defense Advanced Research Projects Agency will all be contributing hundreds of millions of dollars over the next several years to
research various ramjet/scramjet technologies to prove that they
have practical applications for a variety of military weapon systems. Those costs are not included here.

Contractor fees are another large component of other

support costs. For that estimate, CBO assumed that contractor fees would be 15 percent of contractor costs in
both the research and development and procurement
phases.

The Cost of Bomber Aircraft

CBO also estimated the cost of developing and producing four types of bomber aircraft to provide long-range
strike capability: a medium-range subsonic bomber, a
medium-range supersonic bomber, a long-range subsonic
bomber, and a long-range supersonic bomber (see
Table A-4 for a listing of the notional characteristics of
those aircraft).
In Alternative 2, the Air Force would develop and purchase a stealthy medium-range subsonic bomber with an
empty weight of 60,000 pounds that would be able to
carry a payload of 20,000 pounds to a range of 1,500
nautical miles without refueling.10 It would be able to
cruise at speeds of around Mach 0.85 using two engines
that would be derived from existing engine designs, such
as the General Electric F118 currently used to power the
B-2 bomber. By CBOs estimates, the cost to develop
such an aircraft would total approximately $19 billion
and the cost to purchase 275 of those bombersenough
to replace the current fleet of F-15E and F-117 strike aircraftwould total $52 billion.
10. Empty weight comprises the weight of the airframe, engines, avionics and electronics, weapons, and other support equipment. It
excludes the weights of the crew, payload, fuel, and oil.

APPENDIX: THE METHODOLOGY BEHIND THE COST ANALYSIS

Table A-4.

Characteristics of Bomber Aircraft

Alternative 2

Alternative 3

Alternative 4

Alternative 5

60

59

128

165

Mach 0.85

Mach 1.5

Mach 0.85

Mach 2.4

20

10

40

40

1,500

1,500

2,500

2,500

Derivative

Derivative

Derivative

New

Number of Engines

Afterburning Engine

No

Yes

No

Yes

Empty Weight (Thousands of pounds)

Maximum Speed
Payload (Thousands of pounds)
Range (Unrefueled, in nautical miles)

Engine Design

Source: Congressional Budget Office.

Note: Alternative 2 = Medium-range subsonic bomber
Alternative 3 = Medium-range supersonic dash bomber
Alternative 4 = Long-range subsonic bomber
Alternative 5 = Long-range supersonic cruise bomber
a. The ranges shown are the maximum distance from an air base or launcher location to the target. For the aircraft alternatives, the total
distance flown on a maximum-range mission would be double the values shown.

In Alternative 3, the Air Force would develop and purchase a stealthy medium-range supersonic bomber conceptually based on either the F-22 or YF-23 fighter.11 It
would have an empty weight of 59,000 pounds and be
able to carry a payload of 10,000 pounds to a range of
1,500 nautical miles without refueling. Using two engines based on an existing design such as the F-22s afterburner-equipped Pratt and Whitney F119, it would be
able to cruise at subsonic speeds of around Mach 0.85
and dash to speeds in excess of the speed of soundup to
Mach 1.5although the bombers range would be reduced substantially when flying at speeds faster than
Mach 1.0. According to CBOs estimates, the cost to develop such an aircraft would total approximately $23 billion and the cost to purchase 275 of those bombers
would total $61 billion.
In Alternative 4, the Air Force would develop and purchase a highly stealthy long-range subsonic bomber with
an empty weight of 128,000 pounds that would be able
11. The F-22 fighter was designed by Lockheed Martin for the Air
Forces advanced tactical fighter program and was selected by the
Air Force for further development and production. The YF-23
fighter was the design submitted by the team of Northrop
Grumman and McDonnell-Douglas.

to carry a payload of 40,000 pounds to a range of 2,500

nautical miles without refueling. It would be able to
cruise at speeds of about Mach 0.85 using improved versions of existing engines, such as the General Electric
F118. Those improvements and other changes would allow the use of two engines on this new aircraft, compared
with the four used to power the B-2 bomber. By CBOs
estimates, the cost to develop such an aircraft would total
approximately $31 billion and the cost to purchase 150
of those bombersenough to replace the current fleet of
B-52 and B-1 bomber aircraftwould total $61 billion.
In Alternative 5, the Air Force would develop and purchase a long-range supersonic cruise bomber with an
empty weight of 165,000 pounds that would be able to
carry a payload of 40,000 pounds to a range of 2,500
nautical miles without refueling. It would be capable of
cruising at Mach 2.4 using four newly designed engines.
Although this aircraft would employ some stealth features, it would not rely on stealth as extensively as the
slower bombers described above. The cost to develop
such an aircraft would total approximately $69 billion
and the cost to purchase 150 of those bombers would total $137 billion, CBO estimates (see Table A-5 for a summary of the costs to acquire the aircraft described above).

43

44

ALTERNATIVES FOR LONG-RANGE GROUND-ATTACK SYSTEMS

Table A-5.

Acquisition Costs of Aircraft in Alternatives 2, 3, 4, and 5

Alternative 2

Research and Development

Concept and technology development
Systems development
Airframe
Engine
Avionics
Other development
Subtotal, development
Procurement
Flyaway
Airframe
Engine
Avionics
Other procurement costs
Subtotal, procurement
Military Construction
Total Estimated Acquisition Costs

Memorandum:
Quantity of Aircraft Procured

Alternative 3

Alternative 4

Alternative 5

5
1
3
5
__

7
1
4
7
__

9
4
7
8
__

25
4
15
21
__

19

23

31

69

24
2
5
21
__

27
4
6
24
__

28
2
6
25
__

63
5
13
56
___

52

61

61

137

72

85

93

207

275

275

150

150

Source: Congressional Budget Office.

Note: Alternative 2 = Medium-range subsonic bomber
Alternative 3 = Medium-range supersonic dash bomber
Alternative 4 = Long-range subsonic bomber
Alternative 5 = Long-range supersonic cruise bomber

The models and equations CBO used to develop the aircraft characteristics were discussed in Chapter 2. CBO estimated the costs to develop and procure each of the
bomber aircraft described above using a common set of
estimating methodologies (see Table A-6 for a summary
of those methods).

Research and Development Costs

Research and development costs include several components: concept and technology development; engineering
design; and other development costs such as prototype
manufacturing, testing and evaluation, systems engineering and program management, integrated logistics
support, contractor fees, and government program
management.

Concept and technology development includes the exploration of early approaches and designs needed to provide
certain capabilities, as well as the development of new
technologies and components needed to enable the new
system to meet performance parameters. CBO estimated
those costs using analogies to other aircraft development
programs such as the B-2 bomber, F/A-18 fighter, and
the F-22 fighter. In developing those estimates, CBO also
considered whether the concept development effort
would most likely be awarded to multiple contractors, as
well as whether flying prototypes would be built. (For instance, contractors Boeing and Lockheed Martin were
both hired to develop technology and designs for the
Joint Strike Fighter and built flying prototypes as part of
the competition to determine which design would be selected for the systems development and demonstration
phase.)

APPENDIX: THE METHODOLOGY BEHIND THE COST ANALYSIS

Table A-6.

Summary of CBOs Cost Estimating Methods for Aircraft in

Alternatives 2, 3, 4, and 5
Research and Development

Procurement

A statistical cost estimating relationship

developed by RAND was used to calculate
engineering and design hours on the basis
of an aircrafts empty weight and
maximum speed, adjusted for the cost of
working with advanced materials.

A statistical cost estimating relationship

developed by RAND was used to calculate
the cumulative average cost of materials,
engineering hours, tooling hours,
manufacturing hours, and quality
assurance hours for 100 aircraft on the
basis of an aircrafts empty weight and
maximum speed, adjusted for the cost of
working with advanced materials.

Engine

A statistical cost estimating relationship

developed by RAND was used to calculate
the cost of engine development on the
basis of engine rotor inlet temperature,
specific fuel consumption, the number of
test hours, and whether the engine was
derived from an existing design.

A statistical cost estimating relationship

developed by RAND was used to calculate
the theoretical cost to produce the first
engine (that cost was based on engine
rotor inlet temperature, the aircrafts
empty weight, and whether the engine
was equipped with an afterburner). Costs
for total production were estimated using
a 90 percent learning curve.

Avionics

Costs were calculated as a percentage of

the cost to develop the engine and
airframe, using as a basis the Joint Strike
Fighter and F-22 development programs.

Costs were calculated as a percentage of

the cost to develop the engine and airframe, using as a basis the B-2, F/A-18E/F,
and F-22 development programs.

Spares and support equipment

Not applicable.

Costs were calculated as 15 percent and

30 percent, respectively, of other hardware costs on the basis of average costs
for other aircraft acquisition programs.

A statistical cost estimating relationship

developed by the Naval Cost Analysis
Division was used to calculate flight-test
costs on the basis of the aircrafts unit
weight, maximim speed, and flight-test
hours plus the cost of producing test
aircraft, which was stepped up from the
production cost estimate.

Not applicable.

Program management

Costs were calculated as a percentage of

the sum of airframe costs and testing and
evaluation costs.

Costs were calculated as a percentage of

airframe costs.

Integrated logistics support

Costs were calculated as a percentage of

the sum of airframe costs and testing and
evaluation costs.

Not applicable.

Hardware, Material, and Fabrication

Airframe

Other Support
Testing and evaluation

Source: Congressional Budget Office.

45

46

ALTERNATIVES FOR LONG-RANGE GROUND-ATTACK SYSTEMS

Engineering design includes the cost to design the airframe, engines, and electronic systems, as well as the cost
to design the special tools and manufacturing equipment
needed to build each of those components. CBO estimated airframe development costs using a cost estimating
relationship formulated by RAND that takes into account the empty weight and maximum speed of the aircraft being designed, as well as the use of advanced materials such as titanium and carbon-thermoset composites
in the design.12 CBO estimated engine development
costs using another RAND cost estimating relationship
that relates engineering development costs to the engines
rotor inlet temperature, the engines specific fuel consumption, and the number of engine test hours in the development program. That method also takes into account
whether the engine design will be derived from an existing engine or will be a brand new design.13,14
Analysts have developed several estimating relationships
for calculating the cost to develop avionics systems, but
the use of those methods requires extensive design specifications and other technology factors. Those specifications
are not usually available for systems in the early concept
phase, such as those considered in this study. Instead,
CBO used the ratio of the total hardware development
costs to the avionics development costs for the F-22 and
the Joint Strike Fighter program to estimate avionics development costs for the bomber-aircraft alternatives. Although not as accurate as detailed cost estimating relationships, that method should produce a reasonable
estimate of costs.
According to a study by the Institute for Defense Analyses, aircraft that feature a significant number of components that reduce the aircrafts radio frequency and infrared signature cost approximately 30 percent more to
12. Younossi, Kennedy, and Graser, Military Airframe Costs: The
Effects of Advanced Materials and Manufacturing Processes.
13. The rotor inlet temperature is defined as the temperature of the
fuel/air combustion products as they enter the first section of
rotating engine blades after leaving the stationary blades just aft of
the combustion chamber. Higher rotor inlet temperatures are
associated with higher thrust-to-weight or power-to-weight ratios.
Specific fuel consumption for turbojet or turbofan engines is the
ratio of the fuel flow rate to thrust generatedthat is, pounds of
fuel per hour to pounds of thrust. Lower ratios indicate moreefficient engines.
14. Obaid Younossi and others, Military Jet Engine Acquisition: Technology Basics and Cost-Estimating Methodology (Santa Monica,
Calif.: RAND Project Air Force, 2002).

design and build than do similarly sized aircraft without

such components.15 Thus, CBO increased the estimated
costs for airframe, engine, and avionics development by
30 percent to account for the additional effort required to
design the stealth features and the additional cost to develop the techniques and tools needed to manufacture
those complex components.
Other development costs include the cost of testing and
evaluation, systems engineering and program management (SE/PM), integrated logistics support, contractor
fees, and government program management costs. Testing and evaluation costs include the cost of the flight-test
program as well as the cost to build prototype airplanes
and other equipment needed for testing. CBO estimated
flight-test costs using a cost estimating relationship developed by the Navy that relates airframe weight, maximum
speed, and the number of flight-test hours to the cost of
testing and evaluation. CBO estimated the cost of prototypes using its aircraft production estimating methodology, which is discussed later in this section.
Both SE/PM and integrated-logistics-support estimates
are based on a cost estimating relationship developed by
the Navy that estimates those costs as a factor of airframe
development and testing and evaluation. CBO estimated
contractor feeson the basis of proposals for other aircraft acquisition programsat 15 percent of the sum of
the development costs discussed above. Government program management costs were estimated at 75 percent of
systems engineering and contractor program management costs plus 20 percent of systems test and evaluation
costs.
Finally, those costs were adjusted to account for cost
growth during the time required to develop the aircraft.
Research by RAND and the Government Accountability
Office has demonstrated that cost analysts systematically
underestimate the cost of developing and procuring major weapon systems. That tendency is particularly pronounced in estimates performed in the concept development and early design phases. In a study of 115 major
defense acquisition systems, RAND calculated that, on
average, estimates of development and procurement costs
were too low by 25 percent and 18 percent, respec15. Bruce R. Harmon and others, Cost Estimating for Modern Combat
Aircraft: Adjusting Existing Databases and Methods to Include LowObservable Cost Consideration (Alexandria, Va.: Institute for
Defense Analyses, 2001).

APPENDIX: THE METHODOLOGY BEHIND THE COST ANALYSIS

tively.16 To account for that systematic bias, CBO increased the estimate of development costs by 25 percent.

Procurement Costs
Procurement cost estimates include flyaway coststhe
recurring costs to manufacture and assemble the airframe,
engine, and avionics systemsand other procurement
costs, both recurring and nonrecurring. Other recurring
costs include program management expenses, contractor
fees, and other government costs. Nonrecurring procurement costs include the cost to set up the production facilities, tools, and manufacturing equipment. They also include the cost of support equipment that is unique to the
airplane; simulators for pilot training; maintenance manuals and technical publications; and an initial complement of spare parts.
CBO estimated airframe costs using cost estimating relationships developed by RAND that predict costs on the
basis of inputs such as the aircrafts empty weight, its
speed, and the increased cost of working with advanced
materials such as titanium and carbon thermosets. Engine
costs were estimated using another RAND cost estimating relationship that considers factors such as the engines
weight and its rotor inlet temperature. That equation also
accounts for the cost of producing engines equipped with
an afterburner.17 CBO calculated avionics costs as a percentage of total hardware costs on the basis of the ratio of
those costs from the F-22 and F/A-18 programs.18 On
the basis of the study by the Institute for Defense Analyses, CBO increased its estimate of the costs of the airframe, engine, and avionics by 30 percent to account for
the additional effort of working with special stealth materials and structures designed to reduce the aircrafts radar,
thermal, visual, and electromagnetic signature.
16. J.A. Drezner and others, An Analysis of Weapon System Cost Growth
(Santa Monica, Calif.: RAND Project Air Force, 1993).
17. Younossi and others, Military Jet Engine Acquisition: Technology
Basics and Cost-Estimating Methodology.
18. Estimates of costs for avionics and electronics are often developed
through the use of complex proprietary cost models like PRICE or
through cost estimating relationships with inputs such as avionics
weight or the level of technology integration. CBO did not define
the avionics systems in the detail needed to use those cost estimating relationships. Rather, CBO estimated avionics costs as a percentage of total airframe and engine costs on the basis of recent
production programs for high-performance aircraft such as the F22 and the F/A-18 E/F.

CBO used two factors to estimate other recurring procurement costs. Combined program management costs
for the contractor and the government are approximately
11 percent of airframe procurement costs, whereas the
contractors fee is estimated as 15 percent of other production costs for the contractor.
CBO used two methods to estimate other nonrecurring
procurement costs. On the basis of actual costs from
other aircraft acquisition programs, CBO estimated costs
for other support equipment and initial spare parts at 30
percent and 15 percent of flyaway costs, respectively.
CBO used a cost estimating relationship developed by
the Navy to estimate the cost of special tools and test
equipment at roughly 1.5 percent of the costs of airframe
procurement.
On the basis of the RAND study on cost growth in
weapon-system acquisitions, CBO increased all procurement costs by 15 percent to account for that factor.

Military Construction
Because the bombers in CBOs alternatives would replace
existing planes, CBO assumed that many of the facilities
used to house and maintain those aircraft would be available to support the new aircraft. However, the introduction of any new weapon system will require some new facilities or significant modifications to existing facilities.
For instance, new buildings for simulators are often
needed, and existing hangars may require modifications if
they are not of the proper dimensions to accommodate
new aircraft. On the basis of actual military construction
costs for other aircraft acquisition programs, CBO estimated that such costs could range from $500 million to
$1 billion, depending on the similarity between the new
system and the aircraft it would replace. For that estimate, CBO included $1 billion in construction costs for
each alternative.

Results
The costs to develop and acquire the four aircraft platforms considered for providing long-range strike capability would range from $72 billion to $207 billion.
Alternative 2. Acquiring 275 medium-range subsonic
bombers would cost $72 billion, CBO estimates. In
CBOs estimate, developing the aircraft accounts for $19
billion, or 26 percent of the total acquisition costs. Concept development accounts for $5 billion of those costs,
including the cost to build several flying prototypes,

47

48

ALTERNATIVES FOR LONG-RANGE GROUND-ATTACK SYSTEMS

whereas the cost to develop the airframe, a derivative engine, and avionics account for another $9 billion. Other
development costs account for the remaining $5 billion.
Procurement costs, estimated at $52 billion, would be 72
percent of the acquisition costs. By CBOs estimates, it
would cost $31 billion to produce 275 aircraft. Other
production costs would total $21 billion. Military construction costs would account for the remaining 2 percent of acquisition costs. On average, each aircraft would
cost $261 million to acquire.
Alternative 3. Acquiring 275 medium-range supersonic
dash bombers would cost $85 billion, CBO estimates.
Developing the aircraft would cost $23 billion, or 27 percent of the total acquisition costs. Including the cost to
build several flying prototypes, concept development
would account for $4 billion of those costs, slightly less
than those for Alternative 2 because CBO assumed that
some of the necessary development work had already
been accomplished under the F-22 and the YF-23 programs. The cost to develop the airframe, a derivative engine, and avionics would account for another $12 billion.
Other development costs would account for the remaining $7 billion.
Procurement costs of $61 billion would account for 72
percent of the acquisition costs, including $37 billion to
build the aircraft, CBO estimates. Other procurement
costs would total $24 billion. Military construction costs
would account for the remaining 1 percent of acquisition
costs. On average, each aircraft would cost $307 million
to acquire.
Alternative 4. By CBOs estimates, it would cost $93 billion to acquire 150 long-range subsonic bombers. Developing the aircraft would cost $31 billion, or 34 percent of
the total acquisition costs. In CBOs estimate, concept
development would account for $3 billion of those costs,
which CBO assumed would not include flying prototypes because of the high cost of building larger airplanes.
The cost to develop the airframe, a derivative engine, and
avionics account for $20 billion, whereas other development costs account for the remaining $8 billion.
Procurement costs of $61 billion would constitute 65
percent of the acquisition costs. It would cost $36 billion
to produce 150 aircraft, CBO estimates. Other procurement costs would total $25 billion. Military construction
costs would account for the remaining 1 percent of acqui-

sition costs. On average, the aircraft would cost $627 million to acquire.
Alternative 5. Acquiring 150 long-range supersonic
bombers would cost $207 billion, CBO estimates. Developing aircraft would cost $69 billion, or 33 percent of the
total acquisition costs. Concept development would represent for $4 billion of those costs, which CBO assumed
would not include flying prototypes. The cost to develop
the airframe, a new engine, and avionics would account
for $44 billion, whereas other development costs would
account for the remaining $21 billion.
Procurement costs of $137 billion would make up 66
percent of the acquisition costs. It would cost $81 billion
to produce 150 aircraft, CBO estimates. Other procurement costs would total $56 billion. Military construction
costs would account for less than 1 percent of acquisition
costs. On average, the aircraft would cost $1.38 billion to
acquire.

Costs of the Common Aero

Vehicle System
The common aero vehicle (CAV) is a maneuverable reentry vehicle that delivers a weapon from space. CBO assumed that the CAV would carry a 1,000-pound warhead
designed to penetrate hardened targets. CBO estimated
the costs of three alternatives for employing CAVs in the
long-range strike missiontwo that would deploy CAVs
aboard ground-based ballistic missiles and one that would
deploy them aboard orbiting satellites.
In Alternative 6, the Army would deploy the CAVs
aboard ground-based solid-fuel boosters powered by
rocket motors similar to the Orion 50 built by Alliant
Techsystems. CBO assumed that the Army would purchase a total of 48 CAVs, 48 rocket boosters, and 24 mobile launchers. Each booster would carry a single CAV,
and two boosters would be deployed on each launcher.
Four mobile launchers, eight boosters, and eight CAVs
would be procured as spares.
In Alternative 7, the Air Force would deploy the CAVs
aboard ground-based Peacekeeper missiles that are being
retired from use as strategic nuclear ballistic missiles.
CBO assumed that the Air Force would purchase a total
of 48 CAVs and convert 24 Peacekeepers to launch them.
Each missile would carry two CAVs. CBO assumed that

APPENDIX: THE METHODOLOGY BEHIND THE COST ANALYSIS

Table A-7.

Characteristics of Common Aero Vehicles in Alternatives 6, 7, and 8

Alternative 6

Alternative 7

Alternative 8

Solid rocket booster

Peacekeeper missile

EELV Heavy

Number of CAVs per Platform

Total Number of CAVs Deployed

40

40

40

Number of Spare CAVs

Number of Replacement CAVs

80

1,000-pound warhead

1,000-pound warhead

1,000-pound warhead

Launch Platform

Payload per CAV

Source: Congressional Budget Office.

Note: EELV = Evolved Expendable Launch Vehicle

Alternative 6 = Medium-range surface-based common aero vehicle
Alternative 7 = Long-range surface-based CAV
Alternative 8 = Space-based CAV

10 Peacekeeper missiles would be based at Cape Canaveral Air Station in Florida and 10 missiles would be based
at Vandenberg Air Force Base in California. Four missiles
and eight CAVs would be used as spares.

aero vehicles (see Table A-9 for a summary of those

methods).

Research and Development Costs

In Alternative 8, the Air Force would deploy the CAVs in

space aboard satellites that would be placed into orbit by
heavy launch vehicles such as the Evolved Expendable
Launch Vehicle (EELV). The orbiting CAVs would use a
rocket motor to bring them out of orbit to attack targets
on Earth. CBO assumed that the Air Force would purchase 128 CAVs and 16 satellites. A constellation of five
satellites, each carrying eight CAVs, would be continuously maintained in orbit. Because orbits decay and the
harsh environment of space eventually degrades orbiting
equipment, CBO assumed that the constellation would
be replaced twice during a 30-year period, requiring a total of 120 CAVs and 15 satellites. Eight additional CAVs
and one additional satellite would be purchased to replace
any that might be lost in a launch failure (see Table A-7
for a description of the technical characteristics of those
alternatives).

CBO estimated that the cost to develop the CAV weapon

system and its various supporting components would total about $2.5 billion for the ground-based alternatives
and about $4 billion for the space-based alternative.
Those estimates include the costs to develop technology
and produce designs for the various system components
such as the common aero vehicle, booster vehicles, mobile launchers, rocket motors, orbital support system
(protective satellite), and launch vehicles. They also include the cost to test and evaluate those components, as
well as the cost to integrate them into a functioning system. There is significant uncertainty about the maturity
of the technologies associated with the common aero vehicle, which results in a substantial risk that costs could
exceed initial estimates. In its estimates, CBO used research by RAND for comparable systems to account for
that risk.

By CBOs estimates, acquisition costs would total about

$4 billion for the ground-based alternatives and about
$12 billion for the space alternative (see Table A-8 for a
summary of the costs of those alternatives). CBO used
several methods, discussed below, to estimate the costs of
development, procurement, and facilities construction
for those alternative methods of deploying the common

Common Aero Vehicle. CBO used DoDs estimate of the

development costs for a program called FALCON (Force
Application and Launch from the Continental United
States) as the basis for estimating development costs for
the CAV. In the program, the Defense Advanced
Research Projects Agency and the Air Force are developing hypersonic technologies needed for weapons that can

49

50

ALTERNATIVES FOR LONG-RANGE GROUND-ATTACK SYSTEMS

Table A-8.

Acquisition Costs of Common Aero Vehicles in Alternatives 6, 7, and 8

(Billions of 2006 dollars)
Research and Development
Common aero vehicles
Booster vehicles
Mobile launcher
Rocket motors
Protective satellites
Launch vehicles
Test and evaluation
System integration
Subtotal, research and development
Procurement
Common aero vehicles
Booster vehicles
Mobile launcher
Rocket motors
Protective satellites
Launch vehicles
Subtotal, procurement
Military Construction
Total Estimated Acquisition Costs

Alternative 6

Alternative 7

Alternative 8

0.6
0.1
0.3
n.a.
n.a.
n.a.
1.0
0.4
___

0.6
0.5
n.a.
n.a.
n.a.
n.a.
1.0
0.4
___

0.6
n.a.
n.a.
0.4
0.6
0
1.7
0.7
___

2.4

2.5

4.0

0.5
0.4
0.3
n.a.
n.a.
n.a.
___

0.5
0.3
n.a.
n.a.
n.a.
n.a.
___

1.4
n.a.
n.a.
0.1
0.7
5.4
___

1.2

0.9

7.7

0.6

3.7

4.0

11.7

Source: Congressional Budget Office.

Note: n.a. = not applicable.

rapidly strike targets over global ranges from launch platforms in the United States. Such technologies would give
the United States the capability to build a weapon that
could be used to promptly strike distant targets by traveling through space in a suborbital trajectory aboard a
hypersonic vehicle. CBO assumed that development efforts for the CAV would be similar, given the common
purpose and performance characteristics of those two
systems.
Over the next several years, DoD plans continued investment in the program to evaluate the performance of
hypersonic technologies in realistic operating environments. The department hopes to launch three test vehicles as part of that program over the 2007-2010 period.
On the basis of funding plans for those programs, CBO
estimates that it could cost $620 million to develop an
operational CAV weapon. That estimate is higher than
the one in DoDs plans for two reasons. First, systems en-

gineering and program management efforts of formal acquisition programs typically cost about 33 percent more
than such efforts in technology development programs
like FALCON. Second, RANDs research on cost growth
in defense programs indicates that cost estimates performed early in the development phase tend to understate
total development costs for space and missile programs by
as much as 40 percent.
Booster Vehicle. In Alternative 6, the Army would use
solid-fuel rocket boosters to launch ground-based CAVs
into a suborbital trajectory toward the target. On the basis of information from the Army, CBO estimated that it
would cost about $140 million to modify existing booster
designs to accommodate the CAVs.
In Alternative 7, the Air Force would use Peacekeeper
missiles to boost ground-based CAVs onto a suborbital
trajectory. Development estimates include the costs to
provide a new guidance and navigation system, the costs

APPENDIX: THE METHODOLOGY BEHIND THE COST ANALYSIS

to modify the missiles propulsion system, and the costs

to design a lighter deployment module and shroud.19
Materials to justify the budget indicate that the Air Force
spent about $400 million to design and test similar upgrades to the Minuteman III missile. CBO assumed that
the costs of the Peacekeeper modification effort would be
somewhat more expensiveabout $490 millionto account for cost growth.
Mobile Launcher. Under Alternative 6, two mediumrange CAV missiles would be placed aboard one mobile
launcher. Mobile-launch vehicles are produced today, although CBO assumed that the CAVs and the solid-fuel
rocket booster would need a newly designed canister to
provide a stable environment for the missiles and to contain them on the vehicle. On the basis of estimates that
were developed in CBOs earlier study on missile defense,
CBO estimated that it would cost about $280 million to
develop those canisters.20
Rocket Motor. CBO used a cost estimating relationship
developed by Tecolote to estimate development costs for
the rocket motors needed for the space-based alternative.21 That method uses several factors, including the
duration of the development phase, the number of prototypes built during that phase, and the total procurement
cost of the rocket motors. CBO estimated that the procurement cost of the rocket motors would total about
$140 million. (Derivation of that procurement estimate
is discussed in the subsequent section on procurement.)
Assuming that the development program for the CAV
rocket motor would require 16 prototypes and that the
development phase would last 60 months, CBO estimated that developing the rocket motor would cost about
$400 million.
Protective Satellite. The protective satellite is the orbital
support component of the space-based system that includes a propulsion system to control the satellites position in orbit and shielding to protect the CAVs from the
harsh environment in space. CBO used a cost estimating
19. The deployment module provides structural support and carries
the electronics needed to activate and deploy the CAV, and the
shroud protects the CAV during flight.
20. Congressional Budget Office, Alternatives for Boost-Phase Missile
Defense (July 2004).
21. Tecolote Research, Inc., The Unmanned Space Vehicle Cost Model,
8th ed. (Goleta, Calif.: Tecolote Research, June 2002).

relationship derived by Technomics that gauges development costs for the protective satellite on the basis of its
weight.22 Assuming that the protective satellite would
weigh about 10,900 pounds, CBO estimated that the development of the protective satellite would cost about
$610 million.
Launch Vehicles. CBO assumed that the space-based
CAVs would be put in orbit aboard rockets that are currently used to launch other military payloads; therefore,
no additional development costs for those vehicles were
included in this estimate.
Testing and Evaluation. For both ground-based alternatives, CBO assumed that DoD would conduct 16 integrated flight tests of the CAV and the launch vehicle over
a four-year period. On the basis of cost estimates for production of CAVs and launch vehicles, CBO estimated
that the hardware for those tests would cost almost $200
million. Additionally, information provided by a contractor indicates that support equipment at the test sight, target sets, and test data processing would cost about $400
million. After accounting for cost growth of about 40
percent, CBO estimated that testing and evaluation
for either ground-based alternative would cost about
$1 billion.
In the flight-test program for the space-based alternative,
the Air Force would launch two heavy launch vehicles,
each carrying one satellite and eight CAVs. By CBOs estimates, the flight-test program would cost about $1.7
billionalmost double the costs for the ground-based alternatives because of the additional expense of two heavy
launch vehicles.
System Integration. For all of the alternatives, CBO assumed that system integration would add 20 percent to
the total costs for common aero vehicles, booster vehicles,
and mobile launchersconsistent with the costs for existing boosters such as the Minuteman and Peacekeeper
programs and the estimated costs for the kinetic energy
interceptor program. System integration would cost
about $400 million for the ground-based alternatives and
about $670 million for the space-based alternative, CBO
estimates.
22. Technomics, Inc., National Missile Defense Propulsion Cost Estimating Relationships (Santa Barbara, Calif.: Technomics, August
2000.)

51

52

ALTERNATIVES FOR LONG-RANGE GROUND-ATTACK SYSTEMS

Table A-9.

Summary of CBOs Cost Estimating Methods for Common Aero Vehicles in

Alternatives 6, 7, and 8
Research and Development

Procurement

Ground-Based Alternatives

Hardware, Material, and Fabrication

Common aero vehicles

Costs were estimated using the Air Forces

FALCON program with adjustments for
additional management and cost risk.

Costs were estimated using the Air Forces

Mk 21 reentry vehicle with adjustments for
more insulation, heavier mass, and the
vehicles payload and navigation
equipment.

An Army cost estimate was used to project

costs.

A statistical cost estimating relationship

was used that relates production costs to
a boosters impulse.

Costs were based on those for the

Minuteman III missile system.

Costs were based on those for the

Minuteman III missile system.

An Army cost estimate was used to project

costs.

Costs were based on a contractors

estimate of the costs for a mobile launcher
proposed for use with boost-phase
interceptors.

Testing and Evaluation

Costs were based on those for hardware

and data processing for 16 tests.

Not applicable.

Systems Integration

A percentage (20 percent) was added to

the costs of hardware design and testing
and evaluation activities.

Not applicable.

Booster vehicles
Medium-range booster

Peacekeeper
Mobile launchers

Continued

Procurement Costs
Procurement costs would total between $870 million and
$1.2 billion for the ground-based alternatives and about
$7.7 billion for the space-based alternative. CBO
calculated total procurement costs using a two-step
approachfirst estimating the costs of producing the
first unit of each of the major components and then projecting those costs for the remaining purchases. CBO assumed that the unit price of subsequent CAVs would not
decline appreciably because the procurement quantities
would be small. In CBOs estimate, procurement costs
include the cost of producing the CAVs themselves, as
well as the cost of booster vehicles and mobile launchers
for the ground-based alternatives and the costs of rocket
motors, protective satellites, and launch vehicles for the
space-based alternative.

Common Aero Vehicle. CBO used actual procurement

costs for the Air Forces Mk 21 reentry vehicle (developed
for the Peacekeeper missile) to estimate the procurement
costs for the CAV in all three alternatives. However, the
CAV would require more thermal protection than would
the Mk 21. According to the Air Force, the average cost
of the Mk 21 reentry vehicle is about $1 million (assuming a purchase of about 50 of them), divided evenly between the cost for the vehicles outer shell and the fusing
and firing assemblies. CBO increased that cost to account
for several technical differences between the notional
CAV and the Mk 21 reentry vehicle.
First, CBO increased costs to account for two differences
in the physical characteristicsthe size of the CAV and
the need for greater thermal protection. The Mk 21s reentry vehicle has a conical shape that measures about two
feet wide at its base by about six feet tall. CBO assumed

APPENDIX: THE METHODOLOGY BEHIND THE COST ANALYSIS

Table A-9.

Continued
Research and Development

Procurement

Space-Based Alternatives

Hardware, Material, and Fabrication

Common aero vehicles

Costs were estimated using the Air Forces

FALCON program with adjustments for
additional management and cost risk.

Costs were estimated using the Air Forces

Mk 21 reentry vehicle with adjustments for
more insulation, heavier mass, and the
vehicles payload and navigation
equipment.

Rocket motors

A statistical cost estimating relationship

was used that relates design costs to
several factors, including procurement
costs.

A statistical cost estimating relationship

was used that relates production costs to
a rocket motors impulse.

Protective satellites

A statistical cost estimating relationship

was used that relates design costs to a
satellites weight.

A statistical cost estimating relationship

was used that relates production costs to
a satellites weight.

Launch vehicles

This cost was assumed to be a sunk

investment.

Costs were based on those for the Air

Forces EELV heavy launch vehicle.

Testing and Evaluation

Costs were based on those for hardware

and data processing for 16 tests.

Not applicable.

Systems Integration

A percentage (20 percent) was added to

the costs of hardware design and test and
evaluation activities.

Not applicable.

Source: Congressional Budget Office.

Note: FALCON = Force Application and Launch from the Continental United States; EELV = Evolved Expendable Launch Vehicle.

that the CAV would measure three feet by 15 feet and

would have three to four times more surface area than
would the Mk 21 reentry vehicle. CBO estimated that
the difference would increase costs by a factor of 3.4, assuming that costs were proportional to surface area.
Whereas the Mk 21 follows a ballistic trajectory toward
the target, the CAV would glide and maneuver toward
the target after it reentered the atmosphere, subjecting it
to greater mechanical stresses and higher temperatures for
a longer period of time. On the basis of information provided by a contractor, CBO estimated that the CAV
would require about two inches of insulation beneath the
surface coating used on the Mk 21 reentry vehicles. On
the basis of discussions with industry analysts, CBO estimated that the price of the insulation would be about
one-third the cost of the coating. After accounting for
those physical characteristics and the cost for the surface

coating and insulating materials, CBO calculated that the

cost of the thermal protection for the CAV would be
about 2.6 times more per square foot than the cost of
thermal protection for the Mk 21 reentry vehicles. After
increasing the average cost of the Mk 21 shell by both the
surface area and thickness factors and increasing the average cost of the fusing and firing assemblies by the surface
area factor only, CBO estimated that those components
would cost between $4 million and $5 million per CAV.
Second, CBO increased the costs of the CAV to account
for an onboard navigation and guidance system. Such systems are components of the Minuteman missile itself;
thus, their costs are not included in the price of the Mk
21 reentry vehicle. CBOs estimate of those costs was
based on the actual procurement costs of currentgeneration guidance and navigation equipment for the
Minuteman III missiles. According to budget materials,

53

54

ALTERNATIVES FOR LONG-RANGE GROUND-ATTACK SYSTEMS

the most recent purchase of guidance and navigation

equipment for the Minuteman missile cost between $2
million and $3 million.
Third, CBO included in the estimate the costs for a
1,000-pound warhead that would be deployed on the
CAV. CBOs estimate of about $6,000 per warhead is
based on the average cost for such devices according to
the Air Forces budget materials.
Finally, CBO increased costs to account for historical cost
growth for comparable systems. On the basis of a report
by RAND, CBO estimated that production costs for the
CAV could grow by 38 percent.23 Thus, CBO estimated
that the average cost of the CAV would total about $11
million. For the two ground-based alternatives, CBO estimated that it would cost about $540 million to buy 48
CAVs. That estimate includes the cost to integrate the
CAV with the other components of the weapon system
about 12 percent of hardware costs based on experience
with other comparable programs. For the space-based alternative, CBO assumed that DoD would buy a total of
128 CAVs at a cost of $1.4 billion.
Booster Vehicle. In Alternative 6, CBO assumed that
each solid rocket booster would carry a single CAV and
that two boosters would be loaded on each mobile
launcher. On the basis of a cost estimating relationship
that relates booster thrust to cost, CBO estimated that
each booster would cost about $9 million, including
about $1 million for the canister to contain the booster
on the mobile launcher. That estimate also includes a factor of about 38 percent for cost growth. In total, CBO estimated that buying 48 boosters would cost about $420
million.
In Alternative 7, the Air Force would deploy the CAVs
aboard Peacekeeper missiles. CBO estimated that each
Peacekeeper missile would be able to carry two CAVs;
therefore, a total of 24 missiles would need to be modified20 for operations and four for spares. (On the basis
of information from the Air Force, CBO estimated that
there are about 60 Peacekeeper missiles in the U.S. inventory.) CBO assumed that the modifications would include upgrading the missiles propulsion and guidance
23. Jeanne M. Jarvaise, Jeffrey A. Drezner, and Daniel M. Norton,
The Defense System Cost Performance Database: Cost Growth Analysis Using Selected Acquisition Reports, MR-625-OSD (Santa Monica, Calif.: RAND, 1996).

system, buying a new deployment module, and buying a

new shroud. As a point of comparison, actual costs for
upgrading the propulsion and guidance system for the
Minuteman III missile averaged about $6 million per
missile. Assuming that the costs for upgrading the Peacekeeper missiles propulsion and guidance system would
be comparable with those costs, and adding $4 million
for the costs to procure a new deployment module and a
new shroud, CBO estimated that the costs to modify
each Peacekeeper missile would total about $14 million.
That estimate includes a cost-growth factor of 38 percent.
In total, CBO estimated that modifying 24 Peacekeeper
missiles would cost about $330 million.
Mobile Launcher. In Alternative 6, the CAVs would be
placed aboard a mobile launcher that would carry two
medium-range CAV missiles. For this analysis, CBO used
a cost estimate that was developed in an earlier study on
boost-phase interceptors.24 In that study, CBO estimated
that a mobile launcher capable of firing two interceptors
would cost a little more than $11 million in 2004 dollars.
After adjusting for inflation, CBO estimated that each
mobile launcher would cost slightly less than $12 million
in 2006 dollars, or about $280 million for 24 mobile
launchers.
Rocket Motor. In Alternative 8, the space-based CAVs
would use a rocket motor to reenter the Earths atmosphere. CBO applied a cost estimating relationship developed by Technomics that relates a rocket motors procurement costs to the product of its total impulse expressed in
newtons of thrust and its burn time expressed in seconds.
CBO estimated that a rocket motor with a thrust of 113
kilo newtons and a burn time of five seconds would cost
just over $1 million to produce.25 That estimate also includes a factor of 19 percent to account for growth observed in the cost estimates of comparable space-based
systems.
CBO assumed that the rocket motors would be bought
over the same time period as would the CAVs and that
the cost per rocket motor would remain essentially unchanged over the production period. Buying 128 rocket
24. Congressional Budget Office, Alternatives for Boost-Phase Missile
Defense.
25. Technomics, Inc., National Missile Defense Propulsion Cost Estimating Relationships.

APPENDIX: THE METHODOLOGY BEHIND THE COST ANALYSIS

motors (one for each of the CAVs) would cost a total of

about $140 million, CBO estimates.
Protective Satellite. CBO applied a cost estimating relationship developed by Tecolote that uses the weight of
the protective satellite to estimate the cost to procure the
protective satellite. CBO assumed that the protective satellite would weigh about half as much as the combined
weight of the eight CAVs. Eight CAVs and their rocket
motors would weigh about 21,800 pounds; therefore, the
protective satellite would weigh about 10,900 pounds.
CBO estimated that each protective satellite would cost
about $45 million.
CBO assumed that the protective satellites would be
bought over the same time period as would the CAVs and
rocket motors. Buying the 16 protective satellites (one for
every eight CAVs) would cost about $720 million, CBO
estimates.
Launch Vehicle. The space-based CAV system would be
put in orbit by a launch vehicle. CBO estimated that each
satelliteconsisting of one protective satellite and eight
CAVswould weigh about 16 tons. CBO assumed that
the CAV system would be launched aboard a heavy
launch vehicle currently planned for the Air Forces EELV

program and assumed that DoD would place one CAV

satellite aboard each launch vehicle. CBO estimated the
procurement costs for the launch vehicle on the basis of
budget data provided by the Air Force. Those data indicate that the costs for a launch vehicle would total about
$340 million, including about $190 million for the vehicle hardware and about $150 million for the launch services. CBO estimated that the costs for 16 launch vehiclesone CAV satellite on each heavy launch vehicle
would total about $5.4 billion.

Military Construction Costs

Under Alternative 7, CBO assumed that the Air Force
would make improvements to facilities at Cape Canaveral
Air Station and Vandenberg Air Force Base, the bases
where the Peacekeeper/CAV systems would be deployed.
At each base, the Air Force would install new equipment
to support ground operations and build 10 new silos to
house the Peacekeeper missiles. Support equipment
would cost about $50 million per set, and the missile silos
would cost about $15 million each. After applying a factor of 50 percent to account for cost growth, CBO estimated that it would cost $600 million for silos and support equipment at the two bases.

55

CONGRESS OF THE UNITED STATES

CONGRESSIONAL BUDGET OFFICE
WASHINGTON, DC 20515

INSIDE MAIL