Propulsion Economic Considerations for Next Generation

Space Launch

Chris Y. Taylor*
Jupiter Research and Development, 1919 W. Sam Hou. Pkwy. N. #310, Houston, TX, 77043

This paper describes the major costs associated with new space launch vehicles, and
analyzes how they are affected by propulsion and other design choices. These costs are
broken down into hardware, operations, risk, development, and propellant categories. The
amortized cost of development at likely near term launch rates, and the cost of risk
(especially for manned systems) seems to favor evolutionary and multi-stage rather than
revolutionary and single stage designs. Examples are given of how cost optimized and
performance optimized launch concepts differ. Lastly, potential developments that could
drastically alter the conclusions of the above analysis are discussed.

Nomenclature
a = amortization factor
cH = cost of vehicle hardware per unit weight
cL = cost of labor per man-hour, including overhead
cP = cost of propellant per unit weight
cPL = cost of payload per unit weight
cT = specific launch cost = CTOT/MPL
c1 = specific launch cost of stage 1
c2 = specific launch cost of stage 2
CTOT = total launch cost
f = fraction of launch vehicle expended in one mission
j1 = stage 1 payload to vehicle payload ratio = (MPL+MS2+MP2)/MPL
L = labor intensity
MP = vehicle propellant mass
MPL = vehicle payload mass
MS = vehicle structure mass
MP2 = stage 2 propellant mass
MS2 = stage 2 structure mass
PFAIL = probability of a mission failure
q = propellant-structure mass ratio = MP/MS
r = structure-payload mass ratio = MS/MPL
r1 = stage 1 structure-payload mass ratio (stage 1 payload includes both upper stage and vehicle payload)
r2 = stage 2 structure-payload mass ratio
η = propellant mass fraction = MP/( MP+MS)
λ = structure cost = CTOT/MS
λ1 = stage 1 structure cost
λ2 = stage 2 structure cost
λH = vehicle hardware structure cost
λK = risk structure cost
λL = flight operations structure cost
λP = propellant structure cost
λR = recurring structure cost
λNR = non-recurring structure cost

*
Principal, AIAA Senior Member

1
American Institute of Aeronautics and Astronautics
 I. Introduction

S pace launch vehicle economics is a subject of considerable interest and study to businesses in the space launch
industry. Space launch cost calculations done by these businesses are usually not published, however, because
they contain trade secrets. The reluctance of space launch economics experts to publish their cost information leaves
many aerospace engineers with a poor understanding of how their work affects the cost of space access. This paper
describes a simple but useful cost model for space launch vehicles and a “back of the envelope” estimate of current
space launch costs. This information is then used to determine the likely characteristics of economical next
generation launch vehicles. The cost model can also be combined with rocket performance models, allowing
researchers to determine the approximate specific launch cost and minimum cost configuration for a given launch
vehicle concept.

II. Cost Model

The cost model used in this analysis is based on the equation:

cT = rλ (1)

where

cT = specific launch cost = total launch cost / payload mass

r = structure-payload mass ratio = structural mass / payload mass
λ = structure cost = total launch cost / structural mass.

This cost model is similar to previously published models, such as those by Griffin1, Claybaugh2, Kalitventzeff3,
and Carton4. It has the benefit of dividing the problem of launch cost into two
variables, r and λ, that reflect different areas of study. The value of the variable Table 1. Values of r
r is driven primarily by technology and the physics of the orbit desired. The
calculation of r is beyond the scope of this paper, but it can be determined from Vehicle r (to LEO)
historical trends, simple rocket equation analysis, or detailed engineering Atlas V 400 2
studies depending on the level of accuracy desired. Sample values of r for Proton M 2.2
several current launch vehicles are shown in Table 1. The value of the variable Ariane 5 5.2
λ is driven by economics and program management. Determining the value of λ Space Shuttle 12
will be discussed in greater detail below.
Structure cost, λ, can be expressed as:

λ = λR + (λ NR / a) (2)

where λR is the recurring launch cost per unit of vehicle structure, λNR is the non-recurring cost of the launch system
per unit of vehicle structure, and a is an amortization factor that determines how much of the non-recurring system
costs are charged to each launch. The amortization factor a is typically proportional to the launch vehicle flight rate.
Recurring structure costs can be further broken down into the categories of vehicle hardware cost, operations
cost, risk cost, and propellant cost.

λR = λH + λL + λK + λP (3)

Vehicle hardware cost per unit of structure mass is2 the product of the fraction of the vehicle expended in a
launch, f, and the cost of vehicle hardware, cH.

λH = fcH (4)

For a fully expendable launch system f would have a value of 1. For a reusable system f would represent the fraction
of the vehicle that is worn out during each mission. Some researchers3,4,5 prefer to separate the hardware cost into

2
American Institute of Aeronautics and Astronautics
more detailed categories such as “engine” and “tankage” costs, but that additional level of detail is not required for
the analysis in this paper.
Operations cost per unit of structure mass is2 the product of the cost of labor including overhead, cL, and a labor
intensity parameter, L.

λL = LcL (5)

For expendable launch systems L is defined as the man-hours of launch operation labor divided by the structure
mass of the vehicle. For a reusable launch system L would also need to include recovery and refurbishment labor.
It is difficult to determine the cost of risk per unit of vehicle structure because launch vehicles have many
different failure modes, each with their own potential cost. For private launches the price of insurance must also be
a consideration. For the simple “back of the envelope” analysis done in this paper, the cost of risk is assumed to be:

λK ≈ PFAIL [(cPL / r ) + (1 − f )cH ] (6)

where PFAIL is the probability of a mission failure resulting in vehicle and payload loss, and cPL is the value per unit
mass of the payload. This equation will likely produce a low estimate of the risk cost because it only estimates the
direct cost of the lost vehicle and payload. A real mission failure would also incur indirect costs resulting from
schedule delays, public relations problems, accident investigation, and similar factors..
The cost of propellant per unit of structure mass is a product of the propellant-structure mass ratio, q, and the
cost per unit mass of the propellants, cP.

λP = qcP (7)

where

q = η /(1 − η ) (8)

A computer spreadsheet based on this cost model is available free on Jupiter Research and Development’s website at
http://www.jupiter-measurement.com/research/rocketcost.xls.

III. Current State of Launch Costs

Table 2 lists values typical of the economic characteristics for current generation space launch vehicles. The
range of values for cH is from Worden6, the range of values for L is from Claybaugh2 and Griffin7, and the range of
PFAIL is from information presented by Chang8. The value for cP
assumes a liquid fuel vehicle. Rocket propellant prices are available Table 2. Current Generation Launch
from the Defense Energy Support Center, and are provided online at Vehicle Economic Characteristics
http://www.desc.dla.mil/DCM/DCMPage.asp?LinkID=DESCCuto Characteristic Values
merService under the heading “missile fuels.” The range of values f 1
for λNR assumes that non-recurring costs are limited to vehicle R&D
cH $1100/lb.to $2300/lb.
costs, and is from Claybaugh2 supplemented by information given
L 1 to 20
by Isakowitz9 and at the website http://www.astronautix.com. In
cL $100/hr
order for the data in Table 2 to be useful in calculating the current
state of launch costs, the amortization factor, a, and values for the PFAIL 2% to 5%
launch vehicle parameters r and η must also be known. For this c PL $10,000/lb.
analysis a will be assumed to be 27. This value for a is based on a c P $0.1/lb. to $0.25/lb.
10 year payback of the non-recurring costs with a 4 year λ NR $20,000/lb.to $120,000/lb.
development program followed by 27 flights over the next 6 years,
a flight rate typical of current American space launch vehicles8, and neglecting interest or inflation. Wertz10
presents a description of the effect of interest and inflation on launch costs that may be useful to researchers wanting
a more detailed treatment of amortization. The value of structure-payload mass ratio, r, used for estimating current

3
American Institute of Aeronautics and Astronautics
launch costs will be 2, and η will be assumed equal to 0.9. These values are not conservative, in order to represent
the best case for current space launch costs.
Table 3 gives the estimated current state of launch vehicle costs from this cost model using the above
information and assumptions. This estimated cost information is also presented graphically in Fig. 1.

Table 3. Estimate of Current Figure 1. Estimate of Current Space Launch Structure Costs to LEO
Space Launch Structure
Costs to LEO $5,000
Cost Value

Costs/Structure Mass
Non-Recurring, $750/lb. to $4,000
λNR $4,500/lb.
Vehicle $1,100/ lb. $3,000
Hardware, λH to $2,300 lb.
Operations, λL $100/lb. to $2,000
$2,000/lb.
Risk, λK $100/lb. to $1,000
$250/lb.
Propellant, λP $1/lb to $0
$2.5/lb.
R&D Vehicle Ops Risk Prop.

The costs given in Table 3 and Fig. 1 are in dollars per pound of vehicle structure mass. To convert this to launch
cost per pound of payload mass you must multiply these costs by the value of r, which in this example is 2. The
resulting estimate of current space launch costs to low earth orbit (LEO) suggests that the two largest components of
space launch cost are the amortized non-recurring cost, which in this analysis is limited to R&D costs, and the cost
of vehicle hardware. This conclusion contradicts the “conventional wisdom” that the two biggest components of
space launch cost are vehicle hardware and flight operations. Flight operation costs may not even be the third
biggest contributor to space launch costs with the current generation of launch
technology. Using the best current practices the cost of operations and the cost of
risk were both estimated at approximately $100/lb. As discussed above, however,
the cost model used probably produces a low estimate of the cost of risk because
it only includes direct costs. Once indirect costs, such as accident investigations
and schedule delays, are considered the third largest contributor to cargo space
launch costs would likely be the cost of risk, with the cost of operations close
behind as the fourth largest cost component. For manned launches, where the
cost of failure is much higher than for cargo, the cost of risk would certainly
overshadow the current state of operations costs. The cost of propellant is
insignificant and can be ignored in a “back of the envelope” analysis.

IV. Reducing Next Generation Launch Costs

A. Amortized Non-Recurring Costs

Even if all other cost components could be eliminated, the current amortized
non-recurring costs would still keep space launch costs above $1000/lb of
payload to LEO. If truly cheap space access is desired, then some way must be
found to reduce the amortized non-recurring costs of the next generation of
launch systems.
One important step in reducing amortized non-recurring space launch costs is
to make that goal a design consideration from the very beginning of the launch
system development project. Space launch vehicle concepts that are designed
with low development costs in mind look different from performance optimized
concepts, or even concepts designed for just minimum recurring costs. A good Fig. 2. Microcosm’s
example of how considering amortized non-recurring costs can affect a launch Space Freighter Concept

4
American Institute of Aeronautics and Astronautics
vehicle design is Microcosm’s Scorpius Space Freighter11 concept shown in Fig. 2. Rather than use a few large
rocket engines to propel the Space Freighter, Microcosm chose to use a large number of small (20,000lb. thrust)
engines. Because a small engine is cheaper to develop, the R&D cost of the vehicle is much lower and the vehicle
economics are improved over a similar vehicle with more conventionally sized engines. This arrangement has
disadvantages from a purely performance standpoint, but the increase in structure-payload mass fraction from those
performance losses are more than compensated by the reduced structure cost achieved through minimizing
development effort. The Scorpius Space Freighter vehicle in Fig. 2 also illustrates similar economic optimization at
the expense of performance in the use of identical clustered propellant tanks.
Another way to reduce amortized non-recurring costs is to eliminate development costs by using off-the-shelf
components. To illustrate the cost benefit of using off-the-shelf components, consider the example of developing a
new engine versus purchasing an existing engine for the sample current generation launch vehicle. Assuming that
the sample launch vehicle achieved the lowest costs in each category of Table 3, it would have a structure cost of
approximately $2,050/lb. These sample costs were calculated assuming a structure-payload mass ratio of 2 and η of
0.9. This performance could be achieved with a two stage rocket having a stage 1 engine ISP of 331 s and a stage 2
engine ISP of 462 s. Figure 3 shows the result of a simple trade study to develop a new engine design with a higher
ISP to replace the existing stage 1 engines. This analysis assumes that the mass and the cost of the new engine are
identical to the existing engine. The solid line in Fig. 3 shows the increase in non-recurring cost that can be justified
vs. the resulting improvement in ISP of the new engine. The dashed line in Fig. 3 shows the size of the engine that
can be developed with the increased non-recurring cost assuming a vehicle structure mass with the original engine of
100,000 lbs. and basing engine development cost on the Spacecraft/Vehicle Level Cost Model from the NASA
Johnson Space Center website at http://www.jsc.nasa.gov/bu2/SVLCM.html. Since rocket engines for launch
vehicles of this size frequently weigh over a ton, it can be seen from Fig. 3 that a new engine development program
must promise tens of seconds of additional specific impulse in order to economically justify its cost. At the present
state of engine technology, off-the-shelf engines will frequently be the most cost effective choice.
Even when off-the-shelf components cannot be used, savings can still be made by using off-the-shelf technology
through evolutionary, rather than revolutionary, design. All of the launch vehicles proposed for the Evolved
Expendable Launch Vehicle program illustrate this strategy. Another advantage of using an evolutionary
development path for launch vehicles is the reduction of risk. As explained above, the cost of the risk of a launch
failure is probably the third largest contributor to cargo space launch costs. Using evolutionary designs minimizes
the development unknowns that could result in a launch failure.
When large technology development efforts are conducted, the space launch system’s non-recurring costs can
still be kept low if some or all of those R&D costs are paid for by other applications of the technology. Sharing the
development cost of new technologies with non-space
1800 3500 launch applications is the most likely way for a large
technology development program benefiting space
launch vehicles to be justified economically. This
1500 2750
means that technologies, such as rocket-based-
Budget, millions of US$

combined-cycle engines, scramjet engines, or materials

Engine Mass, lbs.

developments which have many potential applications

are easier to justify economically than technologies,
1200 2000 such as skyhook tethers, which may provide greater
space launch benefits for less effort but have few other
potential applications.
An alternative strategy to reducing the
900 1250
amortized non-recurring launch costs is to increase the
amortization factor. This usually requires increasing
the launch vehicle’s flight rate so that the non-recurring
600 500 costs are spread over a larger number of launches.
10 30 50 70 Ways of dramatically increasing a launch vehicle’s
Specific Impulse Improvement, s flight rate include: using on-orbit assembly to launch
Engine Budget Engine Mass
large payloads with multiple flights12, using clustered
identical vehicles to launch large payloads with one
Figure 3. Economically Justified Rocket Engine multi-vehicle flight, or spreading out into new markets
Development Budget and Engine Size vs. ISP that provide additional launch opportunities beyond the
Improvement Produced current space launch industry13. The construction of the
International Space Station is an obvious example of

5
American Institute of Aeronautics and Astronautics
on-orbit assembly. Bimese14 and trimese15 rocket designs are examples of cluster vehicle concepts. Unfortunately
cluster vehicles usually require very non-optimal staging that increases the structure-payload mass ratio enough to
overcome the benefit of reduced amortized non-recurring costs. Whether launching a cluster of identical vehicles
should be modeled as an increase in the amortization factor (because you are flying more than one vehicle per
launch) or as a decrease in non-recurring costs (where the cluster is considered to be one larger vehicle whose
development is made cheaper by copying stages) is debatable, but for this analysis it is modeled as an increase in a.
Potential new markets for space launch vehicles would include space tourism13, suborbital weapons delivery16, fast
suborbital cargo delivery14, and fast suborbital passenger delivery17. Most of these applications do not require
orbital flight, and provide a potential evolutionary development path where a suborbital vehicle could be
incrementally improved into an orbital vehicle. As discussed earlier, evolutionary vehicle development could be
used to reduce both development cost and risk cost.

B. Vehicle Hardware Costs

Lowering vehicle hardware costs has been considered an important part of making space launch more
economical since the beginning of the space age. As shown in Eq. 4, vehicle hardware cost per unit of structural
mass is a product of two parameters: the fraction of the vehicle expended, f, and the cost of the vehicle hardware, cH.
There are two popular schools of thought on improving vehicle hardware economics, each focusing on a separate
parameter. Reusable vehicles attempt to lower λH by reducing the amount of the vehicle that is expended in each
flight. The expendable big-dumb-booster approach18 attempts to lower λH by reducing the cost of vehicle hardware.
There are, of course, hybrid concepts that include some component of each approach. Both approaches typically
improve structure cost at the expense of additional weight, increasing the vehicle’s structure-payload mass ratio.
The optimum amount of either reusability or “dumbness” for a given launch concept can be determined by
combining the cost model with the vehicle performance model so that the minimum cost condition can be
calculated. Whitehead5 suggests plotting iso-cost graphs with structure cost on one axis and vehicle mass ratios on
the other to help visualize the effect of trading cost and weight against each other.
Analysis from Wertz10 argues persuasively that reusable space launch vehicles are still not economically
competitive with expendable ones. One of the major reasons for this conclusion is that the vehicle hardware savings
promised by reusable systems are offset by not just additional weight, but also by higher amortized development
costs. Based on this analysis it appears that the next generation of economical cargo launch vehicles will likely be
expendable systems. For reusable launch systems to be economically justified, something must occur to reduce the
amortized non-recurring costs. This could be a much higher launch rate than currently projected or a technological
development that provides a large improvement to reusable launch vehicle performance but is paid for, at least
partially, by non-space launch applications. An evolutionary vehicle development path, particularly if combined
with a lean organization, might be another way that the amortized non-recurring cost could be reduced enough to
make reusability an economically viable choice for launch vehicle concepts. This evolutionary path might lead to
the next generation launch vehicles initially flying as expendable vehicles with reusability added later in the design’s
life.
Staging has a noticeable effect on the economics of vehicle design, with different stages on the same vehicle
having different minimum cost configurations. This can be illustrated with the example of a two stage space launch
vehicle where the costs of the stages are considered separately. The total vehicle cost is the sum of both stage costs.

cT = c1 + c2 (9)

This can be rewritten as:

cT = j1r1λ1 + r2λ2 (10)

where
j1 = the stage 1 lifted mass to vehicle payload mass ratio = (stage 2 mass + payload mass) / payload mass.
The parameter j1 appears in the first term of Eq. 10 because cT is in dollars per unit of vehicle payload but the
variable r1 is the ratio of the first stage’s structure to the first stage’s payload. Since the payload of the second stage
is also the same as the vehicle’s payload, no similar parameter is needed in the second term. Researchers preferring
consistency may want to add a j2 parameter to the second term of Eq. 10, and then set j2 equal to 1. Assuming that
all the stage mass can be divided into structure and propellant allows the definition of j1 to be rewritten into a more
useful form.

6
American Institute of Aeronautics and Astronautics
 j1 = ( M PL + M S 2 + M P 2 ) / M PL = 1 + r2 + r2 q2 (11)

Combining Eq. 10 with Eq. 11 gives the vehicle structure cost as a function of the stage structure costs and the stage
structure-payload mass ratios.

cT = r1λ1 + r2 r1λ1 + r2 q2 r1λ1 + r2λ2 (12)

The effect of stage location on economic decisions becomes obvious from examining Eq. 12. The variables r1 and r2
appear in the same number of terms, so that if λ1 and λ2 have approximately the same value then a percentage change
of either r1 or r2 will have the same effect on the vehicle’s specific launch cost. The variable λ1 appears in three of
the terms but λ2 appears in only one term, so for typical cases a percent change to λ1 will have a larger effect on the
vehicle’s specific launch cost than a similar change to λ2. This can be understood by considering the normal
configuration of a two stage rocket, with a large first stage topped by a much smaller second stage. If the first stage
is ten times larger than the second, then a dollar per pound reduction in the first stage’s structural cost will have ten
times the effect on the vehicle’s specific cost than a similar dollar per pound reduction in the second stage’s
structural cost. These relationships between r1 and r2 and between λ1 and λ2 in Eq. 12 mean that technologies used
to reduce stage structural cost at the expense of increased structure-payload mass ratio should be put in the first stage
because they will produce more savings at the same weight penalty. Technologies that reduce the stage’s structure-
payload mass ratio, but increase stage structural cost should be used in the second stage to get the performance
increase for the least possible cost. A similar effect occurs on rockets three or more stages. These trends will drive
economical launch vehicles to have relatively high performance, high cost upper stages on relatively cheap, but
heavy lower stages.
The idea that weight savings on the upper stages of a multi-stage rocket is more valuable than weight savings on
the lower stages is hardly a new discovery; rockets have been built that way since staging was developed. What Eq.
12 does do, is to illustrate this effect of a stage’s position in a rocket on its minimum specific cost configuration.
Because the goal of adding reusability to a stage is to decrease its structural cost, usually at the expense of increasing
the stage’s structure-payload mass ratio, Eq. 12 also demonstrates that economical cargo launch vehicles should
develop reusable lower stages before they develop reusable upper stages. Having an expendable upper stage on a
reusable lower stage seems to contradict the usual trend of reusable vehicle concepts. From the Reuse Crew
Module19 design for Apollo, through the Space Shuttle, to current Crew Exploration Vehicle designs, high profile
reusable space vehicle concepts typically have a reusable upper stage on expendable lower stages. These high
profile reusable concepts are intended to be manned vehicles, not economical cargo launch vehicles. Because a
manned system requires you to recover the upper stage anyway, in order to recover the crew, there is less penalty to
adding reusability to the uppermost stage of a manned system. The original Boeing EELV concept with a partially
reusable first stage (described at http://www.fas.org/spp/military/program/launch/eelv_b.htm) is an example of an
economical cargo launch vehicle concept incorporating reusability on a lower stage, while using an evolutionary
vehicle development path to reduce non-recurring costs.

V. Conclusions
Despite the roughness of the cost assumptions used, the cost model described in this paper allows several
conclusions to be drawn about economical next generation space launch vehicles. To achieve low cost access to
space, methods of reducing the high amortized non-recurring costs of space launch vehicles must be found.
Economical next generation space launch vehicles will make high use of off-the-shelf components and evolutionary
technological development to reduce non-recurring costs. Major new developments in economical space launch
technology will be partially or fully paid for by non-space launch applications, such as aircraft or weapon system
applications, to reduce non-recurring costs.
The cost of vehicle hardware must also be reduced to provide low cost space launch capability. Economical next
generation cargo launchers will likely have a high performance expendable upper stage on a relatively cheap
expendable lower stage. The expendable lower stage may evolve into an even lower cost reusable stage if flight rate
and technological development permit. Also, suborbital reusable vehicles developed to take advantage of high
launch rates in new space markets may economically evolve orbital capability by adding a high performance
expendable upper stage. For the next generation manned launcher, the need to have a recoverable upper stage for
the crew will continue to make reusable upper stages economically attractive.

7
American Institute of Aeronautics and Astronautics
 Lastly, an economical launch vehicle does not resemble a performance driven space launch vehicle in
configuration, component sizing, or mission design; to be successful an economical launch vehicle must be designed
for minimum specific launch cost from the beginning of the project. Economic performance is not something that
can be added as an afterthought.
The cost model described in this paper should be of interest to researchers wanting to study the cost of space
access or the design of minimum cost launch vehicles. Care should be taken, however, in using this model or the
analysis presented here to predict future events in launch vehicle development. In the real world, economics is not
the only driver of launch vehicle design. Public launch vehicles may be developed for non-economic reasons, such
as national security or political goals. A profitable launch vehicle for a private company may not necessarily be an
economical launch vehicle. For example, if a private launch company can get government funding to conduct the
research that they need for their own vehicle development program then they can solve the difficult amortized non-
recurring cost problem and turn a profit, by pushing that cost onto a government agency, even though the vehicle
may still not be economical. Future launch vehicle flight rates and vehicle development costs are notoriously
difficult to predict and easy to misrepresent, so that even an organization that thinks it is funding an economical
launch vehicle may discover otherwise once the vehicle goes into operation. The engineers who do succeed in
developing real world economical space launch vehicles, however, will be the ones who make launch vehicle
economics an obsession from the beginning.

References
1
Griffin, M. D., and Claybaugh, W. R., “The Cost of Access to Space,” JBIS, Vol. 47, 1994, pp. 119-122.
2
Claybaugh, W. R., AIAA Professional Study Series Course: Economics of Space Transportation, Oct. 12-13, 2002,
Houston TX.
3
Kalitventseff, B., “Various Optimization Methods for Preliminary Cost and Mass Distribution Assessment for
Multistage Rocket Vehicles,” JBIS, Vol. 20, 1965, pp. 177-183.
4
Carton, D.S., and Kalitventzeff, B., “Effect of Engine, Tank, and Propellant Specific Cost on Single-Stage
Recoverable Booster Economics,” JBIS, Vol. 20, 1965, 183-196.
5
Whitehead, J.C., “Launch Vehicle Cost: A Low Tech Analysis”, presented at the 36th AIAA/ASME/SAE/ASEE
Joint Propulsion Conference and Exhibit, AIAA2000-3140, Huntsville, AL, 2000.
6
Worden, S., “Perspectives on Space Future,” presented at the 2003 NIAC Fellows Meeting, Atlanta, GA Nov. 6,
2003, http://www.niac.usra.edu/files/library/fellows_mtg/nov03_mtg/pdf/Worden_Simon.pdf.
7
Griffin, M.D., “Heavy Lift Launch for Lunar Exploration,” presented at the U. of Wisconsin, Madison, WI, Nov. 9,
2001, http://fti.neep.wisc.edu/neep533/FALL2001/lecture29.pdf.
8
Chang, I.S., “Overview of World Space Launches,” Journal of Prop. And Power, Vol. 16, No. 5, 2000, pp. 853-
866.
9
Isakowitz, S. J., Hopkins, J., and Hopkins, J. P., International Reference Guide to Space Launch Systems, 4th ed.,
AIAA, Reston, VA, 2004.
10
Wertz, J. R., “Economic Model of Reusable vs. Expendable Launch Vehicles,” presented at the IAF Congress,
Rio de Janeiro, 2000, http://www.smad.com/scorpius/IAFPaper.pdf.
11
Chakroborty, S., Wertz, J.R., Conger, R., Kulpa, J., “The Scorpius Expendable Launch Vehicle Family and Status
of the Sprite Small Launch Vehicle,” presented at the 1st Responsive Space Conference, Redonodo Beach, CA,
2003, http://www.smad.com/scorpius/9005p.pdf.
12
Wingo, D., “Transforming Spacecraft Economics Via On-Orbit Assembly,” SpaceDaily, Jan. 30, 2002,
http://www.spacedaily.com/news/satellite-tech-02a.html.
13
Foust, J., “Is There a Business Case for RLVs?,” The Space Review, Sept. 2, 2003,
http://www.thespacereview.com/article/44/1.
14
Olds, J. R., and Tooley, J., “The Bimese Concept: A Study of Mission and Economic Options,” NASA Langley
Grant NCC1-229 Final Report, Atlanta, GA, Dec. 22, 1999.
15
Dore, F. J., “Aircraft Design and Development Experience Related to Reusable Launch Vehicles,” Reducing the
Cost of Space Transportation: Proceedings of the American Astronautical Society 7th Goddard Memorial
Symposium, edited by George K. Chacko, American Astronautical Society, Washington, D.C., 1969.
16
Daehnick, C., “Spacelift: Suborbital, Earth to Orbit, and on Orbit,” Airpower Journal, Summer 1995, pp. 42-64,
http://www.airpower.maxwell.af.mil/airchronicles/apj/spacast3.html.
17
Hunter, M. W., “The Hypersonic Transport- The Technology and the Potential,” presented at the AIAA 7th
Annual Meeting and Technical Display, Houston, TX, 1970.

8
American Institute of Aeronautics and Astronautics
18
London, J. R., LEO on the Cheap, Air University Press, Maxwell AFB, AL, 1994,
http://www.dunnspace.com/leo_on_the_cheap.htm.
19
Raymes, F., and Dodds, J. I., “Reusable Space Transportation Systems: 1970/1980- An Evolutionary View,”
Reducing the Cost of Space Transportation: Proceedings of the American Astronautical Society 7th Goddard
Memorial Symposium, edited by George K. Chacko, American Astronautical Society, Washington, D.C., 1969.

9
American Institute of Aeronautics and Astronautics