An argument can be made for the effectiveness of a shuttle-derived heavy-lift booster like Ares. (credit: Encyclopedia Astronautica)

The cost of medium lift

by Michael K. Robel
Tuesday, June 1, 2004

The recent article “The myth of heavy lift” asserts the US does not need a heavy-lift launch vehicle because of their expense and that the capabilities of existing medium-lift vehicles are sufficient for our current needs.

In order to sustain a manned space flight program, the US needs to pursue a balanced program with a family of vehicles providing several capabilities, a built-in modernization program, and not committing itself to a start and stop method of development that results in several non-compatible types of vehicles and spacecraft (Mercury, Gemini, Apollo, and Shuttle) that do not directly build on the strengths of the predecessor programs. Just as the Air Force does not limit itself to a single fighter or bomber, NASA should not limit itself to a single booster to support manned space flight.

First, examine the costs of the pursuit of the Earth orbit rendezvous (EOR) strategy that relied on medium-lift rockets as initially studied for the Apollo program. EOR was considered for a long time as the best means of moving US astronauts to the moon and a whole series of Saturn launchers were designed to fulfill the needs of the various methods studied to get to the Moon.

EOR called for multiple launches and assembly of the spacecraft in orbit. One method required the use of no less than five Saturn C-3 boosters. The Saturn C-3 was about midway between the capability of a Saturn 1B and a Saturn 5 with about 8.9 million newtons of thrust and the ability to lift 13,600 kilograms to low Earth orbit (LEO) at an estimated cost of nearly $80 million in 2004 dollars. As plans advanced, the Saturn C-4 would likely have been used, requiring only two launches at a cost of $105 million each.

Of all the vehicles considered, only the Saturn C-1 (later named the Saturn 1 and modified into the Saturn 1B) and the C-5 (Saturn 5) were built. The Saturn 1B was capable of lifting the 30,329 kg Apollo CSM or the 14,696 kg Lunar Module, but could not send these vehicles to the Moon.

Five Saturn C-3 launches would cost nearly $400 million in present-day dollars. On the surface, this is a fraction of the $2.8-billion price of a Saturn 5 launch. However, given the cost estimate for the Saturn C-3 was for only for a conceptional vehicle, the actual launch cost would have probably been closer to the Saturn 1B’s $700 million (in 2004 dollars) for a total of around $3.5 billion for five launches.

This flyaway cost does not count the need to launch many flights to rehearse and perfect the EOR procedures, the cost of developing the infrastructure to have such a rapid-fire launch capability, nor the risk and time involved in assembling the spacecraft in orbit as compared to the comparative ease of assembling a Saturn 5 on the ground and launching it all at once.

The most interesting problem would have been refueling the booster in orbit. While it was not thought to be particularly difficult, being analogous to air-to-air refueling between aircraft, it is probably significant that this method has not yet been used for any manned or unmanned program on an interplanetary mission. It does appear that the Soviet Union was planning to use this approach for its own Moon program, but it met with failure for a variety of reasons.

Present-day applications

How much would it cost to fly the Apollo spacecraft to the moon using the Delta 4 Heavy booster? The Delta 4 Heavy is capable of carrying 25,800 kg to a185 km orbit at an estimated cost $190 million per launch. It is comparable to a Saturn 1B and could carry the individual components of Apollo to LEO. However, like the Saturn 1B, it cannot launch the entire Apollo spacecraft in one lift and lacks an injection stage that can send the craft on a trans-lunar trajectory. It would therefore be necessary to use three to five launches to configure the spacecraft and refuel the booster resulting in a cost of at least $510-680 million and to develop a new injection stage. This cost is greater than the cost of a shuttle launch.

It is likely that political and public support of such an extended program, with a high developmental cost, entailing significant risk and complexity due to orbital assembly would be at least as difficult to sustain as the commitment to a heavy-lift booster. Remember, everything used in orbit to assemble a spacecraft will have to be launched from the Earth. Is it really more effective to do this than to build and launch spacecraft from Earth?

A shuttle-derived launch vehicle, as proposed in 1992 by Martin Marietta (now Lockheed Martin), together with associated cost for the development of manned spacecraft was possible at an estimated cost of $20-30 billion. (The estimated 2004 cost would be about $35-40 billion dollars; however, even if the cost doubled, as was the case with Gemini and Apollo, this is still not prohibitive.). Prior to this, there was serious consideration given to the development of the Shuttle C that would have had near the capability of the Saturn 5. At the time, however, there was no real need for such a vehicle.

Martin proposed stripping the shuttle off the stack, making a four-engine pod (perhaps recoverable, perhaps not), strengthening the external tank into a load bearing structure, and topping it off with an injection stage, based on the Space Shuttle Main Engine (SSME), and a spacecraft to be delivered directly to Mars weighing an awesome 121,000-135,000 kg. While there are undoubtedly significant engineering challenges to this approach, this proposal represents an effective approach to fielding a new booster, since the individual components are already man-rated and there is little new technology to invent, and would build on and further amortize the shuttle’s development cost.

To execute Martin’s proposed manned Mars program utilizing this hardware would require the launch of three vehicles every two years—scarcely a budget buster—and well within our demonstrated capability. Cost data for various boosters is summarized in the table below. While none approach the inflated cost of launching a Saturn 5, neither are they significantly less expensive than the shuttle when incorporating the need for multiple boosters in support of a single mission. (I have assumed the same launch costs for the Space Shuttle, Shuttle C, and Ares.)

A rational program of modernizing the booster would also spread development costs over a longer period, as well as bringing on planned increases in payload capacity and safety:

• First replace the 4 SSMEs in the engine pod with 3 RS-68s used in the Delta 4. The RS-68 is a simpler engine with 50% more thrust than the SSME, and by the time it would be incorporated into the booster, would be proven and easier to man-rate, further reducing expenses as well as providing common components between the heavy and medium lift vehicles.
• Second, convert the stack into an inline booster configuration using three or more RS-68s. Industry has also made several studies of similar boosters with a variety of engine combinations and payload capabilities that never came to fruition.
• Third, replace the Solid Rocket Boosters with a liquid booster of similar capability (requiring a auxiliary booster, up to four Delta 4 Common Core Boosters, or a rather more esoteric vehicle such as the Starbooster.) These could be made reusable.
• Finally, if needed, replace the conventional injection stage with a nuclear rocket stage, if events determine we need more payload capacity for further development of bases.

The use of medium-lift vehicles is dependent upon the ability to assemble large complex structures in space (demonstrated with the ongoing assembly of the International Space Station), the ability to demonstrate a rapid launch capability (not demonstrated since Gemini 6 and 7 and the first Skylab missions), and the ability to refuel a spacecraft in orbit (not demonstrated), and sending the entire assemblage off to its destination, at a reasonable cost.

This examination shows there is no significant cost savings by pursuing the use of numbers of medium-lift vehicles when compared to the development of a new, shuttle-derived heavy lift booster. The development of such a heavy-lift booster supports the President’s space vision by providing the capability of lofting heavy payloads to the Moon in support of the construction of a lunar base as well as providing the capability to conduct other missions. I believe the development of a heavy booster in conjunction with the appropriate use of medium-lift boosters and modular spacecraft represents the most effective strategy for the US manned space program.

Michael K. Robel a retired Army officer currently working for Northrop Grumman Information Technology as an operations researcher, developing the next generation of Army battle simulations.

Home