Habitat and Logistic Support Requirements for the Initiation of a Space Manufacturing Enterprise


A detailed scenario for the initiation of a space manufacturing enterprise using lunar materials to construct solar power satellites (SPS) has been developed, with particular attention to habitat design and logistic support requirements. If SPS's can be constructed exclusively from lunar materials, the entire enterprise can be initiated in a 7-year period of launch activity (beginning as early as 1985) using the Space Shuttle and a low-cost, Shuttle-derived heavy lift vehicle. If additional chemical feedstocks must be imported from Earth in significant quantities, it may be necessary to bring the next-generation launch vehicle (single-stage-to-orbit) into operation by 1991. The scenario presented features use of the mass-driver reaction engine (MDRE) for orbit-to-orbit transfer of cargos and makes extensive use of the expendable Shuttle external propellant tanks.


During the last 3 years, a number of studies have shown the technical feasibility of manufacturing solar power satellites (SPS) at a space manufacturing facility (SMF) using lunar materials during the next two decades (refs. 1-5). These and related studies have provided conceptual designs and analyses for large-scale habitats in the Earth-Moon system, for chemical and metallurgical extraction and processing from lunar materials, and for a mass driver, an electromagnetic catapult to be installed on the lunar surface - to launch raw materials into space at rates up to 600,000 metric tons per year (refs. 6-10). In addition, a bench-top prototype mass driver has been designed, built, tested, and publicly demonstrated (ref.11).

Here we report on a study of the logistical feasibility of starting such a space manufacturing enterprise within the constraints of the Space Shuttle and of a simple Shuttle-derived heavy lift vehicle (SD/HLV) of higher payload capacity during the period 1985 through 1991. In addition, we have investigated the logistic support required to sustain such an enterprise in a steady state thereafter. This scenario and the dates given above are purely hypothetical, serving principally to indicate that such an enterprise need not be relegated to a distant future and to provide a detailed baseline for trade-off studies.

A key element in the system is the use of a mass-driver reaction engine (MDRE) as an upper stage for the Space Shuttle to carry large payloads from low Earth orbit (LEO) to high Earth orbits or to low lunar orbit (LLO) (ref. 12). Since the MDRE can accept any solid or liquid substance as a propellant, it can be optimized to use powder fabricated from the expendable external liquid propellant tank of the Space Shuttle. Thus optimized, an MDRE can deliver to lunar orbit the payloads from nearly 30 Shuttle launches in a round trip of 6 months duration. The mature Shuttle system is expected to provide 30 launches every 6 months, beginning about 1985. Each MDRE payload to lunar orbit can thus total 850 metric tons, expending 1050 tons of reaction mass fabricated from 30 external tanks.

The entire space manufacturing system thus consists of (1) a lunar mining and launching base to deliver raw materials into space; (2) a catcher in space to collect the materials launched from the lunar surface; (3) a space manufacturing facility that extracts useful materials from the lunar ores and fabricates solar power satellites and other objects required to support and expand the system; (4) a staging base in LEO that provides an interface between the Earth-launch vehicles and the interorbital vehicles; (5) habitats at each of these locations; and (6) logistic support for the entire operation.

Logistic support includes (1) Earth-launch vehicles (beginning with the Shuttle); (2) a fleet of MDRE's to transfer cargos between various space facilities; (3) chemically propelled orbital transfer vehicles (OTV) to transport people between space facilities; (4) chemically propelled lunar transfer vehicles (LTV) to soft-land cargos on the lunar surface and to transport people between lunar orbit, the lunar surface, and the catcher; and (5) passenger modules to carry large numbers of passengers in the Shuttle and on the OTV. In addition, the enterprise must have a large infrastructure on Earth to provide administration and control, recruitment and training of personnel, coordination of research and development activities, procurement of material and services, and marketing of solar power satellites and maintenance services for them.

A fairly detailed step-by-step consideration of how the entire operation can be initiated has been carried out for the first 8 years of launch activities. The result of this effort has been to show the logistical feasibility of initiating a space manufacturing enterprise within the constraints of the Space Shuttle, a Shuttle-derived heavy lift vehicle, and the MDRE during the period 1985 through 1990, with advanced launch vehicles brought into operational use by 1991; to detail personnel and hardware requirements during this initial period; to determine the habitat requirements for each space facility in the system; to highlight certain key problems for intensive research and development which would facilitate or improve the system; and to provide cost estimates for research and development, for materiel and services procurement, for crew selection, training, and employment, for administration and other overhead costs, and for transportation.

We are grateful for the assistance of Edward H. Bock, Fred Lambrou, Jr., and Michael Simon. Joe W. Streetman of General Dynamics/Convair Division provided extensive assistance in the analysis of chemically propelled vehicles and their missions.


The scenario for the initiation of a space manufacturing enterprise has been developed around three key milestones:

  1. Establishment of the lunar base and operation of the lunar mass driver at a launch rate of 30,000 metric tons/yr.
  2. Establishment of an initial space manufacturing facility with a processing capacity of 30,000 metric tons/yr.
  3. Expansion of the entire system to a capacity of 600,000 metric tons/yr with a work force of 3000 at the SMF producing 2.4 solar power satellites per year, each with a capacity of 10 GW.

To minimize the total time and total costs incurred before investment return, the second milestone is phased to coincide with the first delivery of lunar materials at the SMF. Thereafter, the entire system is bootstrapped to higher levels by diverting some of the lunar materials delivered to the SMF into reaction mass for the MDRE fleet, permitting more rapid expansion of the system than would be possible if the only source of reaction mass were the Shuttle external tanks.

The scenario was developed by working backward from each of the major milestones described above. The total masses required to achieve the first two milestones are sufficiently large to justify establishment of an inhabited station at LEO where MDRE's can be assembled, payloads from Earth-launch vehicles can be reconfigured, and large space structures such as habitats can be built. The establishment of the LEO station is thus a "zeroth" milestone.

The basic working assumptions used to develop the scenario were as follows:

  1. The entire enterprise is self-contained, that is, no assumption is made that an essential piece of hardware (such as the LEO station) will already be in place as a result of some other space program. Only the basic Space Shuttle system is given.
  2. Once launch operations begin, the entire program is to be carried out as rapidly as possible within the constraints of available transportation systems.
  3. Each Shuttle or SD/HLV brings its expendable external tank into orbit. To the maximum extent possible, these tanks are to be utilized for MDRE propellant or for pressure hulls.
  4. Initial masses at each facility are to' be kept as small as reasonably possible.
  5. Research, design, development, and procurement are to be paced by deployment schedules, even if this results in higher technical risk due to high levels of concurrency.
To describe the scenario and its generation, the following sections will describe the initial operations at the Moon, at LEO, and at the SMF.


A site for the initial lunar base must naturally be selected before the first crew arrives at lunar orbit to begin soft-landing equipment and supplies. At the present time, the likely site for the initial base is in the vicinity of the lunar equator and long. 33.1o E. This location permits launch of lunar materials via a family of achromatic trajectories that arrive with minimum dispersion at a moving spot in space in the neighborhood of the second Lagrange libration point L2 some 63,000 km behind the Moon on the Earth-Moon axis (ref. 13). If precise site selection has not been accomplished by prior lunar orbiter missions, the first MDRE payload arriving at lunar orbit (nearly I year before arrival of the first lunar crew) can include one or more lunar orbiters totaling several tons. These could provide detailed spectrographic and topographic mapping of candidate sites and high-resolution maps (down to 1 m or less) of the final site.

The initial lunar crew, arriving at lunar orbit aboard a chemical OTV at the same time as the last MDRE payload of initial lunar base supplies and equipment arrives, begins a series of LTV flights down to the main lunar base shortly after sunrise at the main lunar base on 4 March 1987. (Dates for sunrise and sunset at the lunar base (lunar long. 33.1o E, lat. 0o were computed from ref. 14.) During this period of a few days to a week, the crew continues to live in the OTV passenger module until the lunar base is habitable. The first payload to be landed on the Moon includes the photovoltaic arrays and energy storage systems needed to provide power for the lunar habitats throughout the lunar day and night. The entire energy system must be deployed rapidly to accumulate sufficient energy to support the crew through the first lunar night. At the present time, fused silica flywheels appear to offer the least massive energy storage system, storing about 0.87 kWhr/kg of flywheel, but trade-off studies- against alternatives should be performed. The next task at the main lunar base is to dig trenches for the habitats and to bury them under several meters of lunar soil to provide radiation shielding and thermal insulation. Meanwhile, a smaller contingent of the lunar crew soft-lands equipment and supplies at a second site about 150 km downrange where the velocities of the packages of lunar material launched by the mass driver will be fine-tuned in ballistic flight by electrostatic means (ref. 10). A smaller habitat is established there in the same manner.

During the next two lunar days, the crew surveys both sites, installs the mass driver and the electrostatic trajectory correction equipment, surveys an initial mining site near the mass driver, and tests the mass driver system. The mass driver would be preassembled at LEO into 30- or 40-m-long sections, thereby minimizing installation time. Initial mining activities consist of one small bulldozer and two small trucks working one 8-hr shift every 24 hr during the lunar day, providing 30,000 tons of sieved material per year for launch by the mass driver. The downrange crew returns to the main lunar base for the lunar night, using a fleet (for redundancy) of long range surface vehicles.

Analogies with terrestrial excavation and construction projects suggest that this work schedule can be accomplished by an initial crew of 24. The mass budget for the habitats, energy systems, consumables (food; gases and water to make up for losses in recycling through leaks and by airlock inefficiencies), lunar surface vehicles, and mass driver system components is summarized in table 1. (Mass estimates for habitats, energy systems, etc., are discussed in a later section.)

ComponentMass budget,
metric tons
Main lunar base (24 persons)
Habitat (two LH2 tanks furnished)72
Energy system (photovoltaic arrays, flywheels)128
Surface vehicles and tools25
Spare parts5
Consumables (3kg/day/person)25
Mass driver and power supply (ref.12)625
Downrange station (12 persons, part-time occupancy)
Habitat (one LH2 tank, furnished)40
Energy system45
Surface vehicles and tools12
Spare parts2
Consumables (lunar day only)7
Mass driver trajectory correction system (ref.12)50
Crew and personal baggage (300 kg/person7.2

Soft-landing this equipment on the lunar surface requires a substantial amount of chemical propellants in lunar orbit. No studies have been made to optimize the capacity of a lunar transfer vehicle (LTV); we have arbitrarily assumed a large cargo and personnel LTV as described in table 2. Eleven round trips are needed for the main base and two round trips for the downrange station; allowing three more trips between LLO and the surface for unforeseen contingencies, 896 tons of propellant are needed to start lunar surface operations.

Large lunar transfer vehicle
Stage mass7 tonsa
Propellant mass (LLO to surface and return)56 tons
Stage mass fraction 0.89
Payload down to lunar surface81 tons
Cabin (pressurizable for crew of two)5 tons
Flight time (LLO to surface or return)0.93 hr
N.B. In normal operations, the cabin would be pressurized. All crew
members and passengers would wear EVA suits. Passengers in excess
of the cabin capacity would ride on top of the LTV in lightweight seats
enclosed by a lightweight transparent canopy to protect the EVA suits
from lunar debris thrown up from the rocket exhaust.
Small lunar transfer vehicle
Stage mass2 tons
Propellant mass (LLO to catcher and return)3.5 tons
Propellant mass (LLO to surface and return)8.9 tons
Cabin (same as for large LTV)5 tons
Flight time (LLO to catcher or return)40.1 hr
N.B. the flight controls of both the large and small LTV should be
designed to "feel" the same, permitting concurrent proficiency
in both.
Orbital transfer vehicle
Propellant mass (LEO to LLO or return)55 tons
Stage mass fraction0.89
Propellant mass (LEO to SMF or return)49 tons
Passenger module (up to 30 passengers for 10 days)23.4 tons
Flight time (LEO to LLO or return)122.4 hr
Flight time (LEO to SMF or return)9.8 hr
N.B. Propellant requirements have been computed on the basis of liquid
oxygen and liquid hydrogen at a mixture ratio of 6:1. Specific impulse of
462 sec was assumed for major burns, 400 for midcourse maneuvers,
and 220 for attitude control. (These specific impulses are slightly conserva-
tive, allowing for some propellant losses by boiloff.) A common engine can
be used for all of these vehicles, except that the engine throttling system
required for the LTV's is not installed on the OTV engines.

aAl1 in metric tons.

The packages of lunar materials launched by the mass driver must be collected in space with high efficiency. As described earlier, the dispersion of packages arriving at the catcher can be minimized by selecting a suitable launch site and catcher orbit. The catcher orbit thus selected is approximately a circle of radius 10,000 km around L2 in a plane normal to the Earth-Moon axis. To track this orbit while compensating for the momentum imparted to the catcher by arriving packages, a small fraction of the captured lunar materials can be used as reaction mass. Either a mass driver or a rotary pellet launcher could be used for this purpose (ref. 2), but engineering trade-offs remain to be done. The catcher would thus be designed to operate automatically under normal conditions, with remote monitoring and capabilities for remote override and control.

Nonetheless, maintenance of the catcher and transfer of accumulated lunar materials to a MDRE for transport to the SMF requires human operations on the catcher for a few days at a time. The catcher thus includes a habitat that can accommodate small crews (up to about 12). Crews for this work travel to the catcher from the lunar surface once a lunar day (13 lunar days per year) for routine maintenance when mass-driver operations cease for the lunar night at the main lunar base. A small LTV can be used for this purpose, with characteristics as shown in table 2. The LTV refuels in lunar orbit both on ascent and on descent. Allowing two additional contingency flights per year, the propellants required for this purpose amount to 186 tons/yr, supplied from LEO initially.

One additional system requires chemical propellants in lunar orbit, namely, the OTV for personnel transport between LEO and LLO. Given the advantages of using the MDRE for cargo transportation beyond LEO, the OTV can minimize its propellant requirements by refueling in lunar orbit for the return trip to LEO. One such flight must be provided at the end of a 1-year tour of duty in space for the lunar crew to return to Earth. For emergencies and unforeseen contingencies, we have allowed two additional flights per year. The OTV specifications assumed are given in table 2. For simplicity, we have assumed a single-stage vehicle; no trade-off has been done against a two-stage vehicle with refueling at LEO and at LLO. Total propellants at lunar orbit amount to 165 tons for the first year.

The catcher is delivered to lunar orbit along with equipment and supplies for the lunar surface aboard the MDRE. After delivery of the final payload to lunar orbit, the MDRE carries the catcher out to the vicinity of L2 and waits there with the catcher. Initial launch begins at sunrise of the fourth lunar day (1 June 1987) at a rate of 30,000 tons/yr. At sunset on the eighth lunar day (11 October 1987), the MDRE departs for the SMF site with a cargo of 11,700 tons of lunar materials. The MDRE also carries along one empty propellant storage depot whose contents have been expended in soft-landing the lunar base equipment and supplies. This depot will be refilled at LEO.

Since the chemical propellants must be stored in space for periods of many months, suitable long-term cryogenic depots must be provided. A Shuttle external tank set has a capacity of 703 tons of liquid oxygen and liquid hydrogen propellants. By adding multilayered insulation, reliquefication systems to convert hydrogen and oxygen vapors back into liquid form, and equipment inside the tanks to permit extraction of the propellants from these tanks in zero-gravity conditions, these tank sets can be converted into long-term storage depots. The extra insulation and equipment totals about 4.15 tons for each such depot.

The total mass to be delivered to lunar orbit to initiate lunar operations (as just described) is 2700 tons (table 3). This includes 125 tons of Shuttle external tankage, so that only 2575 tons must be delivered to LEO in the Shuttle cargo bay. Since the MDRE can obtain the reaction mass for the return leg of its last trip from lunar materials at the catcher, the propellant mass required for this leg can be used on the outbound leg for about 150 tons of additional cargo. The total mass of 2700 tons can thus be accommodated on three MDRE flights with payloads of 850, 850, and 1000 tons, respectively. Once the zeroth milestone has been reached, 88.8 Shuttle payloads of 29 tons each are required to assemble the initial lunar orbit payloads at LEO. The liquid hydrogen tanks from four Shuttle external tank sets are used to build habitats; two complete external tank sets are converted into propellant depots; and 85.7 external tank sets (including unused portions of those tank sets used for habitat construction) are ground up for MDRE reaction mass.

The lunar crew must also be transported to lunar orbit to arrive simultaneously with the third MDRE flight, approximately 5 months after departure of the MDRE from LEO. During the intervening time, a chemical propellant depot must be prepared at LEO, chemical propellants electrolyzed and liquefied from water brought up aboard the Shuttle, and two chemical orbital transfer vehicles brought up from Earth. One OTV is sufficient to deliver the initial lunar crew of 24 from LEO to LLO aboard a passenger module that can fit inside the Shuttle cargo bay, but a second OTV is essential for rescue capability. Propellants required for the first year of OTV operations (matching those provided at lunar orbit) total 165 tons. The total mass required in LEO for crew delivery to the lunar base is also shown in table 3.

ComponentMass budget
metric tons
Mass in lunar orbit
Main lunar base880
Downrange station156
Large LTV 896
Small LTV 186
OTV 165
Propellant storage
depots (2)
Shuttle tank set 33.45
Conversion materials 4.15
Conversion ensemble
Habitat 45
Energy system 20
Consumables 7
Catcher 168
Large lunar transfer vehicles (2)24
Small lunar vehicles (2)14
General contingency63.8
Total payloads for MDRE2700
Shuttle external tanks-125
Total payloads for shuttle2575
Mass at LEO
Orbital transfer vehicles (2)13
Propellant depot
conversion materials
Total payloads for Shuttle182


Before launch of supplies and equipment for the initial lunar operations can begin, facilities must be established in LEO to provide a staging base for assembly of lunar orbit payloads. During the year and a half of Shuttle launches required to transport the lunar supplies into orbit, the crew at the LEO station must refurbish four liquid hydrogen fuel tanks into habitats for the lunar surface bases and for the catcher; convert two complete external tank sets into long-term propellant storage depots; grind 85.7 external tanks into powder for MDRE reaction mass; electrolyze nearly 1250 tons of water; liquefy the resulting hydrogen and oxygen; assemble the lunar mass driver and its enclosing tunnel into sections 30 or 40 m long; assemble the catcher ensemble; and bolt all of these payloads together into predetermined configurations designed for ease of transfer to the LTV, with due consideration for weight and balance limits both of the MDRE and the LTV. Assembly and checkout of the MDRE are also essential tasks for the LEO crew during this period.

Some of this work involves assembly of prefabricated components; most of it requires tending and maintaining automated or semi-automated equipment (such as the electrolysis and cryogenics units that produce chemical propellants). Refurbishing one liquid hydrogen tank for a habitat (as seen below in the discussion of habitat design) is comparable to assembling three prefabricated houses of 1800 ft2 each, a task of a few weeks for a few skilled workers. Such considerations suggest that a crew of 24 is sufficient during the 18-month period required to assemble the initial lunar orbit payloads, with an average productivity requirement of 460 kg/day/worker.

As shown below, the work force at LEO must be expanded to about 48 sometime after departure of the third MDRE payload for lunar orbit. Four liquid hydrogen tanks can adequately house this size crew; 6 tanks would be ample, leaving room for further expansion to 84. We have adopted a six-tank configuration. Construction of the LEO station begins by refurbishing a single tank, working out of the Shuttle at first. Five workers can ride aboard the Shuttle as passengers on 6- to 8-day flights, with minimal decrease in cargo bay payload capacity. Using the LEO station's own solar panels to supplement the Shuttle power supply, the Shuttle would provide life support initially until habitability is established for at least one liquid hydrogen tank.

Interior fittings such as antislosh baffles must be stripped out of the tank, and air-conditioning ducts, prefabricated floor and wall panels, and other interior fittings must be moved inside. This can be done through a special hatch installed before launch or cut in the tank in LEO for later installation of an airlock. Additional holes must be cut for attachment of the environmental control/life-support system (EC/LSS) pods; these pods and the airlock welded to the hull; photovoltaic arrays and radiators mounted; and the entire hull pressurized. Installation of floors, walls, air-conditioning ducts, plumbing, electrical wiring, elevators, and furnishings can then proceed in a shirtsleeve environment, with structural attachments made to the interior ribs of the tank. Once minimum habitability for 5 has been reached, the 5 workers aboard the current Shuttle flight can remain in orbit, the remainder of the initial LEO crew of 24 joining them during the next four Shuttle flights. The hydrogen tank can then be separated at leisure from the rest of the external tank set which is ground up into MDRE reaction mass.

During the following weeks, refurbishing of the first tank is completed, and work can begin on the other five tanks for the LEO station, beginning, in these cases, by separating the hydrogen tank from the rest of the external tank set. The hydrogen tanks are connected by a tunnel, a docking hub, and cables, and rotated to provide 1 g in the hydrogen tanks themselves. Completion of this and subsequent habitats would be facilitated by use of a "space caisson," a collarlike structure that encircles a Shuttle external tank to provide a shirtsleeve environment for work outside the pressurized tank simultaneously with work in progress inside. The caisson has inflatable cuffs to form an airtight seal with the exterior of the tank and contains arc-welding equipment and assorted machines and construction tools for cutting holes through the tank wall to install view ports, airlocks, and other apertures.

Other equipment to be launched to the LEO facility includes the pelletizer for chewing up external tanks into MDRE propellant; the electrolysis and - liquefication systems for making liquid oxygen and hydrogen for chemical propellants; consumables for the crew for the first year; and a set of trusses and booms to provide temporary tiedown storage for Shuttle cargo items and external tanks and to serve as a basic construction yoke. In addition, the first MDRE must be brought to LEO before accumulation of lunar cargos can begin in earnest.

Table 4 summarizes the estimated mass budget required in LEO to attain the zeroth milestone.

ComponentMass budget
metric tons
Initial station (24 persons)
Habitat refurbishing142
Energy system16
Electrolysis and
liquefaction systems
Space caisson25
Cargo tiedown,
construction yoke
Spare parts5
Consumables (one year supply
at 3 kg/person/day)
General contingency15
Mass driver reaction
engine (ref.12)
Total payload for Shuttle464
Expansion to 48 persons
Habitat refurbishing100
Energy system11
Consumables (one
year supply for 24)
Total payload for Shuttle138

The total mass of 464 tons can be accommodated in 16 full Shuttle payloads. Six liquid hydrogen tanks and one complete external tank set must be reserved for constructing the LEO station. The mass required to complete the interior refurbishing of tanks for expansion to 48 persons is also shown in table 4, requiring 4.8 additional Shuttle payloads, but no additional external tanks.

The schedule to attain the zeroth and first milestones is as follows:

  1. First Shuttle launch for this enterprise occurs early in January 1985.
  2. Shuttle launch No. 16 completes delivery of equipment and supplies for the LEO station and for the first MDRE late in March 1985.
  3. Shuttle launch No. 106 completes delivery to LEO of supplies and equipment to be delivered to lunar orbit, with departure of the third MDRE flight to lunar orbit in October 1986. (One Shuttle flight in January 1986 has been used to resupply consumables for the LEO station.)
  4. The third MDRE flight to lunar orbit arrives there early in March 1987, simultaneously with an OTV flight carrying the initial lunar crew.
  5. The lunar base begins routine launch of materials into space at the rate of 30,000 metric tons/yr shortly after sunrise at the lunar base on 1 June 1987, completing the first milestone.

Departure of the third MDRE flight from the LEO station does not, of course, imply shutdown of LEO operations. Accumulation of supplies and equipment for expansion of the lunar base, for assembly of the initial SMF, and for expansion of the MDRE fleet begin in mid-November 1986, after nearly five Shuttle payloads have been used to expand the LEO station.

We suggested above that the MDRE would remain at the catcher for four lunar days after launch of lunar materials begins in June 1987 to permit accumulation of a reasonable cargo of lunar materials for delivery to the SMF. Departure to the SMF would then occur on 11 October 1987, with arrival at the SMF site (discussed below) in February 1988. The waiting period is arbitrary since no optimization studies have been made. A shorter waiting period would permit earlier SMF operation with a smaller supply of raw materials; a longer period would postpone initial SMF operation still further by prolonging the time enroute for the MDRE because of its increased payload. The present choice seems reasonable but nonetheless arbitrary.

If the SMF is to begin operation in February 1988, the last cargo of equipment and supplies for the initial SMF must depart LEO aboard an MDRE in October 1987. Thus everything necessary for the initial SMF must be delivered to LEO and assembled in about 11 months (mid-November 1986 to October 1987), with several Shuttle flights reserved for resupply of consumables for LEO, for delivery of the first lunar crew in an OTV passenger module, and for additional MDRE's. As shown below, some 1980 metric tons of payload must be delivered to LEO for the initial SMF alone, equivalent to more than 14 months of Shuttle launch operations. It is thus appropriate to upgrade the surface-to-LEO transportation system at this stage of operations.

Space Shuttle
Configuration: airplane-like Shuttle orbiter
contains main engines. Cabin can accommo-
date flight crew of two plus up to five mis-
sion and payload specialists. Cargo bay can
accomcodate cargos within a cylindrical
volume about 5 m in diameter by nearly
20 m in length. Orbiter is launched verti-
cally, attached to side of a large (expend-
able) external tank carrying liquid oxygen
and liquid hydrogen for Space Shuttle
main engines. Two solid-propellant boosters
(recoverable) are strapped alongside the
external tank
Estimated price per launch: $20 million (1975
Payload: 29 metric tons
Launch price per kilogram: $690
Shuttle-derived Heavy LIft Vehicle
(SD/HLV) (ref.15)
Configuration: Shuttle orbiter replaced by
cargo enclosed in disposable aerodynamic
shroud atop Space Shuttle main engines
enclosed in reentry capsule for reuse. Same
external liquid propellant tank as used in
basic Shuttle. Four solid propellant boosters
used instead of two.
Development time: approximately 5 years,
beginning from operational Shuttle
Development cost: $1.3 billion (1975 dollars)
Estimated price per launch: $19.6 million
(1975 dollars)
Launch price per kilogram: $173
Single-Stage-to-Orbit Vehicle
(SSTO) (ref.16)
Configuration: delta-winged vehicle with
dual-mode engines. Mixed cargo and passen-
ger capability, with vertical launch/
horizontal landing. Overall length in launch
configuration: 90 m. Anterior payload
shroud collapsible for reentry and landing,
with overall length in touchdown configura-
tion of 70 m. Completely reusable system.
Development time: about 10 years, requiring
some advanced technologies
Development cost: $7 billion (1975 dollars)
Estimated price per launch: $3.6 million (1975
Payload: 227 metric tons
Launch price per kilogram: $15.86

A relatively inexpensive upgrading of the Shuttle system can be achieved with a 5-year development program. Some of the basic characteristics of such a Shuttle-derived heavy lift vehicle (SD/HLV) are shown in table 5, along with similar data for the Shuttle itself and for a more advanced reusable single-stage-to-orbit (SSTO) launch vehicle. Assuming development of the SD/HLV begins between 1980 and 1982 as part of the total R&D effort for the space manufacturing enterprise, it is reasonable to expect that a fleet of four SD/HLV's could be operational beginning in January 1987, with 80 launches/yr in parallel with 60 Shuttle flights per year. With this combination of launch vehicles, the mass delivered annually to LEO would increase from 1,740 to 10,780 tons. This sixfold increase would not, however, require a sixfold expansion of the LEO crew, for two reasons: first, some of the assembly work for the SMF can be postponed until the SMF crew arrives at the SMF site; second, the larger payload capacity of the SD/HLV will reduce the assembly work required per ton of payloads delivered to LEO. Should further study show that a crew of 48 is insufficient, the LEO station we have provided can accommodate a total of 84 workers, with some adjustments required in payload scheduling to bring up more furnishings and more consumables per year.

With this expansion in Earth launch capabilities, the number of expendable Shuttle external tanks also increases from 60 tanks/yr to 140 tanks, providing propellants for delivery of somewhat more than 2.4 times as much mass per year to high orbit for reasons that will emerge below. The MDRE fleet thus must be expanded in parallel with assembly of the initial SMF and with expansion of the lunar base. The first priority, however, is to provide the initial SMF.


Although the objective of the entire space manufacturing enterprise is to produce solar power satellites from nonterrestrial materials, as yet no one has developed a credible SPS design optimized for nonterrestrial materials. In this sense, the end product of the SMF is poorly defined so that the designs of the chemical processing plant and of the fabrication plant are subject to major uncertainties. As a point of departure for logistical planning purposes, we have adopted a conceptual design for the SMF factory developed during the 1976 NASA Ames/OAST Summer study on space manufacturing (refs. 7, 8).

In that design, it was assumed that the lunar materials arriving at the SMF would be completely unselected (except for choices between mare and highland soils). Present thinking suggests that it would be much more efficient to perform extensive beneficiation operations on the lunar surface before launch by the mass driver (ref. 17). The initial lunar operations would not include beneficiation; equipment for beneficiation would be added during expansion of the lunar base to full capacity (as discussed below). Raw materials arriving at the SMF would thus consist of separate cargos of (1) unselected (but sieved) lunar materials for radiation shielding or for MDRE refueling, (2) plagioclase concentrates from lunar highland soils, and (3) ilmenite concentrates from mare soils (ref. 17). The chemical processing plant required to extract oxygen, silicon, glass, iron, aluminum, and titanium from these concentrates is far simpler than that described in references refs. 7 and 8; nonetheless, for conservatism, we have used the mass estimates given there. For a processing plant with an annual capacity of 600,000 metric tons, the chemical processing plant totals 2770 tons; the energy system (photovoltaic arrays and radiators), 3852 tons; and the fabrication plant, 5000 to 6800 tons. The entire factory then totals 11,600 to 13,400 tons, exclusive of radiation shielding.

The first lunar materials delivered to the SMF are unselected soil. Approximately 10,000 tons of this material are needed for radiation shielding of the initial habitat, and an additional 4500 tons are required for reaction mass for the MDRE fleet to expand the lunar base and the SMF during the year following initial operation of the SMF. The initial factory includes equipment to extract oxygen from the unselected lunar soil by reduction with hydrogen, with water vapor collected in a cold trap. The hydrogen is recovered by electrolysis, and the oxygen can be used either for life support or liquefied for use in the OTV and the LTV. The factory also includes equipment to sinter the lunar soil into several shapes of large slabs for radiation shielding. Most of the factory mass is taken up by chemical processing equipment and fabrication machinery for the production of structural sheets, plates, and beams and of glass to permit rapid expansion of the SMF factory and habitat as soon as concentrated ores begin to arrive. (This equipment may possibly be used, at much lower efficiencies, to extract metals and glass from the unselected soil provided in the early deliveries.)

To bootstrap the entire enterprise from a minimum initial mass, as much as possible of the final factory and habitat masses should themselves be manufactured at the SMF. Portions of the pressure hulls, photovoltaic arrays, radiators, and massive foundry components could be fabricated readily with the same equipment used to build the SPS's. As a preliminary (and highly uncertain) working estimate, we have assumed that 9 percent of the mass of any finished products manufactured by the SMF must be imported from Earth in the form of complicated components or of raw materials to be used, for example, in alloying structural materials, doping semiconductors, or as special optical coatings (refs. 7, 8 estimate about 9 percent imports for the manufacture of SPS's). Initially, the factory will have only a limited capacity to contribute to its own expansion; as the factory approaches full size over a 3-year period, however, we assume it will approach a 91-percent contribution toward the mass required to complete the SMF factory and habitat. We thus estimate that an additional 4670 tons of factory components must be imported from Earth over the 3-year growth period, with 6900 tons provided by the factory itself from lunar materials. Similar considerations apply to habitat interior furnishings and airlocks. Shuttle external tanks (as shown below) continue to be available for use as pressure hulls until the end of 1990, but detailed trade-offs remain to be made to determine whether the optimum use of these tanks is as reaction mass for the MDRE fleet or as pressure hulls at the SMF.

The estimated masses for the initial SMF and for its expansion over the subsequent 3 years to reach full capacity (600,000 metric tons/hr) are shown in table 6.

ComponentMass budget
metric tons
Initial SMF (30,000 tons/year)
Habitat (14 LH2 tanks,
pressurized and furnished for
crew of 150, expandable to
Energy system for habitat40
Structural cable to support
radiation shielding
Consumables (1 year supply
for 150)
Propellant depot38
Chemical processing plant
(60,000 tons/year)
Fabrication plant
(60,000 tons/year)
Energy systems for factory415
Spare parts20
General contingency50
Total MDRE payload2225
Shuttle external tanks-236.5
Shuttle or SD/HLV payload1988.5
Expansion to full capacity
(600,000 tons/year)
Interior furnishings for
154 more LH2 tanks (or
Environmental control/life
support system (EC/LSS)
units; airlock pumps and
Chemical processing plant
Fabrication plant equipment3000
Miscellaneous components and
special materials
Shuttle or SD/HLV payload7640

An initial crew of 150, including administrative, maintenance, galley, and medical personnel, should be sufficient, with productivity of 100-110kg/hr/person required to process 30,000 tons/yr. A modular habitat design using 14 liquid hydrogen tanks from the Shuttle external tank sets was assumed, with maximum commonality of design with the habitats at LEO and at the catcher. (See the discussion of habitat design below.) Such a configuration readily accommodates 150 persons initially. Completion of interior refurbishing work will allow expansion to 234 to 252 persons, depending on assumptions about the amount of community space required and about sharing of private quarters.

Based on a consideration of the energy and reaction mass needed to transport large masses of lunar soil from the catcher to the SMF and to subsequently transport solar power satellites from the SMF to geosynchronous orbit, a previous study identified the 2:1 resonance orbit as a highly desirable site for the SMF (ref. 13). Additional factors, however, must also be included in optimizing the SMF site, especially in the early stages of a space manufacturing enterprise. Since the MDRE is to be used to deliver cargos from LEO to the initial SMF site, its operating characteristics must be taken into account. A continuous, low-thrust propulsion system (such as the MDRE) is illsuited for orbital transfers between a circular orbit such as the LEO station and a highly eccentric orbit such as the 2:1 resonance orbit. Both time enroute and total propellant mass for the MDRE will be less for a circular orbit SMF site. Optimization of the SMF site (both in the initial stages of the space manufacturing enterprise and in a late, mature stage) will also depend strongly on the ratio of mass that must be imported from Earth to the mass brought from the Moon to produce SPS's, and of these two masses to the mass of the SPS which must be moved to geosynchronous orbit. These factors were not considered in the earlier optimization.

Geosynchronous orbit might seem a natural choice for an SMF; unfortunately, the outer portions of the Van Allen radiation belts are sufficiently intense at geosynchronous distance to create serious problems. Without performing the careful trade-off needed to optimize the system, we have somewhat arbitrarily selected a 2-day circular orbit at 10.6 Earth radii (67,500 km) for the initial SMF site. Should the 2:1 resonance orbit or some other highly eccentric orbit prove to be the optimum site for the SMF in a mature system, the entire SMF can be relocated at leisure, using MDRE "tugs," perhaps 5 or 10 years after, initial operation of the SMF.

On its way to lunar orbit, the MDRE reaches 2-day circular orbit about 4 months after departure from LEO; on the return voyage, it passes through this orbit about 3 weeks before arrival at LEO. An identical MDRE could thus depart from LEO with increased payload and decreased reaction mass, completing a round trip to the SMF in a travel time about 5 weeks shorter than the 6 months required for the round trip to lunar orbit. The propellant requirements for a round trip to lunar orbit to deliver 850 tons amount to 30 external tanks expended more or less uniformly over the duration of the flight. If the total mass at departure from LEO is the same for both missions, the MDRE flight to the SMF site can deliver 1025 tons while expending less than 25 external tanks as reaction mass. Once the MDRE can refuel at SMF (less than 175 tons of lunar materials will suffice to return the unloaded MDRE to LEO), 1200 tons can be delivered to the SMF using fewer than 20 external tanks in the same round-trip time of slightly less than 5 months. Refueling at the SMF can also increase the effective capacity of the MDRE in transporting cargos to lunar orbit, requiring fewer than 20 external tanks at LEO and 175 tons of lunar materials at the SMF of the outbound voyage to deliver 1200 tons at the expense of about 1 week more for the outbound flight. Propellants for the return trip from the catcher to the SMF would, again, be provided from the cargo of lunar materials.

For MDRE flights arriving at the SMF site after February 1988, the above improved performance figures were used. In addition, we have estimated that an MDRE can deliver to LEO a cargo of lunar materials for a reaction mass equivalent to about 30 Shuttle external tanks (1050 tons) in 4 months, expending 850 tons enroute. This procedure is a somewhat inefficient way to supplement the supply of external tanks but is occasionally necessary to avoid excessive bottlenecks in orbital transfer. Detailed study is required to evaluate whether it would be more effective to reoptimize the MDRE for the SD/HLV's higher ratio of payload mass to external tank mass.


The initial lunar base installation was limited in its productive capacity to 1/20 the capacity of the mass driver, both by the limited power available for the mass driver and by the limited mining equipment. Expansion to full capacity (600,000 metric tons launched per year) has been examined in detail by Williams et al. (ref. 17), who estimate that a permanent crew of about 48 will be sufficient to mine, process, and launch lunar materials at full capacity. The initial habitat (two liquid hydrogen tanks from the Shuttle external tank sets) must be replicated. Besides the equipment required for the expanded mining and beneficiation operations, additional photovoltaic arrays must be provided to power the mass driver and the mining and beneficiation equipment. We have added two more items to the list: an oxygen extraction plant to replace oxygen lost from the habitats by leakage, and a small factory to spin fiberglass sacks for the lunar soil to be launched by the mass driver. The total mass requirements are summarized in table 7.

ComponentMass budget
metric tons
Habitat (two LH2 tanks,pressurized
and furnished for crew of 24)
8 mining trucks154
1 bucket-wheel extractor
5ore beneficiation modules385
2 conveyor belt systems96
Maintenance sheds100
Oxygen extraction plant125
Fiberglass sack factory125
General contingency and spare parts75
Photovoltaic arrays1805
Photovoltaic arrays manufac-
tured at the SMF
Two liquid hydrogen tanks2225
Shuttle external tanks-29
Net Shuttle or SD/HLV payloads2271

While most of the mining and beneficiation equipment is very specialized and has thus been assumed to be imported from Earth, more than half the total mass is photovoltaic arrays. As the entire space manufacturing enterprise expands, production of photovoltaic cells at the SMF will grow. We have assumed that about 1225 metric tons (of the 1805 tons total required) can be provided by the SMF, reducing the net payload to be transported to LEO by the Shuttle or the SD/HLV. Giving credit for the liquid hydrogen tanks used to construct the additional habitat facilities, we find that the total equipment to be launched to LEO amounts to 2271 tons. Chemical propellants and propellant storage depots were not included in this figure, but appear elsewhere.

We have not analyzed in detail the relative merits of using lunar oxygen in the lunar transfer vehicles (LTV's) versus continuing to provide all propellants for LTV operations at lunar orbit. Tanking up with hydrogen at the lunar orbit propellant depot and tanking up with oxygen at the lunar surface would provide a major reduction in the total propellant mass required to be delivered to lunar orbit by the MDRE, with hydrogen coming from Earth and oxygen from the SMF beginning in 1990. On the other hand, the payload deliverable by the LTV in this mode of operation would be reduced, so that more LTV flights would be required. We have conservatively assumed that all propellants continue to be provided in lunar orbit.

For the first year of lunar operations, when a crew of 24 was used to establish the main and downrange stations, a total of 32 tons of consumables was provided (25 at the main base; 7 at the downrange station which would be occupied by no more than 12 during the lunar day only). With expansion of the lunar base to 48 people early in 1988, the consumables are increased to 60 tons/yr until oxygen is supplied at the lunar surface, cutting the consumables budget in half. Similarly, provisions for the catcher habitat initially include 7 tons for the first year of operations. After operations at the catcher have become more routine, 5 tons/yr should suffice, with a further reduction to 3 tons/yr when oxygen becomes available at the SMF.

We have assumed that the lunar crews have a 1-year tour of duty in space (i.e., 1 year from launch to LEO until reentry into the Earth's atmosphere). If the lunar base expansion to a crew of 48 is staggered 6 months from the date of rotation of the first 24, the main lunar base need never be unoccupied. A total of three OTV's (with passenger modules large enough for 24 to 30 persons) is sufficient to provide emergency evacuation capability for the entire lunar crew with one of these OTV's parked in LEO for emergency backup. The OTV fleet used to transport crews to and from the SMF provides additional backup. Thus two OTV round trips must be provided for each year, with contingency propellants at LEO and at lunar orbit for two more flights per year. These propellant requirements, together with the propellants required for soft-landing supplies, equipment, and personnel on the lunar surface and the propellants needed for monthly trips with the small LTC to the catcher, are given in table 8.

Year of delivery to LEOPropellants loaded on OTV at LEOPropellants loaded on OTV at LLOPropellants for large LTVPropellants for small LTVTotal Earth-supplied propellants at LLOTotal SMF-supplied propellants at LLO

For the first chemical propellant storage depots, we minimized the time and labor of converting Shuttle external tanks by keeping the basic external tank structure intact. Once oxygen and hydrogen are supplied from different sources (LEO for hydrogen from Earth, SMF for oxygen from lunar soil), it is more practical to separate the oxygen and hydrogen tanks and to discard the intertank section, reducing the mass of each complete depot by about 8 tons.

The expansion of the lunar base to full capacity would be completed by mid-1990, with a total of 4530 metric tons of equipment soft-landed at the two surface stations. Continuation of the lunar surface operations at this level requires soft-landing 30 tons of consumables per year, 14.4 tons of crew and baggage, plus an ample allowance of replacement parts and supplies. We have arbitrarily allotted 195.6 metric tons/yr for this purpose (about 4 percent of the total installed mass), so that three trips per year by the large LTV will suffice. The mass budget we have provided above will provide the lunar base with the capability of launching 600,000 metric tons/yr into space with daytime operations only. Some consideration should be given to ways of extending lunar base operations through the lunar night. The possibility of several large reflector satellites placed in high-inclination orbits around the Moon to provide full daylight illumination for the lunar base should be examined. Such satellites could be manufactured at the SMF and placed in lunar orbit with minimal consumption of propellants, significantly reducing the total mass of the energy systems which must be soft-landed on the lunar surface, and nearly doubling the launch capacity of the lunar base (ref. 18).


The final milestone for the space manufacturing enterprise considered here is to produce 2.4 solar power satellites per year, each with a capacity of 10 GW. Previous studies of Earth-launched solar power satellites estimate that 750 to 800 man-years of assembly work must be done in orbit for each such satellite (H. P. Davis, NASA, Johnson Space Center, Houston, TX, private communication, 1977). In the case of the space manufacturing system considered here, assembly work should be comparable, but additional labor must be provided for chemical processing and for fabricating basic structural components and active elements (such as photovoltaic cells or thermal-cycle radiators). Chemical processing and basic foundry and milling operations are generally much more automated (when measured in tons produced per worker per year) than are assembly operations in terrestrial industries. We have thus assumed that a total of 1250 man-years will be needed at the SMF to produce each SPS. The target of 2.4 SPS's per year thus requires an SMF crew of 3000.

We have assumed that the SMF crew is initially exchanged after a 1-year tour of duty, as were the LEO station crew and the lunar base crew. Once the total crew at the SMF reaches 1000, however, we have assumed that the tour of duty can be extended to 2 years because of the larger physical size of the SMF and because the larger size of the community affords more opportunity for each person to establish satisfactory social interactions. The schedule of crew buildup and exchange and the numbers of OTV's and the masses of propellants required at LEO and at SMF for personnel transportation between LEO and the SMF, are shown in table 9. A number of contingency flights have also been budgeted, as well as a number of spare vehicles at LEO. As a minimum, we have ensured that sufficient propellants and vehicles are available at the SMF at all times to evacuate to LEO all SMF personnel (up to 252) living in one cluster of 14 converted liquid hydrogen tanks.

DateCrew at SMFCrew increaseCrew exchangeTotal outbound passengersNo. of flights outbound (full)aEarth supplied propellants loaded at LEO (metric tons)No. of OTV's in active useNo. of spare OTV's at LEO
1988.5-300030010+391 H2
1989.5-54030084028+5 231 H2
1386 O2
1990.0840---- ---
1990.5-1500840234078+5 581 H2505
1991.02340---- ---
1991.5-6841670235479+5 588 H2515
1991.73024---- --
1992.03024---- ---
1992.5-01512151251+5392 H2515
1993.5-01512151251+5392 H2515

DateTotal inbound passengersNO. of flights inbound(full)aNo. of flights inbound (empty)bEarth-supplied propellants loaded at SMF (metric tons)Earth-supplied propellants delivered to LEO for later use at LEO (metric tons)Earth supplied propellants delivered to LEO for later use at SMF (metric tons)SMF supplied oxygen delivered to LEO for later use at LEO (metric tons)
1988.500+13091 H2
546 O2
1989.530010+58142.7 H2
1990.0----- --
1990.584028.520325 H2495 417.64286
1991.0----- --
1991.5167056+522530.6 H2489 4593528
1991.7----- --
1992.0----- --
1992.5151251+50392 H23923922352
1993.5151251+50392 H23923922352

aEach flight is assumed to carry 30 passengers and to consume 49 tons of propellants. Scheduled flights and contingency flights are indicated before and after the plus sign, respectively. Fuel requirements are computed assuming that all contingency flights are actually used. Unused propellants provide additional launch space to LEO in the following year, reducing the numbers shown in the last two columns year by year.

bTo the extent that outbound flight requirements exceed inbound flights plus increases in the OTV fleet at LEO, some empty flights must be made from the SMF back to LEO, consuming 33 tons of propellants per flight.


A space habitat must provide equipment for extracting potable water from human wastes and washing water; for maintaining atmospheric temperature, humidity, and composition within an acceptable range; and for storing or disposing of waste materials. A modular design for an environmental control and life-support system (EC/LSS) was adopted for use in all space habitats in this space manufacturing system. Carbon dioxide is concentrated by a molecular sieve and then reduced to carbon ash and free oxygen by the Sabatier process. Each EC/LSS unit, with a mass of 10.4 metric tons, supports 12 people with 100-percent redundancy, and fits in a cylindrical pod (3 m in diameter, 2.5 m long) large enough to- provide interior access for repairs and maintenance (E. H. Bock, General Dynamics, Convair Division, San Diego, CA, private communication regarding a design adapted from ref. 19). With efficient oxygen recovery, a space station using such a system would require about 1.5 kg/person resupply per day (including food as well as water, oxygen, and nitrogen losses through leaks and during EVA). Without oxygen recovery, such a system would require 2.05 kg/person/day. We have assumed 3.0 kg/person/day for consumables imported from Earth until lunar oxygen becomes freely available in 1990, when the consumables resupply is reduced to 1.5 kg/person/day. All waste materials are stockpiled for later use in agriculture.

Radiation protection is an important issue in habitat design. For the LEO station (below the Van Allen belts), we have provided no shielding additional to that afforded by the walls and furnishings of the converted Shuttle tanks. The lunar habitats can be shielded readily by burying them under several meters of lunar soil. The habitat at the catcher can be shielded several years after initial operation against cosmic rays and solar flares, but the expense of providing shielding very early would be difficult to justify because of the very limited times during which it will be occupied by small crews.

For the SMF, on the other hand, cosmic-ray shielding must be provided early. This can be done using the unselected lunar soils provided in the first few deliveries from the catcher during the first 6 months of occupancy. Solar flares, however, pose a serious threat to the SMF crew during this early period. Stowing the consumables for the first year around the walls and between flooring in the galley and mess hall areas of the communal tanks can provide an adequate shelter for solar storms lasting several days, should they occur before the entire habitat is shielded properly.

Estimates of power requirements per person in space stations have been extremely variable. In Skylab, electrical power for life support, environmental control, cooking, refrigeration, and lighting was 2.87 kW/person (average power). Most of the astronauts who flew Skylab complained that lighting was inadequate ref. 20). Allowing for the limited EVA of the Skylab missions in comparison with the space manufacturing enterprise considered here, we have allowed 9 kW/person (average power) throughout the system, except that the main lunar base provides only 8 kW/person during the lunar night when little EVA is expected, and the downrange station on the lunar surface provides only 3 kW/person during the lunar night for emergencies, since it is expected to be occupied only during the lunar day. Photovoltaic arrays have been sized accordingly, assuming peak power of 200 W/m2 under normal incidence of full sunlight and a mass of 1.61 kg/m2. Radiator systems to remove heat dissipated by use of electrical power and from human metabolism were assumed to require 170 kg/person with a radiator area of 68 m2 /person. Energy storage is provided by flywheels at the lunar surface habitats and at the LEO station, with an energy storage density of 870 W-hr/kg, and a highly conservative 70-percent overall efficiency. Research and development costs for the photovoltaic arrays, radiators, and flywheels can be minimized by modular designs sized for a few persons, just as the EC/LSS pods were sized for 12 persons, providing maximum commonality throughout the system.


For the space manufacturing system described here, habitats are required in four different sets of environments and operational conditions:

  1. LEO station: permanent occupation, minimal radiation protection needed, prompt return to Earth possible by means of the Shuttle, small crew, frequent visitors
  2. SMF: permanent occupation, full radiation shielding required early, rapid return to LEO possible (but expensive) by means of OTV, large crew, no visitors
  3. Catcher: infrequently occupied for short periods by very small crews, 40 hr travel time to LLO and lunar surface
  4. Lunar surface: permanent occupation at main base, intermittent long occupations of downrange base, small crews, prolonged isolation remote from Earth, 1/6 normal gravity, early protection required against radiation and thermal extremes

Because the Shuttle external propellant tank is an existing piece of hardware that can be delivered into orbit at minimal cost and because it was designed to sustain very heavy loads under launch accelerations, we have based all our habitat designs on the use of the liquid hydrogen tank as the basic structural frame and pressure vessel. This approach also permits a high degree of commonality throughout the system and seems to provide a reasonable compromise between the safety features of small modular habitats in the early stages of growth and the economies of scale of large monolithic structures. The extensive structural ribbing along the tank walls provides easy attachment of internal partitions and structures, permitting simple installation of the habitat interiors. (Some years ago, a proposal to use spent fuel tanks from Saturn rockets for space station shells was considered. The cryogenic insulation for those tanks, however, was on the inside of the tanks, and outgassing of hydrogen from the insulation for prolonged periods seemed to be a sufficiently severe problem that the proposal was rejected. For the Shuttle tanks, there is no such problem since the insulation is on the outside.)

The liquid hydrogen tank is 31 m long and 8.5 m in diameter (96 by 27.5 ft). For the orbital habitats, we have adopted a "condominium tower" configuration, dividing each tank into 10 levels of circular floor plan, with 2.6 m (8 ft) ceiling heights, leaving more than 0.5 m of subflooring space for utilities and storage. (Reducing the subfloor storage space to 0.2 m would allow 11 levels per tank with the same ceiling height.) A central shaft (1.8 m in diameter) runs the entire length of each tank, containing a dumbwaiter-like continuous belt elevator and ladders for access to all levels.

For the LEO station and the SMF habitat, several residential tanks would be clustered around a communal tank. The lowest level (in the bottom hemispherical dome of the tank) of each tank would be used for storage and for maintenance equipment. In the residential tanks, seven of the levels would be divided into three segments surrounding the elevator shaft to provide three studio apartments per level. Each apartment would have 17.5 m2 (188.5 ft2 ) of floor space, sufficient for one or two persons. One level (in the middle of the tank) would provide toilet, bath, and laundry facilities, while the top level (in the upper hemispherical dome) would be used as a leisure and social area (game room, observation deck).

In the communal tanks, the lowest level would be used for storage and maintenance. The next three levels up would provide recreational facilities (gymnasium, games, library, music room). A pantry and a galley would follow on the next two levels, with the following three levels used for dining rooms. (The dining rooms would double as assembly halls and movie theaters as well.) The topmost level would be used for EVA preparation, providing lockers for storage of EVA suits and facilities for recharging oxygen tanks and repairing EVA suits.P> The top level and one lower level of each residential tank would be connected to the corresponding levels of the communal tank and of each adjacent residential tank. When fully occupied, each residential tank (21 persons) requires two EC/LSS pods (12 persons each). These can be nestled between adjacent hydrogen tanks, with short tunnels connecting each pod to both hydrogen tanks to provide access to the EC/LSS equipment for maintenance as well as to provide alternative emergency passageways between tanks. All passageways would be equipped with airtight hatches to permit isolation of any tank.

Two identical clusters of residential tanks and a communal tank can then be assembled with a 140-m-long tunnel between them to form a dumbbell-like configuration. Rotation at 3 rpm then provides Earth-normal gravity at the bottom level of each tank and 0.7 g at the top level. The tunnel is stress-free, the hydrogen tanks being supported from the docking hub at the middle of the connecting tunnel by a set of cables. Each tank has an emergency airlock at the top level, with routine entry into the habitat through two airlocks in the hub. One of the dining room levels (of the six provided in the two communal tanks) would be used instead for a medical clinic/emergency operating room.

The Shuttle tank sets, designed to carry 700 metric tons under launch accelerations exceeding 2 g, are structurally capable of supporting the loads contemplated here with a very large safety factor. The internal loads would be borne by the tank ribbing and by three or more tie bars running the length of the elevator shaft which transmit their load to the cables suspending the entire tank from the hub. With a rotation rate of 3 rpm, crew selection for adaptability to rotation would be essential but would not pose a serious problem for most people (ref. 21).

For the LEO station, two residential tanks and one communal tank would be placed at each end of the dumbbell, providing facilities for as many as 84 persons. For the SMF habitat, each end of the dumbbell would consist of six residential tanks encircling a communal tank. The top level of the communal tank provides access to both the elevator system within that tank and a second elevator system running up to the hub. Because the SMF has a lower ratio of communal space to population than does the LEO station, some reallocation of space different from that described above may be necessary. Several levels among the six residential tanks could be changed from apartment levels into communal facilities. The total occupancy might be reduced from 252 to as low as 234 or the reduction in space could be compensated by some doubling up of crew members. (Alternatively, the 11-level design could be adopted from the start, increasing the total mass somewhat.) Spinning up the SMF habitat dumbbell to 3 rpm requires only 1 metric ton of propellant, a mass sufficiently small to be omitted from the mass budgets shown in the tables. (Figure 1 shows a design for the SMF habitat.)

For the catcher habitat, only one tank need be provided, using a mixture of levels from residential and communal tanks. A crew of 12 would be amply accommodated by four apartment levels; one toilet and laundry level; one pantry/galley/dining level; one recreation level; one level for workshops used to maintain and repair the catcher; one storage level; and one level for EVA preparation. This single tank does not provide pseudogravity, so no elevator is installed in the central shaft, crew members using ladders instead. The consistent vertical orientation throughout the 10 levels, required for the rotating habitats at LEO and at the SMF, would ease adaptation to the weightlessness of the catcher habitat (ref. 20). The habitat tank would be placed directly behind the catcher itself to protect the habitat from stray projectiles of lunar materials.

Furniture and personal belongings for a small one- or two-person apartment in the United States may total 1000 to 3000 lb (including refrigerator, carpets, and drapes). Instead of attempting to give a detailed breakdown of masses for furniture, toilets, shower stalls, laundry equipment, and galley equipment, we have allowed 1000kg (2200 lb) of furniture for each of the nine furnished levels of the residential tanks, and 60 percent more for the communal tanks where galley equipment and chairs and tables for dining rooms would require more mass than average. Table 10 summarizes the mass estimates for refurbishing the liquid hydrogen tanks for the orbital habitats at LEO, at SMF, and at the catcher.

ComponentMass, kg
Central shaft for each tank:
Vertical tie rods (in tension
Continuous belt elevator system400
For each level:
Structural flooring
(12 panels)
Outer edging and attach-
ment to tank
Inner ring and attachments
to tie rods
Radial beams (12)90
Ducting and plumbing100
Lighting fixtures100
Partitions and wall finishings400
Acoustic ceiling120
Carpeting or tiling75
Total internal shell
Central shaft840
9 floors at 1740 kg15,660
Communal tanks14,200 each
Residential tanks9,000 each
Miscellaneous components
Central hub1,700
Airlock1,400 each
Tunnels from hub to
communal tanks
Elevator system to hub800
Cables (per tank)2,300
EC/LSS pod (completely
redundant) for 12 persons
10,4000 each

For the lunar surface habitats, a different design approach is required because of the necessity of burying the habitats. Instead of a condominium tower configuration, the liquid hydrogen tanks must lie flat on their sides. The interior configuration is then a long two-story bungalow. Setting ceiling heights at 2.6 in (8 ft) leaves ample room for storage, ducting, wiring, and plumbing along the full length of the top and bottom of the tank and between the two levels in the subflooring space. Leaving space for a companionway the full length of each level, with spiral stairways between levels, each level would provide six studio apartments approximately 3.1 in by 4.5 m (10 by 14 ft), with about 11 m of tank length on each level for communal facilities (galley, toilets and baths, laundry, dining room, recreation facilities, and EVA preparation). Each tank would have two airlocks, and the two tanks at the initial main lunar base would be connected by a tunnel to make a single habitat unit. When the main base is expanded, the third and fourth tanks can be similarly connected to the first two.

Each tank thus provides private space for 12 persons instead of 21. The floor area per person is somewhat smaller, but this is psychologically compensated (in part) by daily work on the surface. With only 1/6 Earth-normal gravity, structural masses for flooring are smaller than would be the case in the orbital habitats. We have thus estimated only 10 metric tons per tank for structure and furnishings; if two airlocks and one EC/LSS pod are added, the total mass per tank is 37.5 tons. The energy system to support 12 persons totals 62.5 tons (4.5 tons for photovoltaic arrays, 2.2 tons for radiators, 55 tons for flywheels, 0.8 ton for motor/generators and for the control system).


In an earlier section, we discussed the logistics of transporting lunar and SMF crews from LEO by means of a chemically propelled OTV. A passenger module (including EC/LSS hardware and the OTV flight deck) that can fit in the Shuttle cargo bay and that can accommodate 30 passengers enroute to the SMF (a 9.8-hr flight) or 24 passengers enroute to lunar orbit (a 5-day flight) would have a dry weight of about 14.2 metric tons. Allowing 300 kg/passenger (including baggage) leaves room for a 27.5-day supply of consumables for a lunar trip or a 2.2-day supply for the trip to the SMF within the 23.4-metric-ton payload capability of the OTV. Conditions aboard would be distinctly crowded, but such conditions can be tolerated for a voyage of strictly limited duration. One passenger module must be provided for each OTV in the fleet. The passenger modules used for the lunar crews must differ somewhat from those used for SMF crews because of the greater duration of the lunar transfer, with more extensive galley and toilet facilities, consistent with the reduction in the number of passengers. The OTV passenger modules, however, would seldom be used aboard the Shuttle since the Shuttle payload capacity is considerably larger than the total mass of a fully loaded passenger module. Most of these modules would be delivered to LEO aboard the SD/HLV, with the passenger cabin filled with such cargos as consumables for LEO or SMF.

During the first year or two of launch operations before the lunar base begins operation, crews for the LEO station can ride to and from orbit in the Shuttle cabin, 4 or 5 at a time. Later on, when LEO crews and lunar crews are exchanged in lots of 24, and the SMF crews are exchanged in groups of 150 or so, this mode of operation becomes impossible. Two or three Shuttle-bay passenger modules designed for a total flight time of 1.5 to 3 hr would suffice, with doubledeck, high-density seating provided for 60 passengers at a time, plus baggage. These Earth-launch modules should be provided in 1987, and could be expected to weigh no more than the OTV modules since the life-support requirements are considerably simpler.


Once the lunar base has been established, the pace of transporting supplies and equipment into orbit is sharply accelerated. A second MDRE is added to the fleet late in 1986, and four more are added in 1987. Because of the higher payload of the SD/HLV, there is a tendency for payloads for high orbit to accumulate faster than the requisite external tanks for MDRE reaction mass. Late in 1987, in anticipation of the arrival of the first cargo of lunar soils at the SMF, the first of eight large MDRE's (LMDRE) is delivered to LEO. Table 11 compares this vehicle with the original MDRE, making some assumptions about economies of scale possible for the larger machine.

Metric tons
Dry weight at LEO174 tons
Reaction mass consumption rate175 tons/month
Reaction mass required for a 4-month
trip from LEO to SMF with refueling at SMF
700 tons
Payload delivered from LEO to SMF1,200 tons
Dry weight at LEO1,600 tons
Reaction mass consumption rate1,700 tons/month
Reaction mass required for a 4-month
trip from LEO to SMF with refueling at SMF
6,800 tons
Payload delivered from LEO to SMF12,000tons

The first LMDRE is delivered to the SMF as cargo aboard two MDRE flights shortly after initial, operation of the SMF in February 1988. After assembly and checkout are completed, the LMDRE makes a round trip to LEO to deliver 2100 tons of lunar material, enough reaction mass for three MDRE flights to the SMF (fully loaded with 1200 tons of cargo each time). Since the LMDRE is operating far below maximum capacity on this flight, only 2400 tons of reaction mass are required to complete the round trip in about 2 months. The LMDRE is then used to haul larger masses of lunar soil from the catcher to the SMF since the launch rate at the lunar base begins to increase by mid 1988 to rates exceeding the capacity of the MDRE fleet.

During 1988, 2000 metric tons of LMDRE components and sub-assemblies are delivered to the SMF which contributes 1200 tons of manufactured components (solar cells, cables, structural members) to complete two more LMDRE's. During 1989, assuming that the SMF can now contribute 50 percent of the mass of an LMDRE, 4000 tons of imported components and parts permit expansion of the LMDRE fleet by five more units.

The fleet of eight LMDRE's is required to transport lunar materials from the catcher to the SMF; to transport liquid oxygen for chemical propulsion from the SMF to both LEO and LLO; to transport lunar material to LEO for the MDRE and LMDRE fleets; and to transport (if necessary) chemical feedstocks for the manufacture of solar power satellites at the SMF. We have arbitrarily assumed that 9000 metric tons of raw materials must be imported from Earth for each SPS totaling 100,000 metric tons when completed. Annual transportation requirements for these chemical feedstocks would then total 21,600 tons, which can be accommodated in two LMDRE flights. More detailed trajectory analyses are required to optimize the MDRE/LMDRE fleet and to confirm some of the mass and flight time estimates used here.


Table 12 summarizes the total launch requirements for the scenario described here.

YEAREquipment and supplies for:VehiclesChemical propellantsCrew and baggageSubtotalChemical feedstocksTotal

aMetric tons.

The greatest single uncertainty concerns the question of how much mass must be imported from Earth to supplement lunar raw materials in the manufacture of products such as solar power satellites. The lift requirements can be met in many ways; table 13 shows an aggressive launch schedule using the Shuttle, the SD/HLV, and the SSTO to meet the milestone on the schedule described here. After 1987, when a fifth SD/HLV is added, the Shuttle is used only for crew transportation with a passenger module in the cargo bay.

YEARTotal payloada (including chemicals), metric tonsNumber of flightsTotal launch capacity, metric tonsTotal payloadb (excluding chemicals), metric tonsNumber of flightsTotal launch capacity, metric tons

aAssuming 9% terrestrial chemical feedstock requirement.

bAssuming no terrestrial chemical feedstock requirement.

Should it be possible to manufacture SPS's without any imports from Earth, the number of launches by the Shuttle and the SD/HLV would be considerably reduced. The right half of table 13 shows the launch schedule that results from eliminating the chemical feedstocks column in table 12 . Redistribution of the remaining cargos to produce a more even launch schedule has not been attempted, nor have we considered in any detail the changes in MDRE fleet size that would also result. Use of the LMDRE to bring lunar materials to LEO for reaction mass would have to be accelerated because of the reduction in expendable Shuttle tanks.

Table 14 summarizes the allocation of Shuttle tanks over the period 1985 to 1990.

YEARLH2 tanks for habitats at:Equivalent complete tanksChemical propellant storage depotsMDRE reaction massTotal external tanks available
LEOSMFLunar baseCatcher
19856-214.4 352.660
1986-282-13.2 343.860
1987-28--12.4 3124.6140
1988-28--12.4 1480.6107
1989-42--18.6 -117.4136
1990-42--18.6 -142.4161

Each liquid hydrogen tank weighs 14.5 tons, while a complete tank set weighs 33.45 tons. Thus the tank allocations must be converted to "equivalent tanks" to account for use of the rest of a tank set for reaction mass after the hydrogen tank has been converted into a habitat. If the reduced launch schedule shown on the right side of table 13 is attainable, deployment of the single-stage-to-orbit vehicle (assumed to operate in a mixed passenger and cargo mode) could be postponed indefinitely, although costs per ton would be higher in that case. After a number of years, the increased costs of using the Shuttle and the SD/HLV would begin to rival the discounted costs for development of the SSTO.


We have examined the logistic feasibility of establishing a space manufacturing enterprise for the construction of solar power -satellites in high Earth orbit, using lunar materials, operating within the constraints of Space Shuttle and Shuttle-derived launch vehicle technologies. Such a system, it seems, could be deployed in less than 8 years of launch activities maintained at an aggressive pace; the launch activities follow 4 or 5 years of dedicated development and hardware procurement. If the program commences at full speed in 1980 or 1981, the first 10-GW SPS's could be on-line late in 1990; the second, late in 1991; and the third, early in 1992. Crew selection and training requirements and an economic analysis of the program have also been carried through. These results will be reported elsewhere. Depending on a variety of assumptions about interest rates, discount rates, costs for competing sources of electrical powerplants, and research and development costs, the total investment for the period 1980 to 1992 (when the third SPS is sold) lies roughly between $50 billion and $ 100 billion, equivalent (in 1977 dollars) to one or two times the Apollo program. A key technical problem that must be resolved at an early date is the design of an SPS using only lunar materials. Should it prove impractical to design such an SPS or to design a factory to manufacture such an SPS, the use of asteroidal materials (with a much greater diversity of chemical elements available, especially such volatiles as hydrogen, carbon, and nitrogen) would be favored. The use of the mass driver as a reaction engine for cargo transportation beyond LEO provides significant advantages over chemical propulsion systems, as can be seen from an examination of the extensive propellant requirements for personnel transportation alone, a small fraction of the total mass required to be transported beyond LEO. Habitat designs based on use of the Shuttle external tank seem to offer numerous advantages and simplifications in the system considered. The orbital habitats (except for the intermittently occupied habitat at the catcher) all provide near-Earth-normal gravity from the beginning. The location for the space manufacturing facility must be optimized after the SPS design has been settled During the early stages of the enterprise, however, a high circular orbit appears to be strongly favored because of logistic considerations.

The detailed scenario presented here is only one of many possible design approaches, but it serves to show that a space manufacturing enterprise is logistically feasible in the Shuttle era, albeit with significant technical risks in such an aggressive program. The feasibility of undertaking such a program in the public sector or along conventionally organized lines in the private sector has not been addressed here.



Table of Contents