Solar Power Satellites

Chapter 5 ALTERNATIVE SYSTEMS FOR SPS Contents

Page Microwave Transmission ...... 65 LIST OF FIGURES The Reference System...... 65 Figure No. Page Laser Transmission...... 78 9 Solar Power Satellite Reference Laser Generators ...... 79 System...... 66 Laser Transmission ...... 81 10 Satellite Power System Efficiency Laser-Power Conversion at Earth ...... 82 Chain ...... 67 The Laser-Based System ...... 82 11 Major Reference System Program Elements...... 68 Mirror Reflection ...... 86 12 The Retrodirective Concept ...... 69 The Mirror System...... 88 13 Power Density at Rectenna as a Space Transportation and Construction Function of Distance From the Alternatives ...... 89 Beam Centerline ...... 70 Transportation ...... , . . . . 89 14 Peak Power Density Levels as a Space Construction...... 91 Function of Range From Rectenna . . . . 70 15 SPS Space Transportation Scenario . . . 73 SPA Costs ...... 92 16 The Solid-State Variant of the Reference Reference System Costs ...... 92 System...... 78 Alternative Systems ...... 96 17 lndirect Optically Pumped CO/CO2 The Solid-StateSystem ...... 96 Mixing Laser ...... 80 The Laser System ...... 96 18 The CATALAC Free Electron Laser The Mirror System...... 97 Concepts...... 81 19 Optics and Beam Characteristics of Two Types of Laser Power Trans- LIST OF TABLES mission System (LTPS) Concepts...... 82 Table No. Page 20. The Laser Concept...... , ...... 84 6. Projections for Laser Energy Converters 21 Components of the Laser Concept . . . . 84 in 1981-90 ...... 83 22. The Mirror Concept (SOLARES) ...... 87 7. 500 MWe Space Laser Power System. . . 85 23. Reference System Costs ...... 92 8. Laser Power Station Specification. . . . . 85 24 How Cost Could Be Allowed...... 93 9. SOLARES Baseline Systerm ...... 88 25 Elements and Costs, in 1977 Dollars, for 10. Research— $370 Million ...... 93 the Baseline SOLARES System ...... 97 11. Engineering– $8 Billion ...... 93 26 Sensitivity of the SOLARES Mirror 12. Demonstration– $23 Billion ...... 93 System to Variations in System 13. SPS lnvestment– $57.9 Billion ...... 94 Parameters ...... 98 Chapter 5 ALTERNATIVE SYSTEMS FOR SPS

A variety of systems have been proposed for tutional, and public acceptance issues in the collecting, transmitting, and converting solar chapters that follow. power from space. Each system has its advantages and disadvantages, its benefits and draw- In order to estimate reliably and fully the backs. Each alternative system would use one range of costs and potential technical uncer- of three transmission modes — microwave, tainties for a given solar power satellite (SPS) laser, or optical reflector–to transmit power option, it would be necessary to subject it to to Earth where it is collected and converted to the same detailed analysis that the reference electricity or some other highly useful form of system has undergone during the last 5 years. energy. Each system would use numerous sub- Unfortunately, this analysis has not been ac- systems to collect and convert energy in space complished for the alternative systems. Hence, or on the ground. This chapter wiII character- detailed comparisons between systems will not ize the alternative systems and subsystems and be possible. At this stage it is possible only to discuss their potential for generating power compare the major features of each technol- from space. It will also describe four repre- ogy and note the uncertainties that should be sentative systems that serve as the technical addressed as conceptual development of the basis for discussion of the environmental, insti- various alternatives continues.

MICROWAVE TRANSMISSION

Because the atmosphere is highly transpar- and Space Administration (DOE/NASA) as a ent to microwaves, they constitute an obvious basis for study. It consists of a large planar candidate for the SPS transmission mode. In array of photovoltaic celIs located in the geo- addition, microwave technology also is well- synchronous orbit 35,800 km above the Earth’s known and is used today in a number of space Equator (fig. 9). The cells convert solar energy and terrestrial communications and radar ap- into direct-current (de) electricity that is plications. Microwave power transmission was conducted at high voltage to a phased-array first demonstrated experimentally in 1964, ’ microwave transmitting antenna mounted at and tested in 1974.2 3 one end of the photovoltaic array. Klystron amplifiers convert the dc electricity to high- The Reference System4 56 voltage radio-frequency power that is then radiated to Earth by slotted waveguides. A The reference system was selected by the receiving antenna (rectenna) on the ground Department of Energy/National Aeronautics reconverts the electromagnetic radiation into electric current and rectifies it into dc. After 1). F Degenford, M D. Sirkis, and PV H Steir, “Ttle Reflecting Beam Waveguide, ” I E EE Transactions 01 Microwave Theory being converted to high-voltage, low alter- Technology MIT-72, July 1964, pp 445-453 nating current (ac), the power can then be ‘Richard M Dickinson, “Evaluation of a Microwave High- either delivered directly to the conventional ac Power Reception-Conversion Array for Wireless Power Transmis- sion, ” Jet Propulsion Laboratory Technical Memorandum No grid or converted back to dc at high voltage 33-741, Sept 1, 1975 and delivered to a dc transmission network. ~R i chard M Dick InsOn, “Microwave Power Transmitting Phased Array Antenna Research Project Summary Report, ” Jet The amount of power delivered to the grid Propulsion Laboratory publication No 78-28, Dec 15, 1978 by each reference system rectenna has been ‘Department of Energy, “Satellite Power System Concept De- velopment and Evaluation Program Reference System Report, ” report No. DOE/E R-0023, October 1978 bR O Piiand, “SPS Cost Methodology and Sensitivities, ” The ‘C. C. Kraft, “The Solar Power Satellite Concept, ” NASA pub- F/na/ Proceedings of the Solar Power Satellite Program Review, lication No JSC-14898, July 1979 DOE/NASA Conf-800491, July 1980.

66 ● Solar Power Satellites

Figure 9.—Solar Power Satellite Reference System

Solar power satellite reference system

Solar cell arr

Transmitt

SOURCE: C. C. Kraft, “The Solar Power Satellite Concept,” NASA publication No. JSC-14898, July 1979

set at 5 gigawatts (GW)—or 5,000 megawatts The system is designed to deliver baseload, (MW). The microwave transmission frequency i.e., continuous 24-hour power to the electric was chosen to be 2.45 gigahertz (GHz). Max- grid. However, some variations in delivered imum microwave power density at the center power would occur. A seasonal fluctuation in of the rectenna (on Earth) was set at 23 output due to the variation of the Sun’s dis- milliwatts per square centimeter (mW/cm2), tance from Earth would cause variations in and the maximum power density at the edge of both incident insolation and photovoltaic cell the rectenna was set at 1 mW/cm2 (one-tenth temperature, the latter producing a conse- the current U.S. recommended occupational quent change in efficiency. In addition, around limit). The reference design assumes that all the spring and fall equinoxes the Earth’s materials would be obtained from Earth, and shadow would occult the SPS, resulting in a that the system lifetime would be 30 years with short period each night for about 6 weeks at no residual salvage value. local midnight (about 75 minutes maximum, at the equinoxes) where no solar radiation im- The area of the satellite’s photovoltaic array pinges on the satellite and therefore no power would be approximately 55 square kilometers 2 could be delivered to the grid (see ch. 9 for a (km ); the diameter of the transmitting antenna discussion of this effect). 1 km. The total in-orbit mass of the complete system, including a 25-percent contingency factor, would be either 51,000 or 34,000 metric Subsystem Description tons (tonnes), depending on whether silicon or ENERGY COLLECTION AND CONVERSION gallium arsenide photovoltaic cells would be Two photovoltaic concepts were considered used. for the DOE/NASA reference system. One uses

Ch. 5—Alternative Systems for SPS . 67

Figure 10.-Satellite Power System Efficiency Chain

57.81 GW 11.58 GW 10.50 GW 9.46 GW

Ga 63.18 GW 71.77 GW (Solar)

10.79 GW 10.29 GW 9.79 GW 70.81 GW Si Si 62.34 GW

Ga 9.08 GW 8.50 GW 8.50 GW 8.18 GW 6.96 GW 6.72 GW

9.08 GW 6.58 GW 5.79 GW 5.15 GW

overall efficiency = 6.970/. Ga MPTS efficiency = 63.00/. 7.06% Si

Abbreviation: “Ga” indicates the gallium-alum aluminum-arsenide option, “Si” the silicon option.

SOURCE: Department of Energy, “Satellite Power System Concept Development and Evaluation Program: Reference System Report,” DOE report No. DOE/ER-0023, October 1978. single crystal silicon converters that would cells: low mass per unit area, resistance to ther- receive sunlight directly; the other uses mal and radiation degradation, and higher effi- gallium-arsenide (GaAs) photovoltaic cells il- ciency. They have the disadvantages of rela- luminated directly and by mirrors in a 2:1 con- tively high cost, the limited production availa- centration ratio. bility of gallium, and a smaller technology base than for silicon cells. Because of these Silicon cells, currently used in all solar latter characteristics, these cells would be powered spacecraft, have the advantages of used in a 2:1 concentration ratio in the refer- an extensive manufacturing base, abundant re- ence system, trading the relatively expensive source materials, and lower cost per cell, as cells for less expensive Iightweight reflectors well as an R&D program in DOE aimed at ma- to concentrate sunlight on the cells. jor cost reduction for terrestrial cells. How- ever, silicon cells in space suffer degradation The structure that supports the solar cells from radiation effects and from high-operating would be an open-truss framework made of temperatures, and hence would probably re- graphite-fiber reinforced thermoplastic com- quire periodic annealing of the array surface posite (fig. 9). Because the solar array must be (possibly by laser or electron beam techniques) oriented toward the Sun and the transmitting or the development of silicon cells less af- antenna toward the Earth, a massive rotary fected by ionizing radiation. joint is essential in order to provide the nec- Gallium-aluminum arsenide photovoltaic essary mechanical coupling. Sliprings about cells have several advantages over silicon 400 m in diameter would be used in conjunc-

68 ● Solar Power Satellites

Figure 11 .—Major Reference System Program Elements

GEO

COTV

construction depot

Space freighter

SOURCE: R. O. Piland, Cost Methodology and Sensitivities,” The Final Proceedings of the Solar Power Satellite Program Review, DOE/NASA 1980. tion with the rotary joint in order to transfer reference system, these design considerations electric power from the array to the antenna. resulted in a l-km diameter antenna. It would be constructed of 7,220 subarrays each con- POWER TRANSMISSION AND DELlVERY taining from four to thirty-six 70-kW klystron The power transmission and delivery system power amplifiers connected to slotted wave- for the reference system design is common to guides for transmitting power to Earth. KIys- both photovoltaic options. It is composed of trons were chosen because their technology three major elements: the transmitting anten- and operating characteristics at low power na, the rectenna, and the substation. levels are well-known. However, they require a cooling system (probably heat pipes). Klystrons The selection of the microwave transmission of 70-kW continuous power rating have not frequency was based on tradeoffs between at- been built and tested at this frequency, so their mospheric attenuation and interactions with characteristics are not known in detail. the ionosphere as well as the sizes of the antenna and rectenna. The optimal frequen- Each of the more than 100,000 klystrons in cies were found to be between 1.5 and 4 GHz. the antenna must be properly adjusted or The reference frequency was selected to be “phased” to provide a uniform power beam 2.45 GHz, which lies in the center of the inter- and to point it. This adjustment is especially national Industrial, Scientific, and Medical critical at the very high, gross power level of (ISM) band of 2.4 to 2.5 GHz. the SPS beam. Were the antenna a totally rigid The size of the antenna is determined by the array of amplifiers precisely fixed in space, the transmission frequency, the amount of heat it adjustment could be accomplished once and is feasible to dissipate at the antenna, the for all just after the antenna is fabricated in theoretical limits of ionospheric heating, and space. However, because it would be desirable the maximum power densities chosen at for the antenna to be relatively flexible it ground level, i.e., at the rectenna.7 For the would be necessary to use an active system of phase control, a so-called “adaptive electronic ‘Raytheon Corp., “Microwave Power Transmission System Studies,” report No, ER75-4368, contract No NAS3-I 7835, De- control” in which a pilot beam, installed in the cember 1975. center of the rectenna and pointed toward the

Ch. 5—Alternative Systems for SPS ● 69

satellite, establishes a phase reference or beam to very low power (0.003 mW/cm2). The standard clock against which the individual transmission system would therefore require klystrons compare and adjust their phases (fig. continual ground-based guidance to keep it 12).8 operating as a coherent beam. By incor- porating relatively well-known anti jamming An important safety feature inherent in this techniques in the pilot-beam generator, de- system is that loss of the pilot beam from the liberate or accidental diversion or misuse of rectenna would eliminate all pointing and the SPS beam could be prevented. phase control. Without the pilot beam, the klystron subarrays would immediately lose The parameters of the microwave beam are synchronization with one another and al I focus of critical importance in assessing the en- would be lost, resulting in the spreading of the vironmental impacts of the SPS. The peak power density at the transmitting antenna is ‘William C Brown, “Solar Power Satellites Microwaves Deliv- calculated to be 21 kW/m2. By the time the er the Power, ” Spectrum, June 1979, pp 36-42. beam reached the upper atmosphere it would have spread considerably and the intensity reduced to 23 mW/cm2, a power Iimit that was Figure 12.—The Retrodirective Concept set because theoretical studies suggested that at higher power densities, nonlinear instabil- ities could appear in the F layer of the ionosphere (200 to 300 km) as a result of the interactions between the beam and the electrically charged particles in this region. Recent ex- perimental studies indicate that the limit in the lower ionosphere might be able to be set much higher, ’ thereby making it possible to decrease the size of the antenna and/or rectenna signifi- cantIy. With these design constraints, a theoretical beam power distribution was conceived resulting in the radiation pattern at the rectenna shown in figure 13, on which are noted the present U.S. recommendations for public exposure (10 mW/cm2) and the current U.S.S.R. occupational guideline (0.01 mW/cm2). The off-center peaks in figure 13 are called In the retrodirective-array concept, a pilot beam from the “sidelobes;” the level of intensity shown is a center of the rectenna establishes a phase front at the consequence of the 1-km antenna aperture transmitting antenna. Central logic elements in each of the (which is optimized to minimize orbital mass) antenna’s 7,220 subarrays compare the pilot beam’s phase front with an internal reference, or clock phase. The phase and the projected cumulative antenna errors. difference is conjugated and used as a reference to control The first sidelobe would have a peak intensity the phase of the outgoing signal. This concept enables the of 0.08 mW/cm2, less than one-hundredth the transmitted beam to be centered precisely on the rectenna and to have a high degree of phase uniformity. If this phase- current U.S. occupational exposure recom- control system fails, the beam would automatically be mendation, about 8 km from the beam center- defocused, dropping the power density to 0.003 mW/cm2, line; the intensity at the edge of the reference an intensity acceptable by current standards. This feature system rectenna (5 km from the beam center- has been referred to as the “fail-safe” aspect of the 2 microwave transmission system. line) would be 1 mW/cm –one-tenth the U.S. occupational exposure guideline.

SOURCE: William C. Brown, “Solar Power Satellites: Microwaves Deliver the ‘w Cordon, and L M Duncan, Impacts on the Upper Power,” Spectrum, June 1979, pp. 36-42. Atmosphere,” Astronautics and Aeronautics, July/August 1980

70 ● Solar Power Satellites

Figure 14.—Peak Power Density Levels as a Function of Range From Rectenna

USA standard 10

1.0

0.01 o 2,000 4,000 6,000 8,000 Radius from boresight (km) 0.005 I I I III 1 I I1 I I o1,000 2,000 3,000 4,000 5,000 Radius from boresight (miles)

0.001 Grating lobe spikes occur every 245 km for the 18-m sub- 0 5,000 10,000 15,000 20,000 arrays used on simulations although only two grating lobes are shown. The SPS 10-m subarrays have grating lobes Ground radius, m every 440 km.

SOURCE: Department of Energy, “Satellite Power System Concept Develop- SOURCE: Department of Energy, “Satellite Power System Concept Develop- ment and Evaluation Program: Reference System Report,” DOE ment and Evaluation Program: Reference System Report,” DOE report No. DOE/ER-0023, October 1978. report No. DOE/ER-0023, October 1978.

In addition to the relatively strong sidelobes, The rectenna design is quite insensitive both the finite size of the antenna subarrays and to the angular incidence of the microwave their projected misalinements would produce beam (within 100, and to variations in phase or much weaker “grating lobes, ” which for the amplitude caused by the atmosphere. Hence, reference system would occur at 440-km inter- rectennas would be interchangeable; the same vals from the rectenna. The integrated intensi- satellite could power different rectennas, as ty of these grating lobes, even for hundreds of long as they were equipped with the appropri- operational SPSs, would be well below even ate pilot beam needed for phase control of the the U.S.S.R. public-exposure guideline, as transmitting antenna. The reference rectenna shown in figure 14. would be composed of billions of dipole an- Ch. 5—Alternative Systems for SPS Ž 71

tennas placed above a transparent wire grid. sideration of the transportation options. The The microwave energy received by each dipole basis for all projected Earth-to-low-orbit would pass through a rectifier circuit that transportation concepts is the current U.S. would convert it to dc power at high current space shuttle, scheduled to become the opera- and low voltage. Several more conversions tional mainstay of the U.S. (and much of the would be necessary to condition the power for world’s) space program. the grid. The received power would first be Of the many possible shuttle derivatives and converted to ac and then transformed to high- other new transportation prospects, 12 NASA voltage low-current 60-cycle ac power and selected four different types of vehicles to sup- then either fed into ac transmission lines for ply the four basic transportation functions: delivery to the users or reconverted to high- voltage dc for transmission, a relatively new ● carrying cargo between Earth and low- transmission technology. Earth orbit (LEO), ● carrying personnel between Earth and Estimates of overall rectenna conversion ef- LEO, ficiency run from about 80 to 92 percent, and ● transferring cargo between LEO and the the extreme simplicity and repetitive-element geosynchronous orbit (CEO), and construction of the electrical components ● transferring personnel between LEO and would facilitate mass production at extremely CEO. low unit cost. Reliability of the rectenna should be extremely high, because each com- The designs of these four vehicles, called re- ponent would be ultrareliable and could oper- spectively, the heavy-lift launch vehicle ate redundantly. Hence replacement would be (HLLV), the personnel launch vehicle (PLV), the necessary only after a large number of individ- cargo orbital transfer vehicle (COTV), and the ual failures. personnel orbital transfer vehicle (POTV), are based on existing technology, although all None of the substation equipment involves would require considerable development be- technological advances beyond those that are fore reaching operational status. 13 14 15 16 projected through normal development by the electric utility industry. The major concern Both the HLLV and the PLV would utilize that has been expressed is the large scale of fully reusable flyback boosters similar to those the minimum individual power unit. Current originally considered by NASA in early shuttle grid control systems are quite adequate to han- designs in the late 1960’s. Both boosters would dle near-instantaneous switching of single employ methane-oxygen rocket engines for power units as high as 1,300 MW. Single unit (vertical) takeoff and airbreathing (turbofan) variations of 5,000 MW could present major engines for flyback to base for horizontal land- control difficulties to the utilities as they cur- ings. The HLLV orbiter would use oxygen- rently operate10 11 (see ch. 9 for a detailed description of utilities interface problems). “Robert Salkeld, Donald W Patterson, and Jerry Grey (eds ), ‘Space Transportation Systems, 1980-2000, ” VOI 2, AlAA Aero- ipace Assessment Series, A IAA, New York, 1978 SPACE CONSTRUCTION ‘‘G Woodcock, “Solar Power Satellite System Definition The mass and physical size of the space seg- Study, ” Boeing Aerospace Co., Johnson Space Center contract No NAS9-I 5196, pt 1, report No D180-20689, June 1977; pt 11, ment needed for an operational 5-GW satellite report No D180-22876, December 1977, pt I I 1, report No power station are larger by several orders of D180-24071, March 1978 magnitude than any space system heretofore “C Hanley, “Satellite Power System (SPS) Concept Defini- tion, ” Rockwell International Corp., Marshall Space Flight Cen- launched and therefore require careful con- ter, contract No NAS8-32475, report No SD78-AP-0023, April 1 ’378 ‘“J. G. Bohn, J. W. Patmore, and H W Faininger, “Satellite 15 Gordon R Woodcock, “Future Space Transportation Sys- Power Systems: Utility Impact Study,” EPRI AP-1 548 TPS 79-752, tems Analysis Study, ” Johnson Space Center contract No. September 1980. NAS9-I 4323, Boeing Aerospace Co. report No DI 80-20242-1 11 p j, Donalek, and J. L. WhYsong, “Utility Interface Require- (three volumes), Dec. 31,1976 ments for a Solar Power System, ” Harza Engineering Co , “Donald P, Hearth (Study Director), “A Forecast of Space DO E/E R-0032, September 1978 Technology 1980-2000,” NASA SP-387, January 1976. 72 . Solar Power Satellites

hydrogen rockets essentially identical to those area would serve as the transfer point for all of the current space shuttle, and then glide materials and personnel both up to CEO and back to base much like the shuttle does. Un- back down to Earth. Alternative strategies like the shuttle, it would be fully reusable; it have been considered, some of which will be would have no disposable external propellant discussed later. tank. The principal factor that governs the cost The PLV orbiter would be very much like the and effectiveness of in-space construction is current space shuttle, but would employ a pas- generally accepted to be the productivity of senger-carrying module in the payload bay. the construction crew and cost, and require- Like the shuttle, it would also use a disposable ments for shielding. The replacement of some external propellant tank, but a somewhat crew by automated equipment is therefore a smaller one. It couId carry 75 passengers, plus major consideration in alI construction strate- the normal shuttle crew. gies or scenarios, e.g., effort has already been devoted to automatic beam-building sys- A fleet of COTV, all reusable, would make 17 tems. The use of teleoperators and robot ma- the round trip from LEO to CEO, carrying the nipulators for assembly of large structures has cargo payloads up to CEO and returning also been considered. The current growth of empty to LEO for reuse. They would be pro- technology in these areas is extremely rapid, ’8 pelled by efficient but slow electrostatic and incorporation of such techniques would engines. Using low-thrust electric propulsion almost certainly benefit all aspects of SPS con- would require very long trip times, of the order struction. Despite the wide range of construc- of 4 to 6 months. The bases for selecting this tion options, estimated personnel require- propulsion option were essentially minimum ments for them are approximately the same: cost and ready availability of the argon pro- 19 750 & 200. pellant and other materials. Such long trip times, although suitable for cargo, are clearly GROUND-BASED CONSTRUCTION , not acceptable for personnel, so a high-thrust Building the rectenna, although a very large propulsion approach was chosen for the and relatively unique structure, nevertheless POTV. The design utilizes a basic oxygen- would involve far fewer uncertainties than hydrogen propulsion stage now undergoing constructing the space segment. A detailed research evaluation at NASA as part of its Ad- analysis 20 of both the basic structure and vanced Space Engine program. It employs construction aspects concluded that the pri- essentially the same level of “technology as mary structural material should be galvanized that used in the current space shuttIe main or weathering steel rather than aluminum engine. It could carry up to 160 people from (which is more scarce and requires a higher LEO to CEO and back, or 98 tonnes (480 man- energy cost to produce). months) of consumables from LEO to CEO. Because it would be impractical to launch a SYSTEM OPERATION full-sized power satellite by single launch vehi- An active control system would be needed cle, a strategy for constructing the satellite in both to keep the satellite in the proper orbit Earth orbit would be necessary. The basic space construction strategy selected for the ‘Denls j Powell and Lee Brewing, “Automated Fabrication of Large Space Structures, ” Astronautics and Aeronautics, October reference system is to launch all materials, 1978, pp 24-29 components, and people to staging areas in ‘ 8 Antal K Bejczy, “Advanced Teleoperators,” Astronautics LEO (fig. 15). The COTVs, because of their and Aeronautics, May 1979, pp. 20-31 “W H Wales, “SPS Program Review Transportation Perspec- large solar arrays, would be assembled in LEO tive, ” I n The Final Proceedings of the Solar Power Satellite Pro- as well. The main construction base would be gram /?ev/ew, DOE/NASA Conf-800491, July 1980 O located in CEO, although not necessarily at ‘ ’’ Feaslbil ity Study for Various Approaches to the Structural Design and Arrangement of the Ground Rectenna for the Pro- the eventual geostationary-orbit location of posed Satellite, ” NASA contract No. NAS-I 5280, Bovay Engi- the operational SPS. Hence the LEO staging neers, In{ , May 1977

Ch. 5—Alternative Systems for SPS ● 73

Figure 15.—SPS Space Transportation Scenario

SOURCE: W. H. Wales, “SPS Program Review Transportation Perspective,” in The Final Proceedings of the Solar Power Satellite Proqram Review, DOE/NASA Conf-800491, July 1980. -

(stationed above the rectenna) and to maintain costs, partly because of the predictability of the solar array’s orientation to the Sun. The the space environment as compared, for exam- mass of the necessary control system is esti- ple, with the uncertain environment in which mated at 200 tonnes; its average electric power aircraft structures must be designed to oper- consumption would be 34 MW. ate, and partly because of the extensive body of applicable design, testing, and operational Because of its low coefficient of thermal ex- experience with high-performance aerospace pansion and relative stiffness, a graphite com- structures. However, questions of dynamic in- posite structural material was selected for the stability resulting from Iow-probability occur- reference system in preference to the alumi- rences such as major meteor strikes or aggres- num alloys so widely used in aerospace struc- sive military action would have to be eval- tures. Although a complex engineering prob- uated. lem and, furthermore, one not readily subject to testing at an adequate scale prior to deploy- Orientation of the transmitting antenna rela- ment in space, it does not appear likely that tive to that of the solar array would be main- dynamic stability would cause any major unex- tained via the large rotary joint. Physical aim- pected problems in either performance or ing of the antenna itself would be accom-

83-316 0 - 81 - 6 74 ● Solar Power Satellites

plished by gyroscopes, which would feed concept, there are many technical uncertainties trol signals to the mechanical-joint turntable associated with the reference system. This sec- so that it could follow the antenna pointing re- tion identifies specific issues or problems in quirements. However, mechanical pointing of the reference system that would be of impor- the antenna would not have to be performed tance in formulating decisions concerning the with high accuracy, since the electronic phas- research, evaluation, development, demon- ing and pointing of the antenna subarrays stration, and deployment of satellite power would be insensitive to angular deflections of stat ions. the antenna of upto100. ● Performance. A major issue in the reference In addition to the equipment for satellite system design is the tremendous scale of the station keeping and attitude control, it would satellite. The level of 5 GW (net output be necessary to provide routine maintenance power) is based on scaling assumptions that of both the space and ground segments. Poten- could be subject to considerable change tial maintenance problems in the space seg- (e.g., the transmission frequency, the an- ment, in addition to the expected routine re- tenna and rectenna power densities); multi- placement of components, include the effects ple rectennas served by a single satellite also of solar wind, cosmic rays, micrometeoroids, constitute a potential variation. and impacts by station-generated debris. Aside ● The overall efficiency of the entire system from the solar wind and cosmic radiation ef- would be subject to considerable variation fects on solar cells, which would require active either up or down, and would be a key factor annealing of the silicon cells, none of these ef- in all cost and technology tradeoffs. Al- fects would appear to introduce significant though all system elements would involve maintenance problems or costs, based on ex- known technology, there is considerable un- tensive past and current experience with oper- certainty about how their efficiencies might ational satellites powered by photovoltaic add up when assembled together. celIs. ● Powerplant lifetime, assumed to be 30 years Repair and replacement of the solar blan- for the reference system, could actually be kets and more than 100,000 70-kW klystrons in greater or less depending on a number of the transmitting antenna are estimated to re- economically interrelated factors (e. g., ease quire a crew of from 5 to 20 people at the 21 of replacement of damaged components, geostationary orbit construction base, along sudden technological advances in compo- with the necessary transportation, support, nent efficiencies, etc.) This would affect all and resupply (e. g., station-keeping propellant) economic projections, even allowing for services. high-discount rates. Maintenance requirements of the rectenna ● The total mass in orbit, one of the critical and substation are also primarily associated parameters in assessing costs and launch- with repair and replacement of their biIIions of related environmental impacts, depends on components. Although a certain degree of re- a number of factors stilI subject to consider- dundancy is built into the system, a mainte- able variation. The power CoIlection/conver- nance crew would still be required to replace sion system is an obvious factor; the refer- storm-damaged rectenna sections and routine ence system’s two photovoltaic options are failures of both rectenna and substation equip- indicative of the significance of that trade- ment. off. The antenna mass is also important. Technical Uncertainties of Prospects for revising the reference-system’s the Reference System 100:1 ratio of rectenna-to-antenna area could have major impact on the overall sys- Although most observers accept the basic tem cost and performance. The 25-percent scientific feasibility of the SPS system con- contingency factor is another major factor 2’ DOE, op cit subject to revision if R&D mature. Ch. 5—Alternative Systems for SPS ● 75

SPS would require an extensive program of Silicon cells are subject to serious degra- research and testing of the numerous satellite dation by high energy electrons and pro- and terrestrial components of the system tons in the solar wind released by solar before planning for a demonstration satellite flares. One study” estimates that the ac- could be completed. In addition, substantial cumulated particle damage would de- improvements in components and overall tech- grade the output from the cells by 30- nology would have to occur before the SPS percent during the 30-year nominal life of could meet the performance specifications of the satellite. The resulting damage could the reference system. However, the current be repaired periodically by annealing the reference system does not constitute a pre- cells by either a laser or an electron beam. ferred system. It is, perhaps, technically feasi- The beam would sweep across the surface ble but certainly not an optimum design. It was of the cells and heat them briefly to sev- chosen by NASA/DOE as a model and a refer- eral hundred degrees centigrade. Very lit- ence to be used in the assessment process. As tle is known about either process in the such it has the inherent I imitation that as new laboratory and nothing at ail about how information becomes available the design be- they would work in space or how much comes progressively obsolete. energy they would use to anneal the surface of the photovoltaic cells. However, The following items summarize the major experiments have shown that annealing technical uncertainties for the reference sys- by electron beam is much more efficient tem and suggest possible ways to alleviate 23 than laser annealing. Because no long- them. term studies have been done, the suita- . Photovoltaic cells. The reference system bility of silicon cells for extended dura- specifies a silicon solar cell efficiency of tion space applications is in question; 17-percent and a mass of 2 grams per peak however, they have demonstrated ex- watt (g/Wp). Current space-rated single cellent performance over a period of crystal silicon cells operate at 12- to 16- about 10 years in operating spacecraft. percent efficiency. However, they are GaAs cells appear to be a more realistic about nine times as massive (18 g/Wp) as candidate for a reference-type satellite, called for in the reference system and though they have received much less at- they cost about $70/Wp (1980). The refer- tention than the silicon cells. GaAs cells ence system assumes a cell cost of about reach higher efficiencies and can operate $0.17/Wp. Although the issue of costs will at higher ambient temperatures than sili- be addressed in more detail in a separate con cells. Laboratory models of GaAs section, it is clear that meeting all three cells have reached efficiencies as high as goals for the silicon cell blanket would 18 percent.24 Because of their currently present manufacturers of current cell higher unit cost, the GaAs array would technology with an extremely difficult probably require refIectors to concentrate task. Normal advances in cell production the Sun’s rays on the cells and thereby techniques would readily result in the reduce the required cell area. Aluminized necessary efficiency increase. However, Kapton has been suggested as a reflective the burden of achieving a nine times material because of its low thermal coeffi- reduction in weight along with a reduc- cient of expansion and low mass density. tion in costs of a factor of 400 makes it highly unlikely that an SPS could be built 2*C R Woodcock, “SPS Silicon Reference System,” The Fina/ using single crystal silicon cells. Proceedings of the Solar Power Sate//ite Program Review, If efficiency-mass-cost goals were met, DOE/NASA Conf-800491, July 1980, there would still be the problem of cell “B E. Anspaugh, J. A Scott-Monck, R. G. Downing, D W. Moffett, and T. F Miyahira, “Effects of Electrons & Protons on lifetime in space and the related problem Ultra Thin Silicon Solar Cells, ” J PL contract No, NAS7-1OO. of the feasibiIity of annealing the surface. “lbld 76 . Solar Power Satellites

Here, again, whether Kapton and GaAs natives to the klystron may provide better cells can maintain their integrity over the noise and harmonic control (see section 30-year design lifetime of the satellite is on alternatives below). unknown. Considerably more study would ● Space transportation. The problems inher- be needed to determine the feasibility of ent in developing the capability to trans- this option. port SPS components to LEO and CEO are ● Space charge and plasma effects. Because those of extending a mature technology, of the high voltages associated with oper- i.e., there is sufficient understanding of ation of the klystrons, electrical charge the problems to be faced that there is lit- buildup in the satellite components could tle doubt that the appropriate vehicle cause arcing and subsequent failure of could be developed. The most important certain components. question is whether the necessary massive ● Rotary joint/slip rings. Although the basic loads could be transported for sufficiently technology of building a rotary joint and low costs, i.e., would reusable vehicles an associated slip ring (for electrical con- prove economic? In this area, much can tinuity) is well-known, considerable uncer- be learned from experience with the shut- tainty surrounds their construction and tle operation on the scale of the reference I n addition to economic concerns, there satellite in a space environment. Because are additional technical questions relating it would operate in a gravity-free environ- to environmental effects that would re- ment, the design demands would be dif- quire study. For instance, can the launch ferent than they are for terrestrial designs. vehicles fly trajectories that would keep ● Klystrons. Current klystrons last about 10 the effects of ionospheric contamination years, but these are tubes especially se- to a minimum? Would it be possible to lected for their long life characteristics substitute other technologies for the and they operate at much lower power argon ion engine proposed for the refer- levels than the 70 kW required of reference system (see ch. 8). ence system klystrons. High-power klys- ● Construction, operations, and mainten- trons do exist, but they operate in a pulsed ance. There are unresolved questions mode, not continuously as the reference about the productivity of humans and ma- system klystrons would have to. The an- chines in the space environment. Some tenna’s phased array control system automated equipment has been built and would need considerable development tested on Earth, but considerable develop- and testing. Although pilot beams have ment would be needed to choose the best been used in other applications, and the ratio between automated and human technology is therefore known, it is tasks. unclear whether the power beam would leave the ionosphere sufficiently unaf- fected to allow for undisturbed passage of Alternatives to the Reference the pilot control beam. System Subsystems Although harmonics and other noise One of OTA’s goals is to explore the possible produced by the klystron or alternative alternatives to the reference system. Some op- transmitting device would seem unlikely tions improve specific components of the ref- to affect the natural environment adverse- erence system. Others would require signifi- ly, they could cause radio frequency inter- cant redesign of the overall system. This is ference for communications systems (see because the reference system is composed of a the discussion of ch. 8). This problem number of interlocking components, some of might be severe and wouId need extensive which depend heavily on the other elements of study, but most experiments could be car- the system. Thus, a radical change in one com- ried out in ground-based testing. Alter- ponent might require numerous other system Ch. 5—Alternative Systems for SPS • 77

changes in order to create the most efficient ● Photoklystron. This device, which is stilI in overall design. the very early stages of study, both con- verts the sunlight directly to microwave A number of alternative subsystems and sys- power, and transmits it. If successful, it tems were considered in the process of elect- could replace both photovoltaic cell and ing the reference system design. Advances amplifier. have been made in some components that ● Offshore rectennas. For highly populated were previously rejected. In addition, consid- European and U.S. coastal areas, recten- eration of some of the above-mentioned tech- nas mounted in the shallow offshore sea- nical uncertainties has engendered new de- beds offer some advantages over long signs that could alleviate these uncertainties transmission lines from suitable land- or resolve some of the technical problems en- based rectennas. countered in the reference system. The following summary lists a number of THE SOLID-STATE SYSTEM subsystem options that could be considered as Two system approaches using solid-state alternatives to the reference system. A more devices have been considered for the SPS. The detailed discussion of each can be found in ap- most direct of these simply replaces the kyls- pendix A. trons and slotted waveguides in the reference Solar thermal power conversion. Either a system by solid-state amplifiers and dipole Brayton- or Rankine-cycle engine offers antennas maintaining essentially the same higher efficiency energy conversion than basic configuration as that of the reference photovoltaics. However, they currently system (fig. 9); the second approach complete- suffer from limitations on the means for ly revises the satellite configuration by inte- heat rejection. grating the antenna and solar array in the Thermionic, magnetohydrodynamic Earth-facing “sandwich” configuration, using a or wave energy exchanger technologies movable Sun-facing mirror to illuminate the might eventually find use in combination solar array (fig. 16). A number of alternative with the Rankine or Brayton cycle. sandwich configurations have been explored Photovoltaic alternatives. Materials other but at the moment the configuration of figure than silicon or gallium arsenide may even- 16 seems to be the best.25 tually prove more viable for use in the Another related subsystem option uses the SPS. Currently none of the other obvious multibandgap photovoltaic cells discussed options meet the projected standards for earlier, possibly in conjunction with selective efficiency, low mass, materials availabili- filtering to reduce solar-cell temperatures. ty, etc., that would be needed for satellite When such cells are utilized in the sandwich use. Different sorts of concentrator sys- configuration of figure 16, they offer consid- tems are also of interest, as is the possi- erable potential mass reduction. A recent pre- bility of using single cells or a combina- liminary case study26 compared sandwich-type tion of cells that respond to a wide por- systems such as that of figure 16 employing tion of the solar spectrum. A possible ap- single-bandgap GaAs photocelIs similar to proach would be to use a combination of those of the reference system but having high- al I these variations. er concentration ratios (CR) with optimized Alternative microwave power converters. multibandgap photovoitaics. Such a configu- Several devices other than the klystron ration would result in an approximate W-per- have been considered for converting elec- cent increase in power delivered per kilogram. tricity to microwaves and transmitting them to Earth including the magnetron, which offers the principal potential ad- “G M Hanley, et al , “Satellite Power Systems (SPS) Concept vantage of cost and low noise, and the Deflnitlon Study, ” First performance Review, Rockwell international report NO SSD79-01 63, NASA MSFC contract No solid-state amplifier whose reliability NAS8- )2475, Oct 10, 1979 could be very high and mass low. ~bl bld

78 ● Solar Power Satellites

Figure 16.—The Solid-State Variant of the Reference System

Sunlight I

Reflected sunlight Detail of solar cell blanket panel

Solid-state amplifier panel

Microwave / power to Earth /’ Solar array/microwave antenna sandwich panels

SOURCE: G. M. Hanley, et al., “Satellite Power Systems (SPS) Concept Definition Study, “First Performance Review, Rockwell International report No. SSD-79-0163, NASA MSFC contract No. NAS-8-32475, Oct. 10, 1979.

LASER TRANSMISSION

Lasers constitute an alternative to micro- ● The use of low Sun-synchronous rather wave transmitters for the transmission of than high geostationary orbits for the mas- power over long distance.27 They offer the fun- sive space power conversion subsystem damental advantage that at infrared wave- might be possible. (A Sun-synchronous or- lengths, energy can be transmitted and re- bit is a near-polar low orbit around the ceived by apertures over a hundred times Earth that keeps the satellite in full smaller in diameter than the microwave beam. sunlight all the time while the Earth ro- This obviously would reduce the size and mass tates beneath it.) In this suggested system, of the space transmitter and the land-area re- the laser would beam its power up to low- quirement of the ground receiver. But perhaps mass laser mirror relays in geostationary even more important, the great reduction in orbit for reflection down to the Earth aperture area would permit consideration of receiver, an arrangement that might con- fundamentally different systems. For example: siderably reduce the cost of transportation, since the bulk of the system mass is in LEO rather than in GEO. However, sys- W H power Satellites: The Laser Option,” Astronautics and Aeronautics, tem complexity would be increased due March 1979, pp. 59,67, to the need for relay satellites. Ch. 5—Alternative Systems for SPS ● 79

● Because the mass of the laser transmitters beam) from about 1 to nearly 50 percent dur- would not dominate the satellite, as does ing the past decade. the reference-system microwave transmit- Of all the currently operating CW lasers, ter, laser satellites would not benefit near- 29 only the electric discharge laser (EDL) seems ly so much by large scale as the reference a feasible alternative for the SPS. The gas dy- system satellites. The resulting smaller namic laser (CDL) suffers from very low effi- systems would improve the flexibility of ciency if used in the closed cycles necessary terrestrial power demand matching, pro- for space (i.e., the gas supply must be circu- vide high degrees of redundancy, permit a lated, cooled, and reused). Chemical lasers re- smaller and therefore less costly system quire a continuous propellant supply that demonstration project, and might even makes them also unsuitable for long-term use preclude the need for ultimate develop- in space. ment of an HLLV. ● The small size of the receiving station High-power density at 50-percent conversion would make it possible to employ multi- efficiency levels has been achieved for EDLs, ple locations close to the points of use, but only in the open-cycle mode for short time thereby simplifying the entire ground dis- periods. The closed-cycle systems needed for tribution and transmission system. It SPS have yet to be tested, even in the labora- would also open up the possibility of tory. In theory, they should achieve high effi- repowering existing powerplants, regard- ciencies in that mode as well, but considerable less of their size, simply by replacing their improvement in the available technology steam generating units with laser-heated would be required to reach the necessary boilers and/or superheaters. goals. The most important technical disadvantages In addition to using improved designs of cur- of laser-power transmission are the very low rently operating lasers, several advanced con- efficiencies of present laser-generation and cepts have been suggested. Of these, the solar- power-conversion methods, low efficiency of pumped laser and the free electron laser (FEL) laser transmission through clouds and mois- seem most promising for the long term. ture, and the relatively undeveloped status of ● Solar-pumped lasers. Figure 17 illustrates laser power-system technology in general. the concept of a solar-pumped laser. The The laser system would consist of three energy contained in sunlight directly ex- distinct elements: the laser-generation sub- cites a combination of gases confined be- system, the laser-to-electric power-conversion tween two mirrors, which subsequently subsystem, and the laser beam itself. “lase” and transmit the captured energy. It suffers the drawback that because only Laser Generators a part of the solar spectrum is useful in ex- citing any given Iasant gas, its conversion Although the laser has become a well-known efficiency is likely to be fairly low. How- and widely utilized device in industry, the ever, elimination of the need for a sepa- high-power continuous-wave (CW) laser gen- rate electric power-generating system, erators needed for SPS are still in the and the consequent reduction in mass and advanced-technology or, in many cases, the complexity, could more than compensate 28 early research phase. However, the technol- for this drawback. Further, in comparison ogy is improving dramatically as exemplified with other laser systems, the solar- by the growth of laboratory-demonstrated con- pumped laser’s efficiency need be only as version efficiencies (input power to laser good as the combined power-generating

28j Frank Coney bear, “The Use of Lasers for the Transmission “G W Kelch and W. E. Young, “Closed-Cycle Gasdynamic of Power, ” in Progress in Astronautics, vol. 61, A IAA, N Y , Laser Design Investigation, ” Pratt & Whitney Aircraft, NASA )ui~ 1978, pp. 279-310 Lewis Research Center report No CR-135530, Jan 1,1970. 80 • Solar Power Satellites

Figure 17. —Indirect Optically Pumped CO/CO2 Mixing Laser

Q Ps SEP solar n

SOURCE: R. Taussig, P. Cassady, and R. Klosterman, “Solar Driven Lasers for Power Satellite Applications,” in Firra/ Pro ceedings of SPS Program Review, Department of Energy, p. 267

system and laser generator of other laser Free-Electron Lasers (FEL) systems (about 7.5-percent for a photo- An FEL is powered by a beam of high-energy voltaic-powered carbon monoxide (CO) electrons oscillating in a magnetic field in such EDL30). a way that they radiate in the forward direc- Although the information exists to deter- tion (fig. 18). A number of pulses reinforce the mine the applicabiIity of solar-pumped lasers stored light between the mirrors, generating a to SPS, adequate studies have not been done. coherent laser beam. The high-energy density There is as yet little or no realistic basis for the of the relativistic electron beam is theoreti- mass, efficiency, and cost projections pro- cally capable of producing very high-power posed by several authors.31 32 33 34 density lasers, and the emitted frequency is tunable simply by changing the electron ‘“R. E. Beverly, “Satellite Power Systems (SPS) Laser Studies energy. Technical Report, Vol. 1, Laser Environmental Impact Study,” Rockwell International SSD-80-0119-I, August 1980 Although efficiencies are theoretically pro- “W. S. Jones, L. L. Morgan, J. B, Forsyth, and J Skratt, “Laser Power Conversion System Analysis: Final Report, Vol. I l,” Lock- jected to be quite high (around 50 percent for 35 heed Missiles and Space Co., report No LMSC-D673466, NASA the combined FEL and storage ring ), it is not report No. CR-1 59523, contract No NAS3-21 137, Mar 15, 1979 known whether such efficiencies could be 32 Claud N Bain, “Potential of Laser for SPS Power Transmis- reached in practice. In addition, the system sion, ” report No R-1 861, PRC Energy Analysls Co , DOE contract No. EG-77-C-01-4024, September 1978 mass per unit power output and the ability to 3JJohn D. G. Rather, “New Candidate Lasers for Power Beam- ing and Discussion of Their Appl icatlons, ” I bid,, pp. 313-332. ‘5John W Freeman, William B. Colson, and Sedgwick Simons, 34 Daryl J. Monson, “Systems Efficiency and Specific Mass Esti- “New Methods for the Conversion of Solar Energy to R. F. and mates for Direct and Indirect Solar-Pumped Closed-Cycle High- Laser Power, ” in Space Manufacturing ///, Jerry Grey and Energy Lasers in Space,” ref 105, pp 333-345 Chrlstlne Krop (eds ) (New York AlAA, November 1979). Ch. 5—Alternative Systems for SPS ● 81

Figure 18.—The CATALAC Free Electron Laser Concepts

SOURCE: R. Taussig, P. Cassady, and R. Klosterman, “Solar Driven Lasers for Power Satellite Applications,” in Final Pro- ceedings of SPS Program Review, Department of Energy. p. 267 scale to the size and power levels of a laser Transmission of the laser beam through the SPS are impossible to predict reliably at this atmosphere is also affected by a phenomenon time. 36 called “thermal blooming;” i.e., heating of the atmosphere that causes it to act Iike a lens and Laser Transmission distort the laser beam. Scientists are currently divided on the significance of this issue and As in the case of microwave transmission, opinions range from assertions that it is a ma- the fundamental parameter that governs much jor factor38 to suggestions that it could of laser transmission performance is the fre- be avoided altogether by selecting the trans- quency (or wavelength). At ultraviolet or visi- mitting wavelengths carefully.39 Considerable ble wavelengths, absorption losses in the at- classified research is now being carried out on mosphere are higher than for infrared wave- this effect in connection with laser-weapons lengths. The wavelength also affects the effi- research. Some of this work might be applica- ciency of the laser power absorption and con- ble to SPS use, though in general the military version equipment. lasers are pulsed, not CW systems. The differ- At the wavelengths of CO or CO, EDLs, (5 to ence could be critical and should be studied 10 microns), the primary mechanism of beam carefulIy. attenuation is molecular absorption. Scatter- With regard to laser optics, it is important to ing by molecuIes or by aerosols in clear air is develop components capable of low-loss, high- relatively unimportant. Attenuation of the power-density transmission and reflection of beam by aerosols under hazy or cloudy condi- laser light.40 It appears that adequate tech- tions is quite significant and can completely nology for SPS systems has a high probability block the beam if the clouds are thick enough. of being available within the next 20 to 30 Although it is apparently possible to burn a years, due primarily to advances being made in 37 hole through thin clouds, the attenuation of current military laser research and technology energy is appreciable, and because clouds are programs. seldom stationary, the laser would continually encounter new water droplets to vaporize.

‘s Beverly, op. cit. “Jones, et al , op cit 37E. W. Walbridge, “Laser Satellite Power Systems, ” Argonne “Beverly, op. cit National Laboratory report No AN L/ES-92 40 Baln, op cit 82 Ž Solar Power Satellites

Transmission options for SPS lasers are eral lasers making up the beam, and each essentially of two types: a narrow, highly con- beam by itself would transmit far too little centrated beam or a wide, dispersed beam (fig. power to cause any problems. Adaptive optics 19). Advantages of the narrow beam are the systems are being studied for use in military reduced land area needed and the smalI size of directed energy weapons and look promising.” the ground power-conversion system; prob- It should be emphasized that the overall sys- lems include potential environmental and tem constraints might be quite different for safety impacts of the high-intensity beam, con- the large CW lasers needed for SPS than for cerns over military uses, and the need for so- pulsed military examples. phisticated high-temperature receivers and power-conversion equipment. Advantages of Laser-Power Conversion at Earth the dispersed beam are its less severe environmental impact, the possible use of low-per- Several approaches are possible for convert- formance optics, and simplicity of low-power- ing high-energy-density laser radiation to use- density receiving systems. Disadvantages in- ful electric power. The technology of laser clude relatively high atmospheric dissipation, energy converters is relatively new, but prog- larger land area required and the large mass of ress has been rapid. Laboratory models have Earth receptors. It is probably too early to achieved conversion efficiencies of 30 to- 40 make an informed selection between the two percent and designers project eventual effi- options, but the narrow-beam approach ap- ciencies of 75 percent for some versions. Table pears to offer the principal benefit compared 6 summarizes the available technology and 42 to reference-system microwave transmission. projects future potential efficiencies. A final concern is the ability to point and The Laser-Based System control the beam to make sure it would always remain within the designated receiver area and Lockheed 43 has generated one possible laser to shut it off instantly should it stray. The system (fig. 20) that utilizes power satellites in adaptive-optics approach to beam control 4’Claud N Bain, “Power From Space by Laser,” in “High-Pow- (e.g., phased-array) such as would be used for ered Lasers In Space, ” Astronautics and Aeronautics, vol. 17, the microwave beam, appears adequate to March 1979, pp 28-40 provide the necessary pointing accuracy and “(;eorge Lee, “Status and Summary of Laser Energy Conver- sion, ‘ In Progress in Astronautics, VOI 61 Al AA, N Y , July to ensure safety, since any loss of phasing con- 1978 pp 549-565 trol would cause loss in coherence of the sev- 4’Jones, et al , op clt

Figure 19.—Optics and Beam Characteristics of Two Types of Laser Power Transmission System (LPTS) Concepts

Optics Optics m

SOURCE: Claud N. Bain, “Potential of Laser for SPS Power Transmission,” report No. R-l WI, PRC Energy Analysis Co., DOE contract No. EG-77-C-01-4042, September 1978. Ch. 5—Alternative Systems for SPS ● 83

Table 6.—Projections for Laser Energy Converters in 1981-90

Current 1981-90 Photovoltaics...... —30% efficiency —45% efficiency —megawatt power levels —megawatt power levels —wavelengths below 1 micron —wavelengths below 1 micron Heat engines ...... —Piston engine: Otto or diesel cycles —Turbine —50% efficiency —75% efficiency —1-10 kW —megawatt power levels —wavelengths near 10.6 microns —wavelengths near 5 microns Thermionics ...... —40% efficiency —50% efficiency —1-10 kW —megawatt power levels —wavelengths near 10.6 microns —wavelengths near 5 or 10 microns Photochemical cells ...... —Photoassisted dissociation of water —Photoassisted dissociation of water —15Y0 efficiency —30% efficiency —wavelengths near 0.4 microns —wavelengths near 0.6 microns Optical diodes ...... —Evaporated junction arrays — Evaporated junction arrays —not ready to convert power —50% efficiency —megawatt power levels —respond to wavelengths from UV to over 10 microns

SOURCE: George Lee, “Status and Summary of Laser Energy Conversion, “ in Progress in Astronautics, vol. 61, AlAA, N. Y., July 1978, pp. 549-565. low Sun-synchronous orbit and relay satellites the significant difference in space basing (i. e., (laser mirrors) both in LEO and CEO. One geo- LEO rather than CEO) which it presents com- stationary relay serves each power satellite. pared to the reference system. Because of the Based on an analysis of five candidate systems significant uncertainties present in the laser in three power ranges, Lockheed selected a systems concepts and the relative lack of tech- CO, EDL powered by a wave energy exchanger nology base for laser devices, the optimum (EE) binary cycle and a similar binary cycle for laser system would undoubtedly look rather ground power conversion. different from any system so far devised. The specific 500 MW system selected is dia- A laser system that used photovoltaic arrays gramed in figure 21; hardware details of the to collect and convert the Sun’s energy would power satellite appear in table 7, and the Over- . suffer from the fundamental difficulty that the all system characteristics are summarized in overalI efficiency of the system wouId be quite table 8. low compared to projected reference system efficiency .45 The major limiting factors are the A major potential advantage of the laser projected efficiencies of the laser itself (50 per- system is that it could be demonstrated via a cent for an EDL), the atmospheric transmis- subscale 500-kW pilot program using the space sion (84 to 97 percent), and the conversion effi- shuttle to deliver the power and relay satellites ciency of the terrestrial receptor (40 to 75 per- into LEO orbits. cent). When multiplied together with the Other laser systems are possible. For exam- higher efficiency of other system components, ple, Rockwell44 has investigated a geosyn- they result in an overall efficiency of 17 to 36 chronous laser SPS powered by photovoltaic percent after photovoltaic conversion of sun- ceils and using 20 to 24 100-MW CO EDL light to electricity to power the laser. When the lasers. The CO laser was chosen because it has efficiency of the solar cells (17 percent) is greater overall efficiency and is lighter than a taken into account, the overalI system efficien-

C02 laser. cy falls to only 2.8 to 6 percent compared to the projected reference system efficiency of 7 This study will use the LEO-based C0 laser 2 percent. Although this decrease would con- system in its subsequent analysis because of

‘*Beverly, op. cit. 45D0E, op. cit.

Ground site SOURCE: W. S. Jones, L. L. Morgan, J. B. and J. “Laser Power Conversion Analysis: Final Report, Vol. Lockheed Missiles and Space Co., report No. NASA report No. CR-159523, contract No. 137, Mar. 15, 1979.

Figure 21 .—Components of the Laser Concept

Synchronous relays

Occulted , Power

SOURCE: W. S. Jones, L. L. Morgan, J. B. and J. “Laser Power Conversion System Analysis: Final Report, Vol. 11,” Lockheed Missiles and Space Co., report No. NASA report No. CR-159523, contract No. 137, Mar. 15, 1979 Ch. 5—Alternative Systems for SPS • 85

Table 7.—500 MWe Space Laser Power System

Power generation Spacecraft, EE/binary and structure, Transmitter aperture Collector Solar cavity cycle conditioning Laser radiators, etc. and optical train Unit efficiency (%) . . . . . 85 86 73.5 93.1 23 — 98.7 System efficiency (%) . . 85 73.1 53.7 50.0 11.5 — 11.4 Power in (MW)...... 7,913 6,726 5,784 4,251 3,958 — 910 Power out (MW)...... 6,726 5,784 4,251 3,958 910 — 899 Orbital weight (kg) . . . . . 242,850 517,750 1,326,330 717,660 1,809,000 128,653 97,811 Spacecraft 4,108 Telescope (2) 89,812 Structure 94,433 Beam reduction 5,379 Radiators 6,032 Phasing array 1,539 Stabilization Optical train 1,181 24,080

Space Atmospheric Ground Thermal Binary Electrical transmission Space relay transmission receiver cavity cycle generation Unit efficiency (%) . . . . . 95 99 85 96 98 75.5 98 System efficiency (%) . . 10.8 10.7 9.1 8.7 8.5 6.5 6.3 Power in (MW). , ...... 899 854 845 718 690 676 510 Power out (MW)...... 854 845 718 690 676 510 500 Orbital weight (kg) . . . . . – 105,438 — — — — — Transmitter 44,703 Receiver 46,729 Optical train 945 Spacecraft 5,900 Radiators 5,762 Structure 1,023 Miscellaneous 376

SOURCE: W. S. Jones, L. L., Morgan, J.B. Forsyth, and J. Skratt, “Laser Power Conversion System Analysis: Final Report, Vol. 11,” Lockheed Missiles and Space Co., report No. LMSC-D673466, NASA report No CR-159523, contract No. NAS3-21 137, Mar 15, 1979.

Table 8.—Laser Power Station Specification stitute a potential problem for the laser system, it must be emphasized that many other Solar power collected (MW)...... 7,913.0 Collector diameter(m)...... 2,710.0 complex factors (e. g., the smaller terrestrial Electrical power to laser(MW) ...... 3,958.0 receivers, or lower mass in GEO), might com- Laser power output (MW) (20 lasers pensate in complex ways for lower efficiency. at 45.5 MW each)...... 910.0 Transmitter, aperture diameter (m)...... 31.5 When added up, the combination might make Secondary mirror diameter (o)...... 3.0 the laser system more acceptable overall than Transfer mirror size (m) ...... 3.0 x 4.2 the microwave systems. ’b Mirror reflectivity (%)...... 99.85 Optics heat rejection (MW) ...... 11.8 Radiator area (m2)...... 2,656.7 Mirror operating temperature (“C) ...... 200.0 “Abraham Hertzberg and Chan-Veng Lau, “A High-Tempera- ture Ranklne Binary Cycle for Ground and Space Solar AppIica- SOURCE: W. S. Jones, L. L., Morgan, J. B. Forsyth, and J. Skratt, “Laser Power tions, ” m “Radiation Energy Conversion in Space, ” K W, Conversion System Analysis: Final Report, Vol. 11,” Lockheed Missiles and Space Co., report No. LMSC-D673466, NASA report No. Billman (cd,), Progress in Astronautics and Aeronautics, vol. 61 CR-159523, contract No. NAS3-21137, Mar 15, 1979. (New York, AlAA, July 1978), pp 172-185. 86 . Solar Power Satellites

MIRROR REFLECTION

Instead of placing the solar energy conver- increasing the orbit altitude and mirror size, sion system in orbit as in the reference SPS, which increases the size of the illuminated several authors have suggested using large or- ground circle and thereby permits the use of biting mirrors to reflect sunlight on a 24-hour larger ground stations.52 The orbiting mirrors basis to ground-based solar-conversion sys- themselves could probably be quite large (up tems. 47 48 49 50 to 50 km’ each) with very low mass density53 and still maintain their required optical sur- Typically, this option would use plane mir- face flatness in the presence of disturbing rors (fig. 22) in various nonintersecting low- forces. altitude Earth orbits, each of which directs sunlight to the collectors of several ground- A mirror system would offer the following based solar-electric powerplants as it passes potential advantages: over them (the so-called “SOL ARE S“ concept). ● The space segment would be simple and Each mirror would be composed of a thin of low mass. It would consist only of film reflecting material stretched across a sup- planar reflective thin-film mirrors. porting structure made up of graphite-rein- ● It would minimize the need for large-scale forced thermoplastic. As they pass within space operations, since recent designs range of the terrestrial receiving station, the allow terrestrial fabrication and packag- mirrors would acquire the Sun and the ground ing with automatic deployment i n space. station nearly simultaneously. They would ● The system would be modular and highly maintain pointing accuracy by means of built- redundant, i.e., there would be many iden- in reaction wheels. tical mirrors capable of mass production. ● The mirrors would operate at low-orbit al- Two typical “limiting cases” have been iden- titudes, thus not requiring the CEO trans- tified from among several alternatives.51 one portation system of some other alterna- wouId use a 1,196-km circular equatorial orbit 0 tives. (O latitude) serving 16 equatorial ground sta- ● It would eliminate the need for develop- tions each generating about 13 CW (baseload, ing microwave- or laser-transmitting tech- with minimum storage) and another 6,384-km nology. 40 ‘-inclination circular orbit serving four 375 ● The mirrors would reflect ordinary sun- GW ground stations at 300 latitude. Additional light, thus eliminating many of the poten- ground stations in each case (to accommodate tial damaging environmental effects due demand growth) could be achieved simply by to laser or microwave transmission. ● It could be used for a variety of terrestrial 47 Hermann Oberth, “Wege zur Raumschiffahrt, ” Oldenburg- uses where enhanced 24-hour sunlight Verlag, Berlin, 1929; also see “Ways to Spaceflight, ” NASA technical translation TT F-662 wouId be useful. SOLARES couId increase 48 Krafft A Ehricke (for example), “Cost Reductions in Energy the solar product fivefold over the same Supply Through Space Operations, ” paper IAF-A76-24, 27th lrr- system operating on ambient sunlight. ternationa/ Astrorraut;ca/ Congress, Anaheim, Calif , Oct. 10-16, ● 1976. Demonstration would be very inexpensive “K, W. Billman, W, P Gilbreath, and S W Bowen, “introduc- compared to laser or microwave options. tory Assessment of Orbiting Reflectors for Terrestrial Power Gen- eration,” NASA TMX-73,230, April 1977 ‘“K, W. Billman, W. P. Cilbreath, and S W Bowen, “Orbiting Mirrors for Terrestrial Energy Supply, ” in “Radiation Energy Con- version in Space,” K, W, Billman (ed ), Progress in Astronautics ‘2K W Billman, “Space Orbiting Light Augmentation Reflec- and Aeronautics Series, VOI 61 (New York Al AA, July 1978), pp tor Energy System: A Look at Alternative Systems,” SPS Program 61-80 Review, June 1979. “K. W. Billman, W. P. Gil breath, and S W. Bowen, “Solar “John M Hedgepeth, “Ult[ ghtweight Structures for Space Energy Economics Revisited: The Promise and Challenge of Or- Power, ” in “Radiation Energy Conversion in SpaceJ” K W, Bill- biting Reflector for World Energy Supply,” DOE SPS Program man (ed ), Progress in Astronautics and Aeronautics, vol. 61 (New Review, June 8,1979. York Al AA, j uly 1978), pp. 126-135.

Ch. S—Alternative Systems for SPS ● 87

Figure 22.–The Mirror Concept (SOLARES)

Photo credit: National Aeronautics and Space Administration

SOURCE: W. Bill man, “Space Orbiting Light Augmentation Reflector -. System: A Look at Alternative Systems,” Review, June 1979.

On the other hand, mirror systems would ● The mechanisms needed to keep the mir- possess the following potential disadvantages: rors pointed accurately might be compli- cated. ● They would require a large number of satellites each with individual attitude con- ● The mirrors might cause unwanted weath- trol. Maintenance might be expensive and er modifications around the ground sta- difficult to accomplish. tions (see below and ch. 8). 88 ● Solar Power Satellites

● Scattered light from the mirrors and the Table 9.—SOLARES Baseline System light beams in the atmosphere would in- configuration: terfere with astronomical research (see Space system 2 ch. 8). 4,146km inclined orbit, 45,800km total mirror area Ground system ● The large power production per site (10 to 6 sites with DOE 1986 goal solar cells @ 15% efficiency 135 GW) and necessary centralization of 11 0/0 overall system conversion efficiency, ~~-circle the electrical supply from them would not area = 1.168km2 each, 135 GWe each be attractive to the utilities (see ch. 9). Impact: Total system would produce 3.24 times current U.S. con- ● 2 2 The large area of the receiving sites (100 sumption, total area = 84 x 84km (52 x 52 mi ) 2 to 1,000 km ) would be likely to make Baselined costs (in 1977 dollars) land-based siting extremely difficult if not Implementation schedule impossible from a sociopolitical stand- 5-year development, design, test, and evaluation (DDTE) 2-year manufacturing and transport fleet facilities point (see ch. 9). preparation 6-year space and ground hardware construction System complete about 1995 The Mirror System Direct costs estimate (billions of dollars) Facilities ...... $ 47.30 54 The “baseline” Mark 1 SOLARES design Hardware...... 885.65 (table 9) would require a total mirror area of Total direct ...... $932.95 2 2 Indirect costs estimate (billions of dollars) nearly 46,000 km . If each mirror were 50 km , 15% contingency on direct costs ...... $139.94 about 916 of them would be necessary for a Design, development, test, and evaluation ...... 43.80 global power system that would produce a Interest a: Facilities ...... 23.58 total of 810 GW from six individual sites, or Hardware ...... 101.26 about twice 1980 U.S. electric generation. It DDTE ...... 41.01 was chosen for comparative purposes because Total indirect...... $349.59 it demonstrates the potential for large scale Total cost ...... $1,282.54 Indirect cost factor...... 1.38 energy output that might be achieved with mir- Installed cost per rated output ($/kWe)b...... 1,508 rors. It is by no means the optimum SOLARES Capacity factor(%) ...... 95 system. A low-orbit version (altitude 2,000 km) 1995 O&M costs: with 15 smaller ground stations (10,000 to Fixed ($/kW-y)...... 3 Variable (mills/kWh)...... 2 13,000 MW output) might be more feasible or Levelized capital cost (mills/kWh)C ...... 27.2 Levelized O&M cost (mills/kWh)d ...... 4.5 desirable. One of the principal features of the e SOLARES concept is that it could be used for Levelized busbar energy cost (mills/kWh) ...... 31.6 any energy use where enhanced sunlight would Comparison baseload power systems (CIRCA 1995): Conventional coal/nuclear mixf be used to advantage. By using many more Levelized busbar energy cost (mills/kWh)e ...... 45 smaller mirrors, the mass per unit area could Ambient sunlight photovoltaicf g be minimized, and the total mass in orbit for Levelized busbar energy cost (mills/kWh)...... 115 the entire baseline system then becomes about a4Y@ first year, 8% per annum until positive cash flow after Year 11. blncludes all direct costs, 157” contingency, interest during implementation at 4X105 tonnes. Thus, the entire SOLARES 8% per annum. c15% fixed charge rate 30 years at 60/0 annual inflatiOn. baseline system would require only the same d30 years at 6% annual’ inflation, e15y& fixed charge rate. mass in space as eight 5,000 MW reference sys- fsee text; these d. not include their historically eXtenSive R&D costs that are tem satellites. Included, in SOLARES costing. 91Jses same terrestrial costing algorithm as SOLARES that results in indirect Several Earth-based energy production cost factor of 1.37. SOURCE: K. W. Billman, W. P. Gilbreath, and S. W. Bowen, “Solar Energy - methods currently under development might Economics Revisited: The Promise and Challenge of Orbiting Reflector for World Energy Supply,” DOE SPS Program Review, be used in conjunction with orbital reflector June 8, 1979. systems: 1 ) photovoltaic arrays of varying sizes are projected for commercial deployment in plants should become commercially feasible the late 1980’s, and 2) solar-thermal electric in selected locations about the same time, possibly also for “repowering” of existing coal- or oil-fired fossil-fuel plants with solar boilers. 54 Billman, et al., “Solar Energy Economics Revisited. The Promise and Challenge of Orbiting Reflector for World Energy Much of the economic disadvantage of both Supply, ” op. cit types of solar-electric powerplants is associ- Ch. 5—Alternative Systems for SPS Ž 89 ated with the energy storage needed to allow of the accelerated evaporation produced by them to serve as intermediate or baseload the high-intensity solar radiation. plants. Should these plants prove to be even If the orbiting mirrors can disperse clouds of marginally successful, relieving their storage moisture around the SOLARES ground station, needs by keeping them I it for 24 hours a day by what effects may they have on the climate sunlight from orbiting reflectors would en- nearby? Large orbiting mirrors have been sug- hance the attractiveness of these terrestrial op- 56 gested for use in climate modification, but tions. their possible detrimental side effects have not The various benefits of a mirror system must been studied (see ch. 8). However, even if be weighed against the percentage of time the reflected sunlight could be shown to have a ground-based energy production facilities salutary effect on certain regions of the Earth, would be obscured by clouds, smog, fog, and there is no reason to believe, without further other atmospheric obstruct ions. However, study, that regions whose weather patterns there is some evidence” that the concentrated could benefit from enhanced sunlight would sunlight provided by the orbiting mirrors necesssariIy coincide with the SOLARES would tend to disperse water-based obscura- ground stations. tions such as clouds and fog, as a consequence — *’I Bekey and J E Nagle, “Just Over the Horizon in Space,” “Ibid Astronaut/es and Aeronautics, May 1980.

SPACE TRANSPORTATION AND CONSTRUCTION ALTERNATIVES

Space transportation and construction (with shuttle size vehicles at high launch rates could the possible exception of SOLARES) are com- be cheaper than developing and using larger mon to all the options. NASA contractors who launch vehicles (see section on costs). Perhaps developed the transportation, construction, the most obvious approach is to upgrade the and assembly plan for the reference system shuttle-based space transportation system to devoted considerable effort to the process perhaps five times the capability (i.e., total of winnowing out a host of alternative ap- mass to space in a given time as represented by proaches. Nevertheless, several other construc- payload size, launch rate, and turn-around) of tion/assembly schemes have been proposed for the present shuttle.57 various phases of SPS program development. The need to conduct relatively sizable ex- If feasible, they would mostly serve the pur- periments, and possibly prototype or demon- pose of reducing costs by using technology stration projects in geostationary orbits rather developed for other programs or by reconfigur- than in low-Earth orbits, would pose a serious ing the reference system scenario. Because transportation problem. Current space-shuttle transportation costs are a significant percent- upper stages, or “orbital transfer vehicles, ” are age of any systems cost (see section on costs not capable of carrying large payloads to geo- below), it would be important to explore these stationary orbit and are not able to support alternatives fulIy. any servicing operations there, since these units are not reusable. Transportation Several innovative approaches have been Transportation strategy in the early develop- suggested that circumvent the need for devel- ment phase and engineering verification is to oping new vehicles. One such approach em- use the shuttle or an upgraded shuttIe to their ploy; an in-orbit propel ant processing facility maximum capacities. In these, as well as later demonstration and production phases, using ‘7 Salkeld, et al,, op. cit.

83-316 0 - 81 - 7 90 ● Solar Power Satellites

built into one of the shuttle’s big “throwaway” sion. This concept is far more ambitious than propellant tanks to convert water into hydro- the in-space propellant processing scheme; fur- gen and oxygen –the best propellants for high- thermore, it depends on a device that, al- performance rocket engines. The water re- though tested extensively on Earth in experi- quired as the feedstock for this process would mental high-speed trains and in the laboratory, be carried into LEO as an “offload” on every has yet to be demonstrated at the scale and ac- space shuttle flight whose payload is less than celeration levels required by the orbital trans- the maximum shuttle capability. The hydrogen fer application. A modest research effort on and oxygen, after being liquefied and stored in this concept is currently being supported by the propellant processing facility’s tank, are NASA’s Office of Aeronautics and Space Tech- then used as the propellants for a reusable low- nology. thrust “space tug” whose principal component The production phase of the SPS program is also a leftover shuttle propellant tank. The would present a number of opportunities for tug, which replaces the cargo orbital transfer transportation alternatives that could not only vehicle of the reference system, would carry reduce production costs, but could also miti- SPS prototype or demonstration hardware up gate environmental and other impacts. Be- to CEO. Although such a system is rather com- 58 cause of the high proportion of total space seg- pletely defined, considerable technology ad- ment construction costs (both nonrecurring vancement and development would be re- and recurring) taken up by transportation, quired, e.g., for the in-orbit electrolysis and many of the proposed innovations center on liquefaction plants, the space-tug-develop- alternatives to the family of four transporta- ment, and the system logistics and integration. tion vehicles selected for the reference system. Cost estimates have not yet been released. Nevertheless, this concept represents an in- The most direct approach to transportation teresting suggestion for eliminating the de- cost reduction would be to improve the HLLV, velopment of a major new (or upgraded) since it absorbs the bulk of transportation launch vehicle just for an SPS demonstration, development and operations costs. The most thereby reducing the “up-front” costs of any likely technological alternative appears to be sizable SPS prototype or demonstration proj- the use of fully reusable single-stage-to-orbit ect. (SSTO) vehicles. 62 Very advanced winged SSTO vehicles that could reduce LEO payload Another scheme would use an electro- 59 delivery costs to the order of $1 5/km are pro- magnetic propulsion device called a “mass jected as becoming practical in the last decade driver” to provide orbital transfer thrust in- of this century, provided sufficient demand stead of the chemical-rocket-powered space exists. 63 tug. The mass driver is simply a solar-powered linear electric motor, which derives its thrust For orbital transfer the personnel and cargo by accelerating chunks of waste mass (e.g., orbital transfer vehicles selected for the chopped-up or powdered shuttle propellant reference system probably represent the best tanks) into space at high exhaust velocities. 60 61 available technology in the two principal op- Since it uses electricity, its energy could come tions: chemical and electric propulsion. directly from the Sun via photoelectric conver- Alternatives for routine high-mass payload 58 Central Dynamics Corp (Convair Dlvlslon), “Utilization of hauling might include solar sails, laser propul- Shuttle External Tank in Space, ” unpublished presentation, j une sion, and various forms of electric propulsion 1978. 5~F, Chiiton, B, H ibbs, H. Kolm, G K O’Neill, and J. phil lips, other than the ion (electrostatic) rocket de- “Electromagnetic Mass Drivers,” in “Space-Based Manufactur- scribed for the reference system, e.g., elec- ing From Nonterrestrial Material s,” G K C)’Neil I (cd.), Progress in Astronautics and Aeronautics, vol. 57 (New York AlAA, August 62 Beverly Z. Henry and Charles H Eldred, “Advanced Technol- 1977), pp. 37-61. ogy and Future Earth-Orbit Transportation System s,” in Space bochllton, et a]., “Mass-Driver Application s,” ibid , PP. 63-94. Manu(actur;ng Facilities //, jerry Grey (ed ) (New York: Al AA, “Gerard K O’Neill, “The Low (Profile) Road to Space Manu- Sept 1, 1977), pp 43-51 facturing,” Astronautics & Aeronautics, March 1978, pp. 24-32. “lbld Ch. 5—Alternative Systems for SPS ● 91

tromagnetic (plasma) thrusters or the mass teroidal materials could be even more favor- driver discussed above. None of these options able. The primary drawback is the high “up- has been studied in enough detail to make front” cost of establishing the necessary min- choices about them at the present time. ing base on the Moon and the space-based facility needed to construct and assemble the Space Construction SPS. Hence, it is not likely that nonterrestrial materials would be used in the prototype, As currently designed, the space component demonstration, or even the early phases of SPS of the reference system would be constructed production. However, if a commitment is in CEO. However, it may be more cost effec- made to produce a large-scale SPS system in tive to build the necessary facilities and CEO, the lunar materials supply option could satellites in LEO and transport them to CEO well be less expensive than the Earth-launched fully constructed. Such a scenario would re- option (including payback of the initial invest- duce the number of personnel needed in CEO merit) . 64 It has been argued that by “bootstrap- as well as lower the total mass that must be ping” the operation (i. e., using nonterrestrial transported there. material right from the beginning, not only to Introducing one of the LEO scenarios (i. e., build the SPS but to build all the necessary laser or mirrors) would open up significant facilities as well), there is no need for any new changes in the construction and transportation launch-vehicle development (a major element option for the SPS. Even a change in one major in the “up-front” investment); i.e., the present component of the reference system satellite space shuttle can provide all the Earth-launch space transportation needed to implement an could alter the ways in which the transporta- 65 tion and construction components are con- operational multi-SPS network. figured. For example, if the photovoltaic cells Decisions on the nonterrestrial materials op- were to be replaced by solar thermal conver- tion clearly hinge on the results of current and sion systems, it would be attractive to con- projected SPS technology studies and experi- struct satellites in LEO and transport them to ments. Sufficient research on the two techno- CEO on their own power because they would logical factors unique to nonterrestrial materi- suffer less from passage through the Van Allen als development—the mass driver (both for radiation belts. lunar materials transfer and for in-space pro- Of all the alternative options for SPS con- pulsion) and lunar materials mining and proc- —should be done so that a struction in the production phase, the prospec- essing capability tive use of nonterrestrial materials is perhaps decision to proceed with either the Earth or the most innovative and, ultimately, capable nonterrestrial materials options could be prop- of the maximum potential return on invest- erly made. Other study and research require- ment. ments for the nonterrestrial materials option include system analyses (including design of The basic premise of the nonterrestrial ma- an SPS that maximizes the use of lunar materi- terials option is that the cost, energy and mate- als), more intensive searches for appropriate rials requirements, and environmental impact Earth-approaching asteroids, and establishing of lifting the enormous cumulative masses capabilities for the host of space operational needed to establish and operate a system of functions needed for other space programs. many satellite power stations off the Earth can be markedly reduced by utilizing first lunar As is clear from the preceding discussion, it materials, and eventually materials obtained is difficult to establish a priori alternatives to from asteroids. The fundamental physical prin- construction, assembly, and transportation, ciple that supports this premise is that it takes over 20 times as much energy to launch an ob- “Davld L Akin, “Optimization of Space Manufacturing Sys- terns, ” in Space Manufacturing ///, Jerry Grey and Christine Krop ject to geostationary orbit from the Earth as it (eds ) (New York. AlAA, November 1979) does from the Moon, and the situation for as- b50’Nelll, op cit 92 . Solar Power Satellites

since each of the SPS alternative options essentially developed space shuttle; 3) max- would call for a different approach. General imizing the common utilization of technology guidelines can be identified, minimizing and development efforts by other programs transportation and construction costs during having related requirements (e.g., large com- the evaluation, development, prototype, and munications antennas and other large space demonstration phases by: 1) utilizing a phased, structures, spacecraft power generation, con- step-by-step approach (e. g., ground-based ex- trol and transmission, etc.); and 4) developing periments, only then followed by dedicated new transportation vehicles and construction space experiments); 2) maximizing use of the hardware only when economically necessary.

SPS COSTS

Although knowledge of the overall costs of Figure 23.—Reference System Costs a an SPS program will be essential to making a (dollars in billions) decision about developing the SPS, current cost estimates are inadequate. Today’s projections are based on extrapolations from current technology and in most cases assume major advances. Thus, the technical uncertainties of the concept are too great to provide a firm basis for economic analyses. Here, as in most other areas, it is only possible to develop the foundation for future analysis that would seek to reduce the current uncertainties.

Reference System Costs

The most detailed cost estimates have been made by NASA66 for the reference system (fig. 23). According to these estimates, which are based on detailed hardware specifications and associated transportation and industrial in- f restructure, achieving the first complete reference system satellite will require an investment of $102.4 billion over a 20-year period. Figure 24 illustrates one estimate67 of how the costs could be allocated over time. Each additional copy of the satellite and associated terrestrial facilities would cost $11.3 billion. Expenses are divided into the following phases:

● Research — $370 million. This phase of SPS development (table 10) is by far the small- est, constituting less than 0.4 percent of the total SPS program. About half of these

bbPiland, op. cit. “Woodcock, “Solar Power Satellite System Definition Study,” aNASA estimates—1977 dollars. op. cit. SOURCE: National Aeronautics and Space Administration Ch. 5—Alternative Systems for SPS ● 93

Figure 24.— How Cost Could Be Allocated Table 11 .—Engineering—$8 Billion — Millions Percent of dollars of total SPS...... $ 370 5 Test article hardware ...... 210 3 LEO base (8 man) ...... 2,400 30 Manned orbital transfer vehicle. . . 1,200 15 Shuttle flights...... 870 11 Shuttle booster...... 2,900 36 Management and integration . . . . 61 1 Total...... $8,000

NOTE: Percentages do not total 100% due to rounding errors. SOURCE: National Aeronautics and Space Administration.

Table 12.—Demonstration—$23 Billion

Millions Percent o of dollars of total Years Demonstrator: DDT&E ...... $2,700 12 SOURCE: National Aeronautics and Space Administration. Hardware...... 2,500 11 Pilot production facilities ...... 400 2 Shuttle DDT&E and fleet ...... 3,000 13 Construction: Table 10.—Research—$37O Million DDT&E ...... 3,100 13 Hardware...... 3,000 13 Millions Percent Space operations (4 years of dollars of total operations, construct bases, and demonstrations) ...... Power generation ...... $ 79 21 2,800 12 Personnel orbital transfer vehicle Power transmission ...... 40 11 (DDT&E and hardware)...... 1,700 Structures and control...... 22 6 7 Electric orbital transfer vehicle Space construction ...... 25 7 (DDT&E) ...... 1,800 Space transportation ...... 20 5 8 Demonstration rectenna ...... System studies...... 19 5 1,800 8 Management and integration . . . . 200 Research flight test ...... 165 45 1 $370 Total...... $23,000 SOURCE: National Aeronautics and Space Administration. SOURCE: National Aeronautics and Space Administration.

costs are chargeable to the development ciated rectenna and ground facilities to of the transportation system. collect and disperse electrical power to ● Engineering–$8 billion. This part of the the grid. The demonstrator requires a sec- program (table 11) contributes the com- ond generation shuttle and orbital trans- plex engineering knowledge necessary for fer vehicle to provide the transportation creating a useful space structure. The capabiIity to GEO. work includes developing an engineering ● Investment—$57.9 billion. By far the test article in LEO, capable of generating largest percentage (57 percent) of the non- 1 MW of power. It is the direct precursor recurring costs of the reference system are to the demonstrator and provides the test- devoted to this phase (table 13). In addi- ing ground for constructing and using col- tion to providing for the transportation lector and transmitting subarrays, a rotary and construction capabilities for the joint and satellite attitude control. space component, it also includes the ● Demonstration –$23 billion. This phase of costs ($7.8 bill ion) for developing the ter- the reference program (table 12) culmi- restrial factories needed to produce satel- nates in a 300-MW satellite and the asso- lite components. 94 ● Solar Power Satellites

Table 13.—SPS lnvestment—$57.9 Billion alone vary by a factor of 30 ($40 to $1 ,250/kg). Millions of Percent ● dollars of total Photovoltaic cells. GaAs cell cost esti- Heavy lift launch vehicle ...... $16,600 29 mates are extremely optimistic given the Development...... $10,500 18% current state of technology. Break- 0 Fleet (6 boosters, 7 orbiters) ... $ 6,100 11 /0 throughs will be needed to reach the Electric orbital transfer Vehicle (21 x 284)...... 6,000 design goals for mass, efficiency, and Construction bases ...... 17,200 30 costs. Silicon cell cost estimates are less Development...... $ 4,300 8% optimistic but will still require significant Hardware and launch ...... $12,900 22%. SPS development ...... 2,200 4 simultaneous reductions in mass and cost Ground-based factories and an increase in efficiency to achieve (klystrons, solar cells, etc.) . . . . . 7,800 13 the SPS goal (2 g/W, $0.17/Wp, and 17- Launch and recovery sites...... 7,300 13 Program management and percent efficiency). integration...... 800 1 ● Slip ring. It is not well enough defined to Total...... $57,900 appraise the slip ring components or their

SOURCE: National Aeronautics and Space Administration. operational capabiIity. ● Satellite electrical systems. The degree of detail is insufficient to judge the credibili- Though these are the best estimates currently ty of the cost estimates of the subsystem. available, they suffer from an unavoidable lack of specific engineering details, as well as Thus, the $102.4 billion estimate of “front from insufficient manufacturing experience end” costs and the $11.3 billion estimates for for most of the system components. Moreover, each satellite may be an optimistic estimate of in some areas, (e. g., klystrons, slip ring, phase SPS costs. control) current technology is inadequate to On the other hand, if unexpected break- define solutions to engineering problems. throughs were to occur in space transporta- Thus, the estimates could eventually turn out 68 tion, rectenna or satellite technology, the costs to be high or low. The DOE SPS Cost Review of the reference system could be lower than examined five different elements of the SPS now estimated. Since NASA estimates already reference design and concluded that the pro- assume some technological breakthroughs jected costs are “based on optimistic assess- (e.g., in solar cell production, space construc- ments of future technological and manufac- tion, rectenna construction), they are more turing capabilities. ” likely to be low than high. In either case, the ● Rectenna support construction. Projected estimates reflect a troublesome feature of the costs were found to be low by a factor of reference system —the high costs that are nec- 3 to 5. Automated production might essary to demonstrate the feasibility of the SPS reduce costs to a level more in keeping (about $31 billion). A further $71 billion would with the reference system estimates, but be needed to build and use a single reference significant advances over today’s meth- system satellite (investment of $57.9 billion ods would be needed. and a first satellite costing $13.1 billion). ● Graphite fiber-reinforced thermoplastic. Because the initial costs have a direct bearing Currently used for golf clubs, fishing rods, on financing the project, they are more fully and for any other use where low weight discussed in chapter 9. and high stiffness are required, this is the A number of opportunities exist for reducing recommended material for the satellite SPS development expenses. Some involve pur- truss work. The proposed structures are suing alternative concepts; others, revising the insufficiently defined to specify the costs. reference system. Because the reference sys- Estimates of future costs for the materials tem is by no means an optimal design, improvements could lead to significant cost ‘*J. H. Crowley and E J. Ziegler, “Satellite Power Sy5tems (SPS) Cost Review,” DOE/TIC-11190, MaV 1980 reductions. Common to all potential systems Ch. 5—Alternative Systems for SPS ● 95

would be the division of SPS development into private investment in space is strong for other the phases outlined above: research, engineer- reasons. Under these combined circumstances, ing verification, demonstration, and invest- the total risk to the U.S. taxpayer would be ment, with increasing commitment of re- substantialIy reduced. sources in each successive phase. For micro- One interesting option for reducing trans- wave and laser systems, space transportation portation costs of a CEO SPS would be to and construction would constitute a high per- assemble the satellite in LEO and send it to centage of the system costs in all phases. It is CEO under its own power. This might be in these areas that there would be a high particularly applicable to the demonstration potential for reducing overall costs. phase of the reference program, since it would The precise costs of an SPS program would avoid the need for premature investment in an also depend strongly on the nature and scope expensive manned geosynchronous construc- of national and global interest in space. If tion/assembly facility. commercial ventures in space grow at a strong Whatever their potential savings, all of these enough rate (e. g., for telecommunications sat- possibilities could only be evaluated after the ellites, space manufacturing, etc.), the current proper scale of a demonstration satellite had shuttle and its related technology would be in- been determined. This decision, in turn, would adequate, and pressures would be strong for depend on considerable terrestrial and space- developing expanded space capabilities. The based testing, some of which will take place in explosive growth of the domestic airline in- other space programs (see ch. 5). dustry since the 1930’s has been suggested as the appropriate model to use to investigate Because the HLLV would be used later on in this eventuality. 69 the production phase of the reference SPS absorbs the bulk of transportation costs, it is of Much of the technology and experience considerable interest to find less expensive needed for space construction (manned LEO ways of transporting mass to space. Some of and GEO bases, large-scale antennas, studies the alternative high-capacity transportation of space productivity, etc. ) and space transpor- vehicles have been discussed earlier in this tation (manned and unmanned orbital-transfer chapter. The heavy Iift launch vehicles achieve vehicles, shuttIe boosters, HLLVS, etc.) of SPS their cost reductions by economies of scale. It would be developed for other programs as has been suggested that smaller vehicles, well. Of these, the SPS program should bear perhaps only slightly larger than the current only its share. By charging only those costs space shuttle, could be used instead of the that are unique to SPS to the SPS program, its 71 much larger HLLV. The smaller vehicles front end costs would be reduced by a signifi- would use higher launch frequencies to cant amount. Seen in this light, the massive achieve the same or better benefits. According space capability needed for mounting an SPS to this proposal, the minimum-cost individual program would be less of an anomaly (given payload necessary to launch as many as five the future evolution of space technology), ’” reference SPS satellites to orbit is about 50 and SPS would need to shoulder fewer of the tonnes (compared to the Shuttle’s 30 tonnes). development costs for this capability. The prospects for employing routine airline- There is also the possibility that a percent- Iike launch practices opens a whole new ap- age of the investment phase could be shoul- proach to the logistics of major space manu- dered by private investment, thereby reducing facturing enterprises as well as providing the burden to taxpayers. This would be all the potential cost reductions for SPS. more likely to happen in a milieu in which

“C, R. Woodcock, “Solar Power Satellites and the Evolution of Space Technology, “ AIAA Annual Meeting, May 1980. “R. H Miller and D. L. Akin, “Logistics Costs of Solar Power 701 bid. Satellites,” Space So/ar Power Review, VOI 1, pp. 191-208,1980. 96 ● Solar Power Satellites

ALTERNATIVE SYSTEMS

Systems other than the reference system The Laser System might be more or less costly, depending on factors such as the achievable efficiency, the The largest unknowns for the laser system mass in orbit, and the state of development of are the efficiency, specific mass and the cost the alternative technologies that make up of the transmitting lasers themselves. This is these systems. At present, these alternatives because the technology of high-power CW are much less defined and their costs accord- lasers is in a relatively primitive state (current ingly even more uncertain than the reference CW lasers achieve outputs of 20 kw or greater, system costs. The following discussion summa- operated in a so-called loop move, i.e., the rizes available cost data and the greatest cost Iasant is recirculated). Space lasers for SPS uncertainties of the alternative systems. would have to operate at much higher outputs (megawatts) and at higher efficiencies (i.e., 50 The Solid-State System v. 20 percent) for current lasers. Concepts such as the solar pumped laser and the free electron ● The unit cost of the solid-state devices is un- laser are completely untried in a form that known. However, the semiconductor indus- would be appropriate to SPS. Therefore their try has considerable experience in producing costs are even more difficult to ascertain. In large numbers of reliable solid-state com- general it can be said that the cost of the ponents at low cost, and the learning curve system would be tied to the overall efficiency for such production is well-known. In princi- of the system and the amount of mass in ple, it should be possible to make a realistic space, but considerable study and some devel- prediction of costs when the appropriate de- opment would be needed to make suitably vice or devices are well characterized. reliable projections. ● Solid-state efficiencies. Present efficiencies ● Transportation. The laser systems that have are much lower than for the klystron. Cur- been explored project higher mass in orbit rent research is aimed at increasing their than for the reference system, which may operating efficiency (to reach at least 85 per- drive the cost of the laser system up. How- cent). ever, if a substantial portion of this mass is ● Mass in space. Current estimates of the mass in LEO rather than in CEO, the overall trans- per kilowatt of delivered power72 suggest portation costs might not exceed the trans- that the mass in space would be higher than portation costs of the reference system and that of the reference system making the could turn out to be lower. transportation costs higher as well. ● Demonstration. Because the laser system is Since many components of the solid-state intrinsically smaller it should be possible to system are shared with the reference system mount a demonstration project for consid- (e.g., the graphite fiber reinforced thermo- erably less than for the reference system. plastic support structures, the photovoltaic arrays, the rectenna design, etc.), it would be ● Terrestrial component. The ground stations possible to generate realistic relative costs if would have to have a certain amount of re- the above uncertainties are reduced. dundancy in order to accommodate laser transmission when cloudy weather obscures one or more receivers. The precise amount of redundancy would depend on the particu- lar location and would include extra trans- 7*G. Hanley, “Satellite Power Systems (SPS) Concept Defini- tion Study,” vol. 1, Rockwell International SSD-8O-O1O8-I, Oc- mission lines as well as extra ground tober 1980. receivers.

Ch. 5—Alternative Systems for SPS ● 97

The Mirror System ducing the same overall output computed on identical assumptions would cost 115 Figure 25 summarizes mirror system cost m i I Is/k Wehr. estimates for the SOLARES baseline case73 based on the DOE 1986 cost goals for photo- Since electricity production from the mirror voltaic cells. These “up front” cost estimates, system would depend heavily on the use of ter- which include contingency and interest on the restrial solar photovoltaic or solar thermal borrowed money, lead to an estimated level- systems, cost variations of either conversion ized busbar energy cost of 31 mills/kWehr system would have a strong effect on total compared to 1990 estimated costs of nuclear/ system costs. Figure 26 summarizes the effect coal mix of 45 mills/kWehr. I n comparison, a of varying several system parameters on the strictly terrestrial system of photovoltaics pro- cost of electricity delivered to the busbar in the SOLARES system. The three most sensitive 73 et “Solar Energy Economics Revisited: The Promise and Challenge of orbiting Reflector for World Energy parameters are solar cell efficiency, solar cell Supply, ” cit. cost per peak kilowatt and total space cost

Figure 25.—Elements and Costs, in 1977 Dollars, for the Baseline (photovoltaic conversion, 4,146 km, inclined orbit) SOLARES System

Solar cells

NOTE: Total costs are proportional to the areas of the circles. Interest and contingency constitute 33 percent of the total SOLARES costs. SOURCE: K. W. Billman, W. P. Gilbreath, and S. W. Bowen, “Space Reflector Technology and Its System Implications” AlAA paper 79-0545, AIAA 15th Annual Meeting and Technical Display, 1979.

98 ● Solar Power Satellites

Figure 26.—Sensitivity of the SOLARES Mirror (transport, construction, mirrors in space). A System to Variations in System Parameters cost over-run of about 2 times (to $1,000/pk kWe) could be tolerated before a busbar cost of 45 milis/kWehr wouId be reached. Similarly, a space system total cost over-run of a factor of 4.25 could be tolerated. Finally, because of the projected high energy production per unit of mirror mass in space, a twenty-three-fold increase in space transport cost (or $1 ,380/kg) would still result in a production cost of 45 mills/kWehr. For comparison, the charge for transporting mass to space by means of the space shuttle is estimated to be between $84 and $154 (1975 dollars). ”

74 National Aeronautics and Space Administration, “Space Transportation Reimbursement Guide,” JSC-11-802, May 1980 % Variation of parameters

SOURCE: Ken Billman, W. P. Gilbreath, and S. W. Bower, “Space Reflector Technology and Its System Implications” AlAA paper 79-0545 AlAA 15th Annual Meeting and Technical Display, 1979,