llllEING Volume I Solar Power Satellite Phase 1, Final Report Executive Summary System Definition Study 0180-25037-1 NAS_.; (. '.-.· /tc ;> 7c BllEING GENERAL #i ELECTRIC -r-GRUMMAN Arthur D IJttle Inc IRW

,, (!USA-CR-160370) SYSTE~ DEFINITION 1: EIRCUTIVE su"~\8Y 22e GJ/15

)-15636 T-1487 MA-731T : ;TEM 3 Solar Power Satalite System Definition Study Conducted for the NASA Johnson Space Center Under Contract NASY-1 S6.l6

Volume I PHASE I, FINAL REPORT Executive Summary 0180-25037-1

February !6. 1979

Approved By: f!i::t~ Study Manager

Boeing Aerospace Company Ballistic Missiles and Space Division P.O. Box 3999 Seattle. Washington 98 ! 24 01~25037-1

1.0 INTRODUCTION AND BACKGROUND

I. I History --~-o.-.. ·------"______...,_...... ·-- ~...... ,.,.. ·---.... -­ Solar power has long been recogniud as an ideal source of energy for mankind. It is natur-.llly available and plentiful. does not disturb the envi­ ronment. e.g .. by creation of wastes. and is itself free.

About ten years ago. a way of utilizing solar energy to gener-.tte electricity on a 24-hour continuous basis was proposed by Peter Glaser of A. D. Little. His proposal was to place the solar collectors in space. where they can collect sunlight continuously. can readily be aimed at the sun. and Figure 1. Solar Power Satellites: Th~ Principle where very brge collector areas can be obtained with relafo·ely litlle investment in material through the NASA Lewis Research Center to inves­ resources. Energy colleckd by these solar power tigate basic technical feasibility of the SPS concept. sat~llites (SPS's) would be transmitted to by The conclusions of that study were that the system eledromagnetk .nc~ms. The original Glaser pro­ is tel.'hnically feasible and could provide baseload posal. and most of the subsequ1:nt studies. have electricity from solar pow>!r for use on Earth. assumed the LISI.' of radio frequency s~ stems in the Additional studies and experiments. partly funded microwave frequenc~ range. Recently. the possibil­ by NASA mer the period 1973 to 1975. estab­ ity of las-:r bt·;uning has also bc.-tc"n rt•t·ogniud. lished the feasibility of efficient energy transmis­ sion at micr(lwave frequencies. In 1975 a demon­ Thi.' solar power satellill.' principle is illus­ str-Jtion con

~-~~- ...... I lo .. ' - • l -~I ...... 4.. STlllNGS/15m l EllMEOIATE SEGME .. 1111 ! " - [ i j] I GMENT

NTSlllAY 10- y ·ll'"'lrl1-- . Y STRING LENGTH '-I $Tllllll.-.sl15m··- ENO SEGMENT

Figure 2. Reference Photovoltaic System Description

2 1.2 Objectives 4. Defme the most economically prudent path to minimizing these unl.'ertainties to the point The overall intent of the systems defmition that confident decisions can be made to pro­ studie-s, past and present, may be summarized as ceed or not to proceed with development of follows: this energy system.

I. Assess the technical feasibility of solar power The specific objectives of the present study satellites based on forecasts of technical capa­ are: bility in the various applicable . Select the most appropriate technological (I) to verify, maintain and update the presently­ paths leading to the most environmentally defined elements of the system, benign and economically practical systems. (2) complete the definition of the total system. Define the areas of high leverage research. and (3) prepare a series of plans required for technol- 2. Define the characteristics of SPS systems that ogy advancement and SPS program may be derived if a development were to pro­ implementation. ceed. Assess perfonnance, cost. operational characteristics, reliability, and the suitability Phase I of the present study. reported here­ of SPS's as power generators for typical com­ with, has concentrated on the first two objectives. mercial electricity grids. Phase II will emphasize end-to-end operations analyses and the third objective. These Qbjectives 3. Assess the uncertainties inherent in the sys­ and the timing of the study are designed to support tem characteristics forecasts, based on tech­ the NASA program recommendations in fulfill­ nological uncertainties, on cost estimating ment of the DOE/NASA evaluation plan. uncertainties, and on uncertainties that arise from incompleteness in the data base.

2.0 STUDY APPROACH AND STUDY TEAM

The JSC/Boeing and MSFC/Rockwell SPS NASA in developing the reference system report, system definition studies of 1977 and early 1978 beginning with a thorough critique of the JSC/ proceeded largely independently of one another Boeing reference system, followed by analyses of and developed system concepts with a number of options and critique items. The reference design significant differences. NASA then developed a ref­ was updated at the end of Phase I. The overdll erence SPS system description based on NASA ~chedule is shown in figure 3. inhouse studies as well as on Boeing and Rockwell contract results. The Study Contract Team included Boeing as prime contractor and General Electric, Grum­ The present study is divided into two man, Arthur D. Little, and TRW as subcontractors. phases; the first lasted 7 months and the second Principal task areas and the study team leaders for will last 9 months. Phase I accomplished additional each contractor are shown in figure 4. analyses of the options and issues identified by

3 0180-25037-1 _, -· -· -·

IUCllllC ...... &-0.UTIU .....-­.uv«

Figure 4. Study Contract Team Organization (Phase I Tasks Shown) Figure 3. Study Plan Overview

J.O STUDY RESULTS

The preS4!ntation of study results in this be built and tested to verify annealability as executive summary ha~ been grouped under three well as survivability of the rest of the array major headings: under annealing conditions.

(I) Highlights of Trade Studies and Analyses; 3. Several concerns addressed the high voltages (2) Summary of Updated Baseline; planned for SPS operation. Plasma interac­ (3) SPS Devdopment Planning. tions may occur with the high voltage solar array. In particular. the critique singled out The presentation is organized according to the the problem of plasmas produced by the elec­ work breakdown structure shown in figure 5. tric thrusters.

3.1 highlights of Trade Studies and Analyses Wear partides may be produced in the slip ring assembly and could .. track" high voltage WBS Item 1.0 Solar Power Satellite Program: insulators causing arcin!; and damage to the Important Critique Results slip ring assembly or nearby components. A high voltage potential will exist between the The critique effort identified and com­ structure and the solar arrays and power mented on about fifty items. These were grouped conductors. into 22 significant concerns for the discussion in Volume II of this report. The four most important Concern was also expressed that the power are summarized bdow. processor high-voltage transformer life may be too short for SPS application due to failures I. Concern was expressed about the tong-tenn caused by a-c corona within the windings. suitability and stability of graphite fiber com­ (This item was addressed anll corrected in the posites in the space environment. The issues Phase I study.) raised included creep, micro~racking, and suitability of thermoplaslil.:s for some SPS Concern was also expressed regarding design thermal environments. of, and selection of materials for. high voltage insulators and cable insulation for use in the "' It was recommended that experimental sam­ space environment. ples of the lightweight anncalabk solar array

4 0180-25037-1

1.0 SOLAR l'OWER SA TELUT£ PROGRAM

1.1 1.4 1.1 SATELLITE GROUND REC£ !YING llANAGEYENT ANO STATION lltT£GRATION

1.1.1 1.2.1 £NEllGY CXINV£R$ION CQIWSTllUCTION FACli.ITIES

. I STRUCTURE .I WORK SU"'ORT .2 COllCf.NTRATORS .I CRE• SU090RT .3 SOLAR 8LANKET .l OIERATIOHS ·' l'OWER DISTii• .5 THElllAAL CONTROL 1.2.2 .6 llAllllTiNANa LOGISTICS $1.WKlRT [!> FACILITIES 1.1.2 l'OW£R TRAN$111$SION .1 WORK SUPPORT .2 CREW SUPPORT .1 ST"UCTUftE .l OIERATIONS .2 TRANSMITTER SUBAARAYS .l l'OW£ R DISTR. 1.U .4 THERMAL CONTROL 06M~T .5 CONTROL FACILITIES .6 MAINTENANCE .1 MECHANICAL POINTING [!> .1 WORK W'P()RT .2 CREW SUOPOl'T 1.1.l .l OP'E RATIONS -ORMATION MANAC'EMENT A.NO CONTROL

1_1.4 BECOMES GEO 8ASE An1TUO£ CONTROL FOR THE LEO ANO STATIONKEEPING CONSTRUCTION OPTION

1.1.5 COMMUNICATIONS

1.1.6 INTERF.'1:£ !ENERGY CONV POWER TRANSlllSSIONSJ

1.1.7 SYSTEMS TEST

[!:> .-.ODED Figure 5. Work Breakdown Structure for SPS

4. The arrangement of the power supply hookup the present NASA 5 gigawatt baseline with one to the klystrons was criticized. Each power transmitting antenna. Analysis of the \;ontrol processor feeds about 400 klystrons; the requirements for this asymmetric c0nfiguration potential fault currents that could arise from determined that because of the overriding impor­ a high voltage arc at one klystron could tance of solar pressure compensation in the control involve the entire power supply current. Also. thrust scheme, no propellant penalties were for the klystron power that is provided incurred by the lack of symmetry. Also, no packag­ directly from the solar array busses, a fault ing differences have been identified that would could short the entire bus to ground. Electri­ arse from dividing the original configuration into cal and magnetic forces caused by severe tran­ two equal halves. Therefore, th~ only consequence sients could 1.:ause major mechanical damage of this alternative to the original baseline is the to the SPS. A residual concern was also requirement for more positions in geosynchronous expressed that power supply faults frequently orbit to effect a given total installed generating cause failun.>~ in other parts of an electronic capacity. system. The next alternative was also a five giga­ WBS Item 1.1 Satellite: Size and Configuration watt system. but the power is divided into two Effects power transmission links each rated at 2Y2 giga­ watts. In order to minimize land use and recten11:1 Smaller SPS configurations were compared costs. it 1s desirable. when reducing the link power. to the original I 0 gigawatt baseline. The first was to increase the transmitter aperture. in turn redue-

5 0180-25037-1

ing the receiving station area. This design option. howt•ver, has approximatdy 4 times as many trans mitter subarrays as thl' single-transmitter 5 giga­ watt satellite. As a rt'sult. it incurs a significant payload packaging problem becau~ uf the low packaging density of completeJy assembled trans­ mitkr subarrays. The packaging density situation appears to be much improved through use of a solid stak transmitter. In the solid state option all of the active functions art' included in a plan:.ir sheet only about 2 centimekrs thick (including the rl'sonant cavities). Thus. a much higher pack­ aging density pt'r unit of aperture art>a can be achieved.

The final 2- l f :!-gigawatt option. like the second option. results from effectively dividing a Figure 6. Solar Power Satellite symmetric configuration in half. As for the other Structural Bay Configuratio1· case. no added penalties were determined for th.is design option excepting the use of more geosyn­ chronous orbit space.

WBS Item I. I. I. I Satellite Energy Conversion Structure

Changing the solar blanket tensioning to mM NNA ,.,... WfllR ....ollCI UllflR ...""" LOllH ...,ACl M•t.m0•4LL uniaxial caused a revision in loads applied to the L~-- LATIMlMAM 01"1111 LOCAnc.. ... _...... solar array support strncture. The structural bay .., .. LOIOTM .. •T"l tMICltflll9 ... -~~ .. ,,_.. 11t- design was updated to retlect re\'ised load require­ - uon-' 1.•u~ ments for self-power orbit transfer and solar blan­ 1...;·:.,,,...... _.... ··-...... CJHTICA&.\.OM)_ .,....~.OOIDf ..... ,.tmUCll.llMlt Jm~MAMI·-"- I _ •tJllGAll·- 6.n•GM·- ket stretching loads. The structure configuration is ·--- ·- shown in figure 6. For the reference case of geo­ synchronous orbit construction of the SPS the Figure 7. Solar Power Satellite Structural Update type B beams shown can be changed to type C Beam Configurations since orbit transfer loads will not be a considera­ tion. Figure 7 shows the characteristics of the three Al'l aluminum strncture was compared to types of beams. the low-coefficient-of-expansion (CTE) graphite composites baseline. The following main conclu­ Evaluation of various structural arrange­ sions resulted: ment options led to the conclusions that a penta­ hedral truss con figuration would be better than the • Roll formed closed section aluminum struc­ present cubic type (this appears to be true for the tures can be automatically fabricated in orbit: transmitter as wdl). The pcntahedral truss retains the square bay shape. has fewer members and types • Design load requirements for LEO-constructed of joints. and would be lighter. SPS moduk can be satisfied. Tht• aluminum design is 23':~ heavier than compo!>:tc but may A change to thl· pcntahcdral truss will be be lower in 1.:ost: considered during Phase II. Trade studies con­ ducted in Phase I used the cu bil· sl ructurc as • The 10-GW SPS natural frequency with alu;ni­ baseline. num ( AR=4 l is 65 times orbital freq l•l'ncy ~~ instead of I 00 times:

6 0180-25037· 1

• The estimated natural frequency appears ade­ manmade space objects with the SPS. These will be quate for satellite control syskm stability: treated as unscheduled maintenance. Switchgear further analysis needed to wrify: may require replacement of one or two units annually. • Based on initial studies, thermal stresses are within the capability of the aluminum design; Occasional maintenance of the solar arr.1y by ann:.!aling to restore output due to damage of • Satellite detlections are within acceptable the solar cells by solar flare radiation is expected to limits (-2°>: be necessary. It would be possible to make the solar array maintenance-free with regard to • Detlections due to occultations could result in expected radiation by oversizing or providing extra dynamic "ringing.. of the entirt• solar an-ay shielding (in the fonn of thicker coverglasses). support structure. Further analysis is needed Annealing, however, has distinct cost advantages. to evaluate this. It is important to recognize that nearly all WBS Item 1.1.1.2 Sat.ellite Energy Conversion: of the radiation damage to solar arrays at geo­ Solar Blanket synchronous orbit comes from solar tlares, which are a statistical phenomenon. Environment models Effects of array shadowing w1..•re investi­ are used to predict the amount of radiation for gated. If a segment or section of a long solar cell which arrays must be designed. The model used by string is shadowed. it will not generate current. Boeing was originated by the Goddard Space Flight Current tlow in the entire string is therefore inter­ Center. As is typical for such design models. it is rupted. If other strings are connected to the load in roughly a 90% confidence model; an expected­ parallel with the shadowed string. the difference value model would predict less radiation. Use of a hetween their loaded output voltage and the zero­ conservative model is warranted by the fact that a current voltage of the shadowed string, appears as a severe solar flare event will affect all SPS's then in reverse-bias voltage. The reverse hias voltage. if it is orbit. more than a few volts per shadowed solar cell. will destroy the shadowed cells if they are not pro­ Even this conservative model indicates that tected. A small spacecraft flying across the face of degradation more than l 0% from a single large an SPS could severely damage an unprotected solar flar.• is highly unlikely. Much improvement in the array. Reverse-bias protection is therefore manda­ confidence in this result can be expected due to tory. The solar blanket panel design was modified continued accumulation of statistical data from the to include shunting diodes. current solar cycle and with direct observation of proton fluxes in the 2 to l 0 MEV range. WBS Item 1.1.1.3 Satellite Energy Conversion: Maintenance The estimated requirements for annealing are clearly sensitive to the model used and the A summary of repair and replace require­ statistical approach adopted. A comparative study ments for the satellite is presented under WBS Item of the available data for silicon and gallium arse­ l. l .2.6.This section addresses maintenance require­ nide SPS's and solar arrays was conductc:d. This ments peculiar to the energy rnnwrsion system. analysis revealed a significant differcn1..·1..· in the environment model used for the Boeing and The enagy conversion system is designed Rockwell solar blanket degradation analyses. Most to be as nearly mainknance-free as possible. Solar of the differences in degradation prcdi...-ted by the cells. blocking and shun ting diodes. and attach­ studies is due to differences in environment models. ment and tensioning devices all incorporate enough The Rockwell model is less conservative: it would re~undancy to provide a lifetime of more th:m 30 predict that neither a silicon nor a gallium :irSl'nidc years. Exceptional maintenance requirements n~ay SPS would he likely to need annealing in 30 years arise. e.g .. in the event of collisions of natural or at gcosynduonous orbit.

7 018().25037-1

Boeing test data on silicon solar cells are compan•d in figure 8 with the Rockwell projec­ tions for the gallium arsenide solar cell. It is dear ...... that there is no signifo:ant degradation difference. -·- Nole the difference in proton/dedron equiva­ lences between silicon and gallium arsenide. This YOlTAGI IOLAREX II..- ltUCOlll IOl.Aa CILL I'll ... difforence arises because of the difference in mass ....J-.V _,_. MOTOtai of the atoms of the two solar cell constituents. Our t-U a 11111 lllMI' lt.ICTIIOIW EOU1v1Clit2• analysis would predict no significant difference in -.... , ..... degradation between the two systems for the same .. - tluence. Since the gallium arsenide solar blanket ... design has significantly less shielding, we would predict more degradation in the equivalent environ­ ... .__ __ _._~ __ ...______.. ____ ...._ __ __...... _.,. __ ___. ment compan:d to the silicon blanket design. 0 20 ... • • - 120 CUMHlll \fM) .. ... Figure 9. Thermal Annealing of Proton Damage In ... Silicon: Boeing Test Data

07 a-?. ... Annealing of the solar blanket on an SPS u will require a technique tailored to that purpose, as •.. well as a blanket fosign compatible with annealing .., temperatures. Attention has been given to laser ., directed-energy annealing under this contract. Ini­ ... tial tests of laser annealing of thin solar cells with .... '"" glass covers were dedicated to measuring thennal response of solar cells to laser energy density. Figure 8. Degradation Comparison For Proton Resulting energy requirements are less than earlier Irradiation estimates by about a factor of 5 and have been reflected in definition of the reference laser anneal­ ing system.

Recent results reported by Hughes show A test program was conductd to further the radiation degradation of gallium arsenide to be explore the laser annealing of SO-micron solar cells. a strong function of junction depth; gallium arse­ The original intent was to anneal glass-covered cells nide is reiJorted to degrade less with shallow junc­ but a suitable method of glassing was not found. tions. Additional radiation degradation testing is Ten cells were coated with 75 microns of glass by needed for both types of solar cells. The possibility Schott in Germany using electron-beam evapora­ that gallium arsenide cells may anneal at relatively tion of the glass. The coatings were of poor qual­ low temperatures also needs to be further explored ity, e.g., full of bubbles, and contained much by testing. frozen-in strain. When subjected to annealing temperatur(!s, the coated cells curled up like potato Annealing of radiation damage in silicon chips and the glass fractured. Attempts at RF sput­ has been repeatedly demonstrated in the labora­ tering at Boeing yielded glass deposition rates too tory. Illustrated in figure 9 are the results of oven low to be usable. Ion sputtering was tried on a few annealing tests of bare 50 micron silicon solar cells. cells at Ion Tech. Good quality coatings were pro­ Several cells were tested with two irrndiations and duced, but the cells were damaged in handling. two anneals. All cells tested showed recovery on Some damaged cells were subjected to annealing both anneals. temperntures and did not exhibit the mechanical

8 01~25037·1

failures of the &hott-coated cells. Ion sputtering The concept of the SPS annealing system is merits further in-1estigation, ::s

DITM. A OETAll I TYfiieCM...... _., 1'YPICALLAll'R--.U.IR

Figure 11. Laser Annealing Concept

WBS Item 1.1. 2.1 Satellite Power Transmission Structure

Two analyses were conducted on this suh­ ORIGINAL IRRADIATED ANNEALED IRRADIATED ANNEALED ject. The first was an evaluation of the use of alum­ inum as a structural material for the primary and Figure 10 - Laser Annealing Results secondary transmitter structures. Results of a NASTRAN analysis showed that thennal deflec­ tions of an aluminum structure would far exceed These tests achieved moderately successful the required tilt tolerance of l to 2 minutes of arc. restoration of output from SO.micron cells after Therefore, if aluminum were used, the structure degradation by proton irradiation, under labora­ would require active themtal deformation tory conditions. Clearly, much additional technol­ compensation. ogy re..earch is needt d to develop designs and proc­ esses that would b<11g this technique to the point The second analysis evaluated structure of suitability for 1arge-scale applh:ation. Additional design approach options. Early investigations of research sho~.id investigate a wider range of time­ the antenna structure developed the tetrahedral temperature histories and irradiation degradation truss primary and secondary structure concept. and should emphasize development of a suitable This system provides a maximum of structural effi­ cell eni:apsulation technique. Tests should encom­ ciency for such an antenna. However, it constrains pass sufficient numbers of cells to enable statistical the subarrays to a non-square system and presents interpretation of results. difficulties with re.sped to maintenance access.

9 0180-25037·1

The cenkr illustration in figure 12 repre­ Aluminum has a high coefficient of thermal sents an attempt to design the antenna structure expansion compared to the grnphite used in the for maintenance. It pro\'ides easy a

WBS l. l .2.2 Transmitter Subarrays Table 1. Comparison of Losses for Metal and Composite Waveguide

Aluminum Subarrays e AVEllAGI STICIC • Z.19 llETlllS

e &T •ll"t Aluminum was also investigated for appli­ cation to the subarray distribution and radiation PERCENT POWEii LOSS waveguides. The baseline defined in the earlier AltMI,_ COll'OSITE study was an aluminum-plated graphite composite STICK LENGTH .SI .02 waveguide. Significant concerns have been STIClt°WIOTH .42 .12 expressed as to its suitability for high power, long CllOll-LfNGTH .17 .02 life applications where temperature cycling may be CllOSS GUIDE WIDTH .11 .G3 a problem. 1.31ll .1n

TETRAHEDRAL TRUSS A-FRAME PENTAHEDRAL TRUSS

• MAXIMUM EFFICIENCY • GOOD ACCESS • GOOD ACCESS • NO TENSION MEMBERS • SOUARESUBARRAYS • GOOD EFFICIENCY •NON-SQUARE SUBARRAYS • POOR EFFICIENCY • NO TENSION MEMBERS • MAINTENANCE ACCESS • USES TENSION MEMBERS •SQUARE SUBARRAYS DIFFICULT • SECONDARY STRUCTURE IS PART OF PRIMARY STRUCTURE

Figure 12. Antenna Structure Options

10 018~25037-1

Solid-State Transmitter period. It also generally requires resc.nant reactive networks· which make achieving bandwidths of a St>:edion of microwave tubes. e.g., klys· tenth or more of the operating frequency (as trons, for the refert:nce system, has raised issues of desired by many communications users) very diffi­ power transmitter reliability and maintenance. cult. In this sense, the narrow bandwidth required Extrapolation of current trends in growth of tube by SPS is a real advantage. life and reliability indicates that a mean-time· between-failure of 25 years is reasonable. Even Present communications interests in r.1icro· with this figure. however, repair and replacement wave power amplifiers are increased linearitv over of microwave tubes is likely to dominate the main­ V'ide bandwidth. RF output powers equivalent to tenance workload for SPS's. traveling wave tubes, and longer life th«n the TWT's. For SPS, efficiency, lifetime, low cost and Solid state devices exhibit much better narrow bandwidth are desired. It is anticipated that failure statistics. Those shown in figure 13 for with adequate funding the development of high­ gallium arsenide FET's show that at a channel tem­ efficiency switched mode amplifiers for SPS is a perature of l 35°C. 98% of the device~ will still be low risk. However, SPS research may have to take operating after 30 years. This suggests that a no· the initiative because the communications industry maintenance mode of operation may be feasible. can develop and prosper with present DC-RF con­ version efficiencies.

The power per device is an important SPS parameter. The number of devices that can bt: dfi­ ------Z'!lFAU.IOIUNIT GMllill cien tly combined in a module is limited by circuit ---1- -ftfAlllD.-.m.I CMAatiJ losses; the power per module determines the RF I I power density per unit transmitting array area and I I ...... If • •VIAR MIJllTl-.VCI is directly relatable to overall transmitter power. I • 1..0G NDlllllA&.. fAtLUIU cr.cratlUttOM I ... Silicon bi;. ..>lar transistors, GaAs FFTs and multi­ I ... I mesa IMPA TTs can all handle powers above I 0 I watts, adequate for SPS application if efficient I RIFUINCI: LUJCNIMfl AND UDO. combining can be accomplished. ... tw1W llllLN. fi'HYl'CS - ,..JUICTIOlilf ---­ T191MTUM. ·c ln•.•m.291 GaAs FETs were selected as the preferred Figure 13. Solid State Device Lifetime DC-RF conversion devices because of higher gain than silicon bipolars, equivalent expected power­ Various devices were considered for this added efficiencies, roughly equal power capabilities application. GaAs FETs, silicon bipolar transistors at 2.5 GHz and much better expected contact and EBS's have the rest values of power-added metallization reliability. GaAs FETs for SPS appli­ efficiency. cation could be fabricated separately and mounted in hybrid fashion or combinecl with other compo­ All the two-terminal devices have efficien­ nents on larger GaAs chips in integrated circuits. cies less than 36%, too low for SPS. As noted ', i.e latter alterna:ive is preferred because of its above in the aluminum waveguide discussion, expected lower costs in mass production. microwJve link efficiency is of great importanu.' in minimizing SPS costs. A solid st:1tc transmitting antenna configu­ ration was synthesized using design criteria for Switched-mode amplifiers operak with the radiating module size devdoped in previous NASA device power dissipation time integral (over the st•tdies anct several circuit and radiator concepts operating cycle) minimized as much as possible. develcpcd in prior proprietary Boeing IR&D work. This generally requires active device transition The basic elements of the so1'd state transmitting times about a factor of ten faster than the RF array are .59 x .~9 X (X = frl'c spa~e wavdcn~ 1 h)

11 0180-25037-1

radia:ing modules. These are fabricated on 20 mil t:sser.tially the ume because in a series of ampli­ thick alumina dielectrk sheet whit.it is metallized fiets noise figures of prior stages overwhelnt con­ for signal. ""-ontrol and power circuitry. Microstrip tributions from the followi11g stages. Thus even techniq:.aes are used for combining. fdtering, and though a klystron may have a noise figure of 30 db making antenna elements. Ways have bet:n found as opposed to approximately S db tor a solid state to efficiently t.-ombine outputs from up to 6 ;ampli­ amplifier it m~kes no noticable difference in the fiers_ The subarray t.-oncept is shewn in figure 14 system noise fagure, if they both have similar front ends. Since the solid state panels have much less ar... a t:tan the comparable klystron modules, the former will sp~d their noise 'wer a wider solid - I angle. thereby reducing the ground noise power per unit area.

The solid >late transmitter is limited by maximum allowable device temperature to a ther­ mal dissipation of roughly l .S kilowatts per square meter. At a conversion efficiency of 80% wii.h a 10 dB :~aussian taper, the thermal t.-onstraints and •'t•.mt---. ionosphere power density constraints follow char­ ~- tJ-"''iEIU(S SU&llAAAY·-•-LMATlllJI LEAST -TCllAA<.( uran acteristic curves as illustrated in figure I 5. As can P .... L •JS~ 1:.1 P&RAl.l.fl 1 ... KV.IA'JKW_UZKG be seen. the solid state system is constrained to a tf&Sfftfrt_u;t... f tr.HT : .ra•. IA:1V. JOJ 1

cells in generation of the DC power. Aggregate sets •, • t 1.J '.J 1.S ,. ,, •• of microwave power generators can then be sup­ ~TTWllil_,...... -n,.._. plied at compc.ratively high distribution voltages. The minimum-risk option is use of dcidc convert­ Figure 15. Representative Solid State SPS Costs ers but this will result in significantly greater SPS and Sizing mas." and cost. AC power distribution may provide a means of :ninimizing distrihution losses and reduc1•1g solar .may voltage. Mass and cost penal­ More elaborate transmitter illumination ties will be similar to those for full de/de processing. strategics may allow higher power. The effects of tmncat!ng the maximum intensity region of the The noise

12 0180-25037-1

WBS Item 1.1.2.3 Power Transmitter Power Table 2. SPS Satellite Failure Summary- Dmribution 1C GW SPS

ou.nTY.. Malt.at\'11. lifetime expectancy for th"' OC-to-OC con­ ~- I.I ·- -1.U.IA _.,.WlllVICU verters was analyzed ~uring the critique. This indi­ LUU au._.,.MIC:l'WCMATf- ..... __ LLtAa ---- 1UIZ 1 cated a significant rroblem with dielectric material L•.Z I ; - U.U.1 -- ---a•n-.a ,....- life. If the earlier com;erter were derated to reflect Ll..J.l.Z ....,...... --912 Ll,%.U -··'llCAIC:__.... • c;; 20 year life, an increase in mass would be I.~ 912 •J LUU -=--nra_,__.._ Ll.U - • expected. However, a new transfonner technology 1.u.a.1 ..... -·--~llECI- :m.N using liquid-cooled trani.fo, m~rs provide long life l.\.J.1.2 - •z LUU ..... l.LZ.U -~ .... • with less mass than the earlier system. Use of a • LU.U --- .... 1.UU CMLmG ,.... • liquid dielectric/coolanl avc-ids the failure mode oi ------• corona-induced dielectric ffaw growth that limits the life of potted transformers. The annual power lo~ due to these failures is a function of the number of failures and the WBS lrem 1.1.2.4 Power Transmitter Thermal power loss per failure. As indicated in figure 16 the Control principal power loss problem is the OC-to-OC con­ verters followed by klystrons and switchgear. Par­ Failure analyses also indicated a problem tial redundancy can be built into the OC-to-OC with the heat pipe cooled klystrons. The difficulty converters (with a small mass penalty) to minimize was that the S00°C 1..-ooling section would utili1.e a this problem. mercury vapor heat pipe. In the event of a meteor­ oid puncture or other leak. the liquid metal would ~I be released into the high voltage environment of - the transmitter system and lead to arcing damage. I I Plating of liquid metals on insulators might lead . " I Ml I I to permanent contamination that would require I • J repair and replacement. Vought Corporation I; I i a • ! examined a l·irculating fluid cooling option and I .• n a found that a mass reduction was possible and that i i I fluid could be selected that would minimize I 1 i r risk of arcing. Their analysis indicates that a cir­ culating fluid system can be made as reliable as the heat pipe sy!>tem a:id certainly more reliable than the expectzd lifetime of the klystrons themselves. Figure 16. Annual Power Loss due to Failures

WBS 1.1.2.6 Power Transmitter Maintenance WBS ltc!m 1.1.3 Information Management and A general failure analysis was conducted, Control with principal emphasis on failure rates and their effects on power production. (Some of the failure Telemetry and command requirements effects problem areas and their resolutions have were

t3 Estimating th"~ requirements necessitated at 1.45 GHz. In the SPS time frame, a higher fre­ many deci~ions concerning the component level to quency bank. e.g., X-band. may be a more desir­ which each satellite subsystem will be inscru­ able choice. mented and the level to which a ccmm;md capabil­ ity will !>e provided. For the subsystems which are WBS Item 1.2 Space Construction and Support relatively well-defined. this process was accom­ plished on the existing design. For those subsys­ The most a.-onomk·al cons.ruction of tems on whil·h very little design infonnation exists, SPS's. given that many are to be built, will employ requirements of typical current satellites were a semi-:mtomated construction facility that extrapolateJ to the SPS. actiieves high crew productivity. The 1977 Boeing study developed a construction facility concept The large numbers of telemetry pc.lilts and that became known as the ·-c-clamp."' One of the commands would represent a heavy worl:load if all principal objectives of the present study was to telemetry data v.'cre transmitted to the ground. develop and evaluate alternatives to the ·-C-clamp" Automatic processing could be done by a distrib­ with the objective of reducing construction base uted processing syskm on the satellite with proces­ mass. complexity, cost. and reducing construction sors located near the t:quipment being monitored operations crew size and operating cost. and comm;mded. This approach providl"S two other advantages: the amount of data transmitted By the midterm of Phase I the options were thr('lughout the satellite and to the ground is narrowed to three preferred systems: the platform­ reduced. and the delay from detection of an type single-deck. and two- and four-bay end­ anomaly to receipt of a ~orrecting command is builders. Figure I 7 compares these two types. In reduced. In view of these considerations the on­ the last two months of the study. these three board processing system was selected. options were compared in .nore detail. Figure 18 summarizes the results. The mass and cost esti­ The design approach provides numerous mates for the three candidates were within a few microprocessors and memories distributed through­ percent of one another. Selection of the preferred out the satellite. These processors are organized approach therefore included other criteria. into groups. each monitored by a processor that is one of another tier of proce'iSOrs. Tiering continues to a Central Processor Unit which manages the data traffic to 2nd fro•11 the ground. The recommended approa~h employs two systems connected by a lim­ ited data link through the slip rings.

The use of fiber optics is recommended for data transmission because of its reduced mass, fa!Jlt tolerance (because a non-conductor does not prop­ agate faults), wide-band multiplexing capability, immunity to EMI and arc discharges, and available inexpensive materials.

WBS Item 1.1.5 Communications

Analysis of communications system requirements appropriate to the command and data rates described under WBS Item 1.1.3 con­ •GEO clucied chat an S-band link with di-;h antennas • SGWSPS • 3SIZES(2-BAY,4-8AY,8-8AY) about 0.6 m diameter would ht: technically ade­ quate as regards interferenct.• from the lower link Figure 17. Alternative Construction Concepts

14 01~25037-1

...... wn • COST -..J_ ...... Aft .. -.. T - • - • - - z

JIAY DID ana.DEll

Figure 18. Alternative Construction Concepts Mass and Cost Comparison

Table 3 summarizes the .!valuation indud· and equipment will be developmental, and configu­ ing qualitative factors. The single-deck facility ration changes may be expected between the pro­ requires less interdependence of construction oper­ totype a."ld later production units. The four-bay ations and is more adaptable to SPS design changes end-builder was ~lected as the (production) sys­ after the base is constructed. but the end-builders tem baseline for the Phase II 5tudy. have an inherent capability for higher production rates; they represent a higher degree of automation. This supports the view that the cost trends found Table 3. Alternate Construction Concepts are valid and are not just estimating variances. The Summary 4-bay end-builder is capable of a significantly higher production rate (for the solar array part of the SPS) than the groundruled I 0,000 megawatts .. ... ground output equivalent per year. If evaluated for -- tacr • -'" ,,.,...... Al', ..... higher production rates, the 4-bay end-builder --ClllWlllt ., would undoubtedly prove more economic than the _,,,," ~~UllJAL ....i smaller facilities...... AllMtfllAY

Evolutionary increases in degree of auto­ mation will be a normal trend of space construc­ tion technology. It is our judgment that the plat· form approach will be well suited to construction of an SPS prototype when some of the processes

15 0186-25037-1

WBS Item 1.3 Transportation HLLV launch and reentry paths will pro­ duce significant sonic overpressi•res. For a launch WBS Item I .J. l HLLV site on an eastern seacoast (e.g., KSC). the launch and booster entry overpressures occur over the The pros:>ect of heavy lift launch vehide ocean and are of little concern. Return of the (HLLV) operations at the launch rates needed to upper stage could result in overpressurl.'s up to support an SPS construction program has raised :ibout 145 pa (3 psf) in the ~mmediate vicinity of issues n:garding potential environmental effects. the recovery site if no mitigating strategy is used. The main ..-oncerns are atmosphere pollution. and noise and sonic overpressures. Mitigating strategies include ( 1) transition from high angle of attack to iow angle of attack at HLLV's will employ clean fuels. methane supersonic speeds to maintain tile terminal portion or kerosene and hydrogen. burned in the rocket of the trajectory at a higher altitude. and t :!) use of engines with oxygen. The quality of fuel l.'Onsumed a supersonic turn to place the overpressure­ by an H LL V fleet will be very small. roughly 0.1 % generating part of the trajectory over the ocean. of total rnmbustion fuel'> (coal a11d oil) presently These strategies were investigated by trajectory burned in the U.S. llms the gross quantities of simulations. Results are shown in figure 20. pollutants will ~e negligible. However. the HLLV's will spend a portion of their trajectories in the

upper atmos~here and ionosphere. Concern as to ATLANTIC possible redt11.:tion of the ion density in the iono­ OCEAN spheric F-layer had been l.'Xpressed by a prelimi­ nary investigation. Analyses conducted by Boeing indicated that this problem could be minimized by !>Uppressing the trajectory peak altitude to about I 00 km. A number of ascent trnjectories were sim­ ulated using various strategies to minimize trajec­ tory altitude. Results are summarized in figure 19.

It was found that the best trajectories had a peak GULF ascent altitude of about 110 kilometers. Trajec­ OF MEXICO tories could be suppressed to keep the path below 100 kilometers with a slight performance penalty. .. Figure 20. Supersonic Turn Return Trajectory Can Reduce Overpressures to Less Than 2psf

.. Gii ..5 o~o-o-o-c !l WBS Item 1.3.2 Cargo OTV IC .. / O OlllGINAL REFERENCE I 0 ~ 0 Electric Orbit Transfer Studies g.. ~ -a/ 0 IN.IECTKINATeltll The prefe:-red orbit-to-orbit transportation R 0 llUECTIOIWAT-ltll ..IC concept identified in the 1977 Boeing study was l> 1-CTION AT 1111-120 KM the use of electric propulsion systems to convert m'° 100 110 120 130 SPS modules into powered spacecraft that could l'EAll ASClNT TRAIECTORY AL TITUOE IN KM transfer themselves to geosynchronous orbit with a trip time of approximately 150 days. A tradeoff Figure 19. Launch Trajectory Suppression study comparing this to constrndion of SPS al Results gcosynchronous orbit with clk•mically-fuckd

16 0180-25037-1

(LO-,/LH.,) orbit transfer vehicles showed a cost performance. The result of these changes was an saving of-roughly $2 billion per l0.000 megawatt improvement in the effective average integrated SPS. specific impulse for self-power tr.1nsfer from approximately 2 I 00 seconds to approximately During the present Phase I study. an analy­ 3000 ~conds. (The electric specific impulse of sis was conducted to evaluate the use of independ­ 7.500 seconds is significantly degrJded hy the use ent electric orbit transfer vehidcs to allow the of ..:hemical thrust during periods of occultation benefits of de1.:tric propulsion to he L·ombi:ied with and high gravity C!radient torque.) tho! benefits of geosynchronous orbit construction. The operational corn:,· pt is illustrated in figure :! 1. The independent ele..:tric OTV configura­ Electri-.· crbit transfer \·ehides an: constructed in tion is shown in figure :!:!. This orbit transfer low Earth orbit ;,it a hase which also provides slag­ vehicle is sized to deliver 4.000 metri..: tons to ing depot functions. A tleet of approximately 20 geosynd;ronous orbit and return with ~()() metric electric orbit transfer vehicles conveys SPS pay­ tons. The r~turn payload capability provides for loads to geosynl hronous !."'rbit when: SPS construc­ return of packaging equipment and other items tion takes pl;Ke. Ir. order to provide expeditious from the geosynchronous orbit construction site. transfers of crews and supplies. high thrnst Because the electric orbit transfer vehicle is smaller 1.:hemicaliy-propdled orhit transfer vehicles are than the SPS modules discussed on the previous used to provide this service. The electric orbit page, it suffers comparatively little from perform­ transfer whides arc reused I 0 times over a lifetime ance losses indm:eJ by gravity gradients. and of sever.11 years. achieves an integrated effective specific impulse of 6000 seconds .

...... __,. ttc..1

•nK1.-usT•u.t• ...... , ...... _ •'-Ori"-"'i·••l

Figure 21. GEO Construction Concept Electric Orbit Transfer Vehicles Figure 22. Electric OTV Configuration

As a part of this analysis. improvements A comprehensive co:.t comparison was were mad..: in the sdf-power configuration. The developed to evaluate the self-powered electric pro­ most important w;is in the arrangement of the solar pulsion option relative to the independent electri..: array to be used for orbit transf1.:r. The change orbit transfer vehicle. Differences in 1..osts of con­ improves inertia balancing. reduces the effects of stmctioa operations are indudl..'d. TIHl'l' options solar blanket stretching loads. and avoids mah:hing are shown in figure 23: (I) SPS self..tran"port degraded solar -.·di arrays to und:rgraded arrays in modules without re..:overy of tlil' dectric propul­ series. Thruster moduks were relocakd to improve sion equipm~nt, I 2) the same option with use of inertia balancing and thrust moment capability. small clectrk OTV's to recov..:r the orbit transfer Several propellant tank locations were tried to hardware. (3 I independent ekl'lric orhit transkr improve inertia bal;llll'ing. location at the center vehicles for all orbit·to·orbit 1.·argo transportation of the module provides the best overall transfer with construction of SPS'sat geosynd1ronousorhit.

17 0180-25037-1

• scenario developed under this study) shows rela­ tively little sensitivity to hardware cost because th: • wst of the orbit transfer hardware is amortized rlfOISPll over several SPS's. ··'-£0/lftlll01V 15 -GIOl£01V Gallium arsenide solar blankets for the ,. ; independent electric OTV were also examined. No I - advantage was found for the use of gallium arse­ I nide in the orbit transfer vehicles under the assumption that silicon was to be used for the - .n sateilite systems. Oearly, if gallium arsenide is to be ~d for the SPS then it makes sense to also use --1-··-·1-· ..._.UTIWTll it for the orbit transfer system. -- - 1-:>--- 1_--- ...... ,.. Self-power from low earth orbit construc­ tion bases for establishment of SPS's at geosyn­ Figure 23. Construction/Transportation Cost Comparison chronous orbit was recommended as the preferred approach. This preference arose primarily because of the significant difference in front-end cost. It The bars on the left show the total cost of was decided, however, to retain the present NASA preparing to carry out the construction operations, baseline of geosynchronous orbit construction for including the unit cost of the construdion bases Phase JI of the study. and of their transportation. The St.'\:ond set of bars shows the transportation system 1leet investment Effects of Ion Jets on the Magnetosphere required to establish a production rate of i 0,000 megawatts per year. The third set of bars shows the Electric orbit transfer operations will con­ cost of trJnsportation operations for the first sume roughly 10,000 tons of argon per J0,000 year's operation, i.e., construction of two 5,000 megawatts of SPS's emplaced. Although this con­ megawatts SPS's. The fourth set of bars sums the sumption rate is entirely negligible in terms of first three sets showing the total investment resource availability. I 0,000 tons of argon repre­ through the first year's production operation. The sents more than 1032 ions and neutrals injected fifth set of bars shows fully amortized costs for the into the Earth's magnetosphere during orbit trans­ three systems including amortization of all capital fer operations. investments at an interest rate of ?Wlc. A brief investigation of possible effects on The greater capital cost of the independent the magnetosphere indicated that this i.s probably electric orbit transfer vehicle system is offset by its not an environmental risk. Significant questions reduced fuel consumption on a fully amortized remain, and an analytical investigation adequate to basis. However, the difference in front end cost to give high assurance of no risk will require roughly establish a production rate of I0,000 megawatts two years. per year is approximately $7 billion. Conclusions from the exploratory analysis If the independent electric orbit transfer were as follows: vehicle can be reused many times by successful annealing of its solar blankets, it can provide low I. Argon beam from thrusters travel intact for cost. If annealing recovery is less complete. the long distances tangential to orbit (at injection self-power operations which expose solar arrays to angle). the orbit transfer radiation degradation only once, will exhibit lower cost. The independent electric Transition from beam to single ion injection orbit transfer vehicle (with the nominal reuse into geomagnetic field is a major 1111.will'ed proh/em in plasma physics.

18 0180-25037-1

3. Below ~.5 earth radii goo1:entric. all of the would incur significant perfom1ance penalties beam argon is tt~mporarily i:aptured by the relative to a zero-degree or low-inclination low goo111agnetk field. t'Uth orbit. However. with electric propulsion this perfom1ance diffcrenct' in tem1s of cost is minimal: 4. Abow 2.5 c:arth radii {10.000 km altitude) the principal factor is difference in orbit-to-orbit most of beam argon es..:apes in a jet of plasma propellant requirements and electril· system:' (86.6 km/sed: a small fraction is peeled off require comparatively little propellant in any case. into tr.ipix·J orbits. These potential cost ad\·antagcs are approximately negated by increased cost of ground transportati0n 5. A small fra.:tion of the thrustl'r bl'am ~·harge to remote sites. Costs of constructing and opa­ exchanges at the thruster exit pianc: and ating r.~mote sites would be t•xpected to t'Xcecd escapes as nl'utral gas. continental U.S. site costs by a factor of 1.5 to '.:.

6. lnl'ffo.:ienl·ics in the present tl:n.s:cr systems Thercfort~. the prindpal motivation for allow 10-20'; of the luel te l'Scapc from the leaving KSC for a remote site will stem from the exit plane as thc:rmal (0. l -1.0 eV) argon ions c,;\·entuality of SPS operations outgrowing KSC. and neutrJls. Our estimates to date indicate that KSC can handk approxim;1tcly lO gigawatts per year of SPS 7. At low altitudes (bduw I 000 km) this ~·onstrul.."tion. thennal argon diffuses back into thl· upper atmosphere. Remote site possibilities include land-based sites such as the mouth of thl.' Amazon in Brazil 8. At higher ~1ltitudes this thermal argon popu­ and ocean-based sites employing large floating lates the magnetosphere. stnictures such as the western Pacific low latitude sites identified by Jim Alckennan in studies at the 9. Th~ primary interaction between resident Johnson Space Center. large uncertainties plasma and argon from the thruster is l·harge presently exist as to the cost of large tloating struc­ exchange bdween icns and neutrals. tures. Further analysis of the floating site option is recommended. Bdow 1000 km A+ + 0 and A + o+ Above I 000 km A+ + H and A + p+ WBS Item 4.0 Gmund Receiving Station

10. The random walk diffusion of neutral and Rectenna ~·onstruction and structural charged spei.:ies is a second major unsofred design concepts were developed jointly with the prohlcm. complicated by the influence of geo­ aim of minimizing constmction cost. The con­ magnetic field orbit effects on ions. struction equipment concepts developed were generally similar in sophistication to modern WBS Item 3.3 Transportation Ground Support designs for roadbuilding and other field construc­ Facilities tion equipment. but were tailored to the specific task ot recknna construction. Figure 24 presents a Launch Site Selection conceptual overvil·w of the rl."i:tenna design and i:onstrudion approach. The launch site analysi~ task was motivated by the premise that selection of a low-latitude site A lktaikd grass-roots ·~ost e~timaling would offer significant cost advantages with model was dc\eloped and U!>t'd. This model resped to operations from the Kennedy Spa1.:t' ;K~·otmts for matt-rials. i:apital t'quipment and labor Center. where earth-to-low-orbit space transporta­ rnsts by task ;111d by schi:dule for i:ad1 dcment of tion arrives at a 300 indination orbit. A 30° plane the rectrnna syst.•m. This was necl·ssary hl·i:ause change is required to reach a geosynduonous equa­ there werl." no parametric rnst estimating methods torial orbit. It was prt''.'ollllll'd that th!s plane diangc that arl" appiii:abk to this problem.

19 D 180-25037-1

Table 4. SI'S Ground System Cost (Summary of Major Equipment, Buildings, Material and Labor) TOTAl.COSTIM&l !!!!. SENSITIVITY 11977 OOUARSI REC!iNlllA MATEJllAL: !>IOOE$ •~:DION 298 STEEl IN PANELS 1S I ltg!m' 349 ~ BUSeAAS 102" • :> S"l 165.000 M!:TRIC TONS 267 STllUCTURE CONCllET£ II' • 10 MflE~ ...OOVU:1 9 METRIC TOHSIMUOULE 81 STEEL lllEINFOACINGI JOO KgtMOOUlE 100 1100 KglAllOi. 3 AACHEStMOOVLEI l.AllOfl 50 CAPITAi. COST ITYPICAL MACHINE LIFE IS ~ S YE4115. COST TO PROJECT ONLY OORING USEI ...ill TOT"1. RECTENNA COST 1360

~ POWEii OISTllllUTIClN AlllO TRANSMISSIOIW IPEll SI'S PHASE • FINAi. REVIEW) Figure 24. Five GW Rectenna Construction TOT"L SPS GROUND SYSTEM COST Concept IEXCLUOING LANO ANO OEVE~OPMENT cosn The cost results of the analysis are sum­ mariud in table 4. The construction task was estimated to require 25 months; the cost of the rectenna is nearly $1 billion less than rectenna costs previously estimated by (admittedly inappro­ priate) parametric methods. The rectenna cost is dominated by materials cost as is evident from the LABOR pie chart associated with table 4. Materials quanti­ POWER CAPITAL ties are relatively well understood by comparison PROCESSING to equipment costs and labor content. Accord­ SYSTEM ingly, the uncertainty in the presented rectenna Rectenna Cost Distributions cost is .iudged to be on the order of 15% (one sigma).

WBS Item S.O Management & Integration program are within the range expected for the Industrial Infrastructure Department of Energy Terrestrial Silicon Program. Much of the production technology for the terres­ Of the several components that require pro­ trial program will be applicable to SPS. Further in duction rates significantly higher than those in the future, the buildup of production capability to present industrial experience, only the solar blan­ support an SPS production program of 10,000 kets represent a significant problem. Production megawatts per year will require production rates rates of SPS hardware are, in general. not high much higher than those for the prototype system. when compared to production rates in major U.S. industries. The solar blankets, however, will require Mi::sion Control major technological advances in production tech­ niques to meet the production demands of an SPS As a part of the Phase I activity, it was system. desired to develop a concept for mission control operntions to enable studies of mission operations Arthur D. Little's analysis of the photo­ in the Phase II activity. Several concepts were con­ voltaic market growth showed that the production sidered. The organization shown in figure 25 was rates of solar cells needed for an SPS prototype selected for the Phase II analysis ..

20 0180-25037-1

chain included a penalty for outages in the klystron power transmitter. These outages are also accounted for in the prediction of SPS plant factor in the maintenance and senice analysis. This double book'ceeping has been eliminated. The revised efficiency chain and power budget are shown in table 5.

Table 5. Updated Efficienc'/ and Sizing

Figure 25. C&C Center Relationships to Major System Elements and To Each Other 3.2 Summary of System Description

WBS Item 1.0 Solar Power Satellite Program

The selected system description is based on a nominal capability to place 10,000 megawatts WBS Item 1.1 Satellite (ground output nominal generation capacity) of SPS power in orbit per year. in the form of two The reference 5,000 megawatt system is 5 ,000 megawatt SPS's. illustrated in figure 26. The solar blanKd is com­ prised of panel of 50 um silicon solar cells e:-il:apsu­ The SPS efficiency chain was updated to lated in glass front and backside protection as illus­ reflect a slight improvement in intersubarray losses, trated in figure 27. T11e solar blanket desi!_;rt was and the orbit transfer compensation factor was revised to include shunting diodes required to pro­ deleted from the power budget to reflect geosyn­ vide shadowing protection. The shadowing protec­ chronous-orbit construction. The earlier efficiency tion is provided at the blanket panel level. In the

lOTAL -CILL AlllA 10. 1 k•2 TOTAL AllllAY MIA &:l.7 m2 lOTAL IATILUTI _. I 117.3 k•2 --1110k-: 8.2'GW

I ~1- ~CHOllD I .• ~ ...... • 1••·. . ' ~· I

,4 STlllllGS/15"' ..,... lNO SlGMl"T Ill TllllillDIATE SElllll£,. I - l J l"TElllllEDIAT~ SE GM ENT ~l;;1b~, 44-lliMllEGlllE NTS/BAY IOon- 1-- 1111 ITlllNGl/IAy f11 l'iUlfLSNIy ITRI IWG LENGTH 5 STlllNGS/15M END SEGMENT··-

Figure 26. 5000 Megawatt Reference Photovoltaic System Description

21 DlSG-25037-1

changed to reflect the use of I 0 buses independent of one another.

The power transmitter is shown in figure 28. The transmitter includes primary and second­ ary structures supporting 7 ,220 transmitter sub­ arrays and their po Ner processing, distribution, ... uu.1- and control equipment. Ten subarray configura­ tions are employed to effect a 9.5 dB, ten-step an~ ·~•, 1.97• -- approximation to a truncated Gaussian illumina­ ...... _.. tion taper. Each subarray includes phase control tlACllllDl' ...... ~.. :m,... systems, klystron power amplifiers, thennal con­ ...... Tlllffl ...... '·-· trol, instrumentation, and distribution and radiat­ Figure 27. Reference Photovoltaic System ing waveguides as illustrated in figure 29. The Description event of shadowing or a fault within the blanket, each affected panel is bypassed by the shunting diodes to prevent reverse breakdown failure.

The structure was revised to reflect the loading conditions appropriate to geosynchronous construction. The structure is a hexahedral truss made of graphite composite beams fabricated by beam machines. Solar array support beams are 12. 7 meter tribeams with a mass of 7.48 kg per meter length; others are 7.5 meter tribeams of 4 1 1 kg/m.

Failure effects analyses had indicated that the previous three-bus power distribution con­ figuration could be subjected to very large fault currents in the event of certain types of arcs. Because of thi'> problem, the bus configuration was Figure 28. Reference MPTS Structural Approach

RECEIVER RECEIVE HORN RECEIVE FILTER PHASE RECEIVER TELEMETRY CONTROL AX/TX G2JUNCTION

- ARC DETECTOR WAVEGUIDE STICKS (QTY 55) WAVEGUIDE COUPLING LINE (QTY 2)

Figure 29. Mechanical Layout of a Typical Klystron Module in the Outer Ring of the Space Antenna

22 0180-25037·1

phase control syskm employs the retrodirective spread-spect1um design developed by lin('om.

WBS Item 1.2 Space Construction and Support

Space construction operations will employ two bases as illustrated in figure 30. The base is low Earth orbit (L~O) will construct and service the independent electric OTV's and fum~tion as •Mlil.n..... cmsra.Ofl••MYft a staging base to accumulate and transfer payloads ..... • aa.1• •.11-. from HLLV's to electric OTV's (one OTV will :=::::: carry ten HLLV payloads) and transffr crews • AMAW' llODUl..I aa.Tll. IQUIP. between personnel launch vehicles (PL\''s) to per­ ·-•U -~.---~ •tt -~__ ...._err.. ..., sonnel orbit transfer vehicles (POTV's). - ...... ,. ....Of'll9 ••

••11LLITI .... -WLUl..-,.YOlllUllAnml • LmrtelfYUOllU.L The SPS construction base will be located - U..'JUlllUL... .. CDlTINUOUa in geosynchronous orbit. The selected system is a four-bay tnd-builder. The main features of this Figure 31. 4 Bay End Builder Base Features base are summarized in figure 31. The ba~line 8 x 16 bay SPS is constructed in two successive passes.

WBS Item 1.3 Space Transportation WBS Item 1.4 Ground Receiving Station

The integrated space transportation and The previous baseline rectenna was changed operations concept was described earlier in figure in structural design to accommodate the construc­ 28. The space transportation system consists of the tion cost reductions developed by General Electric elements summarized in tabfo 6. A budget for (discussed earlier). These revisions were summar­ annual flights and estimated costs is included in ized on page 19. the table.

ANTENNA CONSTRUCTION PLATFORM

LEO STAGING BASE

•MASS 1320MT •CREW 220 CONST PHASE 130 OPS PHASE

CONSTF'tUCTION GANTRY

CONSTRUCTION PLATFORM

GEO CONSTRUCTION BASE

• MASS~ 6370 MT STAGING DEPOT •CREW= 385 OPERATIONS UNDERNEATH

Figure 30. Orbital Bases-GEO Construction

23 Table 6. Space Transportation Element Summary WBSNO. HLLV OTHER COST COST REMARKS AND FLIGHTS FLIGHTS PER PER PER TRANSPOR- DESCRIPTION PER YEAR YEAR FLIGHT YEAR TATION SPS PROPEL- SUPPLIES ELEMENT CARGO LANT ' 1.3.1 HLLV • Two-Stage Series-Burn Winged Reusable (Heavy lift Rocket, VTOH L Total is 323 flights per Launch • Booster L02/LCH4 (51"'3/1709 Tons) HLLV Total year; does nf"t include (252) (641 (7) N/A $14M Vehicle) • Orbiter L02/LH2 (1976/329 Tens) $4522 SPS maintenanc-. • Payload Net/Gross (380/420 Tons) opcrrations. ! • Gross Liftoff Mass I 11 ,000 Tons I 1.3.2 COTV • Silicon Electric OTV, Annealable 1458 (COTV) 85 HLLV Flights are (Cargo Orbit •Payload up/down 4000/200 Tons 3528 (HLLV 27 required to launch Transfer ll~~ •Trip Time up/down 180/40 days CARGO) 252 39 (Fleet of $54M COTV Fleet; this cost Vehicl1;i • lsp = 8000 sec; Initial power 296 Mw 2 574 (HLLV 23 vehicles) ;s amortized into •Thrust 3345 N; Empty mass 1462 Tons Prop& COTV cost/flight. c -~~ •Argon 469 Tons; L02/LH2 46 Tons Supplies) - 1.3.3PLV • Space Shuttle Orbiter and ET with ~ (Personnel Flyback Booster and 75-passenger 38 Cost per flight assurnM i Launch personnel transfer module in payload N/A NIA N/A (Two Boosters $12M 456 other shuttle traffic as . Vehicle) bay. and 2 Orbiters) well as SPS support. El •Gross Liftoff Mass 2715 Tons - 1.3.4 POTV • Delivers crews and suppiie~ to GEO !Personnel base; two-stage reusable vehicle Orbit • Propellant L02/LH2 (800 Tons evenly 12 I 36 (POT'/) H LL V launch of OT I/ Transfer split between two stages) NIA 25 N/A (Fleet of $3M 350 (HLLV is amortized into cost Vehicle) • Empty mass 28 Tons per stage 2 vehicles) Propellant) •Thrust-Booster 1800 KN; upper stage ~ 900KN 1.3.5 PM ~ • Flight Control Module - Crew= 2, (Personnel Mass 4 Tons 12 (2 sets of 24 (PM) Module) • Passenger Module - 160 passengers The PM files as a NIA N/A 5 hardware, $ 2M 70 (HLLV ®=3 Mass 36 Tons payload for the POTV. • Cargo Module - 96 Tons crew provisions transported by Supplies) 0 (480 man-months), 15 Tons empty mass POTV)

TOTAL 6496 0180-26037-1

3.3 SPS Ma~ and Cost Summaries Table 8. Projected Unit Cost of One 5·GW SPS (Millions of 1977 Dollars) Table 7 presents a mass summary for the ma~' COIT 5,000 megawatt SPS. The is about 3% less RE~ ITEll POINT OF - FOii than that for the equivalent system at the begi..1- DEPARTURE CURRENT CHANGE ning of the study. The update in the selected - IASl-GWI system description has also resulted in a revision to 1.1 SA!'ELLITl 3111 1!!!. the cost estimates. Current values are compared to 1.1.1 SATELLITE ENERGY COlllV£- 2274 2111 REVISED EF•ICIENC'f the value from the 1977 study in table 8, in 1977 CHAIN; GEO CONSTllUCTION dollars. Reasons for significant changes are given in 1.1.2 SA TEL LITE l'OWUI TRANSMISSION $'fSTEll 1227 12U REDUNDANCY ADDED the tablt~. The presentation has been rearranged TO PttASE CONTROL 1.1.3 ll'IFOll!IATION -Ta from earlier results to reflect the current work CONTROL 42 ·~ breakdown structure and SP-paration of indirect 1.1.4 ATTITUDE CONTROL a I STATION •EEPINC .... ,.;.i cost factors from direct outlays. ' 1.1.5 COIWUNICATIONS I 111 111

1.1.e INTERFACE flNCLUOED IN Table 7. Baseline Solar Power Satellite Nominal ENERGY CONVI

Mass :Wmmary 1.2 SPACE CONSTRUl;TION AND ~T •DIRECT SM•• 211 GEO CONSTRUCTION - • CAPITAL CO

4.0 SPS PROGRAM EVOLUTION

Analyses of the programmatic structure of Table 9. SPS Development Phases an SPS program have resulted in the multi-s. ·p

approach summarized in table 9. Each step will ITIP KNOWLIOOI TICHNIC•l .-MJQ .. AM OAINID CONFIDENCE DlCllfON J provide knowledge nnd technical confidence lead­ IXPlOlllATD1'V IVITEMI CONCl" THIJtiMIJrfO ""ocno ITUOIH OPTIC... PIRITOJID'~ TtCHNICAL WITl'f SYSTEMS AND I ICOMl'Lftfl Oft ECONOMIC aAllUUl!q t:VALUATION STUCIH ing to a program dec!sion tc initiate the r.ext step. TO IVUlfTUAL IUCCHI If the appropriate technical confidence from any IYITEllfl CON..:IPTUAl..DHIGJll OHIGN~ACHH INl'l'IATl ITUDtH CHA ..ACTE'41ZATIONIOf E,oOST THAT CAN ..ECMfllOLOGY step is not achieved, then the approach would be llUCTIO IAll LINH, "'0U•LY ACHlfVf "UIAlllCH~D ll:CMNO\.OGY "-"'ORM.A.h_. I TIClWl!tCAL .\ND CONflNUE EVALUATION modified or possibly the program terminated if Dl.llCTIVH ICONC*IC oaJECTIVU STUDtH TICHNOC.OOV ACTUAL TlCMNOloGV ffCHNOLOGV P'llUOllMAHCf INITIAff (NQINlUUNG major cHffii:ulties were encountered. AtaAMH N"'OllMANCf ll#'P'OllTI 9'S DHIQN tfCHNIOUH ~HES t>fVfLCJlllMENl

The next phase of SPS activities beyond t:NGINUllUNU '"I AND •'l'lltllltl 9'I OHIGN ...,,lllOACHH INIT!Afl ,UL-~ ICALt: lt:..:liNIQUfl ENff1• 40 ,.Ell,OllMANCI. YALIOATlD, 'lllfFEltfllfD DEVlLONlfNT the present evaluation phase will be ground-based DfVf.LO'MfNT ""4DIOUJ111•c. IA.HI ,0111 "'"°ACMllHLlCllD ..cmc•-: JNI exploratory research. Recommended research fULL ICAU ...'WOIUtt" ftCAN•t INT£" Ca.fftCIAl DIVILOf'MINT succt••uL t." 'llODUCTION activities for the beginning of this GBER phase are CotlilllllltCIAl.IZtP presented in tabie IO.

25 0100.25037-1

Tatle 10. Recommendations For First Year of Ground-Based Experimental Research

DISCIPLINE RESEARCH ITEM SOLAR • Begin deYelopment of anr>ealable silicon and gallium arsenide arrays. ARRAYS • Improve data base on solar cell/blanket radiation degradation for long-term use . • Begin rciearch and tests on high-voltage array/plasma effects. THERMAL • Begin development ::it neat pipes and circulating system~ technology for thermal control of MPTS & FLUID and power processing systems, including zero11 heat transfer and thermal radiators. SYSTEMS • Study and research techniques/technology for in-space assefrtbly and repair of fluid systems . • Begin development of ln'1Q-life thermal coatings and repair/restorP technology . • Update thermal engine SPS evaluation (as part ot systems studies) . MICROWAVE • Begin development of high-efficiency low-noise klystrons and magnetrons. POWER TRANSMISSION • Expand solid state research: Devices, circuits, radiators, thermal control, power supply, and integration.

• Begin/ext~nd experimental investigation and evaluation of phase control technologies. • Begin experimental evaluation of rectenna concepts. • Begin laboratory test programs of transmitter subarray integration . SPS • Continue integrated structure and control dynamics analyses. STRUCTURES • Begin development or structural verification technology . MATERIALS • Continue/extend critical materials evaluation and accelerated life test technohgy. & PROCESSES • Begin development of high·temperalure composites technology for SPS . • Begin development of low-coefficient-of-thermal-expansion RF waveguide technology . • Explore alloys for thermal engines and heat exchangers . • Begin development of bonding and welding technology. FLIGHT • Extend/develop control theorv and algcrithms including active damping . CONTROL SYSTEMS • Perform design study of data and communicat10ns systems. SPACE • Continue beam builder technology. CONSTRUCTION • Extend crew aids and manipulator technology . • Evaluate and define needs for simul.itor technology . - 01~25037-1

Table 10. Recommendations For First Year of Ground-Based Experimental Research (Continued)

DISCIPLINE RESEARCH ITEM

SPACE • Begin development of critical high-power electric propulsion technology . TRANSPORTA- TION • Explore/t.39in development of reusable LH2 insulation. • Update SPS trar1sportation system definition (emphasis on HLLVI . SPSPOWER • Evaluate and extend high-voltage insulator techaology. DISTRIBUl"ION • Extend high-voltage lightwei!tit transformer and pawer processor technology . • Analyze SPS pawer transients and EMI effects. • Begin technology for SPS power interrupters and circuit breakers. • Begin technology de-1eloprnent for high-voltage cables and insulations. ENVIRONMENT • In addition to environmental and relat'!d research to be conducted by DOE, space plasma effects and spacecraft charging research should be initiated.

Technology verification flight experiments We have identified the need for an SPS may be nee

27