Click for next page ( 32


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 31
4 Spacecraft Electric Power BACKGROUND AND STATUS A spacecraft's electrical power system generally consists of the primary power generating unit (solar or nuclear), the power management and distribution system, and an energy storage unit. The power system typically accounts for 25 to 35 percent of spacecraft dry mass (Herrera and Kuck, 1992; Larson and Wertz, 19921. In solar power systems, the energy storage unit represents about one-third of the mass of the power generating unit. Nuclear radioisotope power systems are independent of sunlight and require little or no energy storage. The choice of power system depends on such considerations as power level (average and peak), mission lifetime, and operating environment. In the vast majority of cases, solar power systems are preferred over nuclear power ones due to their lower cost and simpler launch approval procedures. POWER SOllRCES Solar Arrays For space missions that are sufficiently close to (and with an unobstructed view of the sun, solar cell arrays can meet most near-term space power needs for small, lightweight spacecraft by converting solar energy to electrical power. Solar cells can be mounted directly on the external surface of the spacecraft or on panels that are deployed once the spacecraft achieves orbit. The performance of solar arrays is quantified by (~) the specific power, power delivered by solar array per unit weight (watts per kilogram); (2) the power density, the power delivered by a solar array per unit area (watts per square meter); and (3) the survivability level, the capability of an array to survive hostile attack (for DoD missions) and the space environment. Solar array configurations are either (~) unconcentrated, in which case they operate on the as-received solar flux or (2) concentrated, where the solar cells' output is increased by the use of lenses in order to focus more solar radiation onto the cells. Unconcentrated arrays are not necessarily planar; they can also be cylindrical or spherical 31

OCR for page 31
32 Technology for Small Spacecraft and can be subdivided into rigid and flexible arrays. The flexible arrays include blanket arrays and inflatables. Concentrated arrays are usually planar. Silicon and gallium arsenide cells are currently used on spacecraft arrays. Of the two technologies, silicon solar cell technology is the more mature and has the lower cost per watt. However, gallium arsenide, first flown in 1983, has a higher energy conversion efficiency and is inherently more resistant to radiation Carson and Wertz, 19921. Unfortunately, gallium arsenide costs approximately two to five times more than silicon and is 2.2 times as dense (Chetty, 19911. Other cell types, such as amorphous silicon, aluminum gallium arsenide, indium phosphide, copper indium diselenide, cadmium tellur~de, and multibandgap cells, are uncler development and have been flown experimentally, but have not yet been flight qualified for major U.S. operational spacecraft (Cooley, 1991~. For example, although copper indium diselenide and amorphous silicon cells are both now flying on the UPS- satellite, and a new program has been funded by OACT for a flexible, indium phosphide, thin-fiIm solar array experiment, neither of these systems has been qualified to an acceptable level for deployment in major U.S. flight systems (Landis and Hepp, 1991; NASA, 1993a). NASA and DoD, both with industry support, are developing high specific power solar array systems for small spacecraft applications. The status of research within each organization is summarized below. NASA Programs NASA's research activities on solar cell and solar array technology are currently centered at the LeRC. Until recently, JPL also had a program for high-performance solar array technology, which was complementary to I~eRC's program on high-eff~ciency, radiation-resistant solar cells. LeRC. LJeRC's power systems programs focus on Earth orbital applications. LeRC contributed to the large area silicon cell technologies (with strong contributions in cell technology from the European Space Agency) that were user] on the Hubble Space Telescope and were scheduled for use on Space Station Freedom, prior to the latest redesign (Cooley, 1991~. Flexible, roll-up arrays made from these silicon cells were initiated at LeRC and further developed by the U.S. Air Force at the Wnght-Patterson Air Force Base. These flexible silicon cell arrays were eventually utilized on the Hubble Space Telescope. LeRC is currently working on many advanced solar cells, such as copper indium diselenide cells, amorphous silicon cells, indium phosphide on germanium cells, cadmium telluride cells, and other multibandgap cells. LeRC is also performing research on ~ The multijunction approach utilizes the solar spectrum more efficiently by stacking several bandgap cells (e.g., a thin gallium arsenide cell stacked on top of a silicon cell) in series such that successive junctions convert different frequency ranges of sunlight.

OCR for page 31
Spacecraft Electric Power ultralight flexible panels that could utilize many of these advanced cells in various flexible array designs. One inflatable array design under development at LeRC utilizes flexible silicon cell panels, which can provide total power in the range of 200 to 500 watts at a specific power of 200 watts per kilogram. In the future, substitution of silicon cells with cells of 20 percent indium phosphide on germanium could result in a more radiation-tolerant array with approximately 1 percent degradation over 10 years and an efficiency of greater than 17 percent. Use of arrays with indium phosphide on germanium solar cells is expected to produce a specific power of 130 watts per kilogram and would enable long-life missions in polar orbits and other high-radiation environments (Budinger et al., 1993). At a recent joint LeRC, JPL, GSFC workshop, the participants concluded that most of LeRC's advanced solar cell work could be ready for flight qualification within seven years (NASA/OACT, 19931. JPL. JPL`'s research and technology programs generally focus on planetary exploration. Most of the solar array work at JPL has been carried out under the Advanced Photovoltaic Solar Array (APSA) program, which was cancelled due to funcling limitations. A flexible, lightweight fold-up solar array with a mass of 1 kilogram per square meter of collecting area was developed by JPL antler the APSA program, in conjunction with TRW's Space and Electronics Group. The array design incorporates both thin silicon cells and thin gallium arsenide on germanium substrate cells. The cells are attached to a flexible Kapton polyamide blanket. A fiberglass deployment mast is used. The complete array system produces 135 watts per kilogram (Scala, 19931. Although a full-size flexible blanket array was not flight tested, fabrication, integration, ground vibroacoustic testing, and ground deployment tests of a prototype 6-kilowatt APSA array were successfully completed (Kuriand and StelIa, 19921. The JPL/TRW APSA flexible solar cell blanket array technology is currently scheduled for use on two future missions: the NASA Earth Observing System AM-! mission, using gallium arsenide on germanium solar cells, and a DoD mission supported by TRW, also using gallium arsenide on germanium solar cells. Prior to program termination, the midterm goal of the APSA program was to implement cell fabrication methods and array assembly procedures for thin-film solar cells that could increase array specific power to 190 watts per kilogram. As an example, advanced thin-f~im gallium arsenide or aluminum gallium arsenide cells produced by Kopin Corporations's Cleaved Lateral Epitaxy Films for Transfer (CLEFT) process could have been utilized by 1996. Just prior to termination, lightweight flexible modules utilizing thin-fiIm gallium arsenide solar cells from this process were fabricated for JPL and sent to LeRC for thermal cycling tests. The long-term APSA program goal was to develop flexible solar cell blanket array designs with array specific powers of 300 watts per kilogram (at 12 kilowatts total power) by the year 2000, and of 20 to 25 kilowatts total power ultimately. IPL has been working with industry and universities to improve the performance of silicon cell arrays at distances from two to five astronomical units from the sun. These 33

OCR for page 31
34 Technology for Small Spacecraf! low-intensity, low-temperature conditions degrade the performance of conventional silicon cells due to metallization and silicon interactions. JPL hopes to solve the low- intensity, low-temperature technology problem by 1995, but the program is likely to be terminated due to budget reductions. DoD Programs DoD, including the U.S. Air Force, BMDO, and the U.S. Navy has performed substantial work to increase the specific power in solar arrays for low-power, small spacecraft applications. Prior to the availability of flexible arrays analogous to those developed in the NASA APSA program, DoD platforms generally utilized planar silicon arrays with specific powers ranging from 40 to 60 watts per kilogram. These arrays have beginning-of-life efficiencies of 12 to 15 percent that drop to end-of-life efficiencies of 10 to 12 percent (Russell et al., 19921. For the same specific power, advanced ~ . ~ ~ ~ ~ ~ technology for gallium arsenide on germanium cells could offer a nominal end-of-life efficiency as high as IS percent (Russell et al., 19921. In the past, DoD has supported research on concentrator arrays through venous programs, such as the now-cancelled BMDO/Martin Marietta Survivable Power Subsystem Demonstration program and its predecessor, the Survivable Concentrator Photovoltaic Array program. These technology programs were supported by the Strategic Defense Initiative Organization with plans for future incorporation into the Brilliant Eyes program. Although the design objectives for DoD concentrator technology generally focus on survivability rather than efficiency and high specfic power, there were several DoD technology advancements made in these programs that may hold promise for future NASA spacecraft. For example, the concentrator technology developed for the Survivable Concentrator Photovolta~c Array and the Survivable Power Subsystem Demonstration programs (mini-Cassagrainian arrays developed by TRW, minidome arrays developed by Boeing, and slats being developed by General Dynamics) has the potential to reduce by a factor of two the cost of planar arrays and to eventually provide specific power of around 80 watts per kilogram, at beginning-of-life overall efficiencies of 24 percent. However, with the formal dissolution of the Strategic Defense initiative Organization into BMDO, and the termination of several programs such as the Survivable Power Subsystem Demonstration program, the advancement of the concentrator technology within DoD is uncertain. Currently, the Air Force is trying to maintain work in multibandgap cells and thin-fiIm cells despite the absence of funding from BMDO and overall DoD budget cuts. Production of cells with intermediate capability levels is within a year or two of completion. BMDO, the Naval Research Laboratory, and NASA are jointly sponsoring the Deep Space Program Science Experiment (Clementine) program, which will demonstrate lightweight technology components with a lunar mapping and asteroid flyby mission that was launched in January 1994. The spacecraft is utilizing gallium arsenide on germanium

OCR for page 31
Spacecraft Electric Power cells that are 0. 14 centimeters thick at 40 watts per kilogram. These are the thinnest gallium arsenide solar cells flown to date (Nozette, 19931. Nuclear Technology Nuclear radioisotope power systems convert the heat from a radioisotope heat source into electricity. Current radioisotope power systems are more compact than solar systems and are mass-competitive, but they are quite costly and require a complicated launch approval process. Therefore, solar power is always preferred over radioisotope power, except for deep space or sun-obscured missions, where there is too little sunlight for efficient photovolta~c conversion, and for missions near the sun, where the solar flux is too intense and too variable for practical solar powered systems. The use of nonrechargeable batteries for primary power has been proposed for some missions, but for the power levels and operating times required by those missions, batteries become prohibitively heavy for a small spacecraft. For such missions, radioisotope systems are enabling and are used in spite of their cost and complicated launch approval process. In today's systems, thermoelectric unicouples are used to convert heat into electricity. These systems, while reliable and long-lived, are inefficient. Substantial reductions in cost and mass of radioisotope power systems can be achieved through development of more efficient power conversion technologies. Potential conversion technologies include advanced thermoelectric materials, Stirling engines, alkali metal thermoelectric converters, and thermophotovoltaic systems. The last three options offer the possibility of tripling or even quadrupling the efficiency of thermoelectric converters, with corresponding reductions in the cost and mass of the required radioisotope fuel. Both NASA and the Department of Energy (DOE) have invested in conversion technologies for radioisotope power systems that could be used for small~spacecraft. The status of research within each organization is summarized below. NASA Programs LeRC. LeRC and its contractors, Mechanical Technologies, Inc., Sunpower, STC, and others, have been working on free-piston Stirling engines internally coupled to linear alternators, to increase engine reliability and lifetime by eliminating the need for external seals on moving shafts. Mechanical Technologies, Inc., recently completed a large system of that type, possibly for a second-generation space station or as an alternative conversion system for the since-cancelled SP-lL00 reactor program. The engine produced an electrical output of 12 kilowatts at an overall system efficiency of over 23 percent. The system gave an initial performance that was in excellent agreement with analytical predictions, but it has not undergone life-cycle testing. A scaled-down, Stirling engine was recently designed for possible use in NASA's proposed Pluto Fast Flyby mission. Analytical models showed that a 75-watt engine would have an efficiency of 23 percent. 35

OCR for page 31
36 Technology for Small Spacecraft JPL. For nearly twenty years, JPL and Advanced Modular Power Systems have been developing a highly efficient, static alkali metal thermal-to-electric converter for the direct conversion of heat to electricity. Alkali metal thermal-to-electric converter cells suitable for zero-gravity conditions were recently tested by Advanced Modular Power Systems and yielded an efficiency of 9.6 percent at 700C. Advanced Modular Power Systems predicts efficiencies of 15 to 20 percent through advanced cell designs and higher-temperature operation. Based on that prediction, JPL system studies estimate that an alkali metal thermal-to-electnc converter-generator for the Pluto Fast Flyby mission that uses two standard radioisotope heat source modules would have a system mass of 9.7 kilograms. Such a system has not yet been demonstrated, and one major uncertainty about these devices is their ability to withstand launch vibration. JPL is also working on thermophotovoltaic conversion systems, which are an outgrowth of recent advances in photovoltaic materials developed for high-efficiency solar cells. Instead of converting solar radiation to electricity, they convert infrared radiation emitted by the radioisotope heat source. Since infrared radiation has a very different spectral distribution than solar radiation, different photovoltaic conversion materials are required. One material under test by Boeing for JPL is gallium antimonide, whose bandgap is well matcher! to the infrared spectrum. Relatively high efficiencies have been demonstrated with this material, and extremely high efficiencies (greater than 30 percent) may be achievable through addition of reflective filters or mirrors to return the unconverted radiation to the heat source. DOE Programs DOE has extensive experience with nuclear power technology. In the late sixties and early seventies, thermoelectric unicouples employing silicon germanium materials were developed by RCA and General Electric for DOE. These unicouples were successfully flown in 200- to 300-watt radioisotope thermoelectric generators (RTG) on several NASA missions and are slated for use on the proposed Cassini mission. The RTGs, developed by DOE laboratories and contractors, with assistance from JPL, have rather low efficiencies (less than 7 percent) but have proved extremely reliable and long- lived (approximately 150,000 hours). A typical small (70-watt) RIG for the Pluto Fast Flyby mission, based on silicon germanium unicouples, has a mass of 15 kilograms, with a cost of $51 million estimated by DOE for three fueled flight units. During the 1980s, a modular, radioisotope heat source module (the General Purpose Heat Source) was developed and safety qualified by DOE laboratories and contractors. Being modular, these heat sources are adaptable to a wide range of power levels and conversion systems. RTGs are flying on the Galileo and Ulysses missions and are slate to be flown on the proposed Cassini mission. Thermoelectric multicouples employing silicon germanium materials with additives have been under development by DOE contractors for over ten years, but their development was recently suspended by the department. The multicouples were developed for use in modular RTGs, which are scalable over a wide range of power

OCR for page 31
Spacecraft Electric Power levels with minimal redesign. They are only a little more efficient than unicouple RTGs, but offer a significantly higher specific power. They also make it possible to generate high voltages from small RTGs (28 volts DC has become the accepted industry standard power bus voltage for small spacecraft). Multicouples have been successfully tested for up to 15,000 hours, but their measured degradation rates were about twice as high as those of unicouples. As of this wnting, funding for continuing development has not been allocated. DOE has also been involved in work on other conversion technologies, sponsoring Fairchild Space and Defense Corporation to prepare and analyze detailed system designs for integrating the advanced conversion systems (Stirling engines, thermophotovoltaic systems, and alkali metal thermoelectric converters) with a radioisotope heat source and a heat rejection radiator. A recently completed system design study showed that replacement of the RIG with a thermophotovoltaic generator for the proposed Pluto Fast Flyby mission would reduce the required number of costly heat source modules by 60 percent, reduce the power source mass by over 50 percent, and triple or quadruple the system efficiency. BATTERY TECHNOLOGY FOR ENERGY STORAGE NASA Programs NASA's battery technology research and development activities are located at LeRC and JPL and have focused principally on the following systems. Nickel Cadmium (NiCdJ Batteries. Rechargeable NiCd batteries, which have been used in spacecraft for over 20 years, may be considered the currently available technology, although they are gradually being phased out, due in part to government restrictions on manufacturing processes involving cadmium and in part to the increasing availability of superior alternatives. Individual Pressure Vessel (IPV) Nickel Hydrogen (NiH2J Batteries. In an IPV NiH2 battery, each cell (cathode-anode pair) is individually contained in its own pressure vessel. (Pressure vessels are neec ed to contain the cell's hydrogen gas at high pressures of 6.2 x lo6 pascals to 6.9 x lo6 pascals.) A NiH2 battery is typically composed of 22 cells. The first such battery was flown in 1977 by the Naval Research Laboratory. Today, it has replaced the NiCd battery for defense applications in geosynchronous orbit, and it is quickly becoming the preferred battery technology in low Earth orbit as well. IPV NiH2 batteries are more voluminous than NiCd batteries due to individual cell containment in rounded vessels, but they offer substantially longer cycle (charge/discharge) lifetime, greater depth of discharge, and improved tolerance of abuse (e.g., overcharging). 37

OCR for page 31
38 Technology for Small Spacecraft LeRC has developed a new, lightweight nickel electrode that is usable in IPV NiH2 batteries as well as in other nickel-based ones and that will increase specific power. Nickel Metal Hydride (NiMH) Batteries. NiMH batteries, which are being studied by LeRC and IPL, would be about 10 percent lighter than Nigh batteries and could provide up to about 50 watt-hours per kilogram. They also offer a longer shelf life, lower cost, and higher power density in a reduced volume. NiMH batteries have not yet been flown, and some development work remains to be done, but they represent an attractive near-term option to replace NiCd batteries on the low-power end of the small spacecraft spectrum. Lithium Batteries. Lithium batteries are a highly promising mid- to far-term technology. Lithium titanium disulf~cle (LiTiS2) batteries, for example, are being studied at JPL for low-power (less than I-kilowatt-electric) applications. These batteries have a high power density (100 watt-hours per kilogram), a lifetime of i,000 cycles at 50 percent depth of discharge, a lO-year shelf life, and low volume, all of which would make these batteries well-suited for long-duration planetary missions. Lithium polymer batteries are being investigated by LeRC and JPL to achieve a specific power goal of 150 to 200 watt-hours per kilogram. Substantial effort is still needed in electrolyte research, but the ultralight weight and small size of these batteries would provide important benefits. In the far term, lithium primary (i.e., nonrechargeable) batteries may present an alternative to radioisotope power systems for outer planetary missions, but such batteries would have much lower specific energies than RTGs or other radioisotope systems. High energies are required not only for extended survey or exploration missions but also for brief flybys like Pluto Fast Flyby. The mission's power requirements are determined not by the length of the flyby but by the amount of stored data to be transmitted back to Earth from deep space. In the case of the proposed Pluto Fast Flyby mission, the power demand stipulated by the current design would have to be reduced by orders of magnitude to lower the battery mass to that of the radioisotope power system. Clearly, that would be a very different mission, with a much smaller scientific return. Other Government Programs DoD, particularly the U.S. Air Force Phillips Laboratory, BMDO, and the Naval Research Laboratory, and DOE have several advanced battery concepts that complement NASA's work. Common Pressure Vessel (CPV) NiH2. The CPV NiH2 battery is a logical near-term follow-on system to the individual pressure vessel NiH2 battery. The containment of all! cells in a single pressure vessel allows for a significant reduction in battery volume as compared with the IPV battery concept. As currently designed,

OCR for page 31
Spacecraft Electric Power however, the CPV battery does not permit monitoring of individual cell performance and the ability to switch defective cells off-line. However, the CPV battery offers significantly higher specific power (nearly 50 watt-hours per kilogram), lower cost, and simplified electrical and thermal interfacing. It is a strong candidate for use in the "larger members" of the small spacecraft family. CPV battery technology development is centered at the Naval Research Laboratory. Additional work is underway at U.S. Air Force Phillips Laboratory. A CPV NiH2 battery has been scheduled to fly within a year on a DoD spacecraft as part of a joint effort between the Naval Research Laboratory and industry under a Cooperative Research and Development Agreement. A second CPV battery is being used on the BMDO Clementine mission, launched in January 1994. CPV battery technology is currently planned for use in the }~DiUM_/SM no_ A; A~..~A be., Lockheed (AuCIair et al., 19931. bpd~rd1 ~ Oilily U~V~1~ by Sodium Sulfi'r (NaS) Batteries. NaS batteries have the potential for a further doubling of specific power up to 100 watt-hours per kilogram, but they require additional development work and a flight experiment in order to complete qualification. This appears to be a promising technology, but predominantly for larger spacecraft. LeRC had planned a NaS cell flight experiment in 1995, but this has been cancelled due to funding limitations. (This decision may be subject to reconsideration by NASA.) The U.S. Air Force Phillips Laboratory remains committed to the technology, but it also has only limited resources available. Lithium Batteries. The U. S. Air Force and BMDO are conducting early work on solid-state batteries (including lithium titanium disulfide and lithium polymer), with a long-term performance goal of 200 watt-hours per kilogram. DOE has produced tiny, bench-scale thin-fiIm lithium batteries to power individual chips. Preliminary work is underway to increase production rates and to develop larger rechargeable lithium batteries with a calculated specific Dower of more than 300 watt-hours per kilogram. The Lawrence Livermore National Laboratory has performed life-cycle testing on a lithium ion battery that may be suitable for low-Earth-orbit applications in the near term. FINDINGS AND PRIORITIZED RECOMMENDATIONS Improvements in power generation and storage technology would be beneficial to virtually all classes of small spacecraft missions, reducing mass and cost, as well as enhancing performance. The Pane! on Small Spacecraft Technology believes that developments in low- weight, high-efficiency solar cells and arrays would enhance not only the power generation capability of small spacecraft but also that of solar electric propulsion. The pane! also found that advanced, high specific power battery technologies for space 39

OCR for page 31
40 Technology for Small Spacecraft applications have received insufficient attention and lag considerably behind developments in terrestrial power storage. Successful development of compact, long-lived, high specific power battery systems, coupled with improvements in power generation and management, will significantly enhance the utility and affordability of small spacecraft by offering reductions in mass and launch costs and improved performance. The use of currently available radioisotope power system technology for interplanetary missions and others where the sun is obscured results in higher-than- desired cost or mass. Technologies that would enable more efficient, lighter-weight systems have shown promise in research and development programs at NASA and DOE. Development times are probably too long to permit use in near-term planned programs such as the proposed Pluto Fast Flyby and Mars Pathfinder missions. However, for future deep space missions and Martian planetary surface investigations with small spacecraft and microrovers, especially at high latitudes, these technologies are enabling. Future developments in the radioisotope power technology require the active involvement of DOE as mandated by the Atomic Energy Act of 1954, which is still in effect. In considering the use of radioisotope power generation systems in future spacecraft, special attention must be paid to ensuring that there is a source of plutonium-238. The U.S. reactors capable of its production have been shut down. Arrangements should be made to ensure the availability of a foreign source (e.g., Russia, France, or the United Kingdom). Considering that investments in this technology area will produce returns across the entire spectrum of missions, the pane! recommends the following, in priority order: I. An advanced solar array program should be initiated at a funding level that will allow reaching a goal of 200 watts per kilogram with 5 to 10 kilowatts of total power within the next five years. 2. The development, characterization, and testing of NiMH batteries for low- power small spacecraft should be completed. 3. Building on the work already completed for the Clementine mission, the characterization and testing of CPV NiH2 batteries for mid- to high-power small spacecraft should be completed. 4. The development of lithium alloy (LiTiS2) batteries, particularly for low- energy-demand planetary missions, should be continued. 5. The application of lithium ion batteries developed by DOE should be evaluated for possible use in low-Earth-orbit spacecraft. If found promising, the technology should be adapted for small spacecraft. 6. For mid- to far-term applications, the development of lithium polymer batteries should be accelerated.

OCR for page 31
Spacecraft Electric Power 7. In the long-term, work on other advanced solar cell and solar array technology, including thin-fiIm cell development, inflatable arrays, ant} flexible blanket wing APSA arrays, should continue at an increased funding level, with the goal of achieving a specific power of 300 watts per kilogram. 8. There is a small but important subset of small spacecraft missions that cannot use solar power or batteries and that are enabled by radioisotope power systems. For those missions, development of more efficient conversion systems to reduce heat source mass and cost would be beneficial. Radioisotope power system designs using Stirling, thermophotovoTtaic, and alkali metal thermal-to-electric converter conversion techniques should be jointly evaluated by NASA and DOE, and the ability of these techniques to satisfy various NASA missions should be assessed. Based on the evaluation, NASA and DOE should select one or more of these systems for experimental demonstrations of its performance against specific pre-determinec] criteria that are peculiar to the approach selected. NASA and DOE should then select the most promising approach for further clevelopment. A decision about flight demonstrations should be made contingent on future NASA planning of missions that would utilize the technology. 9. Research on concentrator arrays, with a goal of reaching power densities in excess of 300 watts per kilogram at one-half the cost of existing arrays, should be increased. 41