'.

IAF-02-R.4.06 Future Photovoltaic Power Generation for Space-Based Power Utilities

Sheila Bailey, Geoffrey Landis, Aloysius Hepp, and *Ryne RaffaeIIe NASA Glenn Research Center, MS 302-1 , Cleveland, OH 44135 *Rochester Institute of Technology, Rochester, NY 14623

53rdInternational Astronautical Congress The World Space Congress - 2002 10-19 Qct 2002/Houston, Texas

For permission to copy or republish, contact the International Federation, 3-5 Rue Mario- Nikis, 75015 Paris, France This is a preprint or reprint of a paper intended for presentation at a conference. Because changes may be made before formal publication, this is made available with the understanding that it will not be cited or reproduced without the permission of the author.

IAP-02-R.4.06 FUTURE PHOTOVOLTAIC POWER GENERATION FOR SPACE-BASED POWER UTILITIES Sheila Bailey, Geoffrey Landis, Aloysius Hepp, and “Ryne Raffaelle NASA Glenn Research Center, MS 302-1, Cleveland, OH 44135 *Rochester Institute of Technology, Rochester, NY 14623 Sheila.Bailey@grc.nasa.gov

ABSTRACT This paper discusses requirements for large weight gigawatt (GW) space power generation. earth orbiting power stations that can serve as Investment in solar power generation central utilities for other orbiting spacecraft, or technologies would also benefit high power for beaming power to the earth itself. The military, commercial and science missions. current state of the art of space solar cells, and a These missions are generally those involving variety of both evolving thin film cells as well solar electric propulsion, surface power as new technologies that may impact the future systems to sustain an outpost or a permanent choice of space solar cells for high power colony on the surface of the moon or mars, mission applications are addressed. space based lasers or radar, or as large earth orbiting power stations which can serve as central utilities for other orbiting spacecraft, or INTRODUCTION as in the SERT program, potentially beaming Starting with Peter Glaser’s initial 1968 power to the earth itself. proposal’, many people have discussed use of A significant risk element for any satellite the satellite solar power system [SSPS] as a power system is the photovoltaic array. This means of supplying energy to the Earth to was identified twenty years ago in the National replace fossil fuel sources. The recent Research Council (NRC) review of SSPS2 as prominence of the “greenhouse effect” from one of the most critical areas where burning of fossil fuels has again brought extrapolations from current technology in terms alternative energy sources to public attention, of cost and performance were made. That and the time is certainly appropriate to analysis is still true today. reexamine the SSPS. A recent NASA program, Space Solar Power Exploratory Research and Technology SPACE AND GROUND SOLAR POWER (SERT), investigated the technologies needed to A first step toward space solar power will provide cost-competitive ground baseload be to demonstrate multi-megawatt power electrical power from space based solar energy production with ground-based solar arrays. conversion. This goal mandated low cost, light Proponents of SSPS often disparage the

Copyright 0 2002 by fhc Amcrican institute of Acronautics and Astronautics, inc. No copyright is asscrtcd in thc Uuitcd Statc undcr Titlc 17, U.S. Code. Thc U.S.Govcrnment has a royalty-frcc licensc to cxcrcise all rights undcr thc copyright claimcd hcrcin for Govcrnmcntal purposcs. All othcr rights arc rescrvcd by thc copyright owncr. potential use of ground-based solar energy, the most significant statistic in this time period, possibly considering ground-based systems as a however, is the shift in use of photovoltaic competitor. Nothing could be further from the systems from only 15% grid connection in 1991 truth: ground-based and satellite-based solar to 55% grid connection in 2001. Off-grid use power are complementary technologies, and dropped from 60% to 44% in that same time satellite-based solar power will only be period. economically viable if terrestrial power is also Governments worldwide have encouraged viable. This synergy and diversity in space the development of the photovoltaic industry and terrestrial photovoltaics was recently with new manufacturers in India, Korea and discussed3 at the 200 1 Space Photovoltaic many other developing countries. In addition Research and Technology conference. Europe, Japan and the United States are also Experience with ground-based solar power encouraging both domestic production and is a necessary step to assess the technology, domestic markets. Governments have sought to define and trouble-shoot the manufacturing address the relatively high capital start-up costs technologies, and move photovoltaics along the of the photovoltaic industry by identifying and learning curve to low-cost production. guiding opportunities between producers and Many ground sites exist in the US. with over investors, clarifying barriers, instilling 300 clear days per year. Flat-plate confidence in consumers, and improving photovoltaic systems will also provide interactions with other industries that have significant power during overcast days. The solved similar problems. difficulty with terrestrial solar power is that it The U. S. Photovoltaic Industry published a provides power only during the daytime. Solar Electric Power roadmap in April of 200 1.’ Ground-based solar power could be viable for a The roadmap goals were to achieve flexible significant fraction of the power needs of the high-speed manufacturing to create a 200 MW United States due to the fact that the peak loads factory by 2020. The incremental needs were are typically in the da~time.~An additional listed as a 5-fold reduction in module advantageous feature of terrestrial photovoltaics manufacturing costs by 2010, then a 10-fold is the short construction lead-time required and reduction by 2020 with a 40-fold increase in the ability to add capacity in small, modular module manufacturing by 2020. Required increments. This allows photovoltaic actions to achieve these goals were identified as installations to avoid the uncertainty of designing of lower-cost module packaging, forecasting power requirements far in advance, developing high-volume, high-throughput, high- and also allows rapid progress along the learning efficiency cell processes and moving from curve. company-specific equipment manufacturing In the ten year interval between 1991 and toward equipment design that can be transferred 2001 there was a six-fold increase in worldwide to and used by more than one manufacturer. terrestrial total sales and a cost decrease in Both industry-government and industry- modules from $6rW to $4rW and a decrease in university interactions were seen as critical to installed system cost from $12/W to $SN. reaching these goals. All of these issues are also There was also a shift, and in some cases a relevant for future SSPS needs. merger, from the four largest suppliers in 1991, Current photovoltaic module production is ARCO/Siemens, Solarex, Sanyo, and Kyocera, about 400 MW(pwkjyear,and increasing by to Sharp, Kyocera, BP Solar, and Siemens/Shell about 40% per year, as seen in Figure I. as the four largest suppliers in 2001. Perhaps World PV shipments

400

350

300 . . .

250 -. . -~ . n p. goo .. .. . -...... I 150

100

50

0 1980 1985 1995 '000 lggoyear

Figure 1: Growth in world photovoltaic array shipments, 1980-2001

Cumulative production of tens to hundreds of tolerant and have the potential for being Gigawatts will be required for photovoltaics to manufacturable on thin, lightweight substrates. reach the technological maturity required for Such materials could be ideal for space use. finalizing a SPS design. Several earlier papers have reviewed the Having gained valuable photovoltaic progress of photovoltaic techn~logy.~,~NASA experience with solar energy, when the solar recently formed a technology review committee generation market share begins to saturate that consulted m-ission plaming ofEces, solar demand for peak power, utilities will begin to cell and array manufacturers, and research search for a solar-energy alternative that laboratories to assess solar cell and array provides continuous power. technologies required for future NASA Science missions, beyond 2006. The resulting draft PV TECHNOLOGIES evaluation of the state of the art of solar cells and arrays and their potential evolution over the As yet it is still too early to chose a next five years cross-referenced to the mission technology for SSPS. Among the photovoltaic needs was presented at the 29th IEEE technologies, many different approaches are Photovoltaic Specialist Conferezce in May, still in consideration. One of the leading flat- 2002.* Areas were highlighted where plate photovoltaic approaches is the use of investments in research and technology thin-film photovoltaic materials such as development are required to meet these mission amorphous silicon, cadmium telluride, and requirements. There is also an evaluation of the copper indium diselenide. Coincidentally, such supporting infrastructure required to sustain a thin-film materials are inherently radiation

Copyright 8 2002 by ihc American insiituic of Acronautics ana Astronautics, Inc. No copyright is asscrted in thc Unitcd Starc undcr Titic 17, U.S. Codc. Thc US.Govemmcnt has a royalty-frcc licensc to cxercisc al! rights undcr thc copyright claimed hercin for Govcrnmental purposcs. A11 othcr rights are rcscrvcd by thc copyright owncr. viable research and technology development organichnorganic solar cells may also be of program within NASA, while assessing those interest in the future. developments in the commercial sector, which Single-crystal solar cells have been used to would require very limited NASA involvement. produce electrical power on almost all space Solar cell and array manufacturers provided satellites since 1958. Their scalability, both data on current and near term commerciai reliability, and predictability, have made solar goals and also insight into future development photovoltaics the prime choice for spacecraft requirements specifically for NASA missions. designers. Cell efficiencies have grown from Prosam managers provided mission details and 8%, Air Mass Zero (AMO), on Vanguard I, the the impact on power system needs. There has world’s first solar powered satellite, in 1958 to also been a recent discussion of terrestrial solar 14.2% efficient crystalline Si cells on the cells, past, present, and f~ture.~ International Space Station (ISS). In the last Current state of the art decade there has been a transition to today’s Crystalline silicon dominates (-90%) -27% efficient multi-junction cells, today’s terrestrial market. Even though GaInP/GaAs/Ge, that are commercially -lOOpm of crystalline Si is required for the available fiom several vendors. It is likely that indirect 90% absorption of light versus -lpm production efficiencies will exceed 30% in a few of GaAs, a direct conversion semiconductor, for years. Higher efficiency cells and arrays are the same light absorption, the historical almost always a benefit. Higher efficiency development of silicon technology and high results in smaller arrays that can be used to quality material for the electronics industry has generate the same ainount of power as less insured the dominance of crystalline Si. efficient arrays, or makes higher power levels However, if we look to the future of crystalline feasible for a given configuration. These smaller Si it is recognized that 50% of the cost of a arrays require less volume when stowed for module is due to the cost of processed silicon launch and present less obscuration to sensors. wafers. In the past the PV industry has relied Traded against this benefit are other parameters on reject material fiom the semiconductor such as cost, specific power, and special needs industry that was available at low cost. This such as high temperature, low intensity low dependence on the semiconductor industry has temperature (LILT) and radiation resistance, led to cycles of low supply and high cost but the trade tends to favor more efficient cells. material. The feed stock supply of‘ silicon is Tables 1 and 2 Below list the current currently a significant issue. New technologies status of cell efficiencies for AM1.5 and are hoping to replace the Czochralski grown AM03. Notice that the conversion fiom 100 kg ingots with poly-crystalline, ribbon, or AM1.5 io AM0 is -.88 f .01 for the thin film thin-film crystalline production. Each of these cells and -.95 * .01 for the high efficiency techniques has yielded lower efficiencies than crystalline silicon and 111-V triple junction cell. crystalline Si. However, over time, the efficiencies have been steadily improving. Approximately 10% of today’s terrestrial market is in thin film materials such as amorphous silicon, crystalline and amorphous heterostructures, copper indium diselenide (CIS), cadmium telluride (CdTe), and dye- sensitized cells. Other types of Module Efficiency (%) Area (cm2) Global AM1.5 c-Si 22.7 778 multi-c-Si 15.3 1017 CIGSS 12.1 365 1 CdTe 10.7 4874 a-Si/a-Sua-SiGe 10.4 905 photochemical 4.7 141.4

- ~ ~ ~~ ~ ~ ~

Table 2. AM1.5 Efficiencies for Modules -.-

FUTURE SOLAR CELL DEVELOPMENT a 40% cell. Cost reduction would be achieved if III-V multi-function solar cells these cells could be grown on a silicon rather than a germanium substrate since the substrate Research in the III-V multi- is -65% of the cell cost. The advent of a new junction solar cells has been focused on competitor to the space cell manufacturing fdx-icati~geither !2ttice-mism2t,tched materials market in i998 and other factors combined to with optimum stacking bandgaps or new lattice reduce space cells costs by - 40% of their 1997 matched materials with optimum bandgaps. In cost in 2002. Several possible cell structures for the near term this will yield a 30% commercially future III-V devices are illustrated in Figure 2. available space cell and in the far term possibly

Lattice mismatched cell Quadjunction cell Triple junction on Si Efficiency -30% Efficiency -35% Efficiency - 30%

Figure 2. Proposed structures for 111-V tandem cell development

The problem areas with projected m-V cell large-area fi-esnel lens is used to concentrate development include the material growth sunlight onto a smaller solar array. The difficulty of the InGaAsN 1.05 eV bandgap proposed fresnel lens has a diameter of ten to material, minimizing defect growth in lattice fifteen meters, and is proposed to be supported mismatched material, and current limiting in the by an datable structure. The advantage of Ge subcell. Longer-term projects in the area of this structure is that the photovoltaic array is multi-junction III-V cells would include the much more expensive than the fresnel lens potential of growing these cells on a low cost material, and hence a low-cost plastic film can cermic subsnate and the possibility of substitute for a large area of array. efficiency enhancement by nano-structures. The difficulty of this approach is that solar With a recurring interest in terrestrial cells do not work well at elevated concentrators, both the space and terrestrial temperatures." Analysis shows that the use of communities have common goals for high active cooling for the focal plane array is, in efficiency III-V cells. general, too expensive to be cost competitive. To make this approach work, therefore, will High-temperature cells require development of a new solar cell In one of the proposed design for a solar technology that retains high performance at power satellite, the "Power Tower" concept, a elevated temperatures. Such photovoltaic technology is being considered for NASA missions such as the Solar Probe." Possible semiconductor technologies that could improve thin film cells of less than 15 to 20 % efficient high-temperature performance include Gap, would not be cost effective except for certain and nitride-based solar cells. While such applications which might involve a high solar cells have not yet been developed, radiation environment, or a stowage volume analysis of high temperature photovoltaic problem in the launch configuration, or perhaps iec’moiogy is the subject of a research effort at a unique spacecraft coniiguration. Current NASA Glenn. terrestrial thin film programs will benefit the space community as manufacturing techniques are improved bringing the small area cell Thin film cells efficiencies in Table 1 closer to the large area Thin film cells require substantially less modules in Table 2. The Space community material and have promised the advantage of requires that thin film cells must be produced large area, low cost manufacturing. However, on a lightweight substrate due to the mass space cell requirements dictate a more penalties imposed in launching. The best thin complicated trade space. Until recently the film cells to date have required processing focus in space cells has been on efficiency temperatures in excess of 600°C, which rather than cost. In a several billion-dollar prohibit the use of current polyimide spacecraft the cell cost is relatively small at substrates. Research has focused both on even a thousand dollars per watt, which is finding high temperature tolerant substrates and approximately the current array cost. This has on reducing the processing temperature of the primarily been true for spacecraft with power thin film cell itself while maintaining the higher needs from a few hundred watts to tens of efficiencies. A low cost flexible substrate kilowatts. However, deployment of a large would also benefit the terrestrial community by earth orbiting space power system will require replacing the expensive and fragile heavy glass major advances in the photovoltaic array structures. weight, stability in the space environment, Clearly, the ability to increase thin efficiency, and ultimately the cost of film cell efficiencies would impact both the terrestrial - production and deployment of such arrays. and space cell communities. Semiconductor The development of large space power quantum dots are currently a subject of great systems, and a host of other proposed space interest by both commupities. This is mainly missions, will require the development of viable due to their size-dependent electronic thin film arrays. The specific power required is structures, in particular the increased band gap almost 40 times what is presently available in and therefore tunable opto-electronic commercial arrays. Whde high efficiency ultra properties. A quantum dot is a granule of a lightweight arrays are not likely to become semiconductor material whose size is on the commercially available anytime soon, a6vances nanometer scale. These nano-crystallites in thin film photovoltaics may still impact other behave essentially as a 3-dimesional potential space technologies (Le., thin film integrated well for electrons (i.e., the quantum mechanical power ~upplies’~)and thus support a broad “particle in a box”). To date these nano- range of missions in the next decades. Mission particles have been primarily limited to sensors, examples include micro-air vehicles, dirigibles, lasers, LEDs, and other opto-electronic devices. ultra-long duration balloons (e.g. Olympus), However the unique properties of the size deep space solar electric propulsion (SEP) dependent increase in oscillator strength due to “Tug” Array, Mars SEP Array, and Mars the strong confinement exhibited in quantum surface power outpost. Spacecraft systems dots and the blue shift in the band gap energy of studies which consider the system level quantum dots are properties that can be implications of increased array area indicate that exploited for developing photovoltaic devices that offer advantages over conventional band gap energy allows for engineering an photovoltaic^.^^ The increased oscillator ensemble of quantum dots in a size range that strength of the quantum dots will produce an will capture most of the radiation from the increase in the number of photons absorbed and terrestrial and space solar energy spectrum, see consequently, the number of photogenerated Figure 3. carriers. On the other hand, the biue shift in the

Additional phc ons accessible via quantum C )tS rT I I I ., I: I I 0 500 1000 1500 2000 2500 Wavelength (mm)

Fi-we 3. Air Mass Zero Spectrum showins additional photons accessible with a quantum dot cell structure There have been several proposed ordered array of quantum dots within the methods to improve solar cell efficiency intrinsic region of a p-i-n solar cell (see Figure through the introduction of quantum dots. 4)- One of the main methods is to produce an Ohmic contacb

Fimre 4. Proposed quantum dot solar cell structure

The overlap of the discrete wavefunctions energy bands or "mini-bands." By adjusting the associated with the electronic states of the dot size and spacing, a device can be individual do-cs will produce narrow electronic manufactured such that these mini-bands will lie energetically between the valence and conduction bands of the host semiconductor, or, CONCLUSIONS in other words, within their bandgap. The Clearly both research and manufacturing quantum dots in an intermediate bandgap solar development is required to fulfill the promise of cell can be thought of as an array of light-weight, low-cost photovoltaic power, semiconductors which are individually size- whether on the ground or in space. Significant tuned for optimal absorption at a desired region strides have been made in many areas, but many of the solar energy emission spectrum. This is more areas will be addressed in the future. in contrast with a bulk material where photons Whether in twenty years, or more, there will be are absorbed at the band gap and energies above viable choices for generating power and we will the band gap where the photogeneration of need them. carriers is less efficient. In addition, bulk materials used in solar energy cells suffer from reflective losses at energies about the band gap, whereas for individual quantum dots reflective losses are minimized. REFEREIVCES Two additional desirable features of I P.E. Glaser, “Power from the Sun: Its quantum dot solar cell behaviour are the Future,” Science Vol. 162, 95’7-961 (1 968). expected superior radiation resistance of such U.S. Office of Technology Assessment, Solar devices and the independence of conversion Power Satellites, (1981). efficiency with temperature. To a first S. Bailey, R. Raffaelle, K. Emery. “Space ’ approximation the energy levels of quantum dot and and Terrestrial Photovoltaics: Synergy and structures are temperature independent. In fact Diversity” Progress in Photovoltaics, Vol.IO, thermal energy assists in populating those NO. 6,399-406 (2002). levels. This implies a greater thermal stability in contrast to a normal pn-junction solar cell. G. Landis, “Peak Power Markets for Satellite Unfortunately, it is difficult to estimate the Solar Power,” paper IAC-02-R.3.06, to be potential temperature range due to the presented, 53rd International Astronautical temperature dependence of other cell Congress/2002 World Space Congress, Houston components. TX, 10-19 October (2002). Both dye-sensitized cells and other ’ “Solar Electric Power”, the US Photovoltaic organic/inorganic hybrid cells may be of future Industry Roadmap, produced and printed by interest. Both efficiency and stability are issues the U.S. photovoltaics industry, facilitated by in both dye-sensitized and polymeric cells. the National Center for Photovoltaics, April However, research efforts in those areas have (200 1). been rewarding. A current assessment of space G: Landis, S. Bailey and M. Piszczor, qualification characteristics was presented at a “Recent Advances in Solar Cell Technology,” J. recent meeting.’’ While presenting significant Propulsion and Power, Vol. 12, No. 5, 835-841, challenges, the small amount of needed material Sept-Oct. 1996. Paper AIAA-95-002’7. and the ability to produce these cells on a S. Bailey, G. Landis and D. Flood, variety of substrates make them attractive for ”Photovoltaic Space Power: Status and Future,” future systems. paper AIAA-98-1053, 36th Aerospace Sciences Meeting & Exhibit, Reno hi,Jan 12-15 1998.

Copyright 8 2002 by thc American Institutc of Acronautics and Astronautics, Inc. No copyright is asscrtcd in the Unitcd Statc under Titlc 17, U.S. Code. The US. Government has a royalty-frcc liccnsc to cxcrcisc all rights undcr thc copyright claimcd hcrcin for Govcrnmcntal purposes. Ali othcr rights arc rcscrved by thc copyright owncr. S. Bailey, H. Curtis, M. Piszczor, R. Surampudi, T. Hamilton, D. Rapp, P. Stella, N. Mardesich, J. Mondt, R. Bunker, B. Nesmith, E. Gaddy, D. Marvin, and L. Kazmerski, “Photovoltaic Cell and Array Technology Development for Future Unique NASA Missions”, Proceedings of the 2ghIEEE PVSC, to bepublished, (2002). A. Goetzberger, J. Luther, G. Willeke, “Solar cells: past, present, future”, Solar Energy Materials & Solar Cells 74, pp. 1-11, (2002) lo G. Landis, ”Review of Solar Cell Temperature Coefficients for Space,” Proc. X7II Space Photovoltaic Research and Technology Conference, NASA CP-3278, June 1994, 385- 400. “D. Scheiman, G. Landis, and V. Weizer, “High-Bandgap Solar Cells for Near-Sun Missions,” presented at Space Technology & Applications International Forum (STAlF-99), Conf. on Global Virtual Presence, Albuquerque NM, 30 Jan.- 4 Feb. 1999; ATP Conference Proc. 458,6 16-620. l2 R. Raffaelle, S. Kurinec, D. Scheiman, P.

Jenkins, S. Bailey, “ Silicon Carbide Cells for Space Applications”, Proceedings of the the I 7‘h European PVSEC, Volume 1, 281- 283 (2001). l3 S. Bailey, G. Landis, R. Raffaelle, “Microelectronic Space Power Supplies”, A. Hepp, Proceedings of the Sixth European Space Power Conference”, pp. 153-157 (2002). l4 R. Raffaelle, S. Castro, A. Hepp and S.

Bailey, “ Quantuz Dot Solar Cells”, Progress in Photovoltaics, Vol 10 Number 6, pp. 433- 439, (2002). 15 S. Bailey, J. Harris, A. Hepp, E. Angiin, R. Raffaelle, H. Clark, S. Gardner, and S. Sun, “Thin-Film Organic-Based Solar Cells for Space Power”, , Proceedings of the 37”’ Intersociety Energy Conversion Engineering Conference, p p 233-236, (2002)