United States Environmental Protection Agency Industrial Environmental Resea Laboratory Cincinnati OH 45268 Research and Development EPA-600/S7-82-066 Mar. 1983 Project Summary Photovoltaic Energy Systems: Environmental Concerns and Control Technology Needs Paul D. Moskowitz, Paige Perry, and Israel Wilenitz Technical and commercial readiness for alternate photovoltaic energy sys- tems, and waste streams from three different photovoltaic systems are exam- ined. At present, specific emission standards for this industry do not exist and measurements of wastes produced by existing manufacturers are not avail- able. Thus, emission estimates pre- sented are based upon design engineer- ing studies of hypothetical facilities. Because of the widespread use of many of the materials used in this industry, available control experience and tech- nologies used in other industries may ultimately be applied to photovoltaic plants. This analysis suggests that some uncontrolled waste streams could be declared toxic or hazardous under various provisions of the Clean Air, Clean Water, and Resource Conserva- tion and Recovery Acts. Although some processes could emit large quantities of pollutants, these can be controlled using available technology. Other processes may emit small quantities of more toxic pollutants which will probably not be directly controlled unless significant hearth hazards are identified. Environ- mental problems in installation and operation are probably associated with large central-station applications; no significant effects are expected from small decentralized applications. De- commissioning of broken or degraded photovoltaic systems will generate large quantities of solid waste which can be sirnply disposed of in a landfill or per- haps recycled. Disposal of spent photo- voltaic devices containing cadmium may present unique hazards. This Project Summary was devel- oped by EPA's Industrial Environmen- tal Research Laboratory, Cincinnati, OH, to announce key findings of the re- search project that is fully documented in a separate report of the same title (see Project Report ordering informa- tion at back). Introduction Identification and analysis of environ- mental concerns and ways to mitigate them for any energy industry before it becomes fully commercialized can limit potential investment costs while simul- taneously minimizing environmental and public health risks. This report updates previous studies by the U.S. Environ- mental Protection Agency (EPA) which examine potential health and environ- mental risks and control methodology related to photovoltaic energy systems. This analysis should provide background information about potential health and environmental effects to planners con- cerned with research and regulatory priorities, and federal, state, and county officials engaged in pollution control per- mitting programs. In this spirit, this study reviewed the technological readiness and examined environmental concerns related to com- mercialization of photovoltaic energy systems. The final report describes the technical and commercial readiness of photovoltaic energy systems; reviews re- fining, fabrication, installation, operation and maintenance, and decommissioning alternatives associated with the industry; identifies methods which are likely to be used to reduce pollutant release; de- ------- scribes environmental regulations that are or might be applied to an installed industry; and examines the hazard poten- tial of wastes generated during the refin- ing of specific materials and the fabrica- tion of different photovoltaic cell types. The information presented is based upon an extensive analysis of the literature, supplemented by discussions with dif- ferent individuals in the governmental and private sectors. Of the several phases in the life cycle of photovoltaic systems, the refining of materials and subsequent cell manufac- ture were most intensively examined be- cause they represent the major steps from which pollutants are released. Waste streams and pollution control methods for the following three different photo- voltaic systems were considered: (1.) silicon n/p cells produced by ingot growing; (2.) silicon metal/insulator/ semiconductor cells produced by ribbon growing; (3.) cadmium sulfide/copper sulfide backwall cells produced by spray deposition. These three systems cover a range of manufacturing options and materials likely to be used in near-term commercialization activities. Measurements of waste streams from existing manufacturers of photovoltaic devices are not publicly available, and therefore design engineering studies of expected typical facilities were prepared. From these, sources and types of pollu- tants were estimated for each of the production processes, as well as from the installation, operation and maintenance, and decommissioning of photovoltaic sys- tems. Estimates were based on a 10 MWp/year plant and for a national annual production rate of one gigawatt peak. Plant emission rates are presented on a kg/day basis and provide background information for personnel engaged in pollution control permitting programs. Estimates presented on a per GWp basis reflect national production rates in 1990 and should be more useful for adminis- trators in determining research and regu- latory priorities. Conclusions Large growth in the photovoltaics in- dustry is expected in the next two de- cades. In the year 2000, it is estimated that total installed capacity will range from (0.2 to 2.0) x 105 MWe. A variety of materials and cell concepts are now being examined for use in different mar- kets: small-remote (10 kWp) for non-grid- connected applications, and small (10 kWp) to large (100 MWp) systems for use in residences, commercial and industrial settings, and central-station electricity generation. Single-crystal silicon cells are current- ly produced commercially and serve as the standard of comparison for new materials and concepts being developed. Production of these cell types is costly; several alternatives are being investi- gated. Processes near commercial appli- cation include both ingot casting and ribbon growing for use in semicrystalline silicon solar cells. Inexpensively pro- duced thin films from such specialized materials as cadmium sulfide/copper sul- fide, polycrystalline gallium arsenide, and amorphous silicon have the potential to yield photovoltaic cells at comparatively low production costs. Presently, specific emission standards for this industry do not exist, and regu- lations on existing facilities range from none to specification of methods of hazardous waste disposal. Standards developed for related industries, pro- cesses, or for specific pollutants, may affect control technology requirements in this industry. Potential requirements under review by the EPA include National Emission Standards for Hazardous Air Pollutants (NESHAPs) for arsenic and possibly cadmium, and New Source Per- formance Standards (NSPS) for particu- lates from electric arc furnaces and vola- tile organic compounds from degreasing operations; Clean Water Act effluent limits applicable to the electronics industry; and. Resource Conservation and Recov- ery Act standards for control of a variety of specific toxic and hazardous wastes, in- cluding a numberfrom electroplating and degreasing operations. Estimates of uncontrolled emission rates from production of silicon ingot photovoltaic cells are presented in Table 1. Similar tables of pollutant emission rates are presented in the full report for cadmium sulfide and silicon ribbon cell production processes. Results of the completed analyses suggest that several processes could emit potentially large quantities of pollu- tants. Large quantities of fine particulates are discharged from electric arc furnaces in producing metallurgical-grade silicon (MG-Si) for a number of uses (e.g., in semiconductors). However, silicon de- mand by the photovoltaic industry repre- sents only a small fraction of total prc- ducton and therefore is not expected to result in a significant increase in these emissions. Further refinement of MG-Si by existing techniques consumes large quantities of hydrochloric acid, which must ultimately be disposed of. Etching of single-crystal silicon ingots generates large quantities of spent acids, including hydrofluoric, which may require controll- ed disposal. Plasma etching may eliminate the need to use wet etching processes. Silicon ribbon production, as hypothe- sized, is a much cleaner production pro- cess than ingot growing. Nevertheless, small quantities of solvents and silver- based inks may require careful control. Cadmium sulfide photovoltaic cell pro- duction is characterized by the use and release of cadmium and the production of large quantities of spent plating solu- tions containing various metals. These may require controlled disposal. Table 2 summarizes control alterna- tives that may be used in this emerging industry. Because of the widespread use of many of the materials in other indus- tries, available control experience and technologies may ultimately be applied to the photovoltaics industry. The final degree of control is likely to be more precisely defined by specific design en- gineering studies seeking compliance with specific standards. Other processes may emit small quan- tities of more exotic pollutants (e.g., boron trichloride, phosphorous oxychlo- ride, phosphine) which will probably not be directly controlled unless significant health hazards are identified. To the degree that processes are integrated within plants and automated, controls implemented to reduce major pollutant waste streams may also reduce discharges of these more exotic pollutants. The most significant environmental problems in operation of photovoltaic devices are expected to be associated with large central-station applications. Herbicides may be used to control plant growth near photovoltaic arrays. Also, it has been speculated that these facilities may produce micro- or meso-scale changes in the physical environment. Subsequent effects on species diversity, standing biomass, wind, temperature, and humidity have been hypothesized. Decommissioning of broken or de- graded photovoltaic systems will gener- ate large quantities of solid waste. Most of these wastes will be nonhazardous and can be disposed of in a landfill or recycled. Disposal of spent photovoltaic devices containing cadmium may present unique problems. Centralized collection by a utility owning a central-station array or maintaining a large number of decentral- ized systems will probably require dis- posal in controlled landfills. Decentral- ized disposal by individual homeowners, however, could result in the release of ------- Table 1. Activity Silicon Production Cell Manufacture Pollutant Emission Rates from Process Step Carbothermic Reduction of Silica Silicon Purification by Siemans Process Silicon Purification by Union Carbide Process Ingot Forming and Doping Wafer Cutting and Etching the Production of Silicon Ingot Pollutant SiO as SK)2 Ash CO Si/icon dust loss from size reduction Distillation bottoms (SiC/jJ Noncondensibles (hydrogen) Vapor Deposition by- product (63% SiClj) Silicon dust loss from size reduction Waste settler discharge (79% SiCIJ Filter waste stream (43% H2, balance chlorosilanes) Stripper overhead (73% chlorosilanes) Product melter loss (Silane and hydrogen. Argon not included) Dust loss from crushing BC/3 Crucible scrap (Si) Silicon chips and dust Slurry (oil, clay, and SiC) Photovoltaic Cells Medium Vapor Vapor Vapor Vapor Liquid Vapor Liquid Vapor Liquid Vapor Vapor Vapor Vapor Vapor Solid Solid Liquid (kg/Day*) /Plant 1,377 33 5,895 0.51 - 1.06 1,160-2,000 283 8,753 3.5 369 38.6 27.6 2.3 1.7 0.0009 25.7 197 30.6 (MT/YR) /GWp 48, 180 1,140 206,340 18-37 40,600 - 70,000 9,900 306,370 124 12,921 1,351 965 81 60 0.031 900 6,890 1,070 Etching liquor (12% SiF4 88% mixed acids) Etching vapors (hydrogen rate given, but also some fraction of liquor Liquid 1,647 57,640 Junction Formation Wafer Edge Grinding and Etching Electroless Plating and Soldering Application of Ant/reflective Coating Testing, Interconnecting and Encapsulating vaporized} POC/s dopant Si/icon dust (under water spray) Etching liquor (mixed acids) Etching vapors (SiF4 and H^ Spent Ni plating solution Acetone Photoresist (urethane varnish and titanium dioxide) Rinse water Exhaust (60% HCI, 33% SiH2C/2) Si3N4 deposit on reactor walls Vapor Vapor Liquid Liquid Vapor Liquid Liquid Liquid Liquid Vapor Solid NEGLIGIBLE 7.7 0.0009 1.7 (?) 42.3 85 61 21 L 28 0.15 0.013 270 0.03 61 m 1,480 2,990 2,140 [740 m3] 1,000 5.2 0.45 "Plant Capacity 10 MWp/yr., and 350 working days/yr. ------- Table 2. Review of Control Technology Alternatives Media/Category Pollutants Control Technology Comments 1. Atmospheric ./. 7 Gas - Combustible 1.2 Gas - Other 1.3 Dust 2. Liquid 2.1 Acid 2.2 Metals 2.3 Other 3. Solid 3.1 Hazardous 3.2 Nonhazardous CO, H, Chlorosilanes BCI3, POC/s, PH3, acids, solvents, (HCN)*, F Metals, Si, Cd, HF, HCI, Acetic, HN03, H2S04 Cd, Cu, As, Ni, Si Chlorosilanes, solvents, slurry, cutting oils Cd, F and other compounds Si compounds CO combustion, air dilution, Chlorosilanes combustion to form muriatic acid Air dilution, treat in lime scrub- ber and discharge liquor to lagoon Bag filters or cyclones Neutralization Flocculation (lime, alum, ferric salts, polyelectrolytes) Process revisions, distillation Three alternatives - resale, re- cycle, disposal in controlled landfill Resale or disposal in municipal landfill Technology commonly employed; limits must be identified. Design studies required to achieve limits. 97 to 99% removal. Impervious lagoon, overflow liquid may be hazardous; sludge will be hazardous. Flocculation technique must be identified. Effective standards must be met and sludge will require control. Chlorosilanes could be feedstock. Metals and F noted under RCRA guidelines. *HCN may be a by-product of CdS cell production. small quantities of cadmium to the at- mosphere (from combustion at municipal incinerators) or to terrestrial and aquatic systems (from disposal in municipal land- fills). Recommendations On the basis of this analysis, the fol- lowing topics may require further investi- gation: (i) Measurement of Existing Waste Streams - Analyses of risks from photovoltaic energy systems are based upon design engineering estimates of wastes emitted from expected typical facilities. Sampling and chemical analyses from existing facilities are required to support the findings of these engineering stud- ies or to identify inadequacies. (ii) Regulation of Existing Industry - At the present time, efforts to control wastes from the existing industry vary. Some state and county officials engaged in the regulation of the existing indus- try need information relating to problems likely to be encount- ered, and methods which could be employed to control these prob- lems. A large number of regulations exist, or are being developed, for related industries, processes, and pollutants. The potential applica- bility of these regulations to this industry needs to be examined in greater detail. (iii) Technology Transfer - Manufac- ture of photovoltaic devices may require use and disposal of large numbers of chemicals; some may be toxic or hazardous. Control experience obtained for these ma- terials in related industries should be made available to manufactur- ers of photovoltaic systems before facilities are actually constructed. Design engineering studies and field application of control tech- niques may ultimately be required to demonstrate their cost-effec- tive applicability to this industry. (iv) Evaluation of New Processes - Alternative materials and fabrica- tion processes are being rapidly developed. Environmental data, however, are not being assembled at the same rate. Thus, existing knowledge about processing tech- niques, materials, and potential control methods is limited. On- going efforts are required to elimin- ate this void. Identification of con- trol needs, prior to full-scale com- mercialization of these new tech- nologies, can reduce total design engineering cost while minimizing potential health and environment- al risks. (v) Micro- and Meso-scale Biological and Climatic Effects - A number of analysts have suggested that opera- tion of large central-station photo- voltaic plants in the Southwest may create micro- and meso-scale ef- fects on biological communities and climate. Research and analysis are required to evaluate the mag- nitude of these risks. •&U. S. GOVERNMENT PRINTING OFFICE:1983/65S»-095/590 ------- PaulD. Moskowitz, Paige Perry, and Israel Wilenitz are with Brookhaven National Laboratory, Upton, NY 11973. Benjamin L. Blaney is the EPA Project Officer (see below). The complete report, entitled "Photovoltaic Energy Systems: Environmental Concerns and Control Technology Needs," (Order No. PB 83-137 380; Cost: $10.00, subject to change) will be available only from: National Technical Information Service 5285 Port Royal Road Springfield, VA 22161 Telephone: 703-487-4650 The EPA Project Officer can be contacted at: Industrial Environmental Research Laboratory U.S. Environmental Protection Agency Cincinnati, OH 45268 United States Environmental Protection Agency Center for Environmental Research Information Cincinnati OH 45268 Postage and Fees Paid Environmental Protection Agency EPA 335 Official Business Penalty for Private Use $300 RETURN POSTAGE GUARANTEED Third-Class Bulk Rate IERL0167053 US EPA REGION V LIBRARY 230 S DEARBORN ST CHICAGO IL 60604 ------- |