Project at a Glance
Title: Control techniques for organic gas emissions from fiberglass impregnation and fabrication processes
Principal Investigator / Author(s): Rogozen, Michael B.
Contractor: Science Applications, Inc.
Contract Number: A9-120-30
Research Program Area: Emissions Monitoring & Control
Topic Areas: Stationary Sources
The objectives of this study were (1) to locate and characterize as many sources of polyester resin/fiberglass process emissions in California as possible; (2) to establish an emission inventory based upon realistic emission factors for the pollutants of interest; and (3) to review the technology for controlling organic vapor emissions from this industry.
We conducted a survey by telephone and by written questionnaire, identifying thereby 305 unsaturated polyester resin users. Between August 1980 and May 1981, the industry used 44.4 to 45.5 million kilograms per year (97.9 to 100.4 million pounds per year) of resin and gel coat; to our knowledge this is the only California-specific estimate based upon an actual survey. The industry was in a recession during this period.
The California polyester resin/fiberglass industry consists of a large number of small firms and a few very large firms. The median firm size is 27,500 kg/yr (60,200 lb/yr) and the range is 99.8 kg/yr (220 lb/yr) to 8.8 million kg/yr (19.3 million lb/yr). The largest ten percent of the users consume 72 percent of the unsaturated polyester resin. The industry is centered in Los Angeles, Orange and San Diego Counties, which in combination account for 81 percent of the state's resin consumption.
Fabrication processes used in California include hand and spray layup, marble casting, filament winding, bag molding, pultrusion, continuous lamination and matched metal molding. Almost three-quarters of the firms in the state use hand layup, spray layup or a combination of the two. Continuous lamination and pultrusion use the most resin per plant. Styrene monomer is used as the cross-linking agent (to polymerize the unsaturated polyester resin) in all but three plants, which use methyl methacrylate. The most common catalysts are methyl ethyl ketone peroxide and benzoyl peroxide.
To develop improved emission factors, we first reviewed published and unpublished data from previous field and laboratory tests. We then conducted source tests at a large continuous lamination plant, a medium-sized spray layup facility, and a large synthetic marble casting plant. The last of these used normal and vapor-suppressed resins on successive days. Total emissions during each test run were determined by integrating the recorder trace of the output of a portable flame ionization detector. The detector was calibrated by simultaneously collecting organic vapor samples on charcoal traps and analyzing them by gas chromatography. The emission factors developed from the literature review and our tests were based upon styrene or methyl methacrylate monomer input rather than total amount of resin and/or gel coat.
Organic vapor emissions from the industry statewide were estimated to be 1.41 to 2.55 million kg/yr (1549 to 2805 tons/yr). Only four percent of the firms account for half of the total emissions. The South Coast Air Basin accounts for 1152 to 2042 tons/yr, or about 73 percent of the statewide total. Emissions from Los Angeles, Orange and San Diego Counties are 262 to 512, 856 to 1478 and 143 to 272 tons/yr, respectively.
Estimated emissions from polyester resin/fiberglass fabrication constitute 0.054 to 0.098 percent of the total organic gas (TOG) emissions, and 0.075 to 0.13 percent of stationary source organic gas emissions, as reported in the 1979 statewide emission inventory. It is difficult, if not impossible, to compare our estimates with those reported in various emission inventories by manufacturing category, since polyester resin/fiberglass operations are often ambiguously or erroneously categorized.
Incineration (at two plants) and use of resins with vapor suppressant additives are the only means of organic vapor emission control in this industry. We reviewed the literature on vapor suppressants and performed laboratory tests on emissions from resin coatings placed in a wind tunnel. Under our test conditions, vapor suppressants indeed reduced weight loss; furthermore, long-term weight loss increased with increasing gel time. To determine whether use of vapor suppressant affected material properties, we performed interlaminar shear strength and bending tests on laminates made of normal and vapor-suppressed resin and glass mat and cloth. Use of vapor suppressant did not degrade the properties measured.
Incineration, activated carbon adsorption, and condensation were found to be applicable in principle to controlling emissions from polyester resin/fiberglass fabrication, although each has some drawbacks. Absorption was not found to be practical. Costs of controlling emissions from hypothetical small and large hand- and spray-layup plants were estimated to be $10.3 to $15 per pound of styrene removed for incineration (assuming no heat recuperation), $4.3 to $4.6 per pound for carbon adsorption (assuming no credit for recovered styrene) and $7.3 to $15 per pound for condensation (assuming a credit of 60 cents/pound for recovered styrene). These costs are for new installations.
Any control strategy developed for this industry should take into account the heavy concentration of emissions among a relatively small number of firms. Strategies examined included setting maximum emission levels, requiring a minimum percentage removal of organic vapors, and requiring specific control technology.
For questions regarding research reports, contact: Heather Choi at (916) 322-3893
Stay involved, sign up with ARB's Research Email Listserver