![]() |
||||||
| June 2000 Volume 30, No. 6, 2129. |
Eliminating asbestos from fireproofing materials
Converting asbestos to nonfibrous mineral phases eliminates a regulated substance while leaving the fireproofing properties intact. Jacob Block We have developed a method for chemically digesting chrysotile in asbestos-containing fireproofing materials to levels below the regulatory threshold. The resulting material, no longer defined as asbestos-containing, can remain in place with its fireproofing properties intact. In the treatment process, chrysotile fibers are digested without generating excessive gaseous byproducts, and the foam-based delivery system essentially eliminates the release of airborne fibers. We used a new X-ray diffraction (XRD) method to quantify chrysotile levels with far greater precision than is possible with standard optical microscopic methods. Full-scale field testing confirmed the results of the laboratory phase of the project. Fire testing of the treated fireproofing material showed that the treated material functions as well as the original. In addition, rats exposed to an aerosol of the treated fireproofing material did not show any adverse inhalation effects. W. R. Grace & Co. (Columbia, MD) currently markets the chrysotile digestion agent described in this article under the tradename DMA. Solving some problems, creating others Asbestos refers to two kinds of minerals: serpentines and amphiboles. Serpentines, the most common of which is chrysotile, Mg6[(OH)4Si2O5]2, account for almost 90% of worldwide asbestos production. Amphiboles, especially amosite, (2, 3, 8) account for the remaining 10%. These minerals share the ability to form fibers with high aspect ratios (i.e., high ratios of length to diameter). In general, U.S. regulatory agencies define asbestos fibers as serpentines or amphiboles with lengths >5 µm and aspect ratios In 1973, the EPA banned the spray application of asbestos-containing fireproofing materials (9). In 1986, the U.S. Congress enacted the Asbestos Hazard Emergency Response Act (AHERA) (14), which required EPA to regulate ACMs in school buildings. Compliance with AHERA requirements was estimated to cost school districts $3 billion. The EPA was also required to estimate the amount of ACMs in other public and commercial buildings (5, 15). In 1988, EPA reported to Congress that 750,000 public and commercial buildings in the United States contained ACMs and that the abatement effort could cost $50 billion. Other estimates have put the potential cost for asbestos removal in the United States as high as $100150 billion (16, 17). In the United States, an asbestos abatement industry with about $3 billion in annual revenue has arisen since Occupational Safety and Health Administration (OSHA) and EPA regulations went into effect. Building owners have several options for asbestos abatement using current practices. The first is management in place: The ACM is monitored, and work practices are established to minimize asbestos exposure to occupants of the building. A second option is encapsulation of the ACM. The third and most costly option is complete asbestos abatement. Current removal practices, as specified by OSHA, require erecting a negative-pressure enclosure to completely seal the work area; wetting the ACM; scraping the ACM from the substrate; and disposing of the ACM as a regulated waste. The negative-pressure enclosure is then cleaned and disposed of as a regulated asbestos waste. Finally, new nonasbestos insulation is applied to replace the removed material. Worker protective equipment is designed to maintain airborne asbestos fiber concentrations below the OSHA permissible exposure limit (PEL) of 0.1 fiber/mL of air on an 8-h time-weighted average (TWA) basis. Chemical asbestos removal Chrysotile asbestos is considered relatively nonreactive chemically, but several approaches for the chemical elimination of chrysotile have been reported (1823). These include using sulfuric acid (18) or other strong acids to dissolve the magnesium oxide (8); dissolving both the magnesium oxide and silica portions of the chrysotile using a combination of a strong acid and fluoride ion source (HF) (19 22); and complete fluorination of the chrysotile (23). Fluorosulfonic acid is an effective agent for chrysotile decomposition (24). The strong acid digestion method (8) requires temperatures exceeding 50 °C and long processing times to digest the asbestos completely. These conditions are necessitated by chrysotiles crystal structure (which consists of alternating layers of magnesia and silica) and by the incomplete removal of magnesia from the structure. When magnesia is removed by acid, mass transport of reactants and reaction products is likely to be severely limited within the narrow channels formed between the acid-resistant silica layers. As a result, magnesia is not removed from the structure completely. An additional consequence of silicas poor reactivity with acids is that removing the magnesium-rich layers can leave behind a silica-rich structure that appears fibrous under the microscope. A combination of HF and strong acids can completely dissolve chrysotile by removing MgO and SiO2 (24, 25); however, using large quantities of HF is hazardous and therefore impractical. Mirick et al. have combined an organic acid and a fluoride salt, thereby avoiding the direct use of HF in the formulation (1922). Because all the fluoride salts Mirick et al. propose dissociate very rapidly under acidic conditions, our experience suggests that the salts produce significant amounts of HF in the liquid and gas phases. More recent patents to Block (2630), Block et al. (31), and Hartman (32) describe a process in which inorganic acids and fluoride generators are used. This process is described here. Chemistry. Our two main objectives were
These objectives presented two technical challenges. First, we had to develop effective chemistry that would digest chrysotile fibers without generating large quantities of HF or other substances of concern. Second, the treated fireproofing material should remain fully functional; that is, the compounds used to destroy the chrysotile should preserve the thermal and mechanical properties in the treated fireproofing. The two components of the system we developed are phosphoric acid and a fluoride salt that hydrolyzes very slowly under acidic conditions. The acid was selected to provide optimum removal of the chrysotiles magnesium component and to generate reaction products that improve the mechanical properties of the fireproofing material. In this way, eliminating asbestos from the matrix does not affect its fireproofing properties. It also is critical that the acid be selective for chrysotile over other components in the material. The fluoride compound was selected to generate fluoride slowly, at a rate sufficient to allow the reaction with chrysotile silica to proceed but not so rapidly that significant free HF is generated. The most effective fluorides were the soluble fluorosilicate or fluoroborate salts. Experiments were conducted on a gypsum-based spray-applied fireproofing material, a common commercial product that typically contains 6065% gypsum as a binder and 2327% vermiculite as a lightweight filler. The remaindernominally 12% by massis chrysotile asbestos. Phosphoric acid successfully removed magnesia with minimal attacking of the gypsum and vermiculite in the fireproofing material. The chemical reaction between the phosphoric acid and the chrysotile generates magnesium phosphate and other phosphate salts as reaction products. These reaction products reinforce the gypsum binder, thus distinguishing phosphoric acid from the other acids tested. As a result, the mechanical properties in the fireproofing material are greatly improved. Phosphoric acid attacks only the magnesium-rich layers in the chrysotile structure, leaving the silica-rich layers largely untouched. In choosing the fluoride source needed to overcome this limitation, we found that small quantities of salts effectively eliminated chrysotile while producing extremely low HF emissions. Figures 1 and 2 are images of the fireproofing material viewed with scanning electron microscopy (SEM) before and after treatment. In Figure 2, the treated reaction products are no longer fibrous because the joint action of the acid and the fluoride generator cause a complete reaction. Using XRD analysis, we confirmed the absence of chrysotile asbestos in the treated fireproofing material. Phosphoric acid removes the magnesia from chrysotile to produce magnesium hydrogen phosphate, silica, and water: 3MgO·2SiO2·2H2O + 3H3PO4 However, the reaction does not go to completion. The fluorosilicate or fluoroborate component slowly hydrolyzes to produce fluoride ions SiF62 + 2H2O or BF4 + 3H2O The fluoride ions then can react with the silica of the MgO-depleted chrysotile to form soluble hexafluorosilicate anions: SiO2 + 6F + 4H+ The hexafluorosilicate anions also will hydrolyze to produce more fluoride and amorphous silica: SiF62 + 2H2O The resulting fluoride can further attack the remaining silica from the chrysotile, producing more hexafluorosilicate anions. Therefore, the reaction is catalytic; it requires small quantities of the fluorosilicate or fluoroborate salt to react with large amounts of chrysotile. The resulting reaction products are magnesium phosphates and amorphous silica. The gypsum in the fireproofing material reacts to a small extent with the phosphoric acid to form calcium mono- and dihydrogen phosphates. XRD analysis reveals very small amounts of calcium phosphates, but no CaF2. Vermiculite, an aluminosilicate, does not react with the acidic components of the digestion mixture, except for partial removal of aluminum from its structure (as determined by NMR). XRD of the treated material showed substantial vermiculite and gypsum concentrations. Thus, the digestion mixtures attack chrysotile selectively, leaving the gypsum binder and vermiculite aggregate of the material essentially intact. Furthermore, the phosphate reaction products appear to be beneficial to the mechanical properties of the fireproofing material. The approximate product composition of the treated fireproofing material, obtained by XRD, is listed in Table 1. Typical results. Experiments were run with 1-in. gypsumvermiculitechrysotile cubes and digestion agent. The cube composition was 63.3% gypsum, 25% vermiculite, and 11.7% chrysotile. The digestion agent typically contained phosphoric acid (3040%), a soluble silicofluoride salt (<2%), and surfactants (<5%). The digestion agent was soaked into the cubes until the weight ratio reached 1.5:1 (liquid/solid). After 4 days, the cubes were analyzed by the XRD procedure (33). The residual chrysotile was 0.12 ± 0.04% for one cube and 0.14 ± 0.04% for the other four cubes. Kinetics. Kinetic experiments were run at 22 °C with gypsumvermiculitechrysotile powder mixtures, in the same proportions found in commercial fireproofing materials. The weight ratio of liquid digestion agent to solid fireproofing material was 1.5:1. The residual chrysotile was measured by XRD. Results indicated that the reaction was essentially complete after 2 days; 0.5% residual chrysotile remained after 1 day, and 0.1% remained on each of the next 4 days. In actual fireproofing specimens, liquid transport by capillary action will affect the kinetics. However, data obtained from monolithic samples that had absorbed the solution were similar to data from the powder samples. Foam development. Although the digestion agent can be applied to fireproofing material directly as a liquid, foam application offers several advantages, including better control of dosage rate (by monitoring foam thickness and density), reducing overspray, and virtually eliminating the release of asbestos fibers. Stable foams were slow to absorb, whereas rapidly absorbing foams were relatively unstable and produced much liquid drainage. To overcome this problem, a surfactant that provided good wetting and rapid absorption was combined with a surfactant that produced stable foam. The selected system included an ethoxylated tallow amine and a mixture of amidopropylbetaines. The foam is produced by mixing liquid and air at high pressure in a static mixer and spraying through a plastic nozzle. Several applications of the foam are required to obtain the stoichiometric amount of digestion agent required to reduce the asbestos level to <1%. The number of required applications depends on the foam density and thickness and the thickness and asbestos content of the fireproofing material. Effect of material depth. In monolithic samples, the concentration of digestion agent will be highest at the application surface (front) and lowest at the back. If, however, the required amount of liquid is added (as determined by the back surface being wet) and the material is given at least 4 days to equilibrate, then the asbestos levels in the core samples will be less than the 1% requirement 99% of the time (see section Scaling up the in situ digestion process). Despite the front-to-back concentration gradient, the total asbestos concentration and the asbestos concentration on the back surface generally will be <1%, as long as the back surface has been wetted. Quantitative determination of chrysotile. Because the legal definition of ACM sets an upper limit of 1%, we needed a method to determine quantitatively the amount of chrysotile in bulk materials at levels <1%. We used a quantitative XRD method for determining chrysotile in gypsumvermiculite matrices (33). In this method, ethylenediaminetetraacetic acid (EDTA) is used to remove gypsum, alumina is used as an internal standard, and the sample is carefully milled to eliminate preferred orientation effects. The EPA has endorsed this method as a referee technique in the determination of chrysotile asbestos in building materials that contain gypsum, vermiculite, phosphates, and other inorganic materials (personal communication). Scaling up the in situ digestion process There were conflicting requirements. Large quantities of acidic mixtures and asbestos-containing samples had to be prepared, which required copious ventilation. Measuring the levels of airborne chemicals during application would best be done in still air. Therefore, the laboratory was designed with extensive ventilation, including extra ventilation at individual equipment locations. The lab also included a specially designed and constructed chamber that could be sealed during experiments so that the air surrounding the test assemblies could be controlled and monitored. In addition to monitoring the air surrounding the experiment, extensive air monitoring was done throughout the facility to ensure a safe working environment. In the absence of OSHA or EPA rules specific to the new asbestos digestion technology, the ITF was set up and operated in accordance with the requirements for an asbestos abatement site. The entire laboratory was operated as a negative-pressure environment with high-efficiency particulate air (HEPA) filters on all exhausts. Additional engineering controls, including three walk-in hoods and exhaust vents on asbestos-mixing equipment, were installed to ensure that the airborne asbestos fiber concentrations would remain below the 0.1 fiber/mL PEL (8-h TWA) during all phases of operation. Equipment was designed to handle low-pH acids. Air monitoring equipment was used to monitor for acids, asbestos, and other emissions that theoretically could be generated by the application procedure and digestion process. In addition to the other safety features (e.g., walk-in hoods, exhaust ventilation, and the experimental chamber), workers were provided with personal protection equipment to protect against potential exposure to airborne asbestos fibers and acid solutions. All persons working within the ITF and on all subsequent full-scale field test operations wore hooded, powered, air-purifying respirators equipped with combination acid gasHEPA filter cartridges and acid-resistant coveralls, gloves, and safety boots. Test assemblies consisted of 30 × 30 in. steel panels (flat plain steel, flat galvanized steel, and corrugated galvanized steel) and 30-in. sections of I-beams coated with varying amounts of asbestos-containing fireproofing material. The composition of the ACM used for these experiments was identical to one that was sprayed commercially in the 1960s and 1970s; it contains nominally 12% chrysotile asbestos along with gypsum and vermiculite. The fireproofing material was applied to the steel assemblies using commercial spray equipment. After curing for several weeks, the test assemblies were placed in the experimental chamber, the chamber was sealed, and the air within it was cleaned by recirculation through the HEPA filter system for ~2 h to remove any airborne fibers or chemicals and to establish low background levels. The recirculation system was shut off before the chemical digestion agent was applied. The digestion agent was applied to the assemblies with concurrent air monitoring. After the chemical application, the ventilation system was run to remove any potential residual airborne contaminants before opening the sealed chamber. Twenty-six assemblies were tested at the ITF during the engineering scale-up phase of the project. Several core samples were taken through the fireproofing material from each assembly to assess asbestos levels after treatment. In every case where the fully developed process was used, the residual chrysotile content after treatment by the digestion agent was below the 1% level stipulated by regulatory agencies for a non-ACM. From these data, we concluded that the process could achieve digestion of asbestos to levels <1%. Area air samples and personal air samples (i.e., from sampling units attached to the workers protective clothing) were taken throughout the program to monitor the concentration of airborne fibers, acids, and any potential reaction byproducts within and around the ITF. Airborne fiber concentrations were monitored by using National Institute for Occupational Safety and Health (NIOSH) methods 7400 (phase contrast microscopy) and 7402 (transmission electron microscopy) (34) and a real-time fibrous aerosol monitor. The airborne asbestos fiber concentrations, even in the sealed experimental chamber, always were far below the OSHA PEL. The NIOSH 7903 method (35), which is applicable to both aerosol and vapor, also was used. HF, H2, and H2S gases as well as HF and H3PO4 droplets were monitored during each experiment to fully characterize the possibility of gaseous emissions from the process. Hydrogen never exceeded ambient levels, and the H2S concentration ranged from 0 to 0.7 ppm. HF concentrations during spray tests were measured continuously at three locations with a Tox Array 2000 HF continuous monitoring system (Mil-Ram Technology Inc., San Jose, CA). The HF concentration inside the chamber where fireproofing assemblies were sprayed with digestion agent varied between 0 and 0.7 ppm, well below the OSHA PEL for airborne HF (3 ppm). The airborne HF concentrations around digested panels stored outside the sealed spray hood always were below the limit of detection, indicating that off-gassing did not occur after digestion. Field-testing Air was monitored extensively to ascertain the levels of airborne asbestos fibers, HF, and H3PO4 during the digestion process. Airborne fiber monitoring indicated that fiber levels always were below the OSHA PEL (Table 3). When transmission electron microscopywhich is able to discriminate between asbestos and nonasbestos fiberswas used, asbestos was not detected in five of six samples; one fiber was detected in the sixth sample. This analysis demonstrated that using foam to apply the digestion agent effectively prevented the asbestos from becoming airborne. Airborne acid measurements showed levels generally consistent with those observed in the ITF. None of the measured H3PO4 concentrations reached one-tenth of the OSHA PEL 8-h TWA (Table 4). Although the airborne HF levels were higher at the field test site than in the ITF, they never exceeded the OSHA PEL 8-h TWA. The instantaneous HF concentrations, however, exceeded the PEL for short periods because HF was produced from the digestion agent that collected and dried on polyethylene sheets in the work area. Workers were protected from these short excursions, should they occur, by the same personal protective equipment used at the ITF (Figure 3). Trial by fire
Data from fire tests conducted on ACMs in the 1960s served as the controls for our tests on treated ACMs. Temperature measurements, taken at specific time intervals at various points along the interface between the steel and the fireproofing material over 3 or 4 h, showed almost identical performance to that achieved in the original ACM testing, indicating equivalent thermal performance. We observed minimal delamination of the treated ACM at higher temperatures, identical to that observed with untreated ACM, which indicated that high-temperature mechanical properties are retained. These results demonstrate that the thermal resistance of the treated fireproofing material and the adhesion of treated fireproofed building components are not affected by the chemical reactions that destroy the asbestos. These findings have led UL to grant fire ratings for specific assemblies comprising DMA-treated fireproofing material adhered to columns, beams, and corrugated decks. Animal testing Rats were exposed to these aerosols for 6 h a day for 5 consecutive days. They were killed at 1, 10, and 19 days after exposure; and alveolar lavage, pleural lavage, and histopathological analyses were performed. The particulates deposited in the lungs of the animals exposed to treated fireproofing materials were analyzed at 1 and 19 days after exposure. The analysis of the rats exposed to treated fireproofing materials produced no evidence of adverse chemical reactions in the lung. The animals exposed to crystalline silica showed a significant increase in quantity of neutrophils, protein, and lactose dehydrogenase, whereas the animals exposed to treated fireproofing materials or TiO2 yielded results statistically the same as the animals exposed to pure, filtered air. On the basis of these results, we concluded that the treated fireproofing material, like TiO2, does not present an inhalation hazard. Problem solved Acknowledgments This article is adapted from Block, J.; Petrakis, L.; Dolher, L. E.; Myers, D. F.; Hegedus, L.; Webster, R. P.; Kukacka, L. E. Environmental Science and Technology 2000, 34, 22932298.
References
Jacob Block is a principal scientist at W. R. Grace & Co. (7500 Grace Drive, Columbia, MD 21044; 410-531-4338; jacob.block@grace.com). He received his B.S. in chemistry from Brooklyn College (CUNY) and his Ph.D. in analytical chemistry from Case Institute of Technology (now Case Western Reserve University, Cleveland, OH). His research interests include asbestos chemistry, solñgel chemistry, and the generation of reagents in the aqueous phase. He has been awarded more than 50 U.S. patents. In 1999, he was a corecipient of an R&D 100 award and the R&D Editors Award for Environmental Responsibility for his role in the invention and as the principal developer of W. R. Graceís DMA digestion material for asbestos. In 1999, he was also a corecipient of the Council for Chemical Researchís Collaboration Success Story Award for the development of DMA in conjunction with researchers from Brookhaven National Laboratory. Leonard Dolhert is the venture manager of W. R. Graceís DMA business (410-531-4188; leonard.dolhert@grace.com). He is a corecipient of the R&D 100 award, the R&D Editors Award for Environmental Responsibility, and the Council for Chemical Researchís Collaboration Success Story Award for the development of DMA. He has been involved in developing and commercializing high-temperature superconductors and ceramic electronic packaging. He holds a B.S. degree and a Ph.D. from the Department of Materials Science and Engineering at the Massachusetts Institute of Technology. Leonidas Petrakis is a guest senior scientist and former chairman of the Department of Applied Science, Brookhaven National Laboratory, Upton, NY. He holds a doctorate in physical chemistry from the University of California, Berkeley. His research interests have focused on problems at the interface of energy and the environment. He has authored or edited six books, and he has more than 150 publications in peer-reviewed journals. He is currently a visiting scholar with the College of Chemistry at the University of California, Berkeley. Ronald Webster is a research engineer and manager of the Asbestos Research Facility at Brookhaven National Laboratory (Energy Sciences and Technology Department, Building 526, Upton, NY 11973; 631-344-2845; webster@bnl.gov). He holds a B.S. degree in architectural engineering and an M.S. degree in engineering from the University of Texas, Austin. His research interests include the in situ digestion of asbestos building materials and the development of advanced cementitous and coating systems for use in civil engineering applications. He was a member of the Brookhaven National Laboratory research team that received the R&D 100 award, the R&D Editors Award for Environmental Responsibility, and the Council for Chemical Researchís Collaboration Success Story Award with W. R. Grace & Co. in 1999 for the development of DMA. |
|||||
|
|
||||||