United States Environmental Protection Agency Air and Energy Engineering Research Laboratory Research Triangle Park NC 27711 Research and Development EPA/600/S7-86/011 May 1986 £EPA Project Summary Environmental Assessment of a Commercial Boiler Fired with a Coal/Waste Plastic Mixture R. DeRosier, H. I. Lips, and L. R. Waterland This report describes emission results obtained from field testing of a stoker-fired commercial boiler firing a coal/waste plastic mixture. Two tests were perform- ed: in one (test 1), the unit fired its typical coal fuel; and in the other (test 2), granulated waste polyethylene terephthalate (PET) beverage bottles were added to the coal to about 16 percent by weight in the mixed fuel. Emission measurements performed included con- tinuous monitoring of flue gas emissions; source assessment sampling system (SASS) sampling of the flue gas, with subsequent laboratory analysis of samples to obtain total flue gas organics in two boiling point ranges, compound category information within these ranges, specific quantitation of the semivolatile organic priority pollutants, and flue gas concenta- tions of 73 trace elements; EPA Method 5 sampling for paniculate; EPA Method 8 sampling for SO2 and SO3 emissions; volatile organic sampling train (VOST) testing for volatile organic priority pollu- tant emissions (test 2 only); HCI train sampling (test 2 only); gas grab sampling for N2O emissions measurements; and grab sampling of fuel ash bottom and cyclone collector hopper for inorganic composition determination. NOX, total unburned hydrocarbon (TUHC), and solid participate emissions were relatively unchanged for the two tests, averaging 286 and 214 ppm, <2 and 3ppm, and 54 and 69 mg/dscm, respec- tively, for tests 1 and 2. The emitted par- ticle size distribution was apparently unchanged as well. SOX emissions decreased with the coal/PET fuel in keep- ing with its lowered sulfur content. S02 and SO3 levels were 930 and 640 ppm. and 4.4 and 2.5 ppm (all corrected to 3 percent O2), respectively, for tests 1 and 2. Average CO emissions at 81 ppm were also decreased in test 2, from 184 ppm in test 1, apparently due to higher excess air levels in test 2. HCI emissions were mea- sured at 336 ppm (at 3 percent 02) with the coal/PET fuel; however, this level is almost 3 times that which was accounted for by the fuel chlorine level. Flue gas emissions of most trace elements were comparable for both tests, as were the trace element compositions of corresponding ash streams. However, estimated lead emissions (1,100 ^g/dscm) were significantly increased for test 2 when compared to levels for test 1 (30 ^g/dscm). This increase apparently reflects the increased lead content of the mixed coal/PET fuel, the increased lead coming inexplicably from the PET additive. The cyclone hopper ash for the coal/PET test had consistently lower leachable trace ele- ment and anion content than that for the coal fuel test. Total flue gas organic emissions were comparable for both tests, in the 1 mg/dscm range. Most (90 percent) of the flue gas organic was nonvolatile (boiling point greater than about 300 °C). Of the semivolatile organic priority pollutants, napthalene was detected in flue gas sam- ples for both tests, emissions increased from about 2 (test 1) to 16 ^g/dscm (test 2) with the coal/PET fuel. Phenanthrene, f luoranthene, and phenol were present in test 2 at levels ranging from 2 to 20 Mg/dscm; they were not detected in test 1. In addition, several phthalates were measured in the test 2 flue gas at levels of 0.4 to 6 /jg/dscm, while none were found in test 1. Phthalates present at such ------- levels are often ascribed to sample con- tamination; however, the observed in- creased emissions of these with the coal /PET in these tests may be real. Of the volatile organic priority pollutants, several chlorinated C, and C2 hydrocarbons, chlorobenzene, toluene. and ethylbenzene were measured in the test 2 flue gas at levels of 1 to 25 pig/dscm. This Project Summary was developed by EPA's Air and Energy Engineering Research Laboratory, Research Triangle Park, NC, to announce key findings of the research project that is fully documented in two separate volumes of the same title (see Project Report ordering information at back). Introduction Some states, Vermont included, are now requiring deposits on beverage containers as a means of controlling litter and en- couraging recycling and reuse. Plastic beverage containers cannot be reused as containers; however, one way to reclaim these is to use them as a fuel supplement in an energy recovery operation. In one Table 1, Summary of Boiler Operation and Fuel Analyses Test 1 (coat fuel) Boiler operation (over the SASS sampling period) Steam load, kg/s (1O3 Ib/hrl Steam pressure, kPa (psig) Feedwater inlet temperature, °C (°F) Stack temperature, °C (°F) Firebox temperature, °C (°F) Fuel feedrate, kg/hr (Ib/hr) Excess air, percent Boiler efficiency, percent Fuel ultimate analysis (percent by weight as fired) Carbon Hydrogen Sulfur Nitrogen Oxygen3 Ash Moisture Chloride Higher heating value, kJ/kg (Btu/lb) PET, weight percent aBy difference. 0.38 (3.0) 17 (2.5) 12 (54) 146 (295) 768 (1,415) 110 (240) 47 79.3 69.5 5.2 1.0 1.5 11.4 8.5 2.7 0.2 31,212 (13,450) 0 Test 2 (coal/PET fuel) 0.42 (3.3) 21 (3.0) 13 (55) 165 (329) 882 (1,620) 130 (285) 77 79.1 71.4 5.5 0.6 1.4 11.7 6.8 2.4 0.2 31,270 (13,479) 16.4 such project, the use of waste PET bot- tles is being evaluated as a fuel supple- ment in a commercial boiler at a quarry and stonecutting plant in Barre, Vermont. Past efforts have indicated that PET is a clean-burning, high-heating-value fuel with no apparent environmental problems as- sociated with its use as a solid fuel sup- plement. The tests described in this report were performed to confirm that this is the case by performing a comprehensive eval- uation of the emissions and ash discharge compositions from a boiler burning its typical coal fuel and a coal/PET mixture. The boiler tested is a coal-fired firetube unit with an underfeed stoker. The unit is rated at 0.6 kg/s steam at 100 kPa (4,640 Ib/hr at 15 psig). Nominal full load coal feed rate is 180 kg/hr (400 Ib/hr). The boiler was equipped with a cyclone collec- tor for particulate emissions control. Summary and Conclusions Boiler Operation The test program called for flue gas emission measurements and ash dis- charge stream sampling with the boiler operating at constant, near-rated capacity with both fuels. Two tests were perform- ed: test 1 with the unit firing its typical coal fuel, and test 2 with a coal/PET mix- ture containing about 16 percent by weight granulated PET beverage bottles. Table 1 summarizes the boiler operating conditions and the fuel ultimate analysis for both tests. As indicated, test 2 was run at about the same coal feedrate as test 1 (110 kg/hr); however, the test 2 fuel mix- ture contained about 20 kg/hr of added waste PET. Consequently heat input and boiler load were slightly higher for test 2. Boiler efficiency, calculated by the ASME heat loss method, was unchanged at about 79 percent. The boiler underwent some load excursions during test 2. These excursions could have some effect on emissions over and above the change in fuel composition. Emission Measurements and Results The sampling and analysis procedures used in this test program conformed to an extended EPA Level 1 protocol. All flue gas was sampled in the unit's stack, down- stream of the cyclone collector, except for the continuous flue gas analyzers which were operated both at the stack and at the boiler exit. Emission measurements included: • Continuous monitoring for 02, C02, NOX, CO, and TUHC. • Source assessment sampling system (SASS) for trace elements and semi- volatile and non-volatile organic emissions. • Volatile organic sampling train (VOST) for volatile organic emissions (test 2 only). • Combined EPA Method 5/8 for par- ticulate and SOX. • HCI train sampling. • Grab samples for N2O analysis. In addition, samples of the fuel, bottom ash, and a cyclone hopper ash were col- lected for analysis. The analysis protocol included: • Performing proximate, ultimate, and heating value analyses of fuel sam- ples, and ultimate analyses of bottom and cyclone hopper ash samples. • Analyzing the fuel, ash,ash aqueous leachates, and SASS train samples for 73 trace elements using spark source mass spectrometry (SSMS), supplemented by atomic absorption spectrometry (AAS). • Analyzing ash aqueous leachates for selected anions using ion chromatography. • Analyzing VOST traps for the volatile organic priority pollutants. • Analyzing ash sample organic ex- tracts for total nonvolatile organic content by gravimetry (GRAV). • Analyzing the SASS train organic ex- tract samples for total organic con- tent in two boiling point ranges: 10C to 300 °C by total chromatographic organics (TCO) analysis, and >300°C by GRAV. • Analyzing the SASS train extrad samples for the 58 semivolatile or ganic species, including many of the polynuclear aromatic hydrocarbor (PAH) compounds. ------- • Performing infrared (IR) spectrometry analysis of organic sample extracts. • Performing liquid chromatography (LC) separation of selected sample extracts with subsequent TCO, GRAV, and IR analyses of eluted LC fractions. Bioassay tests were also performed on SASS train and ash samples to estimate their potential toxicity and mutagenicity. Table 2 summarizes flue gas emissions measured in the test program. Emissions are presented as nanograms per Joule heat input and as milligrams per dry stan- dard cubic meter of flue gas. As a measure of the potential significance of the emis- sions levels for further monitoring or evaluation, an occupational exposure guideline for most pollutants is also noted in the table. The occupational exposure guideline noted is either the time- weighted-average Threshold Limit Value (TLV) established by the American Con- ference of Governmental Industrial Hy- gienists, or the 8-hr time-weighted-aver- aged exposure limit established by the Occupational Safety and Health Admin- istration (OSHA). These are noted only to aid in ranking the potential significance of the emission levels. In this respect, pol- lutants emitted at levels several orders of magnitude higher than their occupational exposure guidelines might warrant further evaluation, while species emitted at levels significantly lower than their occupational exposure guidelines might be considered of lower priority. Only elements emitted at levels exceeding 10 percent of their occu- pational exposure guidelines in these tests are noted in Table 2. Table 2 shows that flue gas concentra- tions of chromium, sodium, nickel, iron, lead, and cobalt were significantly higher for test 2 than for test 1. The test 2 flue gas levels for all these (except cobalt) were also greater than their respective occupa- tional exposure guidelines. The higher lead emissions for test 2, are thought to be ac- tual increases. The coal/PET fuel had significantly higher lead content (28 ppm) than did the coal itself (3 ppm). The source of the added lead is apparently the waste PET, although no lead source within the PET is known. Unfortunately, the trace ele- ment emission levels noted in Table 2 for test 2 are based on an estimated 1 to 3 /urn SASS train paniculate emission rate and the assumption that the Mo 3 urn par- ticulate had the same composition as the less-than-1-/dn paniculate (the 1 to 3 urn SASS particulate sample for test 2 was destroyed in transit from the field to the laboratory). Consequently, the significance Table 2. Summary of Flue Gas Emissions8 Concentration Test 1 (coal fuel) Test 2 Icoal/Pet fuel) Component (ng/J heat input) (mg/dscm) (ng/J heat input) (mg/dscm) Occupational exposure guideline (mg/m3)b Major constituents NOX (as NO2) SO2 S03 CO HCI Solid particulate Total semivolatile organics (TCO) Total nonvolatile organics (GRAV) Trace elements 176 800 4.6 69 _c 35 0.049 0.575 270 1,210 7.0 105 _c 54 O.075 0.875 165 690 3.4 38 206 41 0.071 0.641 280 1,160 5.7 64 347 69 0.12 1.08 6.0 5.0 1.0 55 7.0* 10* Chromium, Cr Sodium, Na Nickel, Ni Iron, Fe Lead, Pb Phosphorous, P Arsenic, As Aluminum, Al Lithium, Li Calcium, Ca Beryllium, Be Silicon, Si Potassium, K Cobalt, Co Cadmium, Cd Vanadium, V Copper, Cu Platinum, Pt Silver, Ag Selenium, Se Barium, Ba 0.0065 0.42 0.017 2.05 0.020 1.39 0.011 1.38 >0.017 1.24 0.00063 >1.52 0.59 0.0019 0.0016 0.011 0.020 _h 0.0017 0.0068 0.036 0.010 0.65 0.027 3.18 0.030 2.15 O.017 2.15 >O.026 1.93 0.00097 >2.36 O.91 0.003 0.0025 0.018 0.032 _h 0.0026 0.010 0.056 3.03 63 2.23 16.2 0.65 0.79 0.029 1.43 >0.081 0.28 0.00093 3.00 0.57 0.024 0.010 0.0073 0.021 0.00031 0.0010 0.024 0.012 5.17 107 3.81 27.6 1.10 1.35 0.049 2.45 >0.014 0.49 0.0016 5.12 0.97 0.041 0.018 0.012 0.035 0.00052 0.0018 0.041 0.021 0.05 2.0? 0.10 1.0 0.0503 0.10 0.010 2.0 0.025 2.0 0.002 10° 2.0* 0.10 0.0 5d 0.050 0.109 0.002 0.010 0.20 0.50 aFlue gas O2 and CO2 averaged 7.0 and 12.0 percent (dry) for test 1 and 5.2 and 12.9 percent for test 2. b Time-weighted average TLV unless noted. CHCI sampled for test 2 only. dCeiling limit. eFor nuisance particulate. fNo occupational exposure guideline applicable. g8-hour time-weighted-average OSHA exposure limit. hLess than the method detection limit. of the increased lead emission levels cited deserves further study. Several possibilities exist to explain the higher concentration of chromium, sodium, nickel, iron, lead, and cobalt. These are conjectures, however, so addi- tional sampling and analysis may be merited. Emission levels of the other elements noted in Table 2 were quite comparable between the two tests. In fact, the trace element content of all SASS train samples (except the impinger solution and lead in the fine particulate noted above), the bot- tom ash, and the cyclone hopper ash were quite comparable for both tests. The data in Table 2 show that boiler emissions of NOX, solid particulate, and total semi- and non-volatile organics were relatively unchanged with the addition of the PET to the boiler fuel. Emissions of S02, and S03 were decreased, in keeping with the lower sulfur content of the coal/PET fuel (the apparently small dif- ference noted in Table 2 is amplified con- siderably when corrected in each instance ------- to a common flue gas 02 level; e.g., 3 per- cent). On a mass loading basis, total sulfur emissions were relatively unchanged due to the constant coal firing rate. CO emis- sions were decreased with the coal/PET fuel as well due probably to the higher level of excess air used with the mixed fuel. The HCI emission level noted in Table 2 for test 2 corresponds to almost 3 times the amount which would be expected from the chlorine content of the fuel. The above discussion regarding the possible in- troduction of inorganic chloride into the SASS train suggests that perhaps ex- traneous chloride was introduced into the HCI train as well. This train measures total vapor phase (at stack temperature) in- organic chloride, not strictly HCI. Further analysis of the SASS train organic sorbent module extract via LC fractionation and IR spectroscopy of both the total extract and the eluted LC frac- tions suggests that the organic emissions for both tests were roughly 25 to 35 per- cent aliphatic hydrocarbons, 40 percent less polar oxygenated hydrocarbons, and 25 to 35 percent more polar oxygenates. Table 3 summarizes the organic priority pollutant emission results. Emission levels of the semivolatile compounds detected were significantly increased with the coal/PET fuel, despite the fact that total semivolatile organic emissions were relatively unchanged. The phthalates noted in the table could be sample con- taminants. Phthalates are common con- taminants in SASS train samples at levels comparable to those noted in Table 3. However, that levels detected in the test 2 samples were generally significantly higher than those in the test 1 samples, and that phthalates are expected to be found in PET, suggest that these com- pounds may actually have been present in the test 2 flue gas. Table 3 also notes that several chlor- inated C1 and C2 aliphatic hydrocarbons, chlorobenzene, toluene, and ethylbenzene were emitted during test 2 at levels of from <1 to about 24 ^g/dscm. The levels of the aromatic hydrocarbons noted (toluene and ethylbenzene) are in the range typically encountered in VOST tests of combustion sources. The source of the chlorinated compounds is not clear, al- though these compounds do arise when- ever chlorine, even inorganic chloride, is introduced into a combustion process. It is interesting that these compounds were detected in the test 2 flue gas, the same test with apparent chloride introduction into the SASS train, and unaccountably high (perhaps again due to extraneous flue Table 3. Summary of Organic Priority Pollutant Emissions Emission concentration fog/dscm) Compound Test 1 (coal fuel) Test 2 tcoal/PET fuell Volatile organic priority pollutants: Chloromethane Chloroform 1,2-Dichloroethane Trichloroethylene Tetrachloroethylene Toluene Chlorobenzene Ethylbenzene Semivolatile organic priority pollutants: 1.0 4.5 2.0 24 2.5 14 7.5 0.7 Naphthalene Phenanthrene Fluoranthene Phenol Bis(2-ethylhexyl)phthalate Diethyl phthalate Di-n-butyl phthalate 2.2 >0.4 >0.4 >0.4 4.4 >0.4 >0.4 16 3.0 2.2 20 4.2 6.0 O.4 aVolatiles sampled for test 2 only. gas chloride) chloride emissions measured with the HCI train. Table 4 summarizes the results of the nonvolatile (GRAV) organic analyses of the bottom ash and cyclone hopper ash ex- tracts. For test 1, only the cyclone hopper ash contained measurable levels of non- volatile organics. For test 2, the organic content of the bottom ash increased, although that of the cyclone hopper ash was about half that for test 1. The total ash loading was less in test 2 than test 1. In test 1, total ash was 13.5 percent of fuel fed, whereas ash was only 8.2 percent in test 2. For both tests, cyclone ash was about 3 percent of total ash. Infrared spectra of bottom ash extracts suggest that only aliphatic hydrocarbons were present. Analyses of cyclone ash ex- tracts via LC fractionation and IR spec- troscopy suggest that these contain: 30 to 40 percent aliphatic hydrocarbons, 15 to 20 percent less polar oxygenated hydro- Table 4. Ash Stream Total Organic Content carbons (e.g., aldehydes and ethers), and 40 to 55 percent more polar oxygenates (e.g., carboxylic acids, alcohols, and ketones). Table 5 summarizes the concentrations of the priority pollutant trace metals in aqueous leachates of the ash stream sam- ples prepared according to Level 1 guide- lines. As a measure of the potential signif- icance of these concentrations for further analyses, the water quality criterion for each element is noted in the table. Again, these are noted only to aid in ranking the potential significance of the teachable ele- ment levels noted. Only elements having at least one leachate concentration greater than its water quality criterion are noted. On a mass flow basis, it is noteworthy that the ash mass flow was 40 percent lower for test 2. The data in Table 5 show that the cy- clone ash leachates for both tests con- tained considerably higher levels of the Total nonvolatile organics by gravimetry (mg/kg ash) Sample Test 1 (coal fuel) Test 2 (coal/PET fuelt Bottom ash Cyclone hopper ash >80 1,100 220 5OO ------- Table 5. Inorganic Priority Pollutants in Ash Leachates at Concentrations Exceeding Their Water Quality Criteria Bottom ash leachate concentration (\ng/l) Cyclone hopper ash leachate concentration lug/1) Element Test 1 (coal) Test 2 (coal/PET) Test 1 (coal) Test 2 (coal/PET) Water quality criterion (mg/l) Arsenic 0.030 0.0 JO 4.1 0.20 2.2 x 10 -5a Beryllium Nickel Selenium Chromium Cadmium Thallium Mercury Lead Barium Antimony >O.OOJ 0.01 0.10 0.02 >0.001 >0.001 0.002 >0.004 0.06 >0.004 >0.002 0.02 0.10 0.06 0.004 >0.002 >0.001 0.009 0.30 0.20 0.32 4.0 0.80 3.7 0.17 0.21 >0.001 0.27 2.1 0.15 0.070 0.80 0.2O 1.0 0.10 0.09 >0.001 0.10 0.05 0.05 3.7 x 10' 0.0134 0.010 0.050 0.010 0.013 1.44 x 10 0.050 1.0 0.146 5* -4 a Water quality criterion based on cancer risk; level noted corresponds to increased lifetime risk of 10~5. inorganic priority pollutants than did the bottom ash leachates. The test 2 bottom ash leachate contained slightly higher levels of most of the elements noted. However, the cyclone ash leachate for test 2 contained significantly lower levels of all the elements noted (with the possible exception of mercury). This suggests that the addition of waste PET to the boiler's fuel may have had the beneficial effect of decreasing leachable toxic trace element discharges in the cyclone hopper ash stream. This effect is augmented by the observed 40 percent reduction in ash out- put in test 2 relative to fuel flow. Table 6 summarizes the leachable anion content of the ash stream aqueous leach- ates. As for the inorganic priority pol- lutants, the anion levels in cyclone ash leachates are significantly higher than those of bottom ash leachates. However, no consistent trend in composition be- tween corresponding ash leachates for the two tests is apparent. Interestingly, the chloride and sulfate concentrations in the test 2 ash stream leachates are slightly lower than those from test 1. One might expect that, if a corrosive component were present in the test 2 combustion gas, as implied by the results of the SASS, HCI, and VOST analyses discussed above, evi- dence of this component would appear in the ash leachates, especially the cyclone ash leachate. An alternative explanation would be that the component was present in the vapor phase at the cyclone temper- ature, although this seems unlikely. Bioassay tests were performed on the SASS train organic sorbent module extract and the ash stream samples. The health effects bioassays included the Ames mutagenicity assay and the CHO cytotox- icity assay. The results of these assays are summarized in Table 7. The results sug- Table 6. Ash Stream Leachable Anion Analysis Results Concentration (mg/l) Test 1 (coal fuel) Test 2 (coal/PET fuel) An ions Chloride Fluoride Nitrate Nitrite Phosphate Sulfate Sulfite Bottom ash leachate 2.6 0.25 >1.0 >1.0 >1.0 300 20 Cyclone hopper ash leachate 300 0.46 >10 >10 >10 19,000 5,000 Bottom ash leachate 1.3 0.10 >1.0 >1.0 >1.0 220 20 Cyclone hopper ash leachate 180 1.8 >10 >10 >10 12,000 5,000 gest that the flue gas was of moderate to high toxicity and mutagenicity for both tests. The cyclone hopper ash had nonde- tectable toxicity for both tests. However, this sample's mutagenicity apparently decreased from borderline moderate/high in test 1 to borderline low/moderate for test 2. This observation is consistent with the decreased leachable arsenic and beryl- lium contents noted in Table 5, and con- firm that the addition of waste PET to the unit's fuel may have and the beneficial effect of decreasing the potential environ- mental hazard posed by this discharge. The positive Ames responses for the sorbent module extracts noted above are typical for such extracts from SASS tests of combustion sources. Current studies are investigating if such bioassay re- sponses are due to artifact compounds formed when combustion product gas containing NOX is passed over XAD-2 resin. Quality assurance (QA) performed for these tests to establish the percision and accuracy of the laboratory analyses gen- erally gave results within the project QA objectives for these measures. However, the precision of the SSMS analyses, as determined by analysis of blind duplicate samples, was of only borderline accepta- bility and the accuracy of this analysis, as determined by analyzing a blind audit sam- ple, was not generally within a factor of 3 as specified by Level 1 protocol. How- ever, this failure to achieve reasonably good precision and accuracy has little effect on test conclusions. The conclusion that corresponding samples from both tests had comparable trace element com- positions (except for lead) should be unaf- fected. The high lead levels in test 2 samples were quantified using AAS ana- lyses, which met appropriate QA objec- tives for precision and accuracy. The con- clusions that cyclone ash leachates had significantly higher trace element concen- trations than bottom ash leachates, and that the test 1 cyclone ash leachate had higher trace element concentrations than that from test 2, were based on observed order of magnitude differences. Further- more, these differences were often sub- stantiated by AAS analyses which, again, met appropriate QA objectives. ------- Table 7. Bioassay Results Bioassay response3 Test Sample 1 (coal) XAD-2 extract Bottom ash Cyclone hopper ash 2 fcoal/PET) XAD-2 extract Bottom ash Cyclone hopper ash Ames mutagenicity H ND M/H M ND L/M CHO clonal toxicity M ND ND H/M ND ND aND — Nondetectable mutagenicity/toxicity. L — Low mutagenicity/toxicity. M — Moderate mutagenicity/toxicity. H — High mutagenicity/toxicity. R. DeRosier, H. Lips, andL. Water/and are withAcurex Corp., Mountain View, CA 94039. Joseph A. McSorley is the EPA Project Officer (see below). The complete report consists of two volumes, entitled "Environmental Assess- ment of a Commercial Boiler Fired with a Coal/Waste Plastic Mixture:" "Volume I. Technical Results," (Order No. PB 86-183 811/AS; Cost: $16.95) "Volume II. Data Supplement," (Order No. PB 86-183 829/AS; Cost: $22.95) The above documents will be available only from: (cost subject to change) National Technical Information Service 5285 Port Royal Road Springfield, VA 22161 Telephone: 703-487-4650 The EPA Project Officer can be contacted at: Air and Energy Engineering Research Laboratory U.S. Environmental Protection Agency Research Triangle Park, NC 27711 ' -&U. S. GOVERNMENT PRINTING OFFICE:1986/646-116/20847 ------- United States Center for Environmental Research Environmental Protection Information Agency Cincinnati OH 45268 Official Business Penalty for Private Use $300 EPA/600/S7-86/011 0000329 PS U S ENVIR PROTECTION AGENCY REGION 5 LIBRARY 230 S DEARBORN STREET CHICAGO IL 60604 ------- |