vvEPA United States Environmental Protection Agency Industrial Environmental Research Laboratory Cincinnati OH 45226 Research and Development EPA-600/PS7-80-148 Sept. 1980 Project Summary Environmental Assessment of a Waste-to-Energy Process, Burlington Electric's Wood and Oil Co-Fired Boiler Mark A. Golembiewski, K.P. Ananth, T. Sutikno and Harry M. Freeman In July 1978, Midwest Research In- stitute conducted a series of emission tests at the Burlington Electric Department's power plant in Bur- lington, Vermont. The study was designed to provide multi-media emission data for the purpose of iden- tifying potentially adverse envi- ronmental impacts, and to identify pollution-control technology needs. The No. 1 boiler at Burlington Elec- tric, which was tested for this study, is fueled by a combination of wood chips and No. 2 fuel oil. Approxi- mately 82% of the heat input (9.3 tons/hr) was provided by the wood fuel and the remaining 18% by the fuel oil (175 gal/hr). Electrical power generated from this boiler system was about 8 MW. The air pollution- control system consists of two mechanical collectors in series. Four effluent streams were sampled and analyzed for this assessment pro- gram: bottom ash, primary collector ash, secondary collector ash, and stack emissions. Common to all streams were characterizations for elemental composition and poten- tially hazardous compounds such as polychlorinated biphenyls and polycyclic aromatic hydrocarbons. In addition, the boiler exhaust gases were analyzed for paniculate, NOx, S02, CO, and total hydrocarbon con- centrations. The Source Assessment Sampling System was also used, following guidelines established by EPA's Level 1 environmental assess- ment protocol. This publication is a summary of the complete project report, which can be purchased from the National Technical Information Service. Introduction Under the sponsorship of the U.S. En- vironmental Protection Agency's Fuels Technology Branch in Cincinnati, Midwest Research Institute (MRI) is presently conducting multi-media envi- ronmental assessments of various waste- to-energy conversion systems. This paper will discuss the results of one such effort at the Burlington Electric Department's power plant in Burlington, Vermont. The Burlington plant was selected for study because of the renewed interest in wood and wood waste as a primary boiler fuel in certain regions of the country. The Bur- lington plant is the only facility in the U.S. presently firing oil with wood waste to generate electric power. The sampling and analysis program at Burlington was designed to provide multi- media emission data for the purpose of identifying potentially adverse environ- mental impacts and assessing pollution- ------- control needs. In the following sections of this paper, we will briefly describe the test facility, the sampling and analysis pro- cedures that were used, and the results of the test program. An assessment of the results using EPA's Source Analysis Model (SAM-1A) is also presented. Description of Boiler Test Facility The No. 1 Unit at the Burlington Elec- tric plant was originally a coal-fired boiler that was modified to fire wood chips with supplementary oil injection. Steam pro- duced to power a 10 MW turbine generator is rated at 100,000 Ib/hr. Wood chips are conveyed from the storage bin to four gravity-fed chutes. From the base of the chutes, the chips are injected into the boiler using compressed air. The wood chips fall onto a horizontal, traveling grate which is supplied with underfire air. Because of the high moisture content of the chips, the boiler cannot provide the needed steam output from wood alone; therefore, supplemen- tary fuel oil is used. No. 2 fuel oil, along with overfire air, is introduced above the grate bed from both sides of the firebox, thus insuring adequate steam production. Residual ash is discharged at the end of the grate into a hopper, from which it is removed pneumatically to an outdoor storage silo. The flue gases leaving the boiler are ducted to an emission control system consisting of two, high efficiency mechanical collectors in series. For a flue gas flow rate of 60,000 acfm at 330° F, the collectors were designed for an overall pressure drop of 6.5 in. H20 and a collec- tion efficiency of 97.75%. Sampling and Analysis Methodology The sampling and analysis test matrix that was used is shown in Figure 1. The input fuels, bottom ash, primary and secondary collector ash, and air emissions at collector inlet and outlet were sampled. Three complete sets of samples from each of these streams were taken over a two- day period. The sampling and analysis procedures used were either EPA methods or those approved by the EPA project officer. Results of the sampling and analysis efforts are discussed below. Results Input Fuels—The wood/oil feed ratio during the MRI tests was approximately 80% wood and 20% oil, on a heat input basis. The average composition of the wood was 4.3% ash, 70% volatile matter, and 25.7% fixed carbon (dry basis). The average sulfur content was 0.35% and the average heat of combustion was 5,870 Btu/lb (as received) and 9,480 Btu/lb (dry basis). The sulfur content of the oil was 0.35%, v\/hile its heat of combustion average was 19,500 Btu/lb (138,100 Btu/gal). Bottom Ash—Analysis of elemental composition in bottom ash indicated that most elements were more concentrated in the bottom ash relative to the fuel inputs. Those elements exhibiting the largest in- creases in concentration were Ba, Zr, Sr, and Li. No PCB materials were detected in the bottom ash samples above the 0.05 ^g/g detection limit. One PAH com- pound, phenanthrene, was identified at a concentration of 0.89 ng/g. Analysis for PCB materials at 0.05 pg/g detection limit was negative in the ash samples. No PAH compounds were iden- tified in the primary ash sample extracts. Wood Feed Collector Inlet Collector Outlet Samp/ing- Take Three 5-Liter Samples Each Day. Mix and Extract a 1 -Liter Composite Analysis: Determine H2O Content HHV. Proximate/Ultimate Elemental Analysis by SSMS Oil Feed Sampling: Analysis. Bottom Ash Samp/ing: Analysis: Take One or Two 0.5-L/ter Samples HHV, Proximate/Ultimate Elemental Analysis by SSMS Oil *- Wood *• Take Three 1 kg Samples Each Day, Mix and Extract 1 kg Composite Elemental Analysis by SSMS PCB/PAH Sampling and Analysis' a. Method 5-Particu/ate - 1 Per Day Elemental Analysis by SSMS° b. Orsat (02 & CO2) c. Particle Sizing - 1 Per Day Wickes Boiler 85,000 Ibs/hr r hr r V Sampling and Analysis: a. Method 5-Particu/ate - 1 Run Per Day Elemental Analysis by SSMS* b Orsat (O2 & C02) c. Particle Sizing - 1 Run Per Day d. Opacity (Method-9) - Two 1 Hr Tests/Day e. Continuous Analyzers (Oz, NO*, SOz HC, CO) f. PCB/PAH with Florist I Train - 2 Runs g. SASS - 1 Run Analyze Per Level 1 Requirements Primary & Secondary Collector Ash Sampling. Take a 1 kg Grab Sample Each Hour - Mix & Extract 1 kg Composite Analysis: SSMS PCB/PAH Bottom Ash Primary Secondary Collector Collector Ash Ash * Atomic Absorption Analysis May Be Conducted Based on Results of SSMS Analysis Figure 1. Test matrix for Burlington Electric's wood and oil-fired power plant 2 ------- However, several compounds were con- firmed in the secondary ash samples, in- cluding acenapthylene, phenanthrene, fluoranthene, and pyrene. One sample contained 10 ng/g of phenanthrene, which was the highest PAH concentration observed. Uncontrolled Air Emissions—The average uncontrolled paniculate concen- tration was 2.96 g/dscm (1.30 gr/dscf) as measured by EPA Method-5. On the basis of heat input, uncontrolled paniculate emissions averaged 1.47 g/MJ (3.43 lb/106 Btu). Oxygen and carbon dioxide contents averaged 12.3% and 8.2%, respectively. Filter samples from the Method-5 paniculate tests were analyzed for elemental composition. The elements emitted at concentrations greater that 10 ng/dscm were Pb, Ba, Sr, As, Ga, Zn, Cu, Fe, Mn, Ti, and P. An optical/diffusional particle counting system was used to measure the particle size distribution of the uncontrolled emis- sions. Particles in the range of 0.005 to 0.10 urn were counted by a diffusion bat- tery/condensation nuclei counter ar- rangement, while those in the 0.3 to 2.6 /jm range were counted by an optical counter. Because the dilution system con- sistently became plugged with larger par- ticles during operation, no particle counts could be obtained in the size region above 2.6 urn. Therefore, the mean particle size could not be determined. Within the size range of particles counted (0.005 to 2.6), the majority of the particles appeared to be between 0.3 to 0.5 pirn in diameter. Controlled Air Emissions-Using con- tinuous gas analyzers, concentrations of 02, NOX, SO2, CO, and total hydro- carbons (THC) were measured in the stack gas. NOX, and SOz concentrations P P and V were in the range of 1 to 75 fxg/dscm. The remaining elements had concentrations which were less than 1 ng/dscm. Particle size data was obtained by the same method used at the collector inlet (optical/diffusional particle counter). As with the inlet measurements, data for par- ticles > 2.6 ^m in diameter could not be obtained. The number of small particles « 2.6 jim) appeared to increase in the controlled gas stream relative to the un- controlled emissions. The reasons for this increase are not clear. Plume opacity data, obtained using EPA Method-9, averaged about 20% on both test days. Samples for analysis of PCB and PAH materials were collected in a special sampling train utilizing impingers and a Florisil adsorbent trap. PCB analysis did not produce any responses greater that the 1 ng/sample detection limit of the GC/MS analytical technique. Similarly, no PAH compounds were identified at levels which permitted structural confir- mation. Organic analysis of the Source Assess- ment Sampling System (SASS) com- ponents, in accordance with Level 1 guidelines, revealed low levels of organic constituents. Characterization of the organic emissions was difficult, although they appeared to be composed mainly of carbonyl-containing groups. Source Assessment Model (SAM-1A) The EPA's SAM-1A methodology was applied to the four effluent streams as a means of interpreting the emission results. The SAM-1A analysis results in- dicated that the secondary collector ash contained the highest degree of hazard, although all three ash streams were similar in the magnitude of their hazard values. Stack emissions showed a relatively low degree of hazard. The primary collector ash stream had the highest Toxic Unit Discharge Rate (TUDR) which would seem to indicate that this effluent should receive the first priority for control measures. However, because of certain limitations which are inherent in the SAM-1A methodology, this observation should not be regarded as being definitive. ------- However, several compounds were con- firmed in the secondary ash samples, in- cluding acenapthylene, phenanthrene, fluoranthene, and pyrene. One sample contained 10 ng/g °f phenanthrene, which was the highest PAH concentration observed. Uncontrolled Air Emissions—The average uncontrolled paniculate concen- tration was 2.96 g/dscm (1.30 gr/dscf) as measured by EPA Method-5. On the basis of heat input, uncontrolled paniculate emissions averaged 1.47 g/MJ (3.43 lb/106 Btu). Oxygen and carbon dioxide contents averaged 12.3% and 8.2%, respectively. Filter samples from the Method-5 particulate tests were analyzed for elemental composition. The elements emitted at concentrations greater that 10 ng/dscm were Pb, Ba, Sr, As, Ga, Zn, Cu, Fe, Mn, Ti, and P. An optical/diffusional particle counting system was used to measure the particle size distribution of the uncontrolled emis- sions. Particles in the range of 0.005 to 0.10 fim were counted by a diffusion bat- tery/condensation nuclei counter ar- rangement, while those in the 0.3 to 2.6 ^m range were counted by an optical counter. Because the dilution system con- sistently became plugged with larger par- ticles during operation, no particle counts could be obtained in the size region above 2.6 urn. Therefore, the mean particle size could not be determined. Within the size range of particles counted (0.005 to 2.6), the majority of the particles appeared to be between 0.3 to 0.5 urn in diameter. Controlled Air Em/ss/ons-Using con- tinuous gas analyzers, concentrations of 02, NOx, S02, CO, and total hydro- carbons (THC) were measured in the stack gas. NOX, and SOa concentrations averaged 66 and 138 ppm, respectively. CO readings averaged 213 ppm and THC concentrations were only 9 ppm. Three Method-5 particulate runs were made simultaneously with the sampling runs at the collector inlet. The average particulate concentration was 0.18 g/dscm (0.08 gr/dscf). The particulate emission rate averaged 0.09 g/MJ (0.17 lb/106 Btu), on the basis of heat input. The average efficiency of the two-stage mechanical collection system, as deter- mined from the simultaneous inlet/outlet tests, was 94.2% for total particulate. Elemental analysis of the Method-5 par- ticulate filters indicated moderately high elemental concentrations. Pb, Ba, Sr, Zn, and Ti were present at the highest con- centrations, approaching 100 /jg/dscm, while Hg, Sb, Zr, Br, Se, As, Ga, Cu, Ni, and V were in the range of 1 to 75 ng/dscm. The remaining elements had concentrations which were less than 1 Particle size data was obtained by the same method used at the collector inlet (optical/diffusional particle counter). As with the inlet measurements, data for par- ticles > 2.6 fjm in diameter could not be obtained. The number of small particles « 2.6 ^m) appeared to increase in the controlled gas stream relative to the un- controlled emissions. The reasons for this increase are not clear. Plume opacity data, obtained using EPA Method-9, averaged about 20% on both test days. Samples for analysis of PCB and PAH materials were collected in a special sampling train utilizing impingers and a Florisil adsorbent trap. PCB analysis did not produce any responses greater that the 1 ^g/sample detection limit of the GC/MS analytical technique. Similarly, no PAH compounds were identified at levels which permitted structural confir- mation. Organic analysis of the Source Assess- ment Sampling System (SASS) com- ponents, in accordance with Level 1 guidelines, revealed low levels of organic constituents. Characterization of the organic emissions was difficult, although they appeared to be composed mainly of carbonyl-containing groups. Source Assessment Model (SAM-1A) The EPA's SAM-1A methodology was applied to the four effluent streams as a means of interpreting the emission results. The SAM-1A analysis results in- dicated that the secondary collector ash contained the highest degree of hazard, although all three ash streams were similar in the magnitude of their hazard values. Stack emissions showed a relatively low degree of hazard. The primary collector ash stream had the highest Toxic Unit Discharge Rate (TUDR) which would seem to indicate that this effluent should receive the first priority for control measures. However, because of certain limitations which are inherent in the SAM-1A methodology, this observation should not be regarded as being definitive. Mark A. Golembiewski, K. P. Ananth and T. Sutikno are with Midwest Research Institute, Kansas City, MO 64110. Harry M. Freeman is the EPA Project Officer (see below). The complete report, entitled "Environmental Assessment of a Waste-to-Energy Process, Burlington Electric's Wood and Oil Co-Fired Boiler,"(Order No. PB 80-220627; Cost $12.00, subject of change) will be available from: National Technical Information Service 5285 Port Royal Road Springfield, VA 22161 Telephone: 703-557-4650 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 r L J ------- |