Grantee Research Project Results
2002 Progress Report: Characterization and Minimization of Fine Particulate Emissions from Waste Incinerators by Real-Time Monitoring of Size-Resolved Mass and Chemical Composition
EPA Grant Number: R828192Title: Characterization and Minimization of Fine Particulate Emissions from Waste Incinerators by Real-Time Monitoring of Size-Resolved Mass and Chemical Composition
Investigators: Smith, Kenneth A. , Worsnop, Douglas R. , Boudries, Hacene , Onasch, T. , Zhang, X.
Current Investigators: Smith, Kenneth A. , Worsnop, Douglas R. , Boudries, Hacene
Institution: Massachusetts Institute of Technology
Current Institution: Massachusetts Institute of Technology , Aerodyne Research Inc.
EPA Project Officer: Hahn, Intaek
Project Period: June 1, 2000 through May 1, 2003
Project Period Covered by this Report: June 1, 2001 through May 1, 2002
Project Amount: $335,000
RFA: Combustion Emissions (1999) RFA Text | Recipients Lists
Research Category: Air Quality and Air Toxics , Air
Objective:
The objective of this research project is to perform a real-time analysis and quantification of particulate pollutants in the exhaust of waste incinerators. All of the pollutants we proposed to measure are known to be either toxic or carcinogenic. Our approach to this task involves using two instruments: an Aerosol Mass Spectrometer (AMS) from Aerodyne Research, Inc. (ARI) for volatile and semi-volatile compounds, and a laser-based AMS for selective and sensitive detection of toxic organics (specifically, chlorinated polycyclic hydrocarbons).
The experimental results reported were carried out in two types of municipal incinerators, a Sewage Sludge Incinerator (SSI) located in Manchester, NH, and a Municipal Solid Waste Incinerator (MSWI) located in Massachusetts. During these experiments, an AMS and a laser-based AMS were deployed to characterize, in real time, the chemical and physical composition of aerosols emitted from incinerators.
Incineration is a common method of waste disposal. From 1995-1996, 33.5 million tons of municipal solid waste (MSW) generated in the United States (16 percent of all MSW) was combusted at high temperatures in MSWIs. An additional 4 million tons of hazardous waste (or 2 percent of all hazardous waste) also was destroyed in hazardous waste incinerators (EPA, 1999). Sewage sludge incinerators burn about 25 percent of the sewage sludge generated in the United States (Werther and Ogada, 1999).
Concerns about toxic air pollutants that may be emitted from incinerators remain a major impediment for public acceptance of these systems. These concerns have focused on dioxins/furans, toxic metals, and acidic gases. Comparatively, little attention has been devoted to particulate emissions. This is partially because most of the particulate mass leaving the combustion chamber is efficiently captured in modern baghouses. However, most of this mass is associated with coarse particles.
In the last 10 years, fine particulates have become associated with respiratory health effects leading to increased hospital admissions and mortality (Wilson and Spengler, 1996). Fine particulates are able to penetrate deeper into the human lungs and, if deposited there, may be an efficient vehicle for exposure to an array of different chemicals.
Incinerator emissions are enriched in particle concentrations from about 0.1 to 2 microns because particles that control systems in incinerators are least efficient for this size range (Saxena and Jotshi, 1996; Ruth, 1998). These particulates tend to be enriched in condensable organics and some toxic metals such as As, Cd, and Pb (Linak and Wendt, 1993; Ruth, 1998; Niessen and Porter, 1991). This occurs because the amount of condensation of a given species during exhaust cooling is proportional to the particle surface area, which favors fine particles over coarse particles. These particles also are more efficiently deposited in the lungs (Wilson and Spengler, 1996); therefore, there is a need to better quantify the amount, chemical composition, and toxics content of these particles.
Recent efforts have been extended to real-time, ultra-sensitive monitoring techniques for gas-phase chlorinated aromatics in incinerator exhaust (Heger, et al., 1999; Zimmerman, et al., 1999), but such advances have not yet been incorporated into particulate phase measurements. For some pollutants, particulate phase emissions may actually dominate over gas phase emissions. Therefore, it is critical to extend the real-time, ultra-sensitive monitoring capabilities to the particle phase. Recently, laser ablation/ionization instruments have been developed that can obtain mass spectra of single particles (Hinz, et al., 1996; Card, et al., 1997; Reilly, et al., 1998). However, these instruments provide only semi-quantitative composition information because of the non-linearity in the particle collection and laser ablation/ionization process.
This program is designed to perform a real-time analysis and quantification of particulate pollutants in the exhaust of several waste incinerators, particularly for those pollutants that are known toxins or carcinogens. A resonance enhanced multiphoton ionization (REMPI) time-of-flight (TOF) AMS, referred to as the REMPI-AMS, originally developed under a previous U.S. Environmental Protection Agency grant was improved and used to monitor particulate emissions from incinerators for selective and sensitive detection of some toxic organics (specifically, polycyclic aromatic hydrocarbons (PAHs). An AMS, commercialized by ARI, also was used during this experiment for monitoring all volatile, semi-volatile, and refractory species present on particles emitted from incinerators.
Progress Summary:
Instrumentation Development. During Year 1 of the project, we:
- Developed a computer based optic-electronic system to synchronize laser triggering with a chopper spin. In the design, the optical signal generated when the chopper opens to the particle beam is used to trigger a counter-board installed in the computer (Model CIO-DAS08-AOH, Computer Boards, Inc., Middleboro, MA); after waiting for a programmed delay time in counter pulses, the board sends a signal to trigger the laser.
- Introduced a programmable counter (Model PCI-6601, National Instruments, Austin, TX) to generate the necessary frequencies to control the chopper wheel at three positions: beam-open, chopping, and beam-blocking.
- Upgraded the software for the data acquisition system for the REMPI instrument to control all instrument operations including: laser triggering time, chopper positions, spectra acquisition, analysis, display, and storage of mass spectra according to operating conditions.
- Consolidated all instrument operation systems into two modules: the instrument, including the pumps, the TOF, and the second module regrouping all control units of the instrument. Now, the instrument is very easily portable and can operate on a mobile platform (like a truck). The two modules can be assembled within 2 to 3 hours.
- Quantitatively calibrated the REMPI-AMS instrument using pure Pyrene particles and comparing the calibration results with an AMS.
During the last deployment of the REMPI-AMS and the AMS to the pilot-scale waste incinerator located at the New Jersey Institute of Technology (see the 2001 Annual Report), we have made additional improvements to deal with sampling artifacts that may occur when sampling at real incinerators. These improvements consist of the:
- Design of a new heated sampling line (this allows us to keep the sampling line temperature at the same temperature as the gas in the stack and avoid the problems of condensation in the line).
- Reduction of the length of the sampling line by sampling very close to the stack (a long sampling line may introduce a loss of particles during the transport, or aerosols from the stack to the instrument).
- Better dealing with critical orifice clogging (high number density of big aerosols and water condensation may clog the 100 m critical orifice of the AMS instruments).
REMPI-AMS Calibration. In the calibration, an atmospheric pressure gas-particle suspension passes through a pinhole (~100 m) that controls the inlet mass flow rate (~100 scc/min). The relationship between REMPI-AMS readings (area of mass-spectral peak that is measured in coulombs) were obtained by sampling nominally monodisperse particles preselected by a differential mobility analyzer (DMA, Model 3032, TSI, St. Paul, MN). The inlet particle concentration was monitored by a condensation particle counter (CPC, Model 3022A, TSI, St. Paul, MN). Figure 1A shows the REMPI-AMS reading of pyrene particles versus the number density of particles isokinetically sampled by both CPC and REMPI-AMS. For each particle size, a rough linear relationship is observed, indicating that the total sampled particle mass is proportional to particle number density. As the larger particle carries more mass than the smaller one, lines for larger particles have larger slopes than the smaller particles. With the known particle size and diameter, particle mass concentration is easily calculated. Figure 1B is a plot of REMPI-AMS readings versus sampled particle mass concentration. Eventually, all three lines in Figure 1A roughly collapse into one universal line. This is an indication that the REMPI-AMS is a particle mass detector and the universal line serves as a calibration for pyrene particles.
Figure 1. Calibration of REMPI-AMS Using Pyrene Particles. Figure A: Signal Versus Particle Number, and Figure B: Particle Mass Concentration. Laser Intensity I=1.6 MW/cm2 and Oven Temperature T=800°C.
Field Measurements. The REMPI-AMS-MS, an AMS-Quadrupole (Jayne, et al., 2000), a CPC (TSI, model CPC 3010), and two atomizers, (Poly-disperse Aerosols, DMA, TSI Model 3071 and TSI Model 3076) were deployed at an MSSI operated by the city of Manchester, NH, from August 14-18, 2002, and the second experiment was conducted at a real MSWI located in Massachusetts from August 27-31, 2002.
Aerosols Sampling. In both experiments, aerosol particles were sampled in the stack just before being emitted into the atmosphere, representative of incinerator exhaust directly emitted to the atmosphere. A heated copper tube (1/5 inch o.d., 5 M long) and a cyclone were used to sample aerosol particles below 2.5 m at a flow rate of about 9 L/min. Both the AMS and the CPC were connected to isokinetically sample through the sampling line. During this experiment, the CPC, REMPI-AMS, and AMS were set to sample at 1.5, 0.1, and 0.1 L/min, respectively. A mass spectrum was obtained every 2 seconds and averaged more than 30 seconds, and the size distribution between 20 nm and 2.5 m was measured every second and averaged more than 30 seconds.
Sewage Sludge Incinerator at Manchester. As presented in Figure 2, the SSI at Manchester consists of the following components:
- Hot Fluid Bed Incinerator
- Combustion Air Heat Exchanger
- Economizer With Heat Recovery System
- Venturi and Impingement Tray Scrubber
The sludge is fed into a bed and air is blown into the bottom of the sand bed simultaneously to create a turbulent suspension of sand, sludge, and gas. Inside the incinerator chamber, the suspension of sludge, sand, and gas is maintained at about 1,360°F (750°C). At this temperature, all materials are rapidly evaporated and volatile matter present in the sludge is combusted. The exhaust gas above the bed is cooled down in the Combustion Air Heat Exchanger from 1,500°F to 1,050°F, while the combustion air is heated to 1,200°F. The exhaust gases then enter the Venturi scrubber to reduce particulate emissions. After a significant fraction of particulate emissions are removed in the Venturi scrubber, the exhaust gas is directed into an impingement tray scrubber. Here, a large quantity of water is used to cool the gas to about 43°C, and caustic solution (1 percent of NaOH) is used to reduce acidic gases. Finally, the exhaust gas is emitted into the atmosphere through a 37-m long stack.
Municipal Solid Waste Incinerator. The MSWI is designed to provide disposal and recycling of 3,000 tons per day of MSW. The first stage of MSW treatment consists of shredding the waste, followed by magnetic separation to remove most ferrous materials. The processed fuel then is blown into a special boiler. Light materials burn in suspension, while the heavy portion of the fuel remains on the surface of the grate. After processed refuse fuel is burned, the dry ash recovered in the bottom of the boiler is removed and separated into three main components; ferrous metal group, non-ferrous metals, and boiler aggregate. Combustion gases are passed through a dry scrubber where they are sprayed with a lime reagent to remove acid gas constituents. Gases then are passed through electrostatic precipitators to capture particles. Finally, the exhaust gas is emitted to the atmosphere through a 37-m high stack.
Manchester Sewage Sludge Incinerator. The experiment at the Manchester Sewage Sludge Incinerator was conducted from August 14-18, 2002. Figure 3 shows the temporal trend of total mass loadings of major organic and inorganic components present in aerosols in the stack. A significant variability was observed in the aerosol mass loadings, while the feeding rate and combustion conditions remained stable. The total mass loadings measured at the Manchester Incinerator vary between 50 and 1,100 µg/m3. This variability in mass loadings may be attributed to the variability in the composition of sludge. The low concentration observed on August 16, between 12:00 a.m. and 8:00 p.m. is because the feeding of sludge was interrupted by a pump failure. Ammonium and sulphate represent the major particulate compounds emitted from the Manchester Sewage Sludge Incinerator. They are followed by organics and chloride, with nitrate representing only a minor fraction of total organics. Further analysis is under way to understand why the concentration increased significantly on August 18 at 7:00 a.m.
The average size distributions for these compounds are presented in Figure 4, as a frequency plot of dM/dlogDa versus the aerodynamic diameter in nm. This figure shows that the size distribution is predominantly monomodal with an average aerosol diameter varying between 130 and 200 nm. Both ammonium and sulphate exhibit the same average aerodynamic diameter of about 130 nm. Organics and chloride are found to be associated with a higher and similar diameter of about 200 nm. Finally, nitrate is found associated with an aerodynamic diameter of about 160 nm. This result suggests that the aerosols emitted from the Manchester Incinerator are externally mixed and consist of ammonium sulphate, organics and chlorinated organics, and nitrates.
All the data of the Manchester facility presented are obtained by the normal AMS equipped with a quadrapole mass spectrometer. The PAH concentration was well below the detection limit of the REMPI-AMS (~100 ng/m3) so that no signal was observed.
Results at a Municipal Waste Incinerator. The experiment at the MSWI was conducted from August 27-31, 2002. Temporal variations of major chemical compounds emitted from a MSWI are presented in Figure 5. It can be seen that the organic fraction represents the most dominant fraction, followed by sulfate and chloride in the AMS. The chemical composition is measured by a quadrupole mass spectrometer following particle flash vaporization in a resistively heated surface (oven). During this experiment, the oven temperature was varied between 550°C and 1,200°C. It is important to note that the optimum temperature for measuring non-refractory compounds is about 550°C. Figure 5 clearly shows that total mass loadings are positively correlated with oven temperature. The average mass loading measured at this incinerator varies from 10 to 150 µg/m3, depending on the oven temperature of the AMS. The highest concentrations were measured at high oven temperature. The average mass loading is about 160±25 µg/m3 at a temperature higher than 800°C, and 34 ±3 µg/m3 at a cold oven temperature of about 600°C. Similar to a SSI, significant variability in total mass loadings was observed with constant oven temperature above 800°C. Data analysis obtained at low oven temperature (550°C) contains little information on particle distributions, while an increase of oven temperature greatly enhanced the measured mass signals as presented in Figure 6. This result suggests that the composition of aerosols from this incinerator mostly is composed of refractory compounds. Further, this conclusion is supported by a large increase in measured mass loadings with oven temperature.
Figure 4. Aerosol Size Distributions Observed for Ammonium, Organics, Sulphate, Chloride, and Nitrate
Figure 5. Distribution of Total Mass Loadings, Organics, Sulfate, Chloride Measured in the Stack of a Solid Municipal Waste Incinerator
Several previous studies already have shown the presence of many chlorinated and polychlorinated hydrocarbons in fly ash and in flue gas of MSWI. Previous studies also have shown that chlorinated hydrocarbons can be intermediate in the formation of chlorinated dioxins and furans (Olie, et al., 1998). Contribution from chlorinated hydrocarbons at the MSWI has been accomplished by a detailed analysis of mass spectra. Chemical ionization of all chlorinated hydrocarbons gives a signal at amu 35, 36, 37, and 38. The average concentration at hot oven temperature (AMS oven temperature above 800°C) is estimated to 19 µg/m3, with a standard deviation of 6 µg/m3. This represents about 10 percent of total mass loading.
The isotopic analysis of Cl signature at amu 35, 36, 37, and 38 suggests that the chlorine is a result of decomposition of chlorinated compounds rather than hydrogen chloride. Further analysis of mass spectra is being performed to distinguish between nonchlorinated and chlorinated hydrocarbons. This will allow us to better identify and quantify the total contribution from individual or groups of chlorinated compounds.
Figure 6. Aerosol Size Distribution Observed for Lead, Chlorine, Sulfate and Organics
Figure 7. Average Mass Spectrum for Lead Observed Under "Hot" Oven Temperatures (>800°C) Compared With the Expected Fractionation Pattern for Lead Obtained From the National Institute of Standards and Technology
As stated before, the increase of total mass loading when the AMS oven temperature increased suggests the presence of refractory compounds. Presence of metals in fly ash and flue gas from incinerators already has been established (Biswas and Wu, 1998). Among the most common metals found in incinerators are lead, nickel, cadmium, and mercury. The identification of metals in our spectrum has been accomplished by comparing the fragmentation pattern of each trace metal with its isotopic composition. Figure 7 shows the identification of lead in the aerosols sampled in the stack from a solid waste incinerator. The average mass loading of lead at hot oven is estimated to be 2 µg/m3 and represents between 1 to 2 percent of total mass loading. Further analysis is necessary to identify and quantify the contribution from all possible metals to total mass loadings. The corresponding size distribution of lead presented as a frequency plot of dM/dlogDa versus aerodynamic diameter (Da) in nm is shown in Figure 5. This figure shows that the size distribution of chlorine and lead are predominantly monomodal and internally mixed with an average aerodynamic diameter of about 800 nm.
The only signal observed in the REMPI-AMS is for pyrene. The REMPI-AMS data for pyrene (from 11:00 p.m. August 29, to 6:00 a.m. August 30) are shown in Figure 8. The laser intensity is about 1.6 MW/cm2, and oven temperature is about 800°C. Apparently, the data show a very large scatter. Occasionally, data show negative values. The reason is that background in the REMPI-AMS is large, on the order of 1 µg/m3, and the pyrene concentration emitted from the combustor is much less than the background. Therefore, the data obtained by subtraction between signals and background are highly scattered and sometimes has negative values, though they were averaged for a period of 10 minutes. Data in Figure 8 suggest that the emission level of PAH is below the level of the current REMPI-AMS configuration to make quantitative measurements. The only conclusion one could possibly draw from the data is that pyrene concentration is on the order of 20 ng/m3. Improvements on the detecting sensitivity and minimizing background is required to make a quantitative measurement.
Figure 8. Pyrene Concentration Versus Time. I=1.6 MW/cm2, T=800°C
In this study, a REMPI-AMS and an ARI commercialized AMS were used to monitor, in real time, particulate emissions from an SSI and an MSWI.
Aerosol size distribution and mass loading concentrations were measured for both incinerators. At the sewage sludge incinerator, the size-resolved distribution shows that the aerosols in the stack are externally mixed with an aerodynamic diameter varying between 130 and 200 nm. The average mass loading was about 200 µg/m3, mostly composed of ammonium, sulfate, ammonium chloride, and organics. At the MSWI, a preliminary analysis shows a clear signature of refractory compounds like lead. Contrary to the sewage sludge incinerator, the size distribution shows that the aerosols in the stack are internally mixed and exhibit a higher average aerodynamic diameter of about 800 nm. The emission level of PAH is below the level of the current REMPI-AMS detection limit, though it is shown quantitatively that the pyrene level is on the order of 20 ng/m3.
Future Activities:
Future activities include further experiments to better characterize the aerosol composition emitted by at least two real MSWIs. The effect of stack temperature, excess air, and feed rate on aerosol mass loading will be studied. We will: (1) computerize control of the AMS oven temperature and incorporate it into a data acquisition program to better cycle through low oven temperature (550°C) and high temperature (1200°C); (2) measure low oven temperature to better characterize volatile and semi-volatile compounds, whereas high oven temperature will be suitable for analyzing refractory compounds like mercury and lead; and (3) improve the detection sensitivity and minimize the background on REMPI-AMS to make quantitative measurement on PAH emissions.
References:
Pratim B, Yu Wu C. Control of toxic metal emissions from combustors using sorbents: a review. Journal of Air and Waste Management Association 1998;48(2):113-127.
Gard E, Mayer JE, Morrical BD, Dienes T, Fergenson DP, Prather KA. Real-time analysis of individual atmospheric aerosol particles: design and performance of a portable ATOFMS. Analytical Chemistry 1997;69(20):4083-4091.
Heger HJ, Zimmerman R, Dorfner R, Beckmann M, Griebel H, Kettrup A, Boesl U. On-line emission analysis of polycyclic aromatic hydrocarbons down to ppt concentration levels in the flue gas of an incineration pilot plant with a mobile resonance enhanced multiphoton ionization time-of-flight mass spectrometer. Analytical Chemistry 1999;71(1):46-57.
Hinz KP, Kaufmann R, Spengler B. Detection of positive and negative ions from single airborne particles by real-time laser mass spectrometry. Aerosol Science and Technology 1996;24(4):233-242.
Jayne JT, Leard DC, Zhang X, Davidovits P, Smith KA, Kolb CE, Worsnop DR. Aerosol mass spectrometer for size and composition analysis of submicron particles. Aerosol Science and Technology 2000;33(1):49-70.
Linak WP, Wendt JOL. Toxic metal emissions from incineration: mechanisms and control. Progress in Energy and Combustion Science 1993;19:145-185.
Niessen WR, Porter RC. Methods for estimating trace metal emissions from fluidized bed incinerators using advanced air pollution control equipment. Presented at the 84th Annual Meeting of the Air and Waste Management Association, Vancouver, British Columbia, June 16-21, 1991.
Olie K, Addink R, Schoonenboom M. Metals as catalysts during the formation and decomposition of chlorinated dioxins and furans in incinaration processes. Journal of Air and Waste Management Association 1998;48(2):101-105.
Reilly PTA, Gieray RA, Whitten WB, Ramsey JM. Real-time characterization of the organic composition and size of individual diesel engine smoke particles. Environmental Science and Technology 1998;32(18):2672-267.
Ruth LA. Energy from municipal solid waste: a comparison with coal combustion technology. Progress in Energy and Combustion Science 1998;24(6):545-564.
Saxena SC, Jotshi CK. Management and combustion of hazardous wastes. Progress in Energy and Combustion Science 1996;22(5):401-425.
Werther J, Ogada T. Sewage sludge combustion. Progress in Energy and Combustion Science 1999;25(1):55-116.
Wilson R, Spengler JD. Particles in Our Air: Concentrations and Health Effects. Harvard University Press, Cambridge, MA 1996.
Journal Articles:
No journal articles submitted with this report: View all 5 publications for this projectSupplemental Keywords:
hazardous air pollutants, combustion emissions, dioxins, furans, air, waste, chemical engineering, chemistry, civil engineering, environmental engineering, environmental chemistry, environmental monitoring, incineration, combustion, particulate matter, PM2.5, polycyclic aromatic hydrocarbon, PAH, volatile organic compound, VOC, VOC incinerator, aerosol mass spectrometry, analytical chemistry, chemical contaminants, combustion contaminants, fine particles, medical waste incinerator, municipal waste incinerator, real-time monitoring, sewage sludge incinerators, size-resolved mass., RFA, Scientific Discipline, Air, Waste, INDUSTRY, particulate matter, Environmental Chemistry, Analytical Chemistry, Environmental Monitoring, Industrial Processes, Incineration/Combustion, Environmental Engineering, fine particles, municipal waste incinerator, medical waste incinerator, PM 2.5, hazardous air pollutants, size-resolved mass, chemical contaminants, PAH, sewage sludge incinerators, waste sludge incinerator, VOC incinerator, furans, PM2.5, combustion, dioxins, real time monitoring, incineration, aersol particles, combustion contaminants, aerosol mass spectrometryProgress and Final Reports:
Original AbstractThe perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.