United States Environmental Protection Agency Environmental Sciences Research Laboratory Research Triangle Park NC 27711 Research and Development EPA-600/S3-84-047 May 1984 &EPA Project Summary Chemical and Physical Characterization of Municipal Sludge Incinerator Emissions Roy L. Bennett, Kenneth T. Knapp, and Donald L. Duke Particulate emissions from a group of municipal sludge incinerators, three with multiple-hearth furnaces and one with a fluidized-bed furnace, were characterized in a study covering a period from October 1,1979 to June 30, 1981. Objectives of the investigation were (1) to obtain specific elemental emission concentrations, and (2) to pro- vide source inventories and source sig- natures, especially in terms of particle size, that would assist in developing and evaluating source apportionment mod- els. Three of the plants investigated in this study operated at or near auto- genous burning conditions. Chemical element composition was determined for total and sized emission samples by x-ray fluorescence analysis. During the study, considerable enrichment of sev- eral elements (S, V, Cu, Zn, Cd, Sn and Pb) in the emissions, compared to their content in the sludge feed, was ob- served. The largest average enrichment ratios were found with cadmium (31), zinc (14), lead (9), and sulfur (8). Th/s Project Summary was developed by EPA's Environmental Sciences Re- search Laboratory, Research Triangle Park, NC, to announce key findings of the research project that is fully docu- mented in a separate report of the same title (see Project Report ordering infor- mation at back). Introduction With the production of municipal wastewater sludge on a definite increase, an attendant increase is expected in the use of incineration for sludge disposal manage- ment. A characterization study was con- ducted on paniculate emissions from the stacks of a group of municipal sludge in- cinerators, three with multiple-hearth fur- naces and one with a fluidized-bed furnace. One purpose of the investigation was to pro- vide information on the concentration of chemical elements, especially heavy metals such as cadmium (Cd) and lead (Pb), in the emissions. A related objective focused on obtaining source emission factors and source signatures to help develop and validate source apportionment models. These investigations involved the collec- tion of representative emission samples for: (1) particulate mass emission determination; (2) chemical characterization; and (3) parti- cle size determinations, for both mass emis- sions and individual chemical element emissions. Sulfur dioxide and sulfuric acid emissions were determined; the gaseous concentrations of hydrocarbons and nitrogen oxides were monitored; and a few samples were collected to determine the emissions of high molecular weight organic com- pounds. At each incinerator, samples of the feed sludge were collected and analyzed. In- formation on operating conditions at the four incinerator facilities during the test periods are shown in Table 1. At each incinerator, the sludge was pre- conditioned thermally with the hot exhaust gases from the furnaces to produce a higher solids sludge cake for better burning. At the multiple-hearth Incinerator 0, this process was carried to the point of positive heat balance, resulting in autogenous burning. Some supplemental fuel was burned at all of the sites. At Incinerator 0 a total of only 16 gal of fuel oil was burned. This burning occurred during the final two 29-h sampling periods, and was carried out for shut-down purposes. The remainder of the time the ------- Table 1. Furnace Types, Control Systems, and Sludge Loadings of Tested Incinerators Site designation 0 P Q R Furnace type Seven-chambered multiple- hearth Eight-chambered multiple- hearth Fluidized-bed Seven-chambered multiple- hearth Control equipment Wet, tray-type scrubber Wet, tray-type scrubber Wet, tray-type scrubber Single pass cyclone scrubber Average load of unit tested kg/h (dry) 1940 1740 839 1890 Sludge percent moisture 63.5 49.3 66.2 71.3 sludge burned autogenously, combusting on its own with no need for supplemental fuel. At Incinerator P an average of 40 ft3/h of natural gas was burned. The supplemental fuel oil at Incinerator Q ranged from 0 to 64 gal/h and averaged 17 gal/h over the three- day test period. Incinerator R was not designed for autogenous combustion and never reached this condition. An average of 40 gal/h of fuel oil was consumed. Experimental Sampling Procedures A standard U.S. Environmental Protection Agency (EPA) Method 5 sampling train was used to collect particulate samples for deter- mining mass emission rates. The impingers were adapted for use in determining gaseous sulfur oxides by EPA Method 8. The first im- pinger contained 150 ml of 80% isopro- panol, which was followed by a high-purity glass wool plug inserted in the U-tube be- tween the first and second impingers. The second and third impingers contained 100 ml each of 3% hydrogen peroxide solution; the fourth impinger was filled with approxi- mately 400 g of indicating silica gel. Prior to sampling, a velocity profile of the duct at the sampling location was determined through transversing with a Pitot tube system. A 5-ft heated, Pyrex-lined probe was used to col- lect all Method 5 and Method 8 samples at a single point determined to have the average velocity of the flue gas within the duct. Samples used for chemical characteriza- tion by x-ray fluorescence analysis (XRF) were collected with a modified Method 5 train in which the conventional sample box, filter holder, and glass impingers were re- placed by an EPA-designed heated sample box housing a stainless steel filter holder for 47-mm filters. The particulate characteriza- tion samples were collected in sets consisting of two Gelman A glass fiber filters, two Millipore AA, one Nuclepore 0.8-^im filter, and six Teflon 0.2-^m filters. To provide a variety of loadings on the filters, sampling periods ranged from 15 sec to 10 min. In all cases, an attempt was made to sample iso- kinetically. Collections on all filters were analyzed by XRF. University of Washington Mark III cascade impactors were used to collect samples for measuring the particle size distribution of total mass emissions, as well as individual chemical element emissions. The impactor samples were taken in-stack at the point of average gas velocity; this same sampling point was used for the characterization sam- pling. Samples were collected isokinetically, with a sampling rate between 0.5 and 0.75 ftVmin through the impactor. Composite pre-burn sludge samples and post-burn ash were collected at all in- cinerators except at the fluidized-bed unit, where only sludge samples were taken. The sludge samples were heated to 600°C for 30 min to remove the volatile contents prior to chemical analysis. Analytical Methods Elemental contents of the emission samples, as well as of the sludge and ash samples, were determined with a Siemens MRS-3 multi-channel wavelength dispersive x-ray fluorescence spectrometer. The Siemens MRS-3 has 15 fixed wavelength monochromators and a scanning channel that allows analysis for 11 additional elements. Results and Discussion Sludge Content The elemental composition of the ashed sludge feed material, determined by XRF, is shown in Table 2. The values represent the averages for four to six composite samples taken during two- to three-day testing periods at each incinerator. Elemental Emissions The mean concentrations of the elements, determined by XRF, in the emissions from the four plants are shown in Table 3. The Table 2. Elemental Composition (Percent) of Municipal Wastewater Sludge Element Na Mg Al Si P S Cl K Ca 77 V Cr Mn Fe Co Ni Cu Zn As Se Br Cd Sn Sb Ba Pb Sitt AV 1.0 1.8 34 W.I 64 0.93 0.45 13 17. 084 0.03 0.22 070 67 0.01 003 0 12 0.30 <03 <0.04 <005 002 0.30 0.03 025 034 > O SD 0.09 0.08 0.28 0.46 0.44 0.15 0.01 0.09 0.62 0.06 0.01 0.02 0.11 1.0 0.003 0.01 0.02 0.04 0.005 0.02 — 0.02 0.13 Sitt AV 0.7 1.0 3.6 13.0 4,4 13 008 1.1 6.2 0.76 0.04 0.50 0.23 17 0.03 0.14 0.75 1.1 <0.3 <004 <0.05 0.04 0.25 0.02 0.41 0.48 ; P SD 011 0.17 047 3.3 0.22 0.41 0.18 024 1.2 0.05 001 0.27 0 11 79 002 0.08 043 008 0.05 0.03 0.004 0 11 0.18 Site AV <0.6 0.87 4.0 170 4.1 0.82 <0.5 1.2 7.5 0.77 0.03 0.22 0.34 7.6 0.01 0.06 0.27 1.0 <0.03 <0.04 <0.05 0.08 0.25 0.01 0.45 0.59 > Q Site R SD 0.13 0.23 2.7 0.13 046 0 15 1.0 0.04 0.01 0.24 0.12 6.4 0.02 0.07 0.37 0.07 0.04 0.03 0.001 0.04 0.25 AV <0.2 0.92 5.2 12.0 3.7 047 0.68 1.1 9.8 1.18 <0.01 0.38 0.11 5.4 0.01 0.06 0.44 1.4 <0.03 <0.04 <0.05 003 0.19 001 027 023 SD 0.03 0.14 0.34 0.12 0.06 0.19 0.02 0 15 0.04 0.02 0008 0 10 0.002 0008 002 003 0.005 0.006 0.001 0.01 0.03 Avg. Content 16 cities la) 0.44 0.60 1.83 — 1.56 — 038 1 22 3.62 023 0.004 0.14 0.019 3.06 0001 — 0,13 0.21 0.0014 0.0003 0.005 0.010 0.022 0.001 0.06 0.18 lat Furr, A.K., A.W. Lawrence, S S C. Jong, M.S. Grandolpho, R.A. Hofstader, C.A. Bache, W.H. Gutenmann, and D J. Lisk, Environ. Sci. Techno/., 10,683, 1976 ------- Table 3. Mean Concentrations of Elemental Emissions Ing/Nm3) from Municipal Sludge Incinerators Site O S/te P Site Q Element Site R AV SD AV SD AV SD AV SD Na Mg Al Si P S Cl K Ca Ti V Cr Mn Fe Co Ni Cu Zn As Se Br Cd Sn Sb Ba Pb 350 90 128 590 590 660 230 210 780 28 ~35 97 27 370 <6 <22 85 810 <29 ~45 ~49 42 180 6 10 510 110 32 47 100 120 300 90 90 170 12 39 13 80 29 220 15 51 2 4 190 290 440 1570 4180 1730 930 130 460 2620 450 38 480 82 7070 25 125 810 1840 <29 20 57 34 1230 23 290 1170 230 550 1980 4826 1890 410 110 500 2660 460 14 330 30 8170 a 121 690 1240 3 19 15 450 7 280 530 100 51 160 270 210 730 47 54 440 33 9 14 23 228 <6 <22 14 87 <29 18 39 7 30 <2 20 114 50 18 48 80 69 270 65 22 150 10 2 8 8 71 4 28 17 17 2 9 7 65 240 21 37 620 570 2610 1990 120 160 16 <6 230 10 230 <6 <22 520 1830 <29 26 170 1890 790 43 14 2140 170 6 20 290 350 910 2280 110 88 21 106 1 145 310 1650 1 140 1410 640 40 15 1880 relative standard deviation of the samples analyzed (between 45 and 95 samples per site) to obtain each of these concentrations averaged 33%, 72%, 32% and 68% for O, P, Q and R, respectively. The greater variability of P and R reflect the wider range of operating conditions, e.g., the rates of feed and stack gas flow during the testing oeriods. It is apparent immediately that elements such as Zn, Cd, and Pb had a higher concentration in the emissions than would have been expected based solely on their concentration in the sludge. Enrichment factors, calculated by dividing the concen- tration in the total paniculate emission by the concentration in the sludge, are listed in Table 4. Conditions causing enrichment are volatility and interaction of the volatile elements with the extremely fine particles not efficiently removed by scrubbers. Total Particulate Mass and Sulfur Oxides The emission rates and concentrations of paniculate mass, sulfur dioxide and sulfuric acid were obtained from the Method 5 and Method 8 tests run afte'r the scrubber outlets. Lowest paniculate emissions were observed at the fluidized-bed incinerator, Q. Highest sulfur oxide emissions were found at In- cinerator R, presumably due to the sup- plementary fuel oil burned to maintain combustion. Particle Size Distribution The average mass median diameters ob- tained from the particle size distribution measurements were 0.28, 0.30,1.1 and 0.85 u,m at Incinerators 0, P, Q and R, respec- tively. Individual size distributions for several elements, as well as the total mass distribu- tion, are shown in Figure 1. For convenience, the largest and smallest size fractions have been assigned finite values. The Zn distribu- tion, predominantly submicrometer particles, was characteristic of that of S, Cd, Pb, and other elements exhibiting enrichment in the emissions, probably due to their volatility in the furnace. Phosphorus exhibited a group of mid-range particles around 2 ^m. Iron was distributed through all sizes, with a signifi- cant fraction of large particles evident even in the controlled, post-scrubber emissions. Only a small fraction of the total mass was greater than 2 ^m (see Figure 1). At Incinerator R panicle size measure- ments were also made before the scrubber. Figure 2 shows the distribution at the scrub- ber inlet, where large particles were preponderant; however, a second mode oc- curred near 2 urn and at the scrubber outlet, where the mass concentration of the smaller particles predominated because of the scrub- ber's more efficient removal of larger par- ticles. Evident in these pre-scrubber measurement is the contrast in the distribu- tion of a more volatile element such as Cd with that of calcium (Ca) (see Figure 2). Conclusions The more volatile elements, Cd, Pb and Zn, were enriched in the emissions. Cad- mium had the highest average enrichment ratio, with a 31-fold increase in the emis- sions. The average mass median diameters for the particles emitted at all four sites were small, ranging from 0.28 urn to 1.1 ^m. Few Table 4. Enrichment Ratios of Elements in Emissions Relative to Content in Sludge Incinerator Element Na Mg Al Si P S Cl K Ca Ti V Cr Mn Fe Co Ni Cu Zn Br Cd Sn Sb Ba Pb 0 2.76 0.42 0.30 0.46 0.73 5.64 4.00 1.32 0.36 0.26 9.33 3.50 0.30 0.43 — — 5.67 21.47 _ 16.50 4.73 1.67 0.32 11.97 P 0.81 0.86 0.85 0.63 0.76 1.38 0.63 0.81 0.82 1.16 1.75 1.88 0.70 0.81 1.67 1.71 2.11 3.25 _ 7.0 9.56 2.50 1.36 4.70 Q 1.16 0.78 0.31 1.01 17.6 _ 0.89 1.16 0.84 6.0 1.27 1.35 0.59 — — 7.0 28.7 7.7 1.6 2.36 — 0.89 3.83 Ft , 0.03 0.01 0.08 0.23 8.14 4.28 0.16 0.023 0.02 _ 0.87 0.10 0.06 — _ 1.75 1.89 _ 98.9 6. JO 4.6 0.07 13.65 Average enrichment 0.38 0.49 0.37 0.68 8.19 2.97 0.80 0.59 0.62 5.69 1.88 0.61 0.54 4.13 13.83 31. 5.68 2.92 0.66 8.53 ------- 5 4 3 2 n~^ 1 '„ Q 0 o1 1 00 < <075 050 n ?f\ 0.00 0 • 1 I- - ; I 0.2 1 02 O.i 1 0.5 - 7 ' 1 2 1 2 / 4. Zin i i 5 /C ' ' C Iro 5 1C ) 20 1 ) 20 - 5 - - 5 3 C OC . - ? 7 - - 7 02 0.2 0.5 0.5 ; 1 2 1 1 i B. Phosphorus _ 5 70 20 5C i i i D7/i tftl - 5 10 20 5C u./o 0.50 0.25 n nn 35 30 25 20 15 10 5 Q ) Figure 1. Diameter, D (um) Diameter, D (um) Particle size distribution of zinc, phosphorus, iron, and total paniculate emissions from Incinerator 0. particles were larger than 2 u.m. The volatile elements were found predominantly in the submicrometer range. The enrichment of Cd (especially relative to that of Zn) is much greater at Incinerator R than at the other facilities. A contributing factor to this higher enrichment is Cd's greater predominance in small particles, as well as the lower efficiency of the control device. Incinerator R has a wet, single-pass scrubber system that is considerably less ef- ficient for submicrometer particle removal than scrubbers at the other incinerators. Elemental particle size measurements at In- cinerator R before the scrubber inlet showed that 72% of the Cd was associated with par- ticles smaller than 1 pirn (see Figure 2). Only 28% of the Zn was associated with this size range. 40 35 ~\ 1 1 1 1 r A. Cadmium Scrubber Inlet B. Calcium Scrubber Inlet 75 0.1 0.2 0.5 1 2 5 10 20 Diameter, D (u) 500.1 0.2 0.5 1 2 5 10 20 50 Diameter, D (urn) Figure 2. Particle size distribution of cadmium, calcium, and total paniculate mass before the scrubber and total paniculate emission after the scrubber at Incinerator ft. ------- The EPA authors, Roy L. Bennett (also the EPA Project Officer, see below), Kenneth T. Knapp, and Donald L. Duke are with Environmental Sciences Research Laboratory, U.S. Environmental Protection Agency, Research Triangle Park. NC 27711. The complete report, entitled "Chemical and Physical Characterization of Municipal Sludge Incinerator Emissions," (Order No. PB 84-169 325; Cost: $10.00, subject to change) will be available only from: National Technical Information Service 5285 Port Royal Road Springfield, VA22161 Telephone: 703-487-4650 The EPA Project Officer can be contacted at: Environmental Sciences Research Laboratory U.S. Environmental Protection Agency Research Triangle Park, NC 27711 United States Environmental Protection Agency Center for Environmental Research Information Cincinnati OH 45268 Official Business Penalty for Private Use $300 ''t 1 ft U.S. GOVERNMENT PRINTING OFFICE: 1984-759-102/947 ------- |