EPA/600/A-94/068 1993 INCINERATION CONFERENCE Thermal Treatment of Radioactive, Hazardous Chemical Mixed, Energetic, Chemical Weapon, and Medical Wastes Proceedings of the 1993 Incineration Conference Knoxville, Tennessee, U.S.A. May 3-7,1993 Charlotte Baker Conference Coordinator University of California, Irvine Earl McDaniel Technical Program Chairman Oak Ridge National Laboratory Jim Tripodes Oversight Chairman University of California, Irvine M.E. Wacks Editor University of Arizona Sponsored by University of California, Irvine (UCI) And: American Insitute of Chemical Engineers (AIChE) Air and Waste Management Association (A&WMA) American Nuclear Society (ANS) American Society of Mechanical Engineers (ASME) Coalition for Responsible Waste Incineration (CRWI) Health Physics Society (HPS) U.S. Department of Energy (DOE) U.S. Environmental Protection Agency (EPA) ------- POTENTIAL SURROGATE METALS FOR INCINERATOR TRIAL BURNS L. R. Waterland and D. J. Fournier, Jr. Acurex Environmental Corporation Incineration Research Facility Jefferson, AR 72079 ABSTRACT New and renewing hazardous waste management permits for hazardous waste incinerators and other thermal destruction devices require that the emissions of hazardous constituent trace metals be controlled via established metals feedrate limits. Thus, the trial burn required to obtain a permit must include consider- ation of metals emissions. To preserve a high degree of operating flexibility within the ultimate permit conditions defined, incinerator operators generally spike a mixture of hazardous constituent trace metals into the waste burned during the trial burn to increase the metals feedrates tested in the trial burn. This exercise can significantly increase the cost of the trial burn, as some metal constituents are quite expensive. The question thus arises, can surrogate metals be used as substitutes for select hazardous constituent metals to decrease the cost of a trial burn. Over the past 4 years, the research program at the Environmental Protection Agency's (EPA's) Inciner- ator Research Facility (IRF) has developed an extensive body of metals partitioning data from pilot-scale incineration tests using synthetic hazardous wastes, actual listed hazardous wastes, and contaminated materials from Superfund sites. From these data, augmented by bench-scale studies and some full-scale incinerator tests, it has become apparent that the same metals volatilization/condensation mechanisms, first used to explain metals enrichment in flyash from coal combustion 15 years ago, drive metals partitioning in incinerators. Further, metals partitioning can largely be explained using only vapor pressure/temperature relationships for metal species in thermodynamic equilibrium in the combustion zone. Because metals volatility dominates partitioning, surrogates can be used and the choice of surrogates is simplified. This paper discusses the results of three extensive parametric test programs performed at the IRF using synthetic hazardous wastes containing both hazardous constituent and potential surrogate metals. These results show that surrogates partition in the same manner as selected hazardous constituent metals. Thus, the use of surrogates deserves consideration, if not in actual trial burns, then at least in scoping tests used to guide the formal trial burn. INTRODUCTION In 1988, the EPA's Risk Reduction Engineering Labora- tory initiated a research program at its Incineration Research Facility (IRF) in Jefferson, Arkansas, to investigate the fate of trace metals fed to a rotary kiln incinerator. Three parametric studies of the fate of five hazardous constituent trace metals (arsenic, barium, cadmium, chromium, and lead) and four nonhazardous constituent trace metals (bismuth, copper, magnesium, and strontium) have now been completed. In these tests each metal's partitioning to the incinerator's dis- charge streams (kiln ash, wet scrubber air pollution control system scrubber liquor, and flue gas) was measured, and the effects of kiln temperature, afterburner temperature, and feed chlorine content on metal partitioning were evaluated. The first parametric study, completed in 1988, investi- gated a venturi scrubber, packed-column scrubber combina- tion for particulate and acid gas control. A second parametric study, identical in scope to the first, was completed in 1989. The only difference between the first and second studies was the air pollution control system (APCS), which was a single- stage ionizing wet scrubber. Results of the studies were re- ported in detail in 1991 (1,2). A third parametric study was completed in 1991 (3). This study added mercury to the set of test metals and used a Calvert Flux-Force/Condensation scrubber system as the APCS. The use of surrogates in trial burns and scoping tests could significantly reduce permitting costs and improve incinerator operation. Therefore a major objective of the studies was to evaluate the four nonhazardous constituent metals as surrogates for the hazardous constituent metals. This paper examines the trace metal partitioning and scrubber collection efficiency data in light of this objective. TEST PROGRAMS Test Facility All test programs discussed in this paper were performed in the IRFs rotary kiln incinerator system (RKS). A process schematic of the RKS is shown in Fig. 1. The IRF RKS consists of a primary combustion chamber, a transition section, and a fired afterburner chamber. After exiting the afterburner, flue gas flows through a quench section followed by a primary APCS. Two primary APCSs arc available at the IRF for use on the unit. One consists of a venturi scrubber followed by a packed-column scrubber fabricated by Andersen 2000. The other is a single-stage ionizing wet scrubber fabricated by Air Plastics, Inc. Downstream of the primary APCS, a backup secondary APCS, comprised of a demister, an activated-car- bon adsorber, and a high-efficiency particulate (HEPA) filter, is in place. This secondary APCS is designed to ensure the particulate and organic emissions from the system are accept- able even under upset conditions. The modular design of the APCS permits the installation of the other pilot-scale scrub- ber system, such as the Calvert Flux-Force/Condensation scrubber system used in the third parametric test series. A process schematic of the Calvert scrubber system is shown in Fig. 2. ------- 434 Waterland METALS FOR TRIAL BURNS Synthetic Waste Mixture The parametric tests were performed with a synthetic waste feed mixture prepared by adding a mixture of organic compounds (toluene, chlorobenzene, and tetrachloroethene) to a clay-based oil sorbent material. The clay/organic mixture contained nominally 25 percent by weight organic liquids, though it remained a free-flowing solid. The waste feed chlo- rine content was adjusted by varying the ratio of the three organics compounds. Test trace metals were added to the clay/organic mixture by metering a concentrated aqueous metals solution onto the clay/organic mixture at the head of the screw feeder used to feed the synthetic waste to the kiln. All metals were added as soluble nitrates, with the exception of arsenic, which was added as A52O3, Table I summarizes the average metal con- centrations in the integrated feed mixtures. Test Conditions The test matrix was the same for the first two test series. Table IF summarizes the average achieved values for the three test variables. Each was varied over three levels, with the other variables held nominally constant. Target kiln exit tempera- tures were 816°, 871°, and 927°C (1,500°, 1,600°, and 1.700T). Target afterburner exit temperatures were 982°, 1,083°, and 1,204"C (1,800°, 2,000°, and 2,200°F). Target concentrations for chlorine in the synthetic waste feed were 0, 4, and 8 per- cent). Based on the observations from the first two test series, the test matrix for the third test series was slightly modified. The test variables were kiln exit temperature, waste feed chlorine content, and scrubber pressure drop. In the first two test series, metal partitioning and scrubber collection efficien- cies were not affected by changes in afterburner exit temper- ature, so it was eliminated as a test variable and held constant at 1,094°C (2,000°F). In addition, the range of kiln exit tem- peratures was expanded downward to 538°C (1,000°F) and the TABLE I Average Integrated Feed Metal Concentrations Concentration, mg/kg Venturi/packed-column Single-stage ionizing wet Calvert Scrubber Metal scrubber test series scrubber test series test series Arsenic 44 48 34 Barium 53 390 465 Bismuth 150 330 370 Cadmium 8 10 20 Chromium 87 40 280 Copper 470 380 350 Lead 52 45 74 Magnesium 17,200 18,800 34,500 Strontium 280 410 390 c i„ H ROTARY KJLN mClNDUTO* ' MOOlftJt* MWIARY m I POUOTON COKTflOl OfY)C£» RZOUNOAMT AIM i PouirnoN control I SYSTEM Fig. 1. Schematic of the IRF rotary kiln incineration system. Fig. 2. Schematic of the Calvert Flux-Force/Condensation Scrubber System. ------- METALS FOR TRIAL BURNS Waterland 435 TABLE II Incinerator Operating Conditions Feed Average kiln exit Average afterburner Test mixture CI temperature, exit temperature, Series Test content, % °C (°F) °C (°F) Venturi/packed- 1 0 874 (1,606) 1,093 (1,999) column scrubber 2 3.7 825 (1,517) 1,071 (1,959) 3 4.2 928 (1,702) 1,092 (1,989) 4 3.8 878 (1,612) 1,088 (1,991) 5 3.6 871 (1,599) 1,196 (2,184) 6 3.4 875 (1,607) 983 (1,803) T 4.6 873 (1,603) 1.094 (2,000) 8 8.3 870 (1,599) 1,092 (1,998) Single-stage ionizing 1 0 900 (1,652) 1,088 (1,990) wet scrubber 2 3.5 819 (1,507) 1,096 (2,002) 3 3.5 929 (1,704) 1,092 (1,998) 4 3.5 877 (1,610) 1,096 (2,006) 5 3.7 885 (1,625) 1,163 (2,125) 6 3.6 887 (1,629) 1,017 (1,863) 7b 3.6 881 (1,618) 1,103 (2,018) Sb 3.8 879 (1,615) 1,098 (2.008) 9 6.9 881 (1,617) 1,087 (1,988) Scrubber pressure drop, kPa (in WC) Calvert scrubber 1 0 541 (1,006) 12.9 (52) 2 0 819 (1,507) 12.4 (50) 3 0 909 (1,669) 12.4 (50) 4 0.6 555 (1,031) 12.4 (50) 5 0.6 842 (1,547) 12.4 (50) 6 0.8 919 (1,686) 12.4 (50) 7 3.6 543 (1,010) 12.4 (50) 8 3.4 817 (1,502) 12.4 (50) 9 3.1 944 (1,731) 12.2 (49) 10 2.3 829 (1,524) 8.2 (33) 11 3.4 827 (1,521) 16.9 (68) aTest point 7 is a duplicate of test point 4. ''Test points 7 and 8 are replicates of test point 4. ------- 436 Waterland METALS FOR TRIAL BURNS target concentrations for chlorine in the waste feed were adjusted to 0.1, and 4 percent. All tests were conducted under excess air conditions. Oxygen concentrations were nominally 12 to 14 and 8 to 10 percent in the kiln and afterburner exit flue gas, respec- tively. Solids residence time in the kiln was approximately 1 hour. TEST RESULTS The measured feed and discharge stream metal concen- trations can be combined with measured feed and discharge flowrates, and the fraction of the metals fed accounted for in the respective discharges can be calculated. The sum of these discharge fractions represents the mass balance closure for each metal in each test. Ideally, near 100 percent trace metal mass balance closure would be desirable. However, past ex- perience in tests to determine the distribution of trace metals from combustion sources has shown that typical good results are in the 30 to 200 percent range. The ranges and averages for the metal mass balance closures for three parametric test series are summarized in Table III. Given that variable and less than perfect mass balance closure is invariably experienced, it is difficult to draw conclu- sions regarding the affect of incinerator operation or feed characteristics on metal partitioning using only percent-of- feed fractional distributions. However, a clearer picture of the variation in relative metal distributions is possible when per- cent-of-feed fractional distributions are normalized by the total mass balance closure achieved. These normalized, or percent-of-measured fractions represent fractions that would have resulted had mass balance closure in each case been 100 percent. Use of distribution fractions normalized in this manner allows clearer data interpretation, because variable mass balance closure is removed as a source of test-to-test data variability. The use of normalized distributions repre- sents a best attempt to quantify metal partitioning phenom- ena, given variable and less than perfect mass balance closure. When subjected to incineration conditions, metals are expected to vaporize to varying degrees, depending on their relative volatilities. To characterize a metal's volatility, equi- librium analyses can be performed to identify the metal's volatility temperature for a given set of incinerator conditions. The volatility temperature is defined to be the temperature at which the effective vapor pressure of a metal is 10"6 atm. The effective vapor pressure is the combined equilibrium vapor pressures of all species containing the metal, reflecting the quantity of metal that would vaporize under a given set of conditions. A vapor pressure of 10"6 atm is selected because it represents a measurable amount of vaporization. The lower the volatility temperature, the more volatile the metal is ex- pected to be. Because the volatility temperature is based on vapor pressure/temperature relationships for metal species in ther- modynamic equilibrium it provides a useful parameter for comparing relative partitioning behavior. Table III also notes the volatility temperature for each metal, based on its elemen- tal and oxide forms (4). Metal discharge distributions have been summarized for each test program and presented in Figures 3,4, and 5. These figures show the amounts of metal found in each discharge stream normalized as a fraction of the total found in the three discharge streams — kiln ash, scrubber exit flue gas and scrubber liquor. In these figures, the bar for each metal rep- resents the range in the fraction accounted for by each dis- charge stream over all tests of the respective test series. The average fraction for that test series is noted by the midrange tick mark. Metal discharge distribution data are plotted versus volatility temperature. For all three test series these figures indicate a correlation between the observed metal volatility and the calculated volatility temperatures. With increasing volatility temperature there is a gradual increase in average TABLE III Summary of Metal Mass Balance Closure Around the Kiln Ash and Scrubber Discharges Mass balance closure, % of metal fed Venturi/paeked column scrubber Volatility series temperature, Metal °C Range Average Single-stage ionizing wet Calvert scrubber scrubber series series Range Average Range Average Arsenic 700 39 77 60 47 95 66 70 128 94 Barium 849 57 147 86 17 60 27 60 106 77 Bismuth 621 36 74 53 35 63 50 38 86 58 Cadmium 214 37 120 96 36 68 50 26 98 60 Chromium 1,613 61 94 73 77 204 154 34 171 96 Copper 1,116 46 79 63 30 81 50 67 103 83 Lead 627 8 96 70 47 177 110 54 118 77 Magnesium 1,549 70 134 92 63 123 99 71 105 87 Strontium 1,454 28 71 48 15 60 28 59 99 74 ------- METALS FOR TRIAL BURNS Waterland 437 * (/) Z < ~ UJ z 3 S b 2* $00 >00 1 000 1200 1400 VOLATILITY TEMPERATURE CC) SCRUBBER EXIT FLUE OAS PD If* I 1 * St Ug t Cr - J B* i <• Cu ¦ 0. , Cd i ¦ , , 600 *00 1000 1200 1400 VOLATILITY temperature CC) SCRUBBER UQUOR I Ca 1 P6 I 1 TBi 1 SrT TCf 1 - lJ £ u i i* **i i 4C0 600 too 1000 1200 140C VOLATILITY TEMPERATURE CC) VOLATILITY TEMPERATURE CC) SCRUBBER UQUOR 400 GOO 600 1 000 1200 VOLATILITY TEMPERATURE CC) SCRUBBER-EXrT FLUE OAS § «/> 2 80 => < 60 400 600 800 1000 1200 VOLATILITY TEMPERATURE CO Fig. 3. Normalized distribution of metals in the RKS dis- charge streams in the venturi/packed-column scrub- ber tests. Bar indicates range observed over all eight tests. Average is noted by midrange tick mark. kiln ash fraction and decrease in average scrubber exit flue gas and liquor fractions. This is as expected. The less volatile a metal is, as reflected in its higher volatility temperature, the less likely it will volatilize in the kiln and be carried out of the kiln in the vapor phase in the combustion flue gas. The relationships between metal partitioning and the calculated volatility temperature is particularly useful because it suggests that surrogates can be selected to represent partic- ular hazardous constituent metals of interest based on relative volatilities. To further defend the use of the nonhazardous metals as surrogates, it is necessary to compare the partition- ing data on a test by test basis. This task can be simplified by noting that strontium, magnesium, and chromium were highly refractory for each test. Because the partitioning of these metals did not vary with any of the test variables, little can be gained by looking at their partitioning on a test by test basis. These figures do clearly indicate, however, that strontium, and perhaps magnesium, accurately represent the partitioning behavior of chromium. Test by test comparisons of metal partitioning for the remaining six metals can also be simplified to partitioning to the kiln ash and APCS collection efficiency. Figures 6 through 11 show the partitioning of cadmium, bismuth, lead, arsenic, barium, and copper to the kiln ash as a function of the test variables kiln exit temperature and waste feed chlorine content for the three parametric test series. Fig. 4. Normalized distribution of metals in the RKS dis- charge streams in the single-stage ionizing wet scrub- ber tests. Bar indicates range observed over all nine tests. Average is noted by midrange tick mark. Cadmium is not included in Figures 6 and 7 because it was not found in any kiln ash sample above detection limits, although calculated partitioning fractions to the kiln ash using sample concentrations set to the analytical detection limit were less than 25 percent. For the first two parametric test series, Figs. 6 through 9 show that bismuth and cadmium were relatively volatile, with a maximum kiln ash fraction of about 75 percent and an average partitioning fraction to the kiln ash of less than 40 per- cent. Figures 10 and 11 show that all of the metals were less volatile in the third test series, although cadmium and bismuth were more volatile than the other metals. Lead volatility be- havior differed between the three test series. For the ven- turi/packed-column test series, the average fraction of lead recovered in the kiln ash was 20 percent. For the single-stage ionizing wet scrubber test series, the average fraction of lead recovered in the kiln ash was 82 percent. For the Calvert scrubber test series, the average kiln ash fraction was 94 per- cent. These figures show that of the nonhazardous constituent metals tested, bismuth best represented cadmium on a test by test basis for partitioning to the kiln ash. It was somewhat less volatile than cadmium, but was similarly affected by changes in the test variables. Figures 8 through 11 show that copper partitioning to the kiln ash was more typical of lead, arsenic, ------- 438 Watcrland METALS FOR TRIAL BURNS H - ft r • 3 Sr Cr 1 « ' . Pb l » 1 Paramrt/tc Taata, Van»urVPac*ad Co*umn Scrubby •71*C pWOf) kin ail tamparatura 600 aoo 1000 1700 VOLATILITY TEMPERATURE (*Cl SCnUBBKR-CXlT FLUC OAS 2 4 • Wasta Fa*d CNonna Comant (%) _1L Pb U 8> _StL Fig. 7. Metal partitioning to the kiln ash versus waste feed chlorine content at constant kiln exit temperature for the parametric test series using the venturi/packed- column scrubber. Paramatrtc Taata, lonbmg Wat Scrubbar 3.5% *aad cttorina 400 M0 aoo 1000 1200 VOCATllfTY TEMPERATURE CC) SCRUBBER UOUOA ss IS " •CO • Pfc 1 •Cu - J + *¦ , j :J ,r , : S' r, . i » ± j. tt-Cf WO S7S 900 Kiln ExI Tamparatura fC) Ba 600 aoo 1000 1200 VOl>TlUTY TEMPERATURE (*C) Fig. 5. Normalized distribution of metals in the RKS discharge streams in the Calvert scrubber tests. Bar indicates range observed over all 11 tests. Average is noted by midrange tick mark. Paramatrie Taatt, VanturtPaekad Column Scrubbar 3.9% taad chlorlrw Cd frl Pb Am B» Fig. 8. Metal partitioning to the kiln ash versus kiln exit temperature at constant waste feed chlorine content for the parametric test series using the single-stage ionizing wet scrubber. Parametric Taata, tontring Wat Scrubb«r I71*C (1600T) kin azt tamptrstura WO ITS 900 Kin Ext Tamparatur* fC) y Pfr Am h Fig. 6. Metal partitioning to the kiln ash versus kiln exit temperature at constant waste feed chlorine content for the parametric test series using the ven- turi/packed-column scrubber. partitioning to the kiln ash was more typical of lead, arsenic, and barium behavior. Lead was much more volatile in the first test series. The reason for this behavior is not clear as the test were conducted over the same range of operating conditions as the later parametric tests. Figures 12 and 13 show the apparent scrubber collection efficiency for each of the test metals as a function of the kiln exit temperature and waste feed chlorine content, respec- tively. The apparent scrubber collection efficiencies are cal- culated by assuming that the sum of the amount of metal measured in the two scrubber discharges (the scrubber liquor and the scrubber exit flue gas) was the amount of metal 2 4 Wasta Fa*d Chtorina Contant (%) Cd y PS Am BM Fig. 9. Metal partitioning to the kiln ash versus waste feed chlorine content at constant kiln exit temperature for the parametric test series using the single-stage ioniz- ing wet scrubber. present in the scrubber inlet flue gas. This allows the apparent scrubber collection efficiency to be calculated as: (scrubber liquor fraction)/(scrubber liquor fraction + scrubber exit flue gas fraction). Figure 12 shows the apparent scrubber collection effi- ciency for each of the test metals as a function of the kiln exit temperature at constant waste feed chlorine content for the three scrubber systems tested. Figure 13 similarly shows the apparent scrubber collection efficiencies as a function of the waste feed chlorine content at constant kiln exit temperature for two of the three scrubber systems tested. These figures show that magnesium behavior approximates barium behav- ior, while copper and bismuth behavior is similar to that, of arsenic, cadmium, and lead. ------- METALS FOR TRIAL BURNS Waterland 439 Parametric Tests, Catvert Scrubber no feed chlorine Parametric Teats, Catvert Scrubber 53**C (1000"F) kiln exA temperature 1 2 •16*C (1S00*F) Win exR temperature 600 700 BOO 900 1000 0.7% feed chlorine 1 2 »2TC (1700T) kiln exit temperature 3.4% feed chlorine 700 S00 Kiln ExR Temperature (*C) §!_ B» (frj Fig. 10. Metal partitioning to the kiln ash versus kiln exit temperature at waste feed chlorine concentrations of 0, 0.7, and 3.4 percent for the parametric test series using the Calvert Flux-Force/Condensation scrubber system. CONCLUSIONS Data from three sets of parametric trace metal partition- ing tests performed at the IRF show that metal partitioning among incinerator system discharges can largely be explained using only vapor pressure/temperature relationships for metal species in thermodynamic equilibrium. Given this, the data from the parametric tests performed to date suggest that bismuth behavior in an incineration process is quite similar to that of cadmium, magnesium or strontium behavior is similar to that of chromium, and copper behavior is similar to that of arsenic, barium, and lead. With respect to APCS collection efficiency, magnesium behavior is similar to that of barium, and bismuth and copper behavior is similar to that of arsenic, cadmium, and lead. Use of these nonhazardous constituent metals as surrogates for the corresponding hazardous constit- uent metals warrants consideration for use in scoping tests used to guide trial burn planning. Other potential surrogates can be similarly identified based on equilibrium vapor pres- sure/temperature calculations. 1 * Waste Feed Chlorine Content (%) Cd qi Pfr As to ' Fig. 11. Metal partitioning to the kiln ash versus waste feed chlorine content at constant kiln exit temperature of 538°, 816°, and 927°C (1,000,1,500, and l^OCF) for the parametric test series using the Calvert Flux- Force/Condensation scrubber system. REFERENCES 1. FOURN1ER, JR., D. J., W. E. WHITWORTH, J. W. LEE, andL. R. WATERLAND. "The Fate of Trace Metals in a Rotary Kiln Incinerator with a Venturi/Packed Column Scrubber." EPA/600/2-90/043. NTIS No. PB90-263864/AS and PB90-263872/AS. February 1991. 2. FOURNIER, JR., D. J., and L. R. WATERLAND. "The Fate of Trace Metals in a Rotary Kiln Incinerator with a Single-Stage Ionizing Wet Scrubber." EPA/600/2-91/032. NTIS No. PB91-223388 and PB91-223396. September 1991. 3. FOURNIER, JR., D. J., and L. R. WATERLAND. "The Fate of Trace Metals in a Rotary Kiln Incinerator with a Calvert Flux-Force/Condensation Scrubber System." Acurex Environmental draft report prepared under EPA Contract 68-C9-0038. January 1993. 4. BARTON, R. G., W. D. CLARK, and W. R. SEEKER. Tate of Metals in Combustion Systems." Combustion Sci- ence and Technology. Vol. 74., pp. 327-342,1990. ------- 440 Waterland METALS FOR TRIAL BURNS Parametric Teata, Ventu/^acked Column Scrubber 4.2% feed chJortne Parametric Teats, Venturl Packed Column Scrubber $71'C (160CF) kiln exit temperature I- Parametric Teeta, Ionizing Wet Scrubber 3.9% teed chlorine Parametric Teats, Ionizing Wet Scrubber 671*C (1600T) kiln exit temperature Parametric Teata. Catvert Scrubber 3.4% teed chlorine 2 4 6 Waste Feed Chlorine Content (%) Cd frl Pb Ba C^u Sr Mg Fig. 13. Apparent scrubber collection efficiency versus wastre feed chlorine content for the two scrubber systems. 600 700 600 90 Kiln ExR Temperature fC) Cd jl Pb Aa ttf (frj ^r Mg 1,000 Apparent scrubber collection efficiency versus kiln exit temperature for the three scrubber systems. ------- TECHNICAL REPORT DATA (Please reed Instructions on the reverie before completing) i 1 REPORT NO. 2. EPA/600/A-94/068 3. RE Ci Pi E N 4. TITLE AND SUBTITLE POTENTIAL SURROGATE METALS FOR INCINERATOR TRIAL BURNS 5. REPORT DATE 6. PERFORMING ORGANIZATION CODE 7. AUTMORIS) Larry P. Waterland and Donald J. Fournier, Jr. 8. PERFORMING ORGANIZATION REPORT NO. 9 PERFORMING ORGANIZATION NAME ANO AOORESS Acurex Environmental Corporation Incineration Research Facility Jefferson, Arkansas 72079 10. PROGRAM ELEMENT NO. It. CONTRACT/GRANT NO. 68-C9-0038 12. SPONSORING AGENCY NAME AND ADDRESS Risk Reduction Engineering Laboratory- Cincinnati, OH Office of Research and Development U.S. Environmental Protection Agency Cincinnati, Ohio 45268 13. TYPE OF REPORT AND PERIOD COVERED Published Paper 14. SPONSORING AGENCY CODE EPA/600/14 IS. supplementary notes project Officer = Howard Wall (513) 569-7691; Published in Proceedings of the 1993 Incineration Conference, Knoxville, TO 5/3-7/93, p:434~440 i6. ABSTRAgTgr ^ 4 years, the research program at the Environmental Protection Agency's (EPA s) Incinerator Research Facility (IRF) has developed an extensive body of metals partitioning data from pilot-scale incineration tests using synthetic hazardous wastes, actual listed hazardous wastes, and contaminated materials from Superfund sites. From these data, augmented by bench-scale studies and some full-scale incinerator tests, it has become apparent that the same metals volatilization/condensation mechanisms, first used to explain metals enrichment in flyash from coal combustion 15 years ago, drive metals partitioning in incinerators. Further metals partitioning can largely be explained using only vapor pressure/temperature relationships for metal species in thermodynamic equilibrium in the combustion zone. Because metals volatility dominates partitioning, surrogates can be used and the choice of surrogates is simplified. This paper discusses the results of three extensive parametric test programs performed at the IRF using synthetic hazardous wastes containing both hazarcous constituent and potential surrogate metals. These results show that surrogates partition in tne same manner as selected hazardous constituent metals. Thus, the use of surrogates deserves consideration, if not in actual trial burns, then at least in scoping tests used to guide the formal trial burn. 17. KEY WORDS AND DOCUMENT ANALYSIS a. DESCRIPTORS b.lDENTJFlERS/OPEN ENOEO TERMS c. cosati Field/Croup Incineration Hazardous Waste Trial Burn Incinerator Testing Heavv Metal Emissions 18. DISTRIBUTION STATEMENT RELEASE TO PUBLIC 19. SECURITY Class (This Report/ Unclassified 21. no. of pages 10 20. SECURITY CLASS (This pagej Unclassified 22. PRICE EPA Form 2520 — 1 (R«». 4—77) PREVIOUS EDlTiOK IS OBJOLETE ------- |