IT3'04 Conference, May 10-14, 2004, Phoenix, AZ IT3-011 DESTRUCTION EFFICIENCY OF MICROBIOLOGICAL ORGANISMS IN MEDICAL WASTE INCINERATORS: A REVIEW OF AVAILABLE DATA Joseph P. Wood, Paul Lemieux, Chun-Wai Lee Office of Research and Development, U.S. Environmental Protection Agency Research Triangle Park, NC 27711 USA ABSTRACT After a building has undergone a terrorist attack using a biological weapon such as B. Anthracis, many of the interior building materials will need to be disposed. Although it is likely that these materials will be decontaminated prior to their removal, officials may decide to remove the potentially bio-contaminated materials without first fumigating them. In either scenario, the possibility exists that some of the building materials will retain viable contaminating agent spores. Incineration may be the best option for the disposal of such building materials to completely destroy all potentially remaining bio-contaminants. In the early 1990s, the US Environmental Protection Agency (EPA) conducted microbial survivability tests at several medical waste incinerators (MWIs); these data have now been examined to evaluate microbiological destruction performance. Microorganisms were spiked into the waste feed and in test pipes, and subsequently analyzed for viability in the emissions, residue, and pipes using EPA conditional test methods. The results showed that for the most of the test runs, at least a five log reduction of the spores was achieved, although viable spores were detected in 10 out of a total of 48 air emission test runs, and spores were detected in 10 out of 27 available ash samples. INTRODUCTION In September, 2001, B. Anthracis spores were sent through the US Postal Service to various locations in Florida, New Jersey, New York, and Washington, D.C. Twenty-two cases of anthrax infection (or suspected infection) resulted in five deaths (1). In the Washington D.C. area, the interior of the buildings where spores were discovered were decontaminated with chlorine dioxide and other techniques (2). Although it is likely that these interior materials were completely decontaminated, the possibility exists that trace amounts of B. Anthracis spores remained viable. In addition, some interior building materials exposed to B. Anthracis spores may be removed without first being fumigated, and therefore could have a relatively high level of spore contamination. The majority of the resulting decontamination waste (debris, solid waste, and personal protective equipment) from the Capitol area was disposed of in the MWIs located at Fort Detrick's U.S. Army Medical Research Institute of Infectious Diseases in Maryland. Officials decided to use the MWIs because B. Anthracis-contaminated materials were classified as medical waste in Washington, D.C. and Maryland. The classification of the waste (solid waste, hazardous waste, or medical waste) determines what disposal options are available, and can vary from state to state. From a technical standpoint, the MWIs were selected because they employed a batch process with higher gas temperatures and longer residence times than a typical continuous feed MWI (2). Based on the knowledge gained from the Capitol area waste disposal operation, incineration will likely continue to be the method of choice for disposal of at least some of the waste materials in any future bio-terrorist events. There are no federal standards to ensure the complete destruction of microorganisms when disposing of medical waste in incinerators. Similarly, there are minimal data and literature available on the thermal destruction of microorganisms in an incinerator. This lack of standards and data is due to the conventional wisdom that all microorganisms should be destroyed in the high-temperature environment of an incinerator. However, for various technical reasons, it is conceivable that some of the ------- IT3'04 Conference, May 10-14, 2004, Phoenix, AZ IT-011 microorganisms in the waste feed may not be completely destroyed and may be emitted out the exhaust stack or remain viable in the residue (3). In one paper, Allen, R. J., et al. (4) reported that incineration of a surrogate waste in a hospital incinerator spiked with Bacillus subtilis resulted in no viable Bacillus subtilis in the stack-gas samples. However, viable bacteria other than Bacillus subtilis were found in stack gas samples, and it was suggested that these bacteria entered the incinerator via the indoor air inlet of the secondary combustion chamber. The vast majority of the microbial destruction data available comes as a result of emissions tests conducted on full-scale MWIs by the EPA Office of Air Quality Planning and Standards and the EPA Office of Solid Waste in the early 1990s (5 - 12). These tests were conducted primarily to collect both criteria and hazardous air pollutant data in support of developing emission standards and to comply with the Medical Waste Tracking Act of 1988. However, the microbial survivability data from these EPA test reports have never been summarized and published in a public document or scientific paper. In light of the need to ensure complete destruction of B. Anthracis spores in incinerators, these data have now been reviewed to evaluate incinerator microbiological destruction performance and are presented here. DESCRIPTION OF FACILITY INFORMATION AND DATA The emission test reports were thoroughly reviewed and all microbial survivability data, incinerator operational data, test methods, quality control procedures and results, and other related information were extracted and logged into a spreadsheet. The data reported in the following tables are as they were reported in the test reports, except for the conversion of some of the data to SI units, and the reporting of some average levels. Table I is a summary of the MWI facilities that were tested. Table II is a summary of the MWI operational data for the time period during testing or for that day of testing. The test reports included detailed narrative descriptions of the process operation during the test as well as the operational data. ------- IT3'04 Conference, May 10-14, 2004, Phoenix, AZ IT-011 Table I Summary of Facility Information Facility Location Date of test Air pollution control Incinerator type/description Capacity 1 Sanford, NC Sep-90 none Controlled air; Manual charge, batch operation. Gas-fired primary and secondary chambers. 79 kg/hr pathological (path.) waste; 113 kg/hr med waste 2 Wilmington, NC Aug-90 none Joy Energy Systems, ram fed. Primary comb, chamber 3.85 m3 starved air, NG fired; secondary chamber 2.4 m3 w/ 1 sec gas res. time. 147 kg/hr of waste at 19,796 kJ/kg. Both red bag (infectious) and non-infect. burned. 3 Plymouth, MA Mar-91 heat exchanger and spray cooling, followed by baghouse, then horizontal cross-flow packed bed absorber using NaOH Simonds batch burn manual charge; Starved air. Primary chamber natural gas (NG) fired, 6 m3. Secondary chamber gas fired, 1.16 sec res. time at 1600 F 340 kg/batch Type 0-4 (red bag, including path.) waste. 4 Kinston, NC Mar-91 None Joy Energy ram feeder; Starved air. Primary chamber NG fired, 3.85 m3. Secondary chamber gas fired, 0.4 sec design res. time 145 kg/hr (19,796 kJ/kg) 5 Kalamazoo, MI Apr-90 baghouse w/ dry lime injection. 3 chamber, starved air. 2nd chamber vol. 1.1m3 (res. time = 0.3 s), with burner. 3rd chamber vol. 2.8 m3 (res time = 1.24 s) with no burner. With waste heat boiler ~ 295 kg/hr 6 Ann Arbor, MI Jun-90 variable throat venturi wet scrubber followed by two packed bed absorbers (in series; caustic solution) 3 chamber, continuous feed, pulsed hearth, w/ waste heat boiler. 1st chamber vol = 25.4 m3. 2nd and 3rd chamber vol combined = 16.5 m3, ~ 408 kg/hr 7 Morristown, NJ Nov-91 lime spray dryer, fabric filter. ThermAll NG-fired rotary kiln (8 m long, 2.7 m dia), automatic ram feed, with after burner (18.7 m3, 2 sec gas res. time) and waste heat recovery boiler 454 kg/hr; only permitted for 363 kg/hr ------- IT3'04 Conference, May 10-14, 2004, Phoenix, AZ IT-011 Table II Summary of MWI O )erational Parameters During Testing Test run Date Waste type Avg. Feed rate (kg/hr) Total feed charged (kg/day) Ash (kg/day) Avg. primary chamber temp (C) Avg. second, chamber temp (C) Avg. gas res. time in second chamber (s) Avg. exhaust flow rate (actual m3/min) Avg. flue gas temp (C) Avg. stack 02 level (% vol. Dry) Facilitv 1 1 9/20/90 Datholoaical 48 229 10 770 807 0.12 69 400 15.8 2 9/21 /90 aeneral 74 462 63 758 946 0.11 64 432 15.7 3 9/22/90 Datholoaical 54 267 10 815 815 0.13 62 404 16.1 4 9/23/90 aeneral 70 411 78 846 858 0.11 67 415 15.2 5 9/24/90 Datholoaical 77 307 20 784 805 0.12 69 397 15.6 6 9/25/90 Datholoaical 65 317 19 772 794 0.12 63 385 15.7 8 9/27/90 aeneral 73 330 62 792 922 0.11 67 408 15.6 9 9/28/90 Datholoaical 76 394 35 785 828 0.11 72 411 15.3 10 10/2/90 Datholoaical 45 206 8 785 801 0.12 64 398 15.7 Facilitv 2 1 8/15/90 mixed* 80 558 24 1029 1104 0.92 116 758 8.5 2 8/18/90 mixed* 76 908 NA 1087 1142 1.05 100 773 7.2 3 8/19/90 mixed* 96 704 55 1014 1115 0.98 113 808 9.4 4 8/20/90 mixed* 121 593 96 863 1070 0.79 142 783 7.7 5 8/21/90 mixed* 88 412 63 877 1001 0.81 145 775 9.2 6 8/22/90 mixed* 87 530 42 854 987 0.83 140 758 9.8 7 8/26/90 mixed* 135 612 54 912 892 1.05 110 672 11.8 8 8/27/90 mixed* 132 567 50 892 909 1.10 105 687 9.3 9 8/28/90 mixed* 138 628 47 904 913 1.14 102 701 9 * both infectious and non-infectious wast 0 Facilitv 3 1 3/5/91 red baa NA 327 32 345 976 1.45 78 633 8.8 2 3/5/19 red baa NA 647 1019 1.52 69 594 8.8 3 3/7/19 red baa NA 319 43 316 976 1.67 69 653 9.6 4 3/7/19 red baa NA 652 1008 1.91 57 622 10.3 5 3/9/19 red baa NA 331 27 338 982 1.77 63 623 11.2 6 3/9/19 red baa NA 647 1050 2.16 51 663 8.6 tests 1. 3. and 5 were conducted durina "burn" conditions (Drimarv chamber at low flow air. sec chamb maintains setDoint temDl tests 2. 4. and 6 were conducted durina 'burndown " conditons (¦follows b jrn conditio n on same dav: 1 5/30/90 mixed 100 NA 129 973 964 0.32 125 698 10.9 2 5/31 /90 mixed 114 NA 81 1049 999 0.31 127 727 10.1 3 6/1/90 mixed 134 NA 68 964 973 0.30 136 727 14.3 4 6/2/90 mixed 87 NA 59 818 880 NA NA 11.4 4R 6/4/90 mixed 86 NA 85 983 966 0.33 121 694 14.1 6 6/5/90 mixed 86 NA 66 853 904 0.36 114 679 12 7 6/6/90 mixed 121 NA 111 1005 916 0.36 116 711 12.9 5R 6/6/90 mixed 89 NA 985 973 0.32 126 711 13.9 8 6/7/90 mixed 135 NA 66 1010 894 0.35 119 687 14.6 9 6/8/90 mixed 130 NA 86 993 903 0.36 117 698 13.1 Note: Run 5 invs Idated and th srefore rec eated CR1 Run 4 reD eated beca use the PM /metals test nvalid Facilitv 5 MB3-1 5/5/90 G500 216 NA NA 706 877 901* 37.1 179 11.7 MB3-2 5/7/90 G500 192 NA NA 703 887 896* 39.8 181 11.8 MB3-3 5/8/90 G500 183 NA NA 716 879 889* 38.2 178 12 MB1-1 5/23/90 G500 238 NA NA 688 1016 1005* 45.8 178 13 MB1-2 5/24/90 G500 223 NA NA 707 996 989* 41.0 174 12.1 MB1-3 5/31 /90 G500 250 NA NA 693 1019 1001* 41.8 186 13.1 EB1-1 5/17/90 G500 258 NA NA 700 1125 1101* 43.0 177 12 EB1-2 5/18/90 G500 269 NA NA 693 1128 1104* 43.6 185 12.1 EB1-3 5/21 /90 G500 283 NA NA 689 1131 1096* 45.0 179 11.2 * Tertiarv chamb er temDeratur e (C) Facilitv 6 1 6/23/90 aeneral 358 NA NA 941 1116 946* NA NA NA 2 6/25/90 aeneral test was aborted 3 6/25/90 aeneral 395 NA NA 943 1130 983* NA NA NA 4 6/26/90 aeneral 395 NA NA 996 1129 982* NA NA NA 5 6/28/90 aeneral 384 NA NA 906 1144 984* NA NA NA Note: flow rate and 02 level not available for indivi dual runs, t ut averaae for all runs as follows: fl ow rate =126.5 dscmm * Tertiarv chamber temDeratur e (C) Facilitv 7 1 11/18/91 mixed 329 NA NA 802 981 NA 99.5* 212 11.1 2 11/19/91 mixed 334 NA NA 792 976 NA 96.7* 211 11.1 3 11/20/91 mixed 359 NA NA 784 975 NA 102* 210 10.6 * flow rate drv standard NA= not avail ------- IT3'04 Conference, May 10-14, 2004, Phoenix, AZ IT-011 RESULTS AND DISCUSSION Table III is a summary of the microbial survivability data for the ash and emissions samples. The test methods that were used are described in detail elsewhere (13, 14), and are summarized as follows. In a typical test, a known number of spores - B. Stearothermophilus, a heat resistant microorganism used as a worst-case surrogate bacterium - was spiked with the waste feed at certain intervals throughout the emissions test. Emission samples were collected isokinetically following the secondary combustion chamber, i.e., prior to any air pollution control device, in a buffered solution in impingers. Ash samples (for most of the tests) were taken the day after each test. The recovered samples were cultured and colonies of B. stearothermophilus were identified and quantified. Table III shows that viable spores were detected in the flue gas in 10 of 48 air emission test runs, and in 10 of 27 available ash samples. For facilities 1 to 4, where both ash and air samples were taken and detection limits presented, the log reduction in spores ranged from 1.8 to greater than 8.2. As discussed by Lemieux, et al. (3), there are several reasons that MWIs may not completely destroy all of the spiked microorganisms, including in-bed mass transfer limitations, incomplete bed mixing, bypassing of hot zones due to poor gas phase mixing, dropping through the grate prior to destruction in the bed, or by coming into contact with cool zones within the MWI. Coupled with complex fluid dynamics, these limitations would cause pockets within the combustion chambers that are not exposed to sufficiently high temperatures and residence times. Another explanation is that all of the spiked spores were destroyed but the spores detected were another bacterium. Although the method used to collect the data presented in this paper includes analytical steps to generally assure that the colonies are B. stearothermophilus, the method does not identify each and every colony. Additional contamination may have occurred, such as in the handling of the sampling train, as indicated by the quality control program that EPA implemented during these tests. Out of approximately 30 field blanks (impingers) taken during the testing at 5 out of the 7 facilities, spores were detected in 3. Some of the data may be false negatives or may be biased high or low, depending on how "non-detects" are handled and reported. Although no spores were detected in the flue gas for the majority of the test runs, this may be due to the high detection limit (~ 10,000 total spores), which is based on 1 spore per sample aliquot, aliquot volume, and gas sample volume. It is conceivable that spores were present in the stack gas but went undetected due to these high detection limits. It should be noted that if spores were indeed completely destroyed, but the reported amount is based on the detection limit, this would be a source of high bias. Finally, the data are based on the questionable assumption that no spores are found in the flue gas once sampling is completed. It may be possible that some of the spiked spores surviving the incinerator environment may not become entrained in the exhaust gas until after flue gas sampling has been completed. Table IV shows the results for the microorganism survivability pipe tests. The pipe tests were performed as another technique for assessing microorganism survivability in the MWI residue. These pipes were charged with between 105 to 10s spores. Refer to EPA Conditional Test Method 25 for further details regarding this test procedure (13, 14). Note that out of 163 total pipes that were recovered from the MWIs, there were 59 pipes with spores that were detected or too numerous to count. In one test where internal pipe temperatures were measured, there were several pipes with spores surviving internal pipe temperatures above 816°C (7). ------- IT3'04 Conference, May 10-14, 2004, Phoenix, AZ IT-011 "able II Summary of Test Results for Microbial Survivability in Emissions and Ash Facility test run date total # spores spiked with waste feed total # spores emitted in flue gas total # spores in ash log reduction spore survivability (%) 1 1 9/20/1990 6.00E+12 ND < 3.8E4 ND <9.6E5 >6.8 <1.7E-5 1 2 9/21 /1990 6.00E+12 ND <4.4E4 ND <6.3E6 >6.0 <1.1 E-5 1 3 9/22/1990 6.00E+12 1 38E+07 ND <1.0E6 5.7 2.30E-04 1 4 9/23/1990 6.00E+12 ND <4.31 E4 ND <7.8E6 >5.9 <1.3E-4 1 5 9/24/1990 6.00E+12 <13.5E6 ND <1.0E4 >5.6 <2.3E-4 1 6 9/25/1990 6.00E+12 1 85E+06 ND <1 9E6 6.5 3.10E-05 1 8 9/27/1990 6.00E+12 >15.5E6 ND <6.2E6 <5.6 >2.6E-4 1 9 9/28/1990 6.00E+12 6.94E+06 ND <1.8E4 5.9 1 20E-04 1 10 10/2/1990 6.00E+12 ND <3.22E4 ND <3.9E3 >8.2 <6.0E-7 2 1 8/15/1990 1.54E+12 ND <2.57E4 5.20E+07 4.5 3.30E-03 2 2 8/18/1990 1.54E+12 1 28E+05 NA 7.1 8.30E-06 2 3 8/19/1990 1.54E+12 ND <5.2E4 ND <2.7E4 >7.3 ND <5.1 E-6 2 4 8/20/1990 1.54E+12 1 05E+05 ND <4.8E4 7.2 6.80E-06 2 5 8/21 /1990 1.54E+12 ND <4.6E4 ND <3.1 E4 >7.3 ND <5E-6 2 6 8/22/1990 1.54E+12 4.64E+04 <4.2E6 7.5 3.00E-04 2 7 8/26/1990 1.54E+12 ND <3.8E4 ND <2.7E4 >7.4 ND <4.2E-6 2 8 8/27/1990 1.54E+12 ND <5.4E4 NA NA NA 2 9 8/28/1990 1.54E+12 ND <4.5E4 ND <2.3E4 >7.4 ND <4.5E-6 3 1 3/5/1991 <1.0E8 3 2 3/5/1991 <8.4E8 3 Total 1.71 E+12 <9.4E8 ND <4.6E6 >3.2 <5.5E-2 3 3 3/7/1991 1.10E+05 3 4 3/7/1991 ND <1.2E5 3 Total 1.71 E+12 1.10E+05I CO.6 to 7.71E7 5.3 4.60E-04 3 5 3/9/1991 ND <1.0E5 3 6 3/9/1991 ND <1 52E5 Total 1.58E+12 ND <2.5E5 <2.8E10 >1.8 <1.8 4 1 5/30/1990 2.80E+12 ND <5.2E4 3.00E+09 3 0.11 4 2 5/31 /1990 2.80E+12 ND <6.4E4 >1 6E9 <3.2 >5.8E-2 4 3 6/1 /1990 2.80E+12 ND <3.7E4 ND <3.4E4 >7.6 <2.6E-6 4 4 6/2/1990 2.80E+12 ND <5.5E4 <1 2E9 >3.4 <4.2E-2 4 5 6/4/1990 2.80E+12 ND <5.7E4 >1.7E9 <3.2 >6.1 E-2 4 6 6/5/1990 2.80E+12 ND <6.4E4 >1 3E9 <3.3 >4.7E-2 4 7 6/6/1990 5.60E+12 ND <6.5E4 <2.2E9 >3.4 <7.9E-2 4 8 6/7/1990 2.80E+12 ND <5.5E4 <1 3E9 >3.3 <4.7E-2 4 9 6/8/1990 2.80E+12 ND <6.7E5 ND >6 6 <2.6E-5 data are presented as presented in test report, although some of the test runs listed do not match with correct date, as listed in the process description; Run 7 has double the spores since run 5 was repeated 5 MB3 5/5 to 5/8/90 NA ND 0.36 NA ~ 8 not determined 5 MB1 5/23 - 5/31 NA ND 0.33 NA ~ 8 not determined 5 EB1 5/17- 5/21 NA ND 0.31 NA ~ 8 not determined Emissions results listed above are based on the averaae of 3 test runs since no SDores were detected. SDoreswere charaed via baas at a rate of 718 billion/hr. Numbers of SDores emitted are Der hour.based on detection limits of 1 SDore Der samDle. No ash samDles were taken. 6 1 6/23/1990 8.50E+11 5.30E+04 not determined 7.20E+00 not determined 6 4 6/26/1990 8.50E+11 ND 9.0 E3 not determined > 8 not determined 6 5 6/28/1990 8.50E+11 ND 2.7 E4 not determined >7.5 not determined Number SDores charaed and SDores emitted are reDorted on a Der hour basis not for whole test. 7 1 11/18/1991 1.40E+12 0* 0 not determ. not determined 7 2 11/19/1991 1.40E+12 0* 0 not determ. not determined 7 3 11/20/1991 1.40E+12 0* 0 not determ. not determined * Result as reDorted in test reDort.. No detection I mit was Dresented in test reDort I Overall notes Number of SDores SDiked is the total number SDiked durina the test run: SDores were SDiked tvDicallv 3-4 times Der test loa reduction = loa CsDiked SDoresI - loa CsDores from emissions + SDores from ashl. NA = not available, not obtained Number of SDores exitina stack= SDores/dscm ¦ stack flow rate ¦ total samDlina time. Numbers indicated as "ND <" are for samDles where no SDores were detected but calculated based on the detection limit. Numbers indicated as "<" are for samDles where SDores were detected, but determined to be less than a certain value Numbers of SDores indicated as ">" are samDles w/ SDores too numerous to count, and assianed a value of 200 SDores/filter ------- IT3'04 Conference, May 10-14, 2004, Phoenix, AZ IT-011 Table IV Summary of Microorganism Survivability Pipe Test Results Facility Number pipes Number pipes Number of pipes Number pipes with Number pipes with charged recovered with no detectable spores detectable spores (range of spores detected) spores TNTC 1 54 54 26 25(1-166) 3 2 27 25 11 12(1-53) 2 3 27 27 21 6(1-8) 0 4 27 27 23 1(1) 3 5 9 9 5 4(3-15) 0 6 8 8 5 0 3 7 27 13 13 0 0 TNTC = number of colonies were too numerous to count due to insufficient dilution STATISTICAL ANALYSIS Data from the six facilities were combined into a single dataset reflecting the log reduction of spores, the total number of spores, primary and secondary combustion chamber temperatures, gas-phase residence time, and stack oxygen (02) levels. This combined dataset is shown in Table V. Table V Data Used in Statistical Analysis Facility Run Log # Spores Primary Secondary Secondary Stack 02 (%) Reduction Exiting Stack Chamber T (°C) Chamber T (°C) Res. Time (s) 1 1 6.8 3.80E+04 770 807 0.12 15.8 1 2 6.0 4.40E+04 758 946 0.11 15.7 1 3 5.7 1.38E+07 815 815 0.13 16.1 1 4 5.9 4.31E+04 846 858 0.11 15.2 1 5 5.6 1.35E+07 784 805 0.12 15.6 1 6 6.5 1.85E+06 772 794 0.12 15.7 1 8 5.6 1.55E+07 792 922 0.11 15.6 1 9 5.9 6.94E+06 785 828 0.11 15.3 1 10 8.2 3.22E+04 785 801 0.12 15.7 2 1 4.5 2.57E+04 1029 1104 0.92 8.5 2 2 7.1 1.28E+05 1087 1142 1.05 7.2 2 3 7.3 5.20E+04 1014 1115 0.98 9.4 2 4 7.2 1.05E+05 863 1070 0.79 7.7 2 5 7.3 4.60E+04 877 1001 0.81 9.2 2 6 7.5 4.64E+04 854 987 0.83 9.8 2 7 7.4 3.80E+04 912 892 1.05 11.8 2 8 - 5.40E+04 892 909 1.10 9.3 2 9 7.4 4.50E+04 904 913 1.14 9.0 3 1 3.2 9.40E+08 345 997 1.49 00 00 3 2 5.3 1.10E+05 316 992 1.79 9.9 3 3 1.8 2.50E+05 338 1016 1.97 9.9 4 1 3.0 5.20E+04 973 964 0.32 10.9 4 2 3.2 6.40E+04 1049 999 0.31 10.1 4 3 7.6 3.70E+04 964 973 0.30 14.3 4 4 3.4 5.50E+04 983 966 0.33 14.1 4 5 3.2 5.70E+04 985 973 0.32 13.9 4 6 3.3 6.40E+04 853 904 0.36 12.0 4 7 3.4 6.50E+04 1005 916 0.36 12.9 4 8 3.3 5.50E+04 1010 894 0.35 14.6 4 9 6.6 6.70E+05 993 903 0.36 13.1 6 1 7.2 5.30E+04 941 1116 - - 6 4 8.0 9.00E+03 996 1129 - - 6 5 7.5 2.70E+04 906 1144 - - The data were imported into the SAS JMP software and a stepwise regression was performed to examine whether either the variability of the log reduction or the number of spores exiting the stack could be accounted for by either the temperatures, residence time, or oxygen concentrations, both singly and as ------- IT3'04 Conference, May 10-14, 2004, Phoenix, AZ IT-011 cross products. A statistically significant relationship was found where the log reduction could be accounted for with an R2=0.51 with a 2 parameter model using the primary chamber temperature, the residence time, and their cross product. Figure 1 shows the resultant leverage plots from the JMP model fit. The leverage plots illustrate the effect of a single parameter while holding all other parameters constant. If the 95% confidence intervals (the dotted lines) cross the horizontal line then the parameter is a statistically significant predictor. The slope of the leverage line shows the effect that parameter imposes on the predicted variable. From Figure 1, it shows that when taken alone, residence time (RT, in seconds) shows a positive correlation with log reduction, which is consistent with the hypothesis that increased residence times would result in better spore destruction. The primary chamber temperature (PRIMARYT, °C), when examined alone, has a slightly negative correlation with log reduction, which is counter-intuitive since it would be expected that higher temperatures promote better spore destruction. This seeming inconsistency, however, is resolved when examining the leverage plot of the cross product of temperature and residence times (PRIMARY T • RT), which shows a positive correlation with spore destruction. The cross product term also showed much narrower 95% confidence interval bands and a much lower "P-value", suggesting that it is a more significant predictor of spore destruction. A P-value less than 0.05 indicates a statistically significant predictor. This suggests that the facilities with the longer secondary combustion chamber residence times and a higher primary combustion chamber temperature yielded the highest degree of spore destruction. It must be noted however, that the data showed a significant amount of variability in QA that was achieved, and that many measurements were driven by detection limits. ------- IT3'04 Conference, May 10-14, 2004, Phoenix, AZ IT-011 i 1 1 ' r~ .5 1.0 1.5 RT Leverage, P=0.0124 o 4- i—i—i—i—i—i—i—r .0 3.S 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 PRIMARY_T*RT Leverage, P<,0001 4- o 11 1 1 1 1 r 300 400 500 600 700 800 900 1000 PRIMARY_T Leverage, P=0.0043 4- 3- 2- 1 Fig. 1 Leverage Plots from Statistical Analysis (R2=0.51; overall P Value=0.Q004) " ------- IT3'04 Conference, May 10-14, 2004, Phoenix, AZ IT-011 CONCLUSIONS Microorganisms were spiked with the medical waste feed and for the majority of the test runs at the MWIs tested by EPA, at least a five log reduction in spores was achieved. However, viable spores were detected in the flue gas at the outlet of the secondary combustion chamber and in the incinerator residue for several test runs. That some spores were not destroyed may be due to technical limitations of the MWIs, or to artifacts of the test procedures. There are several potential sources of bias in the data (i.e., the number of spores emitted), including high bias and false positives due to field contamination (as indicated by some of the false positives detected in the field blanks) and low bias due to a high detection limit. The data presented here may not be totally representative of an actual scenario in which building materials potentially contaminated with B. Anthracis spores are incinerated in an MWI. The tests were conducted over 10 years ago on relatively small hospital MWIs with types of combustion and air pollution control equipment and designs that may not be in use today. For example, due to the cost of complying with air emission standards and guidance developed in the 1990s, medical waste disposal has shifted from small hospital MWIs to larger commercial MWIs with state-of-the-art incinerator and air pollution control technology. Further, these tests were worst-case scenarios with billions of spores charged to the MWIs; in an actual disposal scenario, we would expect that very few target microorganisms would be viable in the waste from a decontaminated building. (The exception to this is when potentially bio-contaminated interior building materials are removed without first being fumigated). These data may also be considered somewhat non-representative since these tests used the heat resistant B. Stearothermophilus bacterium and sampling occurred prior to the air pollution control device. (It is hypothesized that an air pollution control system would further contribute to the overall destruction of viable spores in the flue gas stream). Nevertheless, due to the real possibility that bacterial spores may escape destruction in a MWI, it may be prudent to carefully select facilities so as to minimize potential escape of microorganisms into the environment. Further research and development are needed to minimize analytical biases and lower the detection limit for the tests methods used to measure viable microorganisms in incinerator exhaust gases and ash. ACKNOWLEDGEMENTS We would like to acknowledge the personnel from EPA's Office of Air Quality Planning and Standards and EPA's Office of Solid Waste who were involved with the MWI emissions testing and subsequent development of the test reports referenced in this document. REFERENCES 1. Inglesby, T. V.; O'Toole, T.; Henderson, D.A.; Bartlett, J. G.; Ascher, M. S.; Eitzen E.; Friedlander, A. M.; Gerberding, J.; Hauer, J.; Hughes, J., et al. "Anthrax as a Biological Weapon, 2002". Journal of the American Medical Association. Vol. 288, No. 17. May 1, 2002: p. 2236. 2. US EPA Region 3, Philadelphia, PA. Federal On-Scene Coordinator's Report For the Capitol Hill Site, Washington, D.C. (no publication date listed). 3. Lemieux, P.M.; Lee, C. W.; Linak, W. P.; Miller, C. A.; Ryan, J. V.; Serre, S. D.; Stewart, E.S. "Thermal Incineration and Homeland Security." IT3 03 Conference, Orlando, Florida, May 12-16, 2003. ------- IT3'04 Conference, May 10-14, 2004, Phoenix, AZ IT-011 4. Allen, R. J.; Brenniman, G. R.; Lougue, R. R.; Strand, V. A. "Emission of Airborne Bacteria from a Hospital Incinerator." Journal of the Air Pollution Control Association, Vol. 39, No. 2 1989: 164- 168. 5. US EPA Office of Solid Waste. Medical Waste Incineration Emission Test Report for Lenoir Memorial Hospital, Kinston, NC, Volume I-III. EMB Report 90-MWI-3. EPA Office of Air and Radiation Docket A-91-61, Item II-A-81, Washington, D.C. May 1990. 6. US EPA Office of Solid Waste. Medical Waste Incineration Emission Test Report for Cape Fear Memorial Hospital, Wilmington, NC; Volume I-III. EMB Report 90-MWI-4. US EPA Office of Air and Radiation Docket A-91-61, Item II-A- 85, Washington, D.C. November 1990. 7. US EPA Office of Solid Waste. Medical Waste Incineration Emission Test Report for AMI Central Carolina Hospital, Sanford, NC; Volume I-III. EMB Report 90-MWI-5. EPA Office of Air and Radiation Docket A-91-61, Item II-A-86, Washington, D.C. December 1990. 8. US EPA Office of Air Quality Planning and Standards. Medical Waste Incineration Emission Test Report for Jordan Hospital, Plymouth, MA; Volume I-III. EMB Report 90-MWI-6. EPA Office of Air and Radiation Docket A-91-61, Item II-A-87, Washington, D.C. February 1991. 9. US EPA Office of Air Quality Planning and Standards. Michigan Hospital Incinerator Emissions Test Program; Volume II and Appendices A-M: Site Summary Report, Borgess Medical Center Incinerator. EPA Office of Air and Radiation Docket A-91-61, Item II-A-89, Washington, D.C. August 13, 1991. 10. US EPA Office of Air Quality Planning and Standards. Michigan Hospital Incinerator Emissions Test Program; Volume III and Appendices A-K: Site Summary Report, University of Michigan Medical Center Incinerator. EPA Office of Air and Radiation Docket A-91-61, Item II-A-90, Washington, D.C. August 13, 1991. 11. US EPA Office of Air Quality Planning and Standards. Michigan Hospital Incinerator Emissions Test Program; Volume I: Project Summary Report. EPA Office of Air and Radiation Docket A-91-61, Item II-A-91, Washington, D.C. November 1, 1991. 12. US EPA Office of Air Quality Planning and Standards. Medical Waste Incineration Emission Test Report, Morristown Memorial Hospital, Morristown, NJ; Volumes I and II. EPA Office of Air and Radiation Docket A-91-61, Item II-A-94, Washington, D.C. December 1991. 13. Segall, R.R.; Blanschan, GC; DeWees, W. G.; Hendry, K.M.; Leese, K. E.; Williams, L. G.; Curtis, F.; Shigara, R.T.; and Romesburg, L.J.. "Development and Evaluation of a Method to Determine Indicator Microorganisms in Air Emissions and Residue from Medical Waste Incinerators." Journal of the Air and Waste Management Association, Vol. 41, No. 11, November 1991: pp. 1454-1460. 14. U.S. EPA, Emissions Measurement Center, Research Triangle Park, NC. Conditional Test Method- 25 and CTM-026. World Wide Web, http://www.epa.gov/ttn/emc/ctm.html (accessed February 2004). ------- |