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).


-------