This paper was given at the 2003 AWMA/EPA Indoor Air Quality Problems and Engineering
Solutions Specialty Conference and Exhibition, July 21-23, 2003, in RTP, NC.

Observations on Engineering Aspects of the Brentwood
Postal Facility Fumigation

Prepared by: G. Blair Martin, Associate Director, APPCD, U.S. EPA (MC-343-4) Research
Triangle Park, NC 27711 and Frank T. Princiotta, Director, APPCD, U.S. EPA (MC-343-
4), Research Triangle Park, NC 27711

ABSTRACT

The 14,000,000 ft3 Brentwood postal facility in Washington, DC, was contaminated with anthrax
spores contained in a letter addressed to Senator Thomas Daschle. Two postal employees died of
inhalation anthrax. The U.S. Postal Service decided to fumigate the entire building with chlorine
dioxide (C102) to inactivate the remaining spores. This paper provides a summary of the four
tests of the C102 generation and scrubbing system at the building.

BACKGROUND

In October 2001, the Brentwood Postal Facility, located in northeast Washington, DC, was
contaminated with B. anthrasis spores, believed to have originated from a letter addressed to
Senator Daschle. Joseph Curseen, Jr. and Thomas Morris, Jr., two postal employees, contracted
and succumbed to inhalation anthrax. The facility has been renamed the Joseph Curseen, Jr. and
Thomas Morris, Jr. Processing and Distribution Center as a memorial. The facility is a
14,000,000 ft3 (about 700,000 ft2) warehouse-type building with a high ceiling and multiple
loading dock doors opening to the outside. The central portion of the space was generally open
and housed a large number of postal processing machines, including "Line 17" where the
contaminated letter appears to have been processed. The two ends of the building have offices
and other enclosed areas. The building was closed October 21, 2001, and the investigation and
decontamination process commenced.

DECONTAMINATION EVENTS

The U.S. Postal Service (USPS) conducted a comprehensive sampling campaign and found that
the anthrax spores were widely distributed throughout the building. The contamination was
found to be centered around Line 17. To prevent the spores from getting outside the building, all
openings were sealed with some combination of plywood, foam board, caulking, and tape. Initial
decontamination included enclosure and decontamination of certain machines, decontamination
and removal of some metal contents (e.g., mail racks), and spot decontamination. At the
completion of these efforts, spores were still found dispersed throughout the building. This led
to a plan to decontaminate the whole 14,000,000 ft3 volume at one time.


-------
WHOLE BUILDING DECONTAMINATION

The USPS decided to use chlorine dioxide (C102) for the whole building decontamination, based
in part on the successful operation at the Hart Senate Office Building. They engaged Shaw
Environmental to run the overall effort with Sabre and Ashland Chemicals as the subcontractor
for the C102 generation systems. The urban location required extensive safety precautions to
prevent leakage from the building resulting in unacceptable ambient C102 concentration in the
neighborhood. In addition, since C102 is not listed under the Federal Insecticide, Fungicide, and
Rodenticide Act (FIFRA) for anthrax fumigation, a crisis exemption from the EPA Office of
Pollution Prevention and Toxic Substances was required. This section will discuss the events
leading up to whole building fumigation: 1) the approach; 2) fumigation target conditions; 3)
safety precautions; 4) the Office of Research and Development (ORD) role; 5) the ORD
dispersion analysis; and 6) a series of four tests of the negative air unit (NAU) used to scrub C102
from air removed from the building.

Approach

The fumigation approach required three major engineering systems: 1) control of temperature
and humidity in the building; 2) equipment for generation of C102; and 3) NAU equipment to
ensure that C102 was contained within the building during fumigation. Note that many of the
details of this technology are considered confidential by the contractors. Therefore, only
engineering aspects observed during tests of systems 2 and 3 above are discussed.

Fumigation Target Conditions

The fumigations were planned based on the experience at the Hart Senate Office Building and
some previous testing of the effectiveness of C102 to kill anthrax spores. The temperature and
humidity were to be maintained at greater than 80 F° and 75% relative humidity, respectively.
The target fumigation conditions were a cumulative 9000 ppm-hr based on a nominal
concentration of 750 ppm for 12 hours. A concentration time (CT) clock record was kept for
multiple locations in the building, with the provision that no individual clock was started until
the C102 was greater than 400 ppm.

The C102 Generation and distribution consisted of two systems, a small one used during the
NAU tests, and a larger one used during the full building fumigation. Each system consisted of
reagent storage, a generator unit producing a solution of C102 in water, one or more emitters, and
a recycle tank for a residual C102 in water solution returned to the sump of the emitter(s). The
emitters stripped gaseous C102 from the water solution into a forced convection air stream and
collected the remaining solution for return to the recycle tank. For the NAU tests, a single
emitter was installed in the inlet duct from the building to the NAU. For the low level C102 test
and the full building fumigation, the larger system delivered the C102 solution to over 20 emitters
distributed around the building.


-------
Two similar capacity Negative Air Units were available. The North NAU was used for the
NAU tests, and both units were used for the low level C102 test and the full fumigation. Each
NAU drew air from the building through a series of cleanup steps including a HEPA filter to
remove any spores that might be entrained, a demister to remove any condensation due to the
high building humidity, a chemical scrubber to remove C102 to 5 ppb, and a carbon bed to
remove residual C102. The system capacity was also designed to provide sufficient air flow to
maintain the building pressure at least 0.05 inches of water below ambient. This should prevent
any significant exfiltration of C102 unless there was some failure of the sealing.

Safety Precautions

Extensive safety precautions were in place to ensure the safety of workers and people in the
surrounding area. USPS established an alert level of 25 ppb and an action level of 100 ppb at
any monitoring site. The alert level would trigger additional monitoring, and the action level
would cause a shutdown of the C102 generator system. C102 monitors were located on the roof
and around the generator systems and NAU's to provide early leak detection and to protect
worker safety. Fence line monitors continuously evaluated the ambient concentrations. EPA's
Trace Atmospheric Gas Analysis (TAGA) van and Shaw monitoring crews circulated through the
areas outside the fence line. In addition, building pressure was continuously monitored to
evaluate control of C102 exfiltration.

ORD Role

The ORD role began with a review of the Remedial Action Plan (RAP) prepared by the USPS
and its contractors, concentrating on engineering and safety aspects of the fumigation systems.
There was frequent consultation with representatives of the Office of Pollution Prevention and
Toxic Substances (OPPTS) and the Office of Solid Waste and Emergency Response (OSWER)
over the duration of the effort. A dispersion analysis was performed to evaluate the ambient
concentrations that might occur under a reasonable "worst case scenario." The key activity was
observation of all NAU tests leading up to the full fumigation.

ORD Dispersion Analysis

The ORD dispersion analysis was conducted to evaluate a credible worst case scenario for
release of fumigant due to the inability to maintain a sufficiently negative pressure differential
(i.e., - 0.05 inches of water) from ambient inside all areas of the building. (Note that pressure
differential measurements were made continuously at over 20 locations). This pressure
differential should ensure that the building is "under control" and minimal exfiltration will occur.
The assumption was made that this pressure differential would be maintained at the anticipated
flow rate through the NAU, and a building leakage coefficient was calculated by Air Pollution
Prevention and Control Division (APPCD) personnel. Assuming that the C102 concentration in
the building was 1000 ppm and that the building pressure differential rose to + 0.02 inches of
water, a source strength was calculated. The National Exposure Research Laboratory (NERL)
used this as a point source strength in a dispersion model analysis and calculated that the C102
leakage could result in an ambient C102 concentration of 1 ppm at 1 kilometer in the prevailing


-------
wind direction. As a result, the USPS increased efforts to ensure that the required pressure
differential would be demonstrated in the NAU trials.

NAU Tests

Prior to the issuance of the crisis exemption, the USPS conducted a series of four NAU tests to
demonstrate that the system was capable of achieving required destruction of C102 in the air
exhausted from the building to maintain negative pressure differentials. The first three tests were
conducted on the north NAU with 1000 ppm C102 introduced directly into the NAU duct. The
fourth test was conducted with both NAUs operational and a low-level of C102 (400 ppm)
introduced through the emitters into the whole building. Each test in the series is described
below.

First NA U Test

The goal of this test was to demonstrate that the NAU could run for 24 hours, which was the
estimated duration of the fumigation of the building and purging of the C102 at completion of the
fumigation. The NAU was to be operated at the design flow rate with 1000 ppm C102in the gas
stream. The planned amount of reagents to generate the C102 was exhausted in about 10 hours,
and the test was terminated. The reason for the failure was not known at the conclusion of this
test. The pressure differential data were quite limited, but indicated value of atmospheric
pressure plus or minus 0.01 inches of water.

Second NA V Test

The goal of this test was to successfully complete the 24 hours of continuous operation. After
the start of the test, rapid depletion of reagents was noted. The cause was traced to an incorrect
calibration curve being used for the NAU flow measurement instrument. The actual flow was
coincidently almost exactly twice the indicated value. The flow rate was corrected and the C102
scrubber portion of the NAU system worked as expected, achieving a less than 5 ppb C102 exit
concentration. A carbon bed completely removed this trace amount. At the conclusion of the
test, a scrubber failure was simulated to evaluate the capacity of the carbon bed to deal with the
full 1000 ppm in an emergency. The capacity was expected to provide a 2-to 4-hour time to
breakthrough; however, it occurred in about 18 minutes. In addition, the required negative
pressure differentials were not achieved.

Third NA U Test

The goal of this test was to evaluate an enhanced carbon bed and to demonstrate consistent
negative pressure. The carbon bed test was successful, and no breakthrough was observed. The
NAU system was tested in an automatic mode that regulated NAU flow to achieve the required
pressure differentials, which were achieved in all zones. (Note that in the previous two tests the
NAU flow rate was set and the pressure differentials were allowed to vary.)


-------
Final NAU Test: Full Building Fumigation Simulation

The goal of this test was to demonstrate the performance of both the North and South NAUs and
to ensure that the C102 was successfully contained within the building. The C102 level in the
building was ramped up to an equilibrium concentration of about 400 ppm and held for about 6
hours. Although there were minor C102 leaks from the building, which were quickly remedied,
and a mechanical delay, both NAUs performed well. The pressure sensors at all locations
indicated pressure differentials at or below negative 0.05 inches of water. At the conclusion of
C102 introduction, the decomposition rate in the building C102 inventory was observed to be
quite rapid.

Brentwood Fumigation

The full-scale fumigation was conducted in December 2002. The USPS stated that the
fumigation had been successfully carried out with all CT clocks exceeding 9000 ppm hours and
no C102 leakage detected. In March, the USPS also announced that all spore strips and wipe
samples were negative. The Environmental Compliance Committee had to give approval for
reentry.

Costs

While exact costs for all anthrax remediation activities have not been documented, it appears that
they will exceed $200 per ft2 of floor space. Following fumigation, the building is being restored
including painting and replacing the furniture and floor covering. No information has been
provided on potential damage to or replacements of the machinery in place during the
fumigation.

OBSERVATIONS

Based on the experience at Brentwood, the following observations result:

1.	During planning of a fumigation, a plausible worst-case scenario analysis needs to
be conducted. It should include all risks reasonably under control of the
performing organization. Among the things it should include: the likely sources
of leaks and methods to control them, the toxicity of the fumigant, the volume and
concentration of the fumigant present in the building, and the monitoring and
emergency shut down plan to ensure worker and neighborhood safety.

2.	Containment of the fumigant in the building is critical. Approaches can include
sealing potential leaks, maintaining sufficient fumigation pressure differentials,
testing or encapsulating the facility, providing a gas clean-up system for air
intentionally exhausted from the building, and having an adequate clearance
mechanism at the conclusion of the fumigation (probably a combination of
ventilation with scrubbing and in-place fumigant decomposition).


-------
3.	Careful testing of the monitoring and safety systems is essential.

4.	The experiences of Brentwood and other decontaminations need to be
systematically analyzed and documented. This analysis should provide guidance
to make any future fumigation safe and more cost effective.

5.	Be prepared for the unexpected, since it appears that there may be any number of
unanticipated problems that will arise during a fumigation. These preparations
should include but are not limited to: spares for critical systems components,
adequate supplies of reagents, emergency power provisions, and close
coordination with community and local agencies.

6.	A good QA plan for instrumentation is also critical.


-------
Key Words

Observations on Engineering Aspects of the Brentwood
Postal Facility Fumigation

Prepared by: G. Blair Martin, Associate Director, APPCD, U.S. EPA (MC-343-4) Research
Triangle Park, NC 27711 and Frank T. Princiotta, Director, APPCD, U.S. EPA (MC-343-
4), Research Triangle Park, NC 27711

Kev Words

Air Pollution

Biological Warfare Defense

Anthrax

Fumigation

Environmental Engineering
Chlorine Dioxide


-------