EPA/600/R-10/035 | August 2010 | www.epa.gov/ord
United States
Environmental Protection
Agency
Assessment of Fumigants for
Decontamination of Surfaces
Contaminated with Chemical
Warfare Agents
Office of Research and Development
National Homeland Security Research Center
-------
-------
Assessment of Fumigants for
Decontamination of Surfaces
Contaminated with Chemical
Warfare Agents
NATIONAL HOMELAND SECURITY RESEARCH
CENTER
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NC 27711
Office of Research and Development
National Homeland Security Research Center
-------
-------
Disclaimer
The United States Environmental Protection Agency, through its Office of Research and Devel-
opment's National Homeland Security Research Center, funded and managed this investigation
through U.S. EPA STREAMS contract (Contract Number EP-C-05-059) with work performed by
the Eastern Regional Group, Inc. This report has been peer and administratively reviewed and
has been approved for publication as an Environmental Protection Agency document. It does not
necessarily reflect the views of the Environmental Protection Agency. No official endorsement
should be inferred. This report includes photographs of commercially available products. The
photographs are included for purposes of illustration only and are not intended to imply that EPA
approves or endorses the product or its manufacturer. Environmental Protection Agency does not
endorse the purchase or sale of any commercial products or services.
Questions concerning this document or its application should be addressed to:
Emily Snyder, Ph.D.
National Homeland Security Research Center
Office of Research and Development (E-343-06)
U.S. Environmental Protection Agency
109 T.W. Alexander Dr.
Research Triangle Park, NC 27711
(919)541-1006
snYder.emilv(@,epa.gov
If you have difficulty accessing this PDF document, please contact Kathy Nickel
(Nickel.Kathy@.epa.gov) or Amelia McCall (McCall. Amelia(@.epa. gov) for assistance.
-------
Foreword
Following the events of September 11, 2001, EPA's mission was expanded to address critical needs
related to homeland security. Presidential Directives identify EPA as the primary federal agency
responsible for the country's water supplies and for decontamination following a chemical, biologi-
cal, and/or radiological (CBR) attack.
As part of this expanded mission, the National Homeland Security Research Center (NHSRC) was
established to conduct research and deliver products that improve the capability of the Agency
to carry out its homeland security responsibilities. One specific focus area of our research is on
decontamination methods and technologies that can be used in the recovery efforts resulting from a
CBR contamination event. In recovering from an event and decontaminating the area, it is critical
to identify and implement appropriate decontamination technologies. The selection and optimal
operation of an appropriate technology depends on many factors including the type of contaminant
and associated building materials, temperature, relative humidity, fumigant concentration,
fumigation time, and others. This document provides information on how two fumigant-based
technologies performed in treatment of CWAs deposited on interior industrial building materials at
various operational conditions.
These results, coupled with additional information in separate NHSRC publications (available at
www.epa.gov/nhsrc), can be used to determine whether a particular decontamination technology
can be effective in a given scenario. With these factors in consideration, the best technology
or combination of technologies can be chosen that meets the clean up, cost and time goals for a
particular decontamination scenario.
NHSRC has made this publication available to assist the response community to prepare for and
recover from disasters involving chemical contamination. This research is intended to move EPA
one-step closer to achieving its homeland security goals and its overall mission of protecting human
health and the environment while providing sustainable solutions to our environmental problems.
Cynthia Sonich-Mullin, Acting Director
National Homeland Security Research Center
-------
Notice
This report is submitted by CUBRC to Eastern Research Group, Inc. in fulfillment of Task Order
47 of the U.S. EPA STREAMS contract (Contract Number EP-C-05-059). This report covers
work completed from January 2008 through June 2009.
-------
Acknowledgments
The authors wish to acknowledge the support of those who helped plan and conduct this study and
prepare this report: Joe Cappello, Meg Stapleton, Richard Fitzpatrick, Bob Ambrusko, and Jac-
queline Hill. We would like to acknowledge Melanie Gooldy and Dave Mangino of CUBRC, Roy
Sieber of ERG, Leroy Mickelsen of the EPA's Office of Emergency Management, Adam Love of
Johnson Wright, Inc., and Lukas Oudejans of the EPA's National Homeland Security Research Cen-
ter for their review of this report.
-------
Abstract
The threat of a chemical agent release in a building or transportation hub is driving the U. S. Envi-
ronmental Protection Agency's (EPA's) National Homeland Security Research Center (NHSRC)
Decontamination and Consequence Management Division (DCMD) to conduct a research pro-
gram that systematically evaluates available decontamination technologies against chemical
agents. A program was designed to answer specific questions regarding the effectiveness of two
decontamination technologies (steam and vaporous hydrogen peroxide modified with ammonia -
mVHP®) against four selected chemical warfare agents (HD, GB, VX and thickened GD) applied
to four different indoor building material surfaces (decorative laminate, industrial-grade carpet,
galvanized metal, and ceiling tile). The technical objectives were to investigate the effects of en-
vironmental conditions (temperature and relative humidity), fumigant concentration, and contact
time on decontamination efficacy as well as to determine the agent vapor concentration in the test
chamber. A secondary objective was to make a qualitative visual assessment of the compatibility
of the building materials with decontaminants: do the building materials decompose, dissolve,
corrode, etc., when exposed to the decontaminants? A test chamber with appropriate controls
and interfaces was designed and fabricated to accommodate the two decontamination systems
under investigation. Known quantities of chemical warfare agent (CWA) were applied to sample
coupons (with replicates, blanks, and positive controls) prior to treatment with the appropriate
decontamination technology. Samples were removed from the chamber at specified time periods
and analyzed for the amount of residual agent remaining on and/or within the sample. Chemical
persistence as a function of time (without decontamination) was determined experimentally for
HD to establish baseline information on the natural degradation of the CWA on the selected ma-
terials under specific operational conditions. Extraction methods were developed and extraction
efficiencies were measured for the agent-material combinations.
Results from the efficacy testing indicated that the steam technology for both feed rates (1.5 and
3 kg/hr) removed the CWA surface contamination to below the method detection limit, on all
of the building materials tested. The presence of GB, TGD and VX in the condensate, however,
indicated that these agents may be re-deposited on other surfaces if the technology were used to
fumigate a building or section of a building. Additionally, the steam impacted both the carpet and
the ceiling tile materials, most significantly dissolving the ceiling tile.
The mVHP® results appear to indicate that increasing fumigant concentration slightly improved
the HD decontamination efficacy for most of the material-exposure time combinations in the test
matrix. The best mVHP® decontamination efficacies were observed for the full flow conditions,
yielding efficacies of 99% or better for all materials, except for ceiling tile, at the 350 ppmv target
concentration conditions.
The most significant findings of the mVHP® study were related to the effect of the mVHP® gener-
ator output flow on the decontamination efficacy. Increased output flow (100% versus 10% of the
generator output flow) resulted in increased efficacy for fumigant (vaporous hydrogen peroxide
and ammonia) concentrations that were the same as or lower than the 10% flow test fumigant con-
centrations. This effect was seen for all HD and VX-material combinations. For example, at the
10 % flow condition, decontamination efficacies were all less than or equal to 32% for the VX-
material combinations at the 400 min exposure time while the efficacies at the full flow condition
were 81-89% for these same material-agent combinations and exposure time. Finally, the mVHP®
fumigant did not permanently impact the appearance of most of the materials, only causing a
white residue to form on the galvanized metal ductwork.
-------
-------
Contents
Disclaimer iii
Foreword iv
Notice v
Acknowledgments vi
Abstract vii
1.0 Introduction 1
1.1 Objectives 1
1.2 General Approach 1
1.3 lest Facilities 1
2.0 Experimental Methods 3
2.1 Chemical Agents 3
2.2 Equipment and Instrumentation 3
2.2.1 lest Chamber 3
2.2.2 Steam Fumigant Instrumentation 3
2.2.3 Modified Vaporous Hydrogen Perioxide (mVHP®) Fumigant Instrumentation 4
2.3 Interior Building Materials 4
2.4 Experimental Design 5
2.5 Experimental Procedures 6
2.5.1 Sample Treatment 6
2.5.2 Chemical Agent Application 6
2.5.3 Efficacy Experiments - General Method 6
2.5.4 Reference Samples (Dose Confirmation) 7
2.5.5 Ambient Positive Control Experiments 7
2.5.6 GC/MS Method for the Analysis of CWAs in Coupon Extracts and Vapor Samples.. 7
2.5.7 Chemical Warfare Agent Purity 9
2.5.8 Extraction Efficiency Determinations 9
2.5.9 Agent Persistence on IBMs 9
2.5.10 Vapor Collection Method Characterization 9
2.6 Detailed Procedures 10
2.6.1 Steam 10
2.6.2 Steam Tests 10
2.6.3 Modified Vaporous Hydrogen Peroxide (mVHP®) 10
2.6.4 Determination of Fumigant Flow Rate 11
2.6.5 mVHP® Diffuser 11
3.0 Results and Discussion 13
3.1 Analytical Method Development Results - Determination of Extraction Efficiency
of CWAs from IBMs 13
3.2 Determination of Method Detection Limits (MDLs) 13
3.3 Persistence of GB on Galvanized Metal Ductwork and Decorative Laminate
without Decontamination 13
-------
3.4 Persistence of HD on IBMs without Decontamination 13
3.5 Ambient Positive Controls for Determination of Fumigant Efficacy 14
3.6 Recovery over Time of CWAs on IBMs with Steam Fumigant Technology 15
3.7 Steam Efficacy Results 17
3.8 C WAs in Condensate Samples Collected during Steam Decontamination 20
3.9 Steam - IBM Compatibility 20
3.10 Recovery over Time of CWAs on IBMs Using mVHP® Fumigant Technology 20
3.11 mVHP® Efficacy Results 24
3.12 Discussion of Effect of mVHP® Decontamination Diffuser Configuration on Test Results
and Additional Positive Controls 26
3.13 CWAs in Chamber Vapor Samples Collected during mVHP® Decontamination 27
3.13.1HD Vapor Sample Results 28
3.13.2 VX Vapor Sample Results 28
3.14 mVHP®-IBM Physical Compatibility 28
4.0 Quality Assurance 29
4.1 ISO 9001 Audit 29
4.2 Deviations from the QAPP 29
4.3 Quality Assurance Indicators 29
4.4 Temperature and Relative Humidity 29
4.5 Air Exchange Rate 30
4.5.1 Steam Air Exchange 30
4.5.2 mVHP® Air Exchange 30
4.5.3 Hydrogen Peroxide Concentrations 30
4.6 Ammonia Concentration 31
4.6.1 Ammonia Concentration Control 31
4.6.2 Verification of Ammonia Concentration 31
4.7 Agent on Positive Controls 31
4.8 Agent on Laboratory Blanks 32
4.9 Agent on Procedural Blanks 32
4.10 Equipment Calibration 32
5.0 Conclusions 33
6.0 Appendices A-D
Appendix A - Determination of Extraction Efficiency of CWAs From IBMs A-
AppendixB -Determination of Method Detection Limits (MDLs) B-
AppendixC - Steam Results Graphs C-
AppendixD - mVHP® Recovery Graphs D-
7.0 References R-
-------
List of Figures
Figure 2.2-1 - Test Chamber. 3
Figure 2.4-3 - Coupon Configuration in Sample Tray 6
Figure 2.6-1 - Test Final Diffuser Configuration 12
Figure C.I -Bar graph of ug of VX recovered per sample for the 3 kg/hr steam tests C-2
Figure C.2 - Bar graph of ug of VX recovered per sample for the 1.5 kg/hr steam tests C-2
Figure C.3 - Bar graph of ug of HD recovered per sample for the 1.5 kg/hr steam tests C-2
-------
List of Tables
Table 2.3-1 - Sample Thickness Measurements 4
Table 2.4-1 - Test Matrix for Steam Fumigant Experiments 5
Table 2.4-2 - Test Matrix for mVHP® Fumigant Experiments 5
Table 2.5-1 - Description of Gas Chromatograph/Mass Spectrometer Conditions for the Analysis
of Liquid Extracts 8
Table 2.5-2 - Description of Gas Chromatograph/Mass Spectrometer Conditions for
Vapor Samples 8
Table 2.5-3 - Description of Thermal Desorption Unit Conditions 9
Table 2.6-1 - Peroxide Diffuser Configurations 12
Table 3.3-1 -Persistence of GB as Indicated by Agent Recovery 13
Table 3.4-1 -Persistence of HD as Indicated by Agent Recovery 14
Table 3.5-1 - Ambient Positive Control Data Summary 14
Table 3.6-1 -Recovery of HD on IBM Decontaminated with Steam at 1.5 and 3 kg/hr 15
Table 3.6-2 -Recovery of GB on IBM Decontaminated with Steam at 1.5 and 3 kg/hr 16
Table 3.6-3 -Recovery of VX on IBM Decontaminated with Steam at 1.5 and 3 kg/hr 16
Table 3.6-4 - Recovery of TGD on IBM Decontaminated with Steam at 1.5 and 3 kg/hr 17
Table 3.7-1 -HD Steam Decontamination (Decon) Efficacy 18
Table 3.7-2 - GB Steam Decontamination (Decon) Efficacy 18
Table 3.7-3 - VX Steam Decontamination (Decon) Efficacy 19
Table 3.7-4 -TGD Steam Decontamination (Decon) Efficacy 19
Table 3.8-1 - Results from the GC/MS Analysis of Condensate Samples Collected during
Steam Decontamination Tests 20
Table 3.10-1 - mVHP® Decontamination of HD, 10%, 250 ppmv Recoveries 21
Table 3.10-2 - mVHP® Decontamination of HD, 10%, 350 ppmv Recoveries 21
Table 3.10-3 - mVHP® Decontamination of HD, 100%, 250 ppmv Recoveries 22
Table 3.10-4 - mVHP® Decontamination of HD, 100%, 350 ppmv Recoveries 22
Table 3.10-5 - mVHP® Decontamination of VX, 10%, 250 ppmv Recoveries 23
Table 3.10-6 - mVHP® Decontamination of VX, 100%, 250 ppmv Recoveries 23
Table 3.11-1 - mVHP® HD Decontamination (Decon) Efficacy 25
Table 3.11-2 - mVHP® VX Decontamination (Decon) Efficacy 26
Table 3.13-1-mVHP®HD Vapor Results 28
Table 4.3-1 - Measurements and Data Quality Indicators for Decontamination
and Persistence Testing 29
Table 4.5-2 - VHP™ Concentrations and Ammonia Concentrations for Each of the Test Runs 30
Table 4.6-1 -Ammonia Generation Data for the Various Tests 31
Table 4.6-2 - Drager Tube Ammonia Measurements 31
-------
Table 4.8-1 -Relative Standard Deviation for Ambient Positive Controls 32
Table 4.10-1 - Equipment Calibration Schedule 32
Table A.I - Results of Solvent Extraction Study for Extracting VX from IBMs A-2
Table A.2 - Results of Solvent Extraction Study for Extracting TGD from IBMs A-3
Table A.3 - Results of Solvent Extraction Study for Extracting HD from IBMs A-4
Table A.4 - Results of Solvent Extraction Study for Extracting GB from IBMs A-5
Table A.5 -Results of GB Persistence Study on Glass A-6
Table B.I- Determination of Method Detection Limit (MDL) - Laminate Coupons B -2
Table B.2 - Determination of Method Detection Limit (MDL) - Galvanized Steel Coupons B-2
Table B.3 - Determination of Method Detection Limit (MDL) - Carpet Coupons B-2
Table B.4 - Determination of Method Detection Limit (MDL) - Ceiling Tile Coupons B-2
Table D.I - HD, GM/DL, 10%, 250 ppmv D-2
Table D.2 - HD, CA/CT, 10%, 250 ppmv D-2
Table D.3 - Temperature, RH and Flow - HD, 10%, 250 ppmv D-3
Table D.4 - HD, GM/DL, 100%, 250 ppmv D-3
Table D.5 - HD, CA/CT, 100%, 250 ppmv D-4
Table D.6 - Temperature, RH and Flow - HD, 100%, 250 ppmv D-4
Table D.7 - HD, GM/DL, 10%, 350 ppmv D-5
Table D.8 - HD, CA/CT, 10%, 350 ppmv D-5
Table D.9 - Temperature, RH and Flow - HD, 10%, 350 ppmv D-6
Table D. 10-HD, GM/DL, 100%, 350 ppmv D-6
Table D.ll-HD, CA/CT, 100%, 350 ppmv D-7
Table D.12 - Temperature, RH and Flow - HD, 100%, 350 ppmv D-7
Table D.13 - VX, GM/DL, 10%, 250 ppmv D-8
Table D.14 - VX, CA/CT, 10%, 250 ppmv D-8
Table D.15 - Temperature, RH and Flow - VX, 10%, 250 ppmv D-9
Table D.16 - VX, GM/DL, 100%, 250 ppmv D-9
Table D.17 - VX, CA/CT, 100%, 250 ppmv D-10
Table D.18 - Temperature, RH and Flow - VX, 100%, 250 ppmv D-10
-------
Abbreviation and Acronym List
AATCC - American Association of Textile Chemists and Colorists
AMC - Army Materiel Command
CA - carpet
CAS ARM - Certified Analytical Standard Agent Reference Material
CAS RN - Chemical Abstracts Service Registry Number
CCV - continuing calibration verification
COTS - commercial off-the-shelf
CT - ceiling tile
CWA(s) - chemical warfare agent(s)
DAIG - Department of Army Inspector General
DC - diffuser configuration
DCMD - Decontamination and Consequence Management Division
DL - decorative laminate
ECBC - Edgewood Chemical Biological Center
El - electron ionization
EPA - U.S. Environmental Protection Agency
GB - Sarin
GC - gas chromatograph
GC/MS - gas chromatograph / mass spectrometer
GD - Soman
GM - galvanized metal
IBM(s) - Interior Building Material(s)
HD - sulfur mustard
HVAC - heating, ventilation and air conditioning
ISO - International Organization for Standardization
LB - laboratory blank
LPM - liters per minute
MDL - method detection limit
mVHP®- modified Vaporous Hydrogen Peroxide
NHSRC - National Homeland Security Research Center
PB - procedural blank
PC- positive control
PFTE - polytetrafluoroethylene
PVC - polyvinyl chloride
QAPP - Quality Assurance Project Plan
r - correlation coefficient
-------
RDECOM - U.S. Army Research, Development and Engineering Command
RDT&E - Research, Development, Test and Evaluation
RH - relative humidity
SD - standard deviation
TGD - thickened Soman
TIC - Total Ion Current
VHP® - vaporous hydrogen peroxide
-------
Unit List
amu - atomic mass unit
cm - centimeter
ft - foot, feet
g-gram
g/m3 - grams per cubic meter
hr - hour
kg/hr - kilograms per hour
L - liter
L/min - liters per minute
L/hour - liters per hour
LPM - liters per minute
m - meter
mm - millimeter
m3 - cubic meter
min - minute
mL - milliliter
ng - nanogram
ppmv - parts per million by volume
°C - degrees Celsius
ug - microgram
uL - microliter
um - micrometer
-------
1.0
Introduction
The U.S. Environmental Protection Agency's (EPA's)
mission is to protect human health and the environment.
Following the terrorist attacks of September 11, 2001,
and the subsequent mailing of anthrax-tainted letters,
EPA's role with respect to homeland security was
expanded. Presidential Directives identified EPA as
the primary federal agency responsible for protecting
public water supplies and remediation following an
attack on indoor or outdoor areas. In recognition of this
expanded role, EPA established a homeland security
research program. This research program is charged
with developing and delivering reliable, responsive
expertise and products based on scientific research and
evaluations of technology. The imminent threat of a
chemical agent release in a building or transportation
hub is driving the EPA's National Homeland Security
Research Center (NHSRC) Decontamination and
Consequence Management Division (DCMD) to
develop a research program that systematically evaluates
available decontamination technologies against chemical
agents on interior surfaces. Exterior surface materials
are also of critical importance and should be studied in
future research efforts.
1.1 Objectives
This test program was designed to answer specific
questions regarding the effectiveness of two
decontamination technologies against four selected
chemical warfare agents (CWAs); 2-(fluoro-
methylphosphoryl)oxypropane (GB; CAS RN
77-81-6), O-ethyl S-[2-(diisopropylamino)ethyl]
methylphosphonothiolate (VX; CAS RN 50782-69-9),
thickened GD 2-(nuoro-methyl-phosphoryl)oxy-3,3-
dimethylbutane (TGD; CAS RN 96-64-0) and bis(2-
chloroethy) sulfide (HD; CAS RN 505-60-2) deposited
onto four different interior building material (IBM)
surfaces. The IBMs were decorative laminate, industrial-
grade carpet, galvanized metal ductwork, and ceiling
tile. The technical objective was to investigate the effects
of environmental conditions (temperature and relative
humidity), fumigant concentration, and contact time on
decontamination efficacy. A secondary objective was to
make a qualitative visual assessment of the compatibility
of the building materials with decontaminants: do the
building materials decompose, dissolve, corrode, etc.,
when exposed to the decontaminants?
This research program addressed the following specific
questions:
• What is the decontamination efficacy of steam for
removal of CWAs on IBMs as compared to controls
at ambient environmental conditions?
• What is the decontamination efficacy of modified
Vaporous Hydrogen Peroxide(VHP®) (mVHP®),
modified in that ammonia is added to VHP®,
for removal of CWAs on IBMs under various
environmental and operational conditions as
compared to controls at ambient environmental
conditions?
• What are the physical effects of the decontaminants
on the IBMs?
1.2 General Approach
A test chamber with controls and interfaces was designed
and fabricated to accommodate the two decontamination
systems under investigation. Known quantities of CWAs
were applied to sample coupons (with replicates, blanks,
and positive controls) prior to subsequent treatment
with the appropriate decontamination technology.
Samples were removed from the chamber at specified
time periods and analyzed for the amount of residual
agent remaining on or within the sample in the pores or
crevices.
Chemical persistence as a function of time (without
decontamination) was determined experimentally to
establish baseline information on the natural decay
of the CWAs on the selected materials under specific
operational conditions. Extraction methods were
developed and extraction efficiencies were measured for
the agent-material combinations.
1.3 Test Facilities
Testing was performed at the CUBRC Chemical Agent
Research, Development, Test and Evaluation (RDT&E)
facility located near Buffalo, New York. The facility is
certified by the U.S. Army Research, Development and
Engineering Command (RDECOM) under a Bailment
Agreement to receive, store, handle, and consume
chemical warfare agents. The facility is inspected for
compliance by the Edgewood Chemical Biological
Center (ECBC), the Army Materiel Command (AMC),
and the Department of Army Inspector General (DAIG).
All chemical agent work performed at this test site
falls under CUBRC's International Organization for
Standardization (ISO) 9001 quality system.
-------
-------
2.0
Experimental Methods
2.1 Chemical Agents
The chemical agents used to evaluate the efficacy of
decontamination were Sarin (GB), thickened Soman
(TGD), VX and sulfur mustard (HD). The purity of
the chemical agents was greater than 85%. In addition,
CASARM-certified agents of higher purity and of a
separate lot were used as analytical reference standards.
The thickened GD was prepared by adding Acryloid
K125 polymer (Rohm and Haas, Philadelphia, PA) to
neat GD to achieve a 4.5% weight percent of GD in the
total mass of thickened agent. All chemical agents were
supplied by the U.S. Army at the request of the EPA.
2.2 Equipment and Instrumentation
2.2.7 Test Chamber
The fumigants under evaluation were passed through
a test chamber (Figure 2.2.1), which consisted of a
commercial glove box (Cole-Parmer®, Vernon Hills,
Illinois) with the physical dimensions of 32"W x
20"H x 24"D. The chamber was modified to allow for
temperature control, sampling and ventilation to meet
the requirements of the test plan. Modifications to the
test chamber included the following:
Heat blankets (SSH-1212-360-120, Omega, Stamford,
CT 06907) were installed on the enamel finished side
walls, back wall and base of the chamber. They were
wired to a six-zone controller (CN616TC1 Six Zone
Temp Controller, Omega, Stamford, CT 06907) and
independently set, feedback-controlled, and monitored
using a temperature controller (CN616, Omega,
Stamford, CT 06907). Each blanket was connected to
a solid state relay (SSr33DC25, Omega, Stamford, CT
06907) and maintained a set point temperature of ± 2 °C.
The temperature controller (CN616, Omega, Stamford,
CT 06907) was equipped with an interface (NIUSB-
232/2 2-port RS232 Serial interface for USB, National
Instruments, Austin Texas) allowing data to be logged to
a computer.
A ventilation fan (90 CFM, Radio Shack, Springville
NY 14141) with a valve (21083 3" gate valve, US
Plastics, Lima Ohio) was installed at the top of the
chamber to allow ventilation at a rate of up to 1840
LPM. After testing the chamber was vented with air
from the laboratory for the steam testing and air from
the mVHP® generator (Vaporous Hydrogen Peroxide
(VHP®) 1000-ARD Biodecontamination Unit modified
for mVHP® Chem/Bio Decontamination, STERIS
Corporation, Mentor, OH 44060) for the mVHP® testing.
The ventilation configurations used during the mVHP®
testing are discussed in more detail in Section 2.4.6.
The chamber, as received, was configured with a
Plexiglas® window (K-34788-00 Economical Glove
Box, Cole-Parmer, Vernon Hills, IL 60061). This
window was replaced with 0.25" thick plate glass
(1/4 " x 26 7/8' x 16" plate glass, Advanced Glass,
Williamsville, NY 14221) to reduce agent absorption
to the window surface. The same gasket that sealed the
Plexiglas® window was used to seal the glass window.
A rack was placed within the chamber to support the
test samples. The rack used during the steam testing was
made of perforated stainless steel. A polypropylene rack
was used for the mVHP® tests.
Figure 2.2-1 - Test Chamber
2.2.2 Steam Fumigant Instrumentation
Steam fumigant was generated using a Reimers® steam
boiler (AB A8ZE1Z 8 KW press steam boiler with AR
series pump, Reimers® Electra Steam, Incorporated,
Clear Brook, VA).
A 0.25" (0.635 cm) stainless steel tube was used to duct
steam into the chamber. During dry runs (i.e., no agent),
the steam flow had a tendency to form water droplets
when injected into a saturated chamber. To mitigate this
problem, a diffuser was designed and installed to remove
water droplets before they could enter the chamber. A
three-inch length of one-inch diameter copper tubing
was flattened at one end to create a flared nozzle. The
interior of the nozzle was packed with copper wool and
a small drain port was built into the nozzle to extract
the collected droplets from the copper wool. The design
allowed enough steam to enter the chamber to maintain
the desired steady-state atmosphere of 600 g/m3 at 100
°C. Excess droplets were drained from the nozzle
-------
through a short section of stock '/i'TO Tygon® tubing
to the floor of the chamber interior. A peristaltic pump
(Masterflex, Cole-Parmer, Vernon Hills, IL 60061) was
used to efficiently remove condensate from the base of
the chamber during the experiments.
A steam-condensing column was fabricated and installed
on the test chamber to serve two major purposes: 1)
to collect water generated by the steam so an output
could be measured; and 2) to condense and collect
the effluent steam from the chamber exhaust for CWA
analysis. The ethylene glycol-jacketed condensing
column was plumbed to a refrigerated circulating bath
(temperature range of-25°C to 150°C) (NESLAB
RTE-10, Thermo Scientific, Waltham, MA) to control
its temperature to roughly 15 °C. A chiller-circulator
bath (Model 1160 Polyscience Circulating Chiller Bath
Recirculating Heater Chilling Bath, VWR, Bridgeport,
NJ 08014) was connected to the liquid jacket to circulate
chilled ethylene glycol(Prestone® antifreeze, Wal-Mart,
Springville NY). This setup allowed the volume of the
all the steam produced by the generator to condense,
collect, and be measured.
The steam generator (maximum output rated at 1.8 L/
min) was calibrated to produce the desired amount of
steam for the experiments.
During the steam fumigant tests the test, chamber
heat blankets were set at 100 °C to minimize water
condensation on the chamber walls.
2.2.3 Modified Vaporous Hydrogen Perioxide
(mVHP®) Fumigant Instrumentation
A VHP® 1000-ARD Biodecontamination Unit (STERIS
Corporation, Menton, OH) modified for mVHP® Chem/
Bio Decontamination was used to generate the mVHP®
fumigant. The unit was modified by STERIS to allow
injection of ammonia gas into the stream of VHP®, and
also to run as an open loop system. This modification
of the 1000-ARD unit affects the system setup, the
decontamination cycle and the system controls.
Additionally a COTS Munter desiccant dehumidifier
(MG90 provided as test equipment by STERIS
Corporation, Mentor, OH 44060) for low air volumes
was integrated into the system to dehumidify input air to
the 1000-ARD unit. STERIS provided CUBRC with a
supplemental operating procedure1 for the modified unit.
The matrix for tests included two fumigant flow rates
through the test chamber, one at 100% of the preset
mVHP® unit output (340 L/min) and one at 10% (34 L/
min) of the mVHP® unit. The 10 % output condition
was created by directing 10% of the preset unit output to
the test chamber.
2.3 Interior Building Materials
The interior building materials selected for the project
were chosen to represent both porous and non-porous
materials used in commercial construction. All test
coupons were cut to dimensions of 3.5 cm x 1.5 cm
from the stock materials supplied by the manufacturer.
Specific information for each material is presented
below:
Decorative laminate (Formica®, Cincinnati, OH), white
matte finish, grade 10, nominal thickness of 1.2 mm);
manufactured by Solid Surface Design (Barcelona,
Spain); no material preparation prior to testing,
(designated as DL).
Industrial grade carpet (style #M7978, color 910; Shaw
Industries, Inc., Ringgold, GA) supplied by Carpet
Corporation of America (Rome, GA); no material
preparation prior to testing (designated as CA).
Galvanized metal duct (standard 24-gauge galvanized
steel HVAC duct; Adept Products, Inc., West Jefferson,
OH) supplied by Accurate Fabrication, Inc. (Columbus,
OH); material cleaned with acetone (99.4% purity
by vendor assay Baker Ultra Resi-analyzed, VWR,
Bridgeport, NJ 08014) prior to testing (designated at
GM).
Ceiling tile (Armstrong® 954; Lancaster, PA), Classic
Fine Textured, or equivalent; manufactured by
Armstrong; no material preparation prior to testing
(designated as CT).
Thickness measurements were made on a randomly
selected set often samples for each material type. A
summary of these data is presented in Table 2.3-1.
Table 2.3-1 - Sample Thickness Measurements
Thickness, mm
standard
in deviation
Decorative Laminate
(DL), n=10
Industrial Grade Carpet
(CA),n=10
Galvanized Metal Duct
(GM),n=10
Ceiling Tile (CT), n= 10
1.19
5.82
0.58
18.64
0.03
0.10
0.03
0.10
-------
2.4 Experimental Design
Tables 2.4-1 and Table 2.4-2 present a summary of the
specific experiments performed using the steam and
mVHP® fumigants, respectively.
The mVHP® test matrix was compressed; no tests were
run with GB or TGD due to the limited availability of
the generator.
GB was not tested against steam for the GM and DL
samples. Previous studies2 have shown GB to be non-
persistent on DL and GM. Studies at CUBRC have
confirmed the non-persistence of this agent on these two
surfaces as detailed in Section 3.3.
The following are designations and descriptions of the
types of samples used to perform the experimental effort.
Test coupons: IBM samples that were contaminated
with CWA and exposed to fumigants for specified
periods of time.
Table 2.4-1 - Test Matrix for Steam Fumigant Experiments
Agent steam Materials Elapsed Exposure CUBRC
output, kg/hr Time, minutes Test ID
HD
GB
VX
TGD
1.5
3
1.5
3
1.5
3
1.5
o
5
GM,DL
CT, CA
GM, DL
CT, CA
CT, CA
CT, CA
GM,DL
CT, CA
GM, DL
CT, CA
GM, DL
CT, CA
GM, DL
CT, CA
120, 180, 240, 400
120, 180, 240, 400
60, 120, 180, 240
60, 120, 180, 240
60, 120, 180, 240
60, 120, 180, 240
60, 120, 180, 240
60, 120, 180, 240
60, 120, 180, 240
60, 120, 180, 240
60, 120, 180, 240
60, 120, 180, 240
60, 120, 180, 240
60, 120, 180, 240
T-l
T-2
T-9
T-10
T-4
T-12
T-5
T-6
T-13
T-14
T-7
T-8
T-15
T-16
Table 2.4-2 - Test Matrix for mVHP® Fumigant Experiments
STERIS Peroxide Elapsed Exposure CUBRC
Agent output target, ppm Materials Time, minutes Test ID
HD
VX
10%
100 %
10%
100 %
250
350
250
350
250
250
GM, DL
CT, CA
GM,DL
CT, CA
GM, DL
CT, CA
GM,DL
CT, CA
GM, DL
CT, CA
CT, CA
GM,DL
CA, CT
120, 240, 400, 468
120,240,400,510,600
120, 240, 400, 468
120,240,400,510
60, 120, 150, 180
60, 120, 180, 400
60, 120, 150, 180
60, 120, 150, 180
120, 180, 240, 400
Aborted
120, 180, 240, 400
120, 180, 240, 400
120, 180, 240, 400
T-18
T-17
T-26
T-25
T-19
T-20
T-28
T-27
T-24
T-23
T-23R
T-21
T-22
-------
Procedural blanks: uncontaminated IBM samples
exposed to the fumigant along with the test coupons;
used to determine if sample-to-sample cross-
contamination occurs during the fumigant testing.
Laboratory blanks: uncontaminated IBM samples not
processed with the decontaminants.
Positive controls: IBM samples that in separate trials are
contaminated but not exposed to fumigant and allowed
to remain at ambient environmental conditions for the
periods of time equal to decontamination exposure
times.
A single test run consisted of 56 total coupons and eight
procedural blanks for each of two material types. For
each material type, five test coupons and two procedural
blanks were evaluated at four different sampling
points. Five laboratory blanks for each material type
were processed at the same time separately from these
samples. The 56 coupons were divided equally into
fourseparate sample trays as shown in the example in
Figure 2.4-3, below.
CAIest samples, replicates 1-5
o
CTtest samples, replicates 1-5
Figure 2.4-3 - Coupon Configuration in Sample Tray
Decontamination efficacy was calculated as:
E=(C0-CF)/C0-100% [1]
where C0 is the average concentration of agent before
decontamination (determined from the positive control
coupons of each surface material) and CF is the average
concentration on the test coupons after decontamination.
Separate efficacy calculations were performed for each
of the material-agent combinations at each exposure
time. In addition, since each of these test matrix points
was represented by multiple sample coupons, a mean
and standard deviation of the efficacy values were
reported. The percent efficacy is an indicator of relative
efficacy and values stated as >99% do not indicate that
the surface is clean. These values simply indicate that
more than 99% of the initial contamination is no longer
present in the extracts of the test samples.
2.5 Experimental Procedures
2.5.7 Sample Treatment
The galvanized metal ductwork samples were cleaned
with acetone (99.4% purity by vendor assay Baker Ultra
Resi-analyzed, VWR, Bridgeport, NJ 08014) prior to
testing. All test coupons were allowed to equilibrate at
room temperature and relative humidity for a minimum
of 30 min before chemical agent was applied.
2.5.2 Chemical Agent Application
Agent droplets were applied to coupons using a gas
tight syringe (80201 25 ul syringe with a 22 gauge
needle for agents HD,VX and GB and a 81020 lOOul
syringe with a 18 gauge needle for TGD , Hamilton
Company, Reno, Nevada) equipped with a repeatable
dispenser (Model PB600-1; Hamilton Company, Reno,
Nevada). A separate syringe was used for each agent
to avoid cross-contamination. A total of 2 uL of agent
was applied to each test sample as four 0.5 uL droplets
for agents HD, GB, and VX; and as a single 2 uL
droplet for agent TGD. This yielded deposition of 2.0
to 2.5 mg of chemical agent on each coupon. Once a
full tray of coupons (10 contaminated test coupons and
four procedural blanks) was prepared, that tray was
immediately placed into the test chamber. The laboratory
blanks, which received no application of agent, were
placed into a 125 mL tall wide-mouth (05-719-54 4,
I-Chem Laboratory Glassware, Fisher Scientific Atlanta,
GA) jar containing extraction solvent (EM-HX0296-6
Omnisolv HR- Pesticide Residue Analysis grade hexane,
VWR).
2.5.3 Efficacy Experiments - General Method
Before the application of agents to coupons was
initiated, the test chamber was pre-conditioned to the
desired operational conditions for use of the steam
or mVHP® system. The first tray placed into the test
chamber corresponded to the longest sampling time
point (exposure time). The second tray placed into
the test chamber corresponded to the next-to-longest
sampling time point and this process was repeated for all
of the sample trays.
At each specified sampling time, the appropriate test
coupons and corresponding procedural blanks were
removed from the test chamber and placed into 125 mL
tall wide-mouth glass I-Chem extraction jars (IR121-
0125 wide mouth jars, VWR) containing extraction
solvent (hexane 99.8% purity by vendor assay). Ten mL
of extraction solvent was used for the galvanized metal,
decorative laminate, and carpet material coupons and
20 mL was used for the ceiling tile material coupons to
effectively cover the material because of the thickness of
the coupon.
-------
Once the final tray was removed from the chamber, the
decontamination system was turned off and the chamber
was ventilated to remove residual decontaminant and/or
agent.
2.5.4 Reference Samples (Dose Confirmation)
Using agent application procedures identical to the
procedures described in Section 2.5.2 above, four
reference dose confirmation samples were prepared
(during the agent application process for each test)
to provide a normalized contamination level (a
100% value) for use in the decontamination efficacy
calculations (Sections 3.7 and 3.11). The appropriate
volume of agent was applied to the inner sidewall of a
125 mL glass jar containing 10 mL of solvent. The jar
was gently swirled to mix the agent and solvent recovery
was calculated based upon the theoretical mass applied.
This calculated mass was normalized and used to
represent the 100% challenge mass for each experiment.
When a test using CT was performed, a second set of
dose confirmation samples was prepared using 20 mL of
solvent to match the specific extraction conditions used
for the ceiling tile coupons.
2.5.5 Ambient Positive Control Experiments
A series of positive control experiments was performed
to allow accurate determination of decontamination
efficacy for the steam and mVHP® decontaminants. This
is necessary because natural attenuation may reduce
the amount of agent remaining on the coupons and this
could be attributed to decontaminant related reduction.
The ambient positive controls were conducted at 24 °C
and 40% RH and at the same airflow, 2.35 LPM (0.016
air exchanges/minute), as the persistence test flow rate
(Section 2.5.9). This flow rate is similar to the flow
rate used in previous EPA testing3. The positive control
coupons resided in the test chamber for times parallel to
the decontamination test contact times.
Both steam and mVHP® produce environmental
conditions (temperature and relative humidity) different
from ambient conditions; additionally, steam generates
condensation. The decontamination efficacy for both
technologies is calculated based upon ambient positive
controls. Ambient positive controls were used with all
four agents.
2.5.6 GC/MS Method for the Analysis of CWAs
in Coupon Extracts and Vapor Samples
The test coupon and the vapor tube extracts were
analyzed and quantified for each CWA using GC/MS
with electron ionization (El), under the conditions listed
in Table 2.5-1. The GC/MS was operated in the full-
scan mode (total ion current, TIC) for a mass range of
50 to 500 daltons. The GC/MS data were acquired and
processed using Agilent ChemStation software (see
Tables 2.5.1 and 2.5.2) (Agilent Technologies, Santa
Clara, CA). The CWAs were identified by comparison
of the retention time and mass spectra against the
retention time and mass spectra of calibration standards.
Each instrument was calibrated prior to the analysis of
samples from each test run using a nine-point calibration
curve spanning the range of 0.98 - 291 nanograms
(ng). Calibration curves were generated in Microsoft
Excel® using a second-order polynomial fit, and
correlation coefficients (r2>0.99) were calculated from
the regression fit. The concentration of the agent was
calculated by external standardization using calibration
standards analyzed with each data set. Analytical
results within the calibration range established for
the instrument were reported in ng. Due to the wide
calibration range, two separate calibration curves
(one with high values and one with low values) were
occasionally employed to generate an improved fit to
the calibration data. Continuing calibration verification
(CCV) standards were inserted into the sample series
every ten samples at a minimum to identify any
calibration drift. The acceptance criterion for the
CCV was ± 25 % of the initial calibrated response or
amount. Maintenance of the instrument was performed
in accordance with manufacturer's recommendations.
All maintenance was recorded in a dedicated GC/MS
Maintenance Log Book for each instrument.
Samples generated during the vapor collection method
characterization (Section 2.5.10) were analyzed using
a thermal desorption unit (Markes Ultra TD Unity ,
Markes International, Ltd, Gwaun Elai Medi Science
Campus, Llantrisant, RCT, CF72 8XL, UK) interfaced
to an Agilent Model 5973/6890 GC/MS (Agilent
Technologies, Santa Clara, CA). For this study, vapor
calibration standards and test samples were generated
by direct injection of liquid calibration standards into
stainless steel thermal desorption tubes (0.25 inch (6.35
mm) OD and 3.5 inches (88.9 mm) long) containing
200 mg of sorbent Tenax TA (35/60 mesh - Markes
International Limited, Llantrisant, RCT, CF72 8XL,
UK). The sorbent tubes were thermally desorbed and
analyzed using the instrumental conditions shown in
Table 2.5-2 and Table 2.5-3.
-------
Table 2.5-1 - Description of Gas Chromatograph/
Mass Spectrometer Conditions for the Analysis of
Liquid Extracts
Table 2.5-2 - Description of Gas Chromatograph/
Mass Spectrometer Conditions for Vapor Samples
Parameter Condition
Instrument
Column
Carrier Gas
Flow Rate
Column
Temperature
(GB)
Column
Temperature
(GD)
Column
Temperature
(HD)
Column
Temperature
(VX)
Injection
Volume/Type
Quad
Temperature
MS Source
temperature
Solvent Delay
Agilent 5973/5975 Mass Spectrometer
Model with electron ionization (El) ion
source, interfaced to a 6890/7980 Gas
Chromatograph equipped with a Model
7673 A Automatic Sampler and Agilent
Enhanced MSD ChemStation Software
version D.02. 00. 275.
30 m x 0.25 mm i.d. Restek RTx-5 MS
(cross-linked methyl silicone), fused
silica capillary column, 0.5 um film
thickness (Restek No. 12638)
1.2 mL/min helium in Constant Flow
Mode
40 °C initial temperature, hold 1 min, 8
°C/min to 90 °C, hold 0 min, 25 °C/min
to 260 °C.
50 °C initial temperature, hold 1 min,
10 °C/min to 240 °C, hold 2 min, 20 °C/
min to 260 °C.
50 °C initial temperature, hold 1 min,
10 °C/min to 150 °C, hold 0 min, 15 °C/
min to 240 °C
50 °C initial temperature, hold 1 min,
25 °C/min to 240 °C, hold 2 min, 20 °C/
min to 280 °C
1 uL splitless injection (4 mm i.d.
double goose neck splitless insert)
(20785-214.5, Restek, Bellefonte,
PA 16823-8812) with 2 min purge
activation time. Split vent flow rate @
50 mL /min.
250 °C
230 °C
5 min
Data were collected from 50 to 550 daltons at a scan rate
of 2.91 scans/sec, threshold of 50 and sampling of 2.
Sampling of 2, a GCMS scan parameter, is the number
of times the abundance of each mass is recorded before
proceeding to the next mass.
Parameter Condition
Instrument
Column
Carrier Gas Flow Rate
Column Temperature
Injection Volume/Type
Quad Temperature
MS Source
temperature
Solvent Delay
Agilent Model 5973
Network Mass Spectrometer
equipped with electron
ionization (El) ion source,
interfaced to a 6890N
Gas Chromatograph and
Agilent Enhanced MSD
ChemStation Software
version E.02.00.
30 m x 0.25 mm i.d. Restek
RTx-5 MS (cross-linked
methyl silicone), fused silica
capillary column, 0.5 um
film thickness (Restek No.
12638)
1.2 mL/min helium in
Constant Flow Mode
60 °C initial temperature,
hold 1 min, 20 °C/min to 280
°C, hold 1 min
Direct interface to Markes
Unity Thermal Desorption
Unit
250 °C
230 °C
5 min
Data were collected from 50 to 550 daltons at a scan
rate of 2.91 scans/sec, threshold of 50 and sampling
of 2. Sampling of 2, a GCMS scan parameter, is
the number of times the abundance of each mass is
recorded before proceeding to the next mass.
-------
Table 2.5-3 - Description of Thermal Desorption Unit
Conditions
Parameter Condition
Instrument
Sorbent Tubes
Tube Desorption
Desorption Flow
Flow Path
Temperature
Cold Trap
Trapping
Temperature
Trap Desorption
Trap Split Flow
Markes Ultra-Unity automated
thermal desorption system
interfaced directly to Agilent
Model 5973/6890 GC/MS.
Tenax TA, (35/60 mesh) 200
mg, 3.5 inch (89 mm) x 1A inch
(6.4 mm) or HaySep D (60/80
mesh), 300 mg, 3.5 inch (89
mm) x % inch (6.4 mm) (Markes
International Limited, Llantrisant,
RCT, CF72 8XL, UK)
280 °C for 4 min
40 mL/min
180 °C
Unity "Chemical Weapons" trap
(U-T10CW) (Markes International
Limited, Llantrisant, RCT, CF72
8XL, UK)
0°C
300 °C for 4 min
10 mL/min
2.5.7 Chemical Warfare Agent Purity
Certified Analytical Standard Agent Reference Material
(CASARM) certified agents were used to prepare
analytical standards for the calibration of instruments.
The agents used for the decontamination testing were
required to have a purity of greater than 85%. These
testing agents were analyzed using GC/MS (injection
of dilute agent) to verify purity and all agents met or
exceeded this criterion.
2.5.8 Extraction Efficiency Determinations
Baseline chemical agent extraction efficiencies were
determined using hexane (EM-HX0296-6 Omnisolv
HR- Pesticide Residue Analysis grade hexane, VWR) for
the four agents (HD, GB, VX and TGD) deposited onto
the decorative laminate and galvanized metal ductwork
materials. Two uL of the appropriate neat chemical
agent was applied using the procedures outlined in
Section 2.5.2 onto triplicates of each sample type. To
minimize any effects of evaporation, individual coupon
samples were placed into 125 mL tall wide-mouth
I-Chem extraction jars with PFTE-lined caps containing
10 mL of pesticide-grade hexane within 30 seconds of
contamination. The jars were placed into an ultrasonic
bath (33995-548 Bransonic Model 5510-DTH, VWR,
Bridgeport, NJ 08014) for 10 minutes. Once the jar
was removed from the ultrasonic bath an aliquot was
removed and transferred to a GC autosampler vial
for GC/MS analysis. These results were compared to
results from a previous EPA effort,2 in which samples
were placed in hexane and soaked overnight (14 hours)
to determine if acceptable extraction efficiencies could
be achieved (i.e., 40-120 % recovery with a <30 %
coefficient of variance).
Extraction method development and validation was
performed to determine optimal procedures for
extracting HD, VX, and TGD from porous materials
(CT and CA). Four solvents/combinations of solvents
were evaluated: 1) hexane; 2) a 1:1 mixture of hexane
and acetone; 3) methylene chloride; and 4) ethyl acetate.
Hexane and methylene chloride have been used in
previous EPA CWA decontamination studies.2 These
recoveries are listed in Appendix A, Tables A. 1-A.4.
2.5.9 Agent Persistence on IBMs
Experiments were performed to generate an agent-
substrate persistence model as a function of time to aid
in the selection of appropriate sampling points (exposure
times) for the fumigant efficacy tests. The first study was
conducted using GB on GM and DL. Triplicate coupons
of each of the two materials were contaminated with GB,
placed into the test chamber, and held at 22 ± 2 °C, 40 ±
10% RH, and with a ventilation rate of one air exchange
per hour. Triplicate samples were removed at 5, 15, 30,
60 and 120 minutes, immediately extracted and analyzed
by GC/MS.
A separate persistence study was performed using HD on
all four building materials under the same experimental
exposure conditions. Three samples were removed and
extracted at time points of one hour, four hours, one day,
two days, and seven days and analyzed by GC/MS. From
these studies, specific exposure times were determined
and are represented in the test matrices presented in
Table 2.4-1 and Table 2.4-2.
2.5.10 Vapor Collection Method
Characterization
A brief investigation was conducted using GD, GB
and HD to determine the feasibility of collecting vapor
samples during mVHP® testing. VX was not evaluated
because the Markes Thermal Desorption /Autosampler
unit does not lend itself to the direct analysis of VX
vapor.
Sorbent tubes were spiked with the respective agent
in triplicate at a level of 5 ug. The spiked tubes were
then connected to the vapor sampling ports on the test
chamber and sampled the decontaminant atmosphere
(1.2 CFM of mVHP®) at a flow rate of approximately
250 mL/min for one hour. Calibration tubes were spiked
at levels of 0.05, 0.5, and 5 ug as a reference. Following
-------
vapor sampling, the spiked tubes were analyzed by direct
thermal desorption to a GC/MS system (Section 2.5.6).
The GD recoveries were less than 20% with one outlier
and the GB recoveries were 0, 0 and 2%. HD recoveries
at one and four hrs were near 100%.
2.6 Detailed Procedures
2.6.7 Steam
The operational conditions for the steam testing
specified in the QAPP4 were based upon previous similar
decontamination work. In a similar study conducted by
Battelle (1995)5, high decontamination efficacies were
observed when a steam generation rate of 20.8 kg/m3/hr
was used for decontamination of stainless steel, concrete
and unglazed porcelain coupons contaminated with
HD, VX or GB. Using this rate and adjusting the rate
proportionally to the volume of the test chamber (5 ft3 =
0.148 m3), a steam generation rate of three kg/hr would
be required to duplicate the conditions in the previous
work. Based upon this calculation, two test conditions
were selected for the steam experiments:
Steam Condition 1-3 kg/hr at time points of 60, 120,
180 and 240 minutes
Steam Condition 2-1.5 kg/hr at time points of 120, 180,
240 and 400 minute.
The rate of steam generation was controlled by an outlet
valve. The steam output was measured as a function
of predetermined valve opening (number of turns).
A series of tests using varying steam injection rates
was performed to establish the outlet valve settings
required to meet the desired experimental steam
conditions. These experiments demonstrated that after
one hour of continuous operation with the outlet valve
open one complete turn, a total of 4 L of water were
collected. Because the largest flow rate for our series
of experiments would only generate 3 L/hr, a second
series of tests with valve settings at 1A, 1A and % were
conducted to determine which valve settings would yield
the desired rates of 3 kg/hr and 1.5 kg/hr. A calibration
curve for outlet valve position versus steam injection
rate was generated and the appropriate outlet valve
positions were selected for the experiments.
During the testing, 1L of steam condensate was
collected, beginning at the start of each test for CWA
analysis. Most of the agent was assumed to be removed
during the first hour of steam decontamination and
potentially recovered in the condensate solution.
2.6.2 Steam Tests
Test coupons were pre-conditioned in the test fume
hood under ambient conditions prior to application of
the chemical agent while the test chamber was heated
to 100 °C and pre-conditioned with steam. Once the
desired chamber conditions were achieved, the chemical
agent was deposited onto the IBM coupons, one tray
at a time, and the tray was placed into the chamber.
The decontamination exposure time was measured
independently for each tray, and started at the time each
tray was placed into the test chamber.
During the steam exposure period, approximately 1L
of water condensate from the heated gas exiting the
chamber was collected for analysis. Approximately 30
minutes was required to collect this volume of water at a
steam generation rate of 3 kg/hr and about an hour was
required for the 1.5 kg/hr tests. A 500 mL aliquot of this
condensate sample was extracted twice with 10 mL of
methylene chloride and analyzed by GC/MS using the
same conditions as the steam test samples presented in
Table 2.5 1. At the appropriate sampling time periods,
the chamber was briefly opened and the test coupons and
procedural blanks were removed, extracted in solvent
with ultrasonication, and the extract was analyzed by
GC/MS.
2.6.3 Modified Vaporous Hydrogen Peroxide
(mVHP®)
A STERIS VHP® 1000-ARD was used to generate the
mVHP® fumigant during this project. The unit was
leased from STERIS Corporation. The 1000-ARD
generates the mVHP® fumigant by injecting a liquid
solution of hydrogen peroxide (Vaprox®, 35% hydrogen
peroxide) (STERIS Corporation, Mentor, OH 44060)
onto a heated vaporizer plate resulting in a heated
hydrogen peroxide vapor. This gas is then mixed
with a low concentration (ppmv) of ammonia. Recent
studies at ECBC6 have shown that the addition of low
levels of ammonia renders VHP® reactive towards GD,
converting GD to pinacolyl methylphosphonic acid. The
study concluded that mVHP® affords broad-spectrum
decontamination of the CWAs including VX, GD and
HD.
The STERIS technology employs four phases of activity:
Dehumidification: The 1000-ARD provides dry, heated
air that was introduced into the test chamber to achieve
a temperature of 30 °C and a relative humidity of less
than 40%. A temperature and relative humidity probe,
provided by manufacturer of the 1000-ARD, was placed
inside the test chamber along with a VHP® sensor and
an ammonia sensor. All sensors operated via a feedback
loop with the 1000-ARD generator to ensure that all
environmental and operational conditions were achieved,
held constant, and recorded.
Conditioning: The 1000-ARD generates VHP® and adds
ammonia to reach the desired concentration as detected
by the sensors. Because the ammonia sensor cannot
operate in the presence of hydrogen peroxide vapors,
ammonia was introduced first and allowed to stabilize
at a fixed percentage of the intended target peroxide
-------
concentration. Once the ammonia concentration was
stable, the ammonia sensor was turned off and the
addition of vaporous hydrogen peroxide was started.
To avoid contaminating the 1000-ARD generator, the
typical practice of vapor regeneration (cycling back
through the 1000-ARD) was not performed. Instead,
all fumigant vapor was exhausted after it was sent into
the test chamber and appropriately filtered. This was
not expected to affect the fumigant concentration or
composition and in turn the efficacy of the fumigant.
Decontamination: The 1000-ARD system supplied
a steady concentration of mVHP® throughout the
experiments.
Because the ammonia concentration could not be
monitored continuously during the experiments, Drager
indicator tubes (Model Ammonia 5/b, SKC Inc.,
Eighty Four, PA) were used to measure the ammonia
concentration during this phase but prior to inserting the
test and procedural blank coupons. A manually operated
bellows pump (EW-86514-14 Drager accuro® Pump Kit
and Gas Detection Cole-Parmer, Vernon Hills, IL 60061)
drew calibrated 100 mL samples through the Drager
Tubes.
Aeration: Once the test was complete, dry heated air
was forced through the test chamber to remove the
hydrogen peroxide and ammonia vapors. The chamber
effluents were scrubbed using a catalytic converter
(supplied by STERIS Corporation, Mentor, OH 44060).
2.6.4 Determination of Fumigant Flow Rate
According to the STERIS web site, the mVHP® ARD
Biodecontamination System can provide high-volume
biological decontamination for enclosures having a
volume up to 10,000 ft3 (-280 m3) depending on the
application. The technology has been tested in several
test programs under various test configurations. A
2006 test program at ECBC utilized a STERIS mVHP®
prototype system with an output of 40 CFM (-1100
L/min) VHP® into a 112 ft3 (3.17 m3) test chamber7
resulting in 2.8 air exchanges per minute.
STERIS training personnel indicated that the minimum
reproducible volumetric output of the 1000-ARD was
340 L/ min. Therefore, in cooperation with EPA, two
test conditions were established for the 0.148 m3 (5.23
ft3) chamber used for this program. The first condition
was at a full flow rate of 340 L/min and the second was
at 10% of full flow, or 34 L/min. The 340 L/min flow
rate equates to an air-exchange rate of 2.4 exchanges per
minute, considerably higher than what is operationally
achievable in the field and close to the values achieved
during the previous programs referenced above. The
10% flow rate (34 L/min) equates to 0.24 air exchanges
per minute and is more representative of real world
facility decontamination applications.
To achieve the 10% output of 34 L/min, the 1000-
ARD generator effluent was sent through a rotameter
(FL 1653 3.17 SCFM, Omega, Stamford, CT 06907)
allowing 90% of the flow to be diverted to vent. In order
to accommodate the lower flow rate, dilute solutions of
the hydrogen peroxide were prepared ranging from 2%
to 5% by volume. Many dry runs were performed to test
the system at the lower flow rates to ensure the system
could generate the target peroxide concentration (250
ppmv ± 10%) under these conditions. Variables that
were investigated and optimized included the hydrogen
peroxide injection rate and concentration. During the
dry runs the temperature within the chamber was not
the same for the 340 and 34 L/min conditions. The heat
output from the 1000-ARD at the 340 CFM flow rate
increased the chamber temperature to almost 40 °C,
while the chamber temperature at the lower flow rate
remained around 24 °C. The sensors were biased high
at the 340 CFM flow rate possibly due to the higher
temperature in the test chamber. All of the results are
presented as a function of hydrogen peroxide target
concentration and not the actual concentrations. The
actual concentrations are reported in Section 4.5.3.
Use of the 340 L/min flow rate caused a pressure
increase in the test chamber that was alleviated by
installing a STERIS auxiliary blower into the exhaust
vent of the chamber and using the auxiliary blower
was controlled by the onboard computer. It was run
at 12 SCFM instead of the ventilation fan originally
installed onto the test chamber. The pressure of one
atmosphere at the 100% flow of 340 L/min ventilation
rate was monitored with a differential pressure gauge
(Series 2000 0-25" Magnehelic®, Dwyer, Michigan City,
Indiana, 46361). The auxiliary blower was not needed
for the 34 L/min flow rate. The original ventilation fan
was, therefore, used for these tests.
Additional positive control sets were run at these two
different flow rates for the entire mVHP® test matrix
and these tests were designated as follows: PC-5 and
PC-10 full flow HD tests, PC-6 and PC-9 reduced flow
HD tests, PC-7 full flow HD tests, and PC-8 reduced
flow HD tests. The sorbent tube data from these tests can
be found in Section 3.1.4 while the coupon extract data
from these tests can be found in Appendix D.
2.6.5 in VHP® Diffuser
Turbulence and non-uniformity of fumigant distribution
were observed during some early experiments. As a
result, it was necessary to build and install a diffuser to
modify the manufacturer-provided chamber in order to
1) improve fumigant distribution; and 2) reduce airflow
velocity over the samples. Several iterations of design
improvements were required to achieve better mixing
and fumigant distribution and to significantly reduce the
flow velocity directly over the surfaces of the samples.
-------
Following the installation of the diffuser and during one
of the positive control tests, uneven evaporation of agent
from the test samples was observed. The evaporation
pattern indicated that the diffuser was the cause of this
problem, so the diffuser design was modified. Table
2.6-1 shows the evolution of the diffuser design and the
tests that were performed under each iteration.
The Configuration 1 design consisted of a section of
PVC tube extending the entire length of the upper
back corner of the glove box with a series of staggered
holes drilled along the entire length pointing downward
at a 45-degree angle from the back of the glove box.
Because many of these holes were located close to the
chamber's exit tube, Configuration 2 was implemented.
Configuration 2 moved the diffuser to the far end of the
fumigant inlet tube and holes were added to the PVC
tube. These holes reduced the fumigant velocities and
yielded a more evenly distributed the flow pattern. In
addition, a capped end piece was added to the terminus
of the inlet tube. This configuration generated a better
fumigant distribution and also reduced the velocity of
the fumigant exiting the diffuser. A third configuration
(Configuration 3) was constructed to further improve
the distribution. In this configuration, the number of
holes along the length of the PVC tubing was reduced
and the end was oriented away from the sensor unit. A
baffle was also added above the diffuser to isolate the
sensor package located above the diffuser and allow the
STERIS fumigant input to be more thoroughly mixed
with the glove box atmosphere prior to measurement.
This configuration further improved the homogeneity of
the fumigant distribution.
Table 2.6-1 - Peroxide Diffuser Configurations
Peroxide Diffuser
Configuration Dates Test(s)
Configuration 1 - open,
pointed straight down
Configuration 2 - new
design, capped and
drilled
Configuration 3 -
reduced holes, different
orientation, baffle
Configuration 4 - holes
drilled in cap (final)
5/14-5/19
5/20 - 5/21
5/22 - 5/25
5/26 - 6/12
PC-5
T17
T18 and
PC-6
T19,T20,
T21,T22,
T23, T23R,
T24, T25,
T26, T27,
and T28
PC-7, PC-8,
PC-9, and
PC-10
Additional experiments were performed with
Configuration 3 to ensure that the following could be
achieved: 1) uniform droplet evaporation; 2) optimal
distribution of fumigant within the test chamber; and
3) stable concentration measurements for mVHP® and
PJ3. The results of these experiments led to one final
modification, termed Configuration 4. Configuration
4 featured holes in the cap. This final diffuser
configuration, with the baffle installed, is shown in
Figure 2.6-1. Although perfect uniformity of droplet
evaporation was not achieved, the evaporation pattern
was greatly improved.
Figure 2.6-1 - Test Final Diffuser Configuration
-------
3.0
Results and Discussion
3.1 Analytical Method Development
Results - Determination of
Extraction Efficiency of CWAs
from IB Ms
Studies were conducted initially to verify the extraction
efficiencies of the CWAs from the IBMs that were to be
used in the decontamination studies. Hexane with a 10
minute ultrasonication was used to extract HD, GB, VX
and TGD from decorative laminate and galvanized metal
ductwork. Hexane, 1:1 acetone hexane (v/v), methylene
chloride and ethyl acetate were used to extract HD, GB,
VX and TGD from ceiling tile and carpeting. These
recoveries are listed in Appendix A, Tables A. 1-A.4.
Acceptable extraction efficiencies (well within the
QAPP criterion of 40-120% with < 30 % coefficient of
variance) were achieved for all agents using 10 mL of
hexane and 10 minutes of ultrasonication for laminate,
galvanized metal ductwork, and carpet. Ceiling tile
required 20 mL of hexane to effectively cover the
material with solvent in the 125 mL extraction jar.
(Appendix A, Tables A.1-A.4). Therefore, hexane was
used at these volumes for the majority of the persistence
testing, positive controls, and decontamination testing.
Better extraction efficiencies were observed using
1:1 acetone hexane (v/v) for VX and TGD deposited
on ceiling tile. This solvent system was therefore
used for all persistence testing, positive controls, and
decontamination testing done with this agent-material
combination.
Relatively lower extraction recoveries were obtained for
GB, possibly due to loss of the GB prior to extraction.
An additional time study was conducted to determine
how quickly the GB was lost either due to reaction
on the IBM surface and/or evaporation. During the
extraction efficiency study, the agent residence time
prior to extraction was approximately 30 seconds for all
agents. For this study, coupon samples were extracted
after residence times of 0, 0.5, 1 and 2 minutes. The
mean results (n=3) were 100%, 79.1%, 67.6% and
34.4% recovery, respectively (Appendix A, Table A.5).
3.2 Determination of Method Detection
Limits (MDLs)
Method detection limits (MDLs) were determined using
the single concentration design estimator recommended
by the EPA. The single concentration design estimator
is defined as the minimum concentration which can be
measured and reported with 99% confidence that the
analyte concentration is greater than zero, determined
from the analysis of a sample in a given matrix
containing the analyte (40 CFR 1361, Appendix B).
This study was performed by spiking replicate low level
matrix samples (7) at identical concentrations 2-5 X the
expected MDL. The MDL was calculated by multiplying
the sample standard deviation by the correct Student's
t-value(3.143).
The resulting method detection limits are shown in
Appendix B, Tables B. 1 - B.4. Based on these results a
detection limit of 5 ug (based on 10 mL extraction) was
established for the recovery of CWAs from the DL, GM
and CA IBMs, and 15 ug (based on 15 mL extraction)
for CT
3.3 Persistence of GB on Galvanized
Metal Ductwork and Decorative
Laminate without Decontamination
A study was conducted to determine the persistence
of GB on galvanized metal ductwork and decorative
laminate using triplicate samples under environmental
conditions of a temperature of 22 °C ± 2 °C, air exchange
rate of one air change/hour (flow rate ± 10% of flow
rate) and RH of 40% ± 10%. The test was conducted in
the fumigation chamber and coupon samples were taken
and extracted after 5, 15, 30, 60 and 120 min of contact
time with the agent. A summary of the results follows in
Table 3.3-1.
Table 3.3-1 - Persistence of GB as Indicated by Agent
Recovery
GB, Mean % Recovery ±SD
T=5 T = 15 T=30 T=60 T=120
Decorative
Laminate
(n=3)
Galvanized
Metal
Ductwork
(n=3)
33
±
20
52
±
2.2
0.48
±
0.51
4.3
±
1.0
O.25
O.25
O.25
O.25
<0.25
<0.25
3.4 Persistence of HD on IBMs without
Decontamination
A study was conducted to determine the persistence
of HD on all the IBMs using triplicate samples under
environmental conditions of a temperature of 22 °C ± 2
°C, air exchange rate of one air change/hour (flow rate ±
-------
10% of flow rate) and RH of 40% ± 10%. The test
was conducted in the fumigation chamber and coupon
samples were taken and extracted after one hour, four
hours, one day, two days, and seven days of contact time
with the agent. A summary of the results follows (Table
3.4-1).
Table 3.4-1 - Persistence of HD as Indicated by Agent Recovery
HD, Mean % Recovery ± SD
Building
Material T=lHr T = 4Hr T=lday T = 2day T = 7day
Decorative
Laminate (n=3)
Galvanized
Metal Ductwork
(n=3)
Carpet (n=3)
Ceiling Tile
(n=3)
97 ± 1
97 ±2
98 ±5
97 ±2
62 ± 2
54 ± 6
67 ± 6
42 ± 5
0.25
0.25
23 ± 3
5 ± 1
0.25
0.25
14 ± 1
1.6 ± 2.E-01
0.25
0.25
7.3 ± 0.3
2.8E-01 ± 8E-02
3.5 Ambient Positive Controls for Determination of Fumigant Efficacy
The positive control tests were run with the test chamber
being held under ambient environmental conditions
(24 °C, 40% RH, 2.35 LPM flow). Three replicates
of each of the IBM and agent combinations were run.
The times the coupons resided in the positive control
chamber paralleled the times the coupons resided in
the fumigation chamber (Table 2.4-1). The ambient
positive control data, shown in Table 3.5-1, illustrate
that the agents persisted throughout the anticipated
decontamination exposure times at levels above the
method detection limits. The G-agents, TGD and GB,
were not as persistent as VX and HD, confirming other
positive control data generated in previous EPA efforts.3
Table 3.5-1 - Ambient Positive Control Data Summary
Ambient Positive Controls, % Recovery ± Standard Deviation (n=3)
Agent IBM 60 min 120 min 180 min 240 min 400 min
HD
GB
VX
TGD
GM
DL
CA
CT
CA
CT
GM
DL
CA
CT
GM
DL
CA
CT
84 ± 2
85 ± 1
83 ± 3
75 ± 3
2.3 ± 0.6
12 ± 1
97 ± 2
93 ± 7
88 ± 8
97 ± 4
48 ± 7
52 ± 3
40 ± 7
83 ± 2
76 ± 3
70 ± 2
75 ± 1
52 ± 3
1.6 ± 0.2
10 ± 1
91 ± 2
90 ± 3
88 ± 6
92 ± 3
18 ± 9
29 ± 5
30 ± 4
59 ± 3
64 ± 3
55 ± 3
54 ± 2
29 ± 2
1.8 ± 0.2
9.9 ± 0.8
91 ± 2
82 ± 2
75 ± 6
85 ± 9
3.6 ± 2.8
8.5 ± 1.9
14 ± 4
55 ± 4
51 ± 2.2
44 ± 2.6
47 ± 0.93
20 ± 1.7
1.3 ± 0.27
7.1 ± 1.8
86 ± 5
82 ± 2
74 ± 2
92 ± 3
0.7 ± 0.1
0.8 ± 0.1
15 ± 4
54 ± 2
24 ± 4
20 ± 3
33 ± 3
7.9 ± 1
0.8 ± 0.1
4.1 ± 0.5
76 ± 6
76 ± 2
76 ± 6
91 ± 4
0.6 ± 0.0
0.4 ± 0.2
9.3 ± 2
40 ± 2
-------
3.6 Recovery over Time of CWAs on
IBMs with Steam Fumigant
Technology
The results from the steam decontamination tests to
assess the recovery of CWAs applied to IBM samples
conducted at both 1.5 and 3 kg/hr are shown in Table
3.6.1.
The IBMs (n=5) with CWAs applied and procedural
blanks (PB) (n=2) were removed at four time periods
as indicated in the tables below. The percent recovery
results are based upon the amount of agent determined in
the dose confirmation reference samples analyzed with
each test (Section 2.5.4).
HD was not detected in any of the IBM extracts from the
samples collected after 120 minforthe 1.5 kg/hr steam
test and after 60 min for the 3 kg/hr steam test.
GB was not detected in any of the extracts from the
IBM samples collected after 60 min from the 1.5 kg/hr
steam test. TGD was not detected in any of the extracts
from the IBM samples collected after 60 min for the 1.5
or 3kg/hr steam tests, except for the carpeting samples
at 60 min for which an average of 0.5% (10 ug) was
recovered. A bar graph plot of the ug/sample of TGD
recovered for the 1.5 kg/hr test condition is shown in
Appendix C.
VX was detected in all extracts from the IBM samples
collected at 60 min, except for the 1.5 kg/hr CT samples.
At 120 min, 1-2% VX was present in the GM, DL, and
CA extracts for both the 1.5 and 3 kg/hr steam tests.
At 180 and 240 minutes, the extracts from the 1.5 kg/
hr decontaminated CA samples were the only sample
extracts containing detectable VX, recovering 1.0%
(20 ug) of the dose reference sample mass. VX was
not recovered in any of the extracts from IBM samples
collected beyond 180 min for the 3 kg/hr steam test.
Statistical analysis of these recoveries could not be
completed because all of test coupon recoveries (n=5)
were calculated using the masses recovered from the
test coupon (n=5) and the agent mass from the single
dose confirmation reference sample analyzed (n=l) with
each test (Section 2.5.4). Bar graphs of the ug/sample
of VX recovered from the IBM coupons can be found in
Appendix C.
Extracts of laboratory blanks (LB) (n=5) and PB (n=2)
were also analyzed for each test and none of these
extracts contained CWAs above the detection limits.
Table 3.6-1 - Recovery of HD on IBM Decontaminated with Steam at 1.5 and 3 kg/hr
Steam HD, Mean Recovery, %
Sample Rate kg/
Description hr 120 min 180 min 240 min 400 min
GM (n=5)
DL (n=5)
CA (n=5)
CT (n=5)
PB GM (n=2)
PB DL (n=2)
PBCA(n=2)
PB CT (n=2)
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
0.25
0.25
O.25
O.75
0.25
0.25
0.25
O.75
0.25
0.25
O.25
O.75
0.25
0.25
0.25
O.75
0.25
0.25
O.25
O.75
0.25
0.25
0.25
O.75
0.25
0.25
O.25
O.75
0.25
0.25
0.25
O.75
60 min 120 min 180 min 240 min
GM (n=5)
DL (n=5)
CA (n=5)
CT (n=5)
PB GM (n=2)
PB DL (n=2)
PBCA(n=2)
PB CT (n=2)
o
J
3
3
o
J
o
J
o
J
3
3
O.25
0.25
0.25
O.75
O.25
O.25
0.25
0.75
O.25
0.25
0.25
O.75
O.25
O.25
0.25
0.75
O.25
0.25
0.25
O.75
O.25
O.25
0.25
0.75
O.25
0.25
0.25
O.75
O.25
O.25
0.25
0.75
-------
Table 3.6-2 - Recovery of GB on IBM Decontaminated with Steam at 1.5 and 3 kg/hr
Steam GB, Mean Recovery, %
Sample Rate kg/
Description hr 60 min 120 min 180 min 240 min
CA(n=5)
CT (n=5)
PB CA (n=2)
PB CT (n=2)
1.5
1.5
1.5
1.5
<0.25
<0.75
<0.25
<0.75
<0.25
<0.75
<0.25
<0.75
<0.25
<0.75
<0.25
0.75
O.25
O.75
O.25
0.75
60 min 120 min 180 min 240 min
CA (n=5)
CT (n=5)
PB CA (n=2)
PB CT (n=2)
3
o
J
o
J
3
0.25
<0.75
<0.25
<0.75
<0.25
<0.75
<0.25
<0.75
0.25
O.75
O.25
0.75
0.25
O.75
O.25
0.75
Table 3.6-3 - Recovery of VX on IBM Decontaminated with Steam at 1.5 and 3 kg/hr
Steam VX Mean Recovery, % ± Standard Deviation
Rate,
IBM Kg/hr 60 min 120 min 180 min 240 min
GM (n=5)
DL (n=5)
CA (n=5)
CT (n=5)
PB
GM (n=2)
PB DL (n=2)
PB CA (n=2)
PB CT (n=2)
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
25 ± 9
3 ± 1
5.8 ± 1.4
0.75
O.25
O.25
O.25
O.75
1 ± 2E-01
1 ± 1 E-01
2 ± 2 E-01
0.75
O.25
O.25
O.25
O.75
O.25
O.25
2 ±8 E-02
0.75
O.25
O.25
O.25
O.75
O.25
O.25
7.0 E-01 ± 6 E-02
0.75
O.25
O.25
O.25
O.75
60 min 120 min 180 min 240 min
GM (n=5)
DL (n=5)
CA (n=5)
CT (n=5)
GM <-*>
PB DL (n=2)
PB CA (n=2)
PB CT (n=2)
3
3
o
J
o
J
o
J
3
3
3
12 ± 1
5.7 ± 0.7
6.0 ± 1.9
O.25
0.25
0.25
0.25
0.75
0.25
0.25
1.2 ± 2 E-01
O.25
0.25
0.25
0.25
0.75
0.25
0.25
8 E-01 ± 9 E-02
O.75
0.25
0.25
0.25
0.75
0.25
0.25
O.25
O.75
0.25
0.25
0.25
0.75
-------
Table 3.6-4 - Recovery of TGD on IBM Decontaminated with Steam at 1.5 and 3 kg/hr
Steam TGD, Mean Recovery, % +/- SD
Sample Rate kg/ fl. . .... . ._„ . „.„ .
_. . . , " 60 mm 120 mm 180 mm 240 mm
Descnption hr
GM (n=5)
DL (n=5)
CA(n=5)
CT (n=5)
PB GM (n=2)
PB DL (n=2)
PB CA (n=2)
PB CT (n=2)
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
<0.25
<0.25
<0.25
5E-01±2E-01
<0.25
<0.25
<0.25
<0.75
O.25
O.25
O.25
0.75
0.25
O.25
O.25
O.75
O.25
O.25
O.25
0.75
0.25
O.25
O.25
O.75
O.25
O.25
O.25
0.75
0.25
O.25
O.25
O.75
60 min 120 min 180 min 240 min
GM (n=5)
DL (n=5)
CA (n=5)
CT (n=5)
PB GM (n=2)
PB DL (n=2)
PBCA(n=2)
PB CT (n=2)
3
3
3
3
3
3
3
3
<0.25
<0.25
<0.25
<0.75
0.25
0.25
0.25
O.75
0.25
0.25
O.25
O.75
0.25
0.25
0.25
O.75
0.25
0.25
O.25
O.75
0.25
0.25
0.25
O.75
0.25
0.25
O.25
O.75
0.25
0.25
0.25
O.75
3.7 Steam Efficacy Results
The computed efficacies for steam decontamination
are shown in Tables 3.7.1, 3.7.2, 3.7.3 and 3.7.4. The
decontamination efficacy for steam decontamination
of HD was greater than 99 % for all materials for all
exposure periods and the decontamination efficacy for
steam decontamination of GB on carpet and ceiling tile
coupons was greater than 99% for all exposure periods.
The decontamination efficacy for steam decontamination
of TGD was equal to or greater than 99% for all
materials and exposure periods.
The decontamination efficacy for steam decontamination
of VX on ceiling tile was greater than 99% for all
exposure periods. For CT, DL and GM samples,
the minimum decontamination efficacy for steam
decontamination of VX was 82% at the 1.5 kg/hr steam
feed rate and 88% at the 3 kg/hr steam rate. Efficacy
increased over time and reached or exceeded 99% after
the first exposure period.
-------
Table 3.7-1 - HD Steam Decontamination (Decon) Efficacy
HD 1.5 kg/hr
Samples,
Ambient Positive
Controls (n=3), % Recovery
HD 3 kg/hr
Samples,
HD
% Recovery
% Recovery ± SD Decon ±SD Decon
±SD (n=5) Efficacy (n=5) Efficacy
GM
DL
CA
CT
60
120
180
240
400
60
120
180
240
400
60
120
180
240
400
60
120
180
240
400
84 ± 2
76 ± 3
64 ± 4
51 ± 2
24 ± 4
85 ± 1
70 ± 2
55 ± 3
44 ± 3
20 ± 3
83 ± 3
75 ± 1
54 ± 3
47 ± 1
33 ± 3
75 ± 3
52 ± 3
29 ± 2
20 ± 2
7.9 ± 0.6
-
0.25
0.25
O.25
O.25
-
0.25
0.25
0.25
O.25
-
0.25
0.25
0.25
O.25
-
0.75
0.75
0.75
O.75
-
>99%
>99%
>99%
>99%
-
>99%
>99%
>99%
>99%
-
>99%
>99%
>99%
>99%
-
>99%
>99%
>99%
>99%
O.25
0.25
0.25
O.25
-
O.25
0.25
0.25
0.25
-
O.25
0.25
0.25
0.25
-
O.75
0.75
0.75
0.75
-
>99%
>99%
>99%
>99%
-
>99%
>99%
>99%
>99%
-
>99%
>99%
>99%
>99%
-
>99%
>99%
>99%
>99%
-
- indicates that an efficacy test was not completed for this material-exposure time combination
Table 3.7-2 - GB Steam Decontamination (Decon) Efficacy
Ambient
Positive
Controls (n=3),
% Recoverv
GB 1.5 kg/hr
Samples,
GB
% Recoverv
GB 3 kg/hr
Samples,
% Recoverv
±MJ Decon * su Decon
(n=5) Efficacy (n=5) Efficacy
60
120
CA
180
240
60
120
CT
180
240
2
2
2
1
12
10
9.9
7.1
±
±
±
±
±
±
±
±
6E-01
2E-01
2E-01
3E-01
1
1
0.8
1.8
O.25
O.25
0.25
0.25
0.75
O.75
O.75
0.75
>99%
>99%
>99%
>99%
>99%
>99%
>99%
>99%
O.25
O.25
0.25
0.25
0.75
O.75
O.75
0.75
>99%
>99%
>99%
>99%
>99%
>99%
>99%
>99%
-------
Table 3.7-3 - VX Steam Decontamination (Decon) Efficacy
VX 1.5 Kg/hr
VX 3 Kg/hr
Ambient
Positive
Controls
Samples,
VX % Recovery
Samples,
VX % Recovery
Decon
Efficacy
Decon
Efficacy
GM
DL
CA
CT
60
120
180
240
60
120
180
240
60
120
180
240
60
120
180
240
97 ± 2
91 ± 2
91 ± 2
86 ± 5
93 ± 7
90 ± 3
82 ± 2
82 ± 2
88 ± 8
88 ± 6
75 ± 6
74 ± 2
97 ± 4
92 ± 3
85 ± 9
92 ± 3
25 ± 9
1 ± 2.E-01
O.25
O.25
3 ± 1
9.E-01 ± l.E-01
<0.25
<0.25
6 ± 1
2 ± 2.E-01
2 ± 8.E-02
7.E-01 ± 6.E-02
<0.75
0.75
0.75
O.75
74%
99%
>99%
>99%
97%
99%
>99%
>99%
93%
98%
98%
99%
>99%
>99%
>99%
>99%
12 ± 1
0.25
O.25
O.25
6 ± 7.E-01
0.25
O.25
O.25
6 ± 2
1 ± 2.E-01
8.E-01 ± 9.E-02
O.25
O.75
0.75
0.75
O.75
88%
>99%
>99%
>99%
94%
>99%
>99%
>99%
93%
99%
>99%
>99%
>99%
>99%
>99%
>99%
Table 3.7-4 - TGD Steam Decontamination (Decon) Efficacy
TGD 1.5 kg/hr
Samples,
Ambient Positive
Controls (n=3),
% Recovery
TGD % Recovery
Decon
Efficacy
TGD 3 kg/hr
Samples,
TGD %
Recovery
Decon
Efficacy
GM
DL
CA
CT
60
120
180
240
60
120
180
240
60
120
180
240
60
120
180
240
48 ± 7
18 ± 9
4 ± 3
7E-01 ± 1E-01
52 ± 3
29 ± 5
9 ± 2
8E-01 ± 1E-01
40 ± 7
30 ± 4
14 ± 4
15 ± 4
83 ± 2
59 ± 3
55 ± 4
54 ± 2
0.25
0.25
O.25
O.25
0.25
0.25
O.25
O.25
5E-01 ± 2E-01
0.25
O.25
O.25
0.25
0.75
O.75
O.75
>99%
>99%
>99%
>99%
>99%
>99%
>99%
>99%
87%
>99%
>99%
>99%
>99%
>99%
>99%
>99%
0.25
0.25
O.25
O.25
0.25
0.25
O.25
O.25
0.25
0.25
O.25
O.25
0.75
0.75
O.75
O.75
>99%
>99%
>99%
>99%
>99%
>99%
>99%
>99%
>99%
>99%
>99%
>99%
>99%
>99%
>99%
>99%
-------
3.8 CWAs in Condensate
Samples Collected during Steam
Decontamination
At the start of the each steam decontamination test,
1L of condensate was collected from the test chamber
and analyzed by GC/MS for CWAs. The elevated
temperature and humidity within the test chamber
were expected to have the combined effects of both
evaporation and hydrolysis of the CWAs. HD was not
detected (<0.02 ug/mL) in any of the 1L condensate
samples collected from either the 1.5 or 3 kg/hr steam
tests. VX was detected in the 1L condensate samples
collected from both the 1.5 and 3. kg/hr steam tests and
the concentration of VX ranged from 0.36 to 1.8 ug/
mL. GB was detected at 5.4 and 0.9 ug/mL in the 1L
condensate collected from 1.5 and 3 kg/hr steam tests,
respectively. TGD (analyzed as GD) was detected in the
condensate collected from both the 1.5 and 3 kg/hr steam
tests and the concentration of GD ranged from 1.5 to 7.5
ug/mL. Due to the lack of replicate samples statistical
analysis of these data could not be completed. Results
are presented in Table 3.8-1.
Table 3.8-1 - Results from the GC/MS Analysis
of Condensate Samples Collected during Steam
Decontamination Tests
Steam
Rate,
Sample Description kg/hr HD, jig/mL
HD Condensate - Test 1
HD Condensate - Test 2
HD Condensate - Test 9
HD Condensate - Test 10
1.5
1.5
3
3
0.02
<0.02
O.02
0.02
GB, ug/mL
GB Condensate - Test 4
GB Condensate - Test 12
1.5
3.0
5.4
0.9
VX, ug/mL
VX Condensate - Test 5
VX Condensate - Test 6
VX Condensate - Test 13
VX Condensate - Test 14
1.5
1.5
3
3
1.8
0.4
0.4
0.4
GD, ug/mL
GD Condensate - Test 7
GD Condensate - Test 8
GD Condensate - Test 15
GD Condensate - Test 16
1.5
1.5
3
3
1.5
2.4
7.5
2.8
3.9 Steam - IBM Compatibility
The coupons exposed to the fumigant for the longest
time were visually inspected and digitally photographed
upon removal from the chamber. Unexposed IBM
coupons were placed side by side with the coupons
exposed to agent/fumigant for comparison. The IBM
coupons were inspected for physical changes such as
discoloration, crumbling, warping, or blistering.
Ceiling tile and carpeting exhibited the most change. The
carpeting samples were saturated with moisture and the
edges of the ceiling tiles also appeared to have absorbed
moisture. The most significant material effect was that
the edges of the tile exhibited some crumbling. Visual
inspection did not identify any discernible difference in
the appearance of the IBM with varying steam fumigant
flow rates (1.5 and 3 kg/hr).
3.10 Recovery over Time of CWAs
on IBMs Using mVHP® Fumigant
Technology
The first set of tests compared the different output
flows of the generator (340 L/min versus 34 L/min) for
two target concentrations (250 and 350 ppmv) against
recovery of HD. These results are shown in Tables 3.11.1
through 3.11.4. The actual VHP® concentration for each
of the tests can be found next to the test number in
Tables 3.11.1 through 3.11.4. For the 250 ppmv tests the
recoveries were lowest for the tests conducted at the 340
L/min output flow. The 350 ppmv full flow (340 L/min)
tests also showed lower recoveries for the test coupons
relative to the 34 L/min flow tests. The higher target
concentration of mVHP® (350 ppmv versus 250 ppmv)
resulted in lower recoveries in both the 34 and 340 L/
min tests. The 34 L/min tests at 350 ppmv showed non-
detectable levels of agent on almost all materials except
CA after 180 minutes.
The presence of HD on the CA and CT procedural
blanks indicate that HD was desorbing from the surface
of the test coupons and adsorbing to the CA and CT
procedural blank coupons in the chamber during the
decontamination testing. The highest masses of HD
(reflected in the recoveries) were detected at the 34 L/
min flow configuration. During these lower flow tests
a more stagnant atmosphere existed in the chamber,
allowing the HD to adsorb to the porous surfaces (CA,
CT) instead of exiting the test chamber. These results
indicate that CA and CT serve as sinks for HD in indoor
environments and, because these materials are sinks, they
could later serve as sources of HD through gas-phase
emission of this agent.
The effect of the two different output flows was also
studied for VX and the results are shown in Tables 3.11-5
to 3.11-6. The actual VHP® concentration for each of the
tests can be found next to the test number in Tables 3.11-
5 to 3.11-6. The full flow (340 L/min) test samples had
lower recoveries than the one-tenth flow (34 L/min) test
samples indicating again that the increased output flow
-------
.**
=
CS
-^^
o
o
^H
O
•
•
^
o
v
«
-H
~
vo
-H
0
-H
?
OH
S
00
II
O
•
•
V~)
o
^
| i -H
CN fN
O O
CD
(N
fN fN
O O
v v -H
-
,
(N (N
O O
V V -H
0>
O
s
(N
(N o
o
-H
s
1
o
t--"
(N
II
O
m
PH
•nb a
a u
B op
> >
BBS,
1-1 <: H
Q o o
o
w
Q
m
PH
O
g
00
1)
1
o
8-
(N (N
-------
.
o
=
I
>•
a.
o
o
C
o
e«
-^^
O
I
•n in
_l
-------
.**
o
I
>
a.
a.
o
=
CS
-*^
o
o
^H
f»5
-H -H -H -H
-H -H -H -H
\o >—i ^H ^r
00 CT\ t~~ t~~
o
CN
-H -H -H -H
o
o
[>
00
-H -H -H -H
ft ft
ft ft
*o *o
CN CN
Pi Pi
CO CO
CN CN
in in
II II
EH CH
S i-l < H
5 Q O O
O O O O
V V V V
O O
V V
o
V
V V
o o o o
V V V V
o
V
111
S i-l < H
5 Q O O
m m m m
g PH PH PH
.**
a.
a.
o
O
O
o
o
1
o
-H -H -H -H
06 """ cK
-H -H -H -H
-H -H -H -H
CO CN '—i
OO CO '—i
-H -H -H -H
<
o
in in in in
(N
O O O O
V V V V
CN CN CN [>
o o o o
V V V V
CN CN l>
O O O
V V V
o o o o
V V V V
CN CN
II II
EH CH
>
ft
5 i-J < H
rH Q O O
m m m m
g PH PH PH
<
I
-------
resulted in less of the agent being recovered from the test agent on the coupon surface.
samples. Statistical analysis of these recoveries could not
be completed because all of test coupon recoveries (n=5)
were calculated using the masses recovered from the test
coupon (n=5) and the agent mass from the single dose
confirmation reference sample analyzed (n=l) with each
test (Section 2.5.4).
Extracts of the laboratory blanks did not contain any
CWAs above the detection limits.
3.11 mVHP® Efficacy Results
Efficacies for application of the mVHP®
decontamination technology were determined using
the ambient positive controls. The efficacies for the
HD tests are shown in Table 3.11-1. This table also
displays the actual VHP® concentrations for the HD
tests. Most of the efficacies for the 350 ppmv test
conditions were higher than those seen for the 250 ppmv
test conditions. These results appear to indicate that
increasing fumigant concentration slightly improved the
HD decontamination efficacy for most of the material-
exposure time combinations in these tests. More testing
is needed to determine if there is a significant difference
in decontamination efficacies for these two different
fumigant concentrations.
Exposure times of 180 min, at the 350 ppmv test
condition, are sufficient for achieving 99% efficacy or
better; however, carpet was the exception with only
90% efficacy being obtained at the 180 min exposure
time. Due to compression of the test schedule, further
decontamination exposure times could not be studied.
Increased contact times at this concentration may result
in a higher decontamination efficacy.
The decontamination efficacy for all HD-material
combinations was affected by the output flow from the
generator. For example, under mVHP® 250 ppmv full
flow conditions, the DL was decontaminated to > 99 %
after 120 min whereas at the 10% output, an efficacy of
only 88% was observed after 400 min. The efficacies for
the non-porous surfaces were better than those for the
porous surfaces with DL and GM having a greater than
99 % decontamination efficacy even at the reduced flow
350 ppmv conditions after 400 min.
These test results, along with the corresponding
environmental data (shown in Appendix D), indicate that
two factors significantly affect test results - output flow
and temperature. The individual effect of these variables
could not be ascertained from this testing. The resulting
chamber temperature differed with each of the output
flows. The 10% flow test chamber temperature was
typically around 24 °C while the chamber temperature
was around 40 °C during the full flow tests. Agent
evaporation is enhanced by elevated temperatures and
increased flow resulting in reduction of the amount of
-------
>n in
2 S
V V
'-H
in
rn
oo vi
'-H
o
-H-H
om
000000
•
• -H i-H-H
'— its'— i
-H-H-H-H-H
n >n
V V V
'-H
(N
^'ri
os
-H-H
"* -n
0<'0<'a<'
• -H
-H-H-H-H-H
H
O
"
a
3 3
00 1)
J3 -fci "-1
-------
Efficacies for the VX tests are shown in Table 3.11-
2. Actual VHP® concentrations are also shown in the
tables. These results indicate that even after 400 min the
greatest efficacy observed was 89% for the VX deposited
on galvanized metal. Due to compression of the test
schedule further decontamination exposure times could
not be studied.
The decontamination efficacy for all VX-material
combinations was also affected by the output flow/
temperature. The decontamination efficacies for the 34
L/min flow were all less than or equal to 32% at 400 min
while the efficacies for the full flow condition were 81-
89%. More research is needed to determine if increased
exposure times and/or increased fumigant concentration
will result in improved decontamination efficacies.
Table 3.11-2 - mVHP® VX Decontamination (Decon) Efficacy
Ambient
Positive
VX 10% output, 250 ppmv,
actual VHP® concentration
GM, DL (272 ppmv) and CT,
CA (261 ppmv)
Samples,
% Recovery
VX 100% output, 250 ppmv,
actual VHP® concentration
GM, DL (ND) and CT, CA (154
ppmv)
Samples,
% Recovery
Controls Decon Decon
IBM Mins (n=3) (n=5) Efficacy (n=5) Efficacy
GM
DL
CA
CT
120
180
240
400
120
180
240
400
120
180
240
400
120
180
240
400
91 ± 2
91 ± 2
86 ± 5
76 ± 6
90 ± 3
82 ± 2
82 ± 2
76 ± 2
88 ± 6
75 ± 6
74 ± 2
76 ± 6
92 ± 3
85 ± 9
92 ± 3
91 ± 4
97 ± 2
87 ± 4
86 ± 16
62 ± 5
95 ± 1
100 ± 20
91 ± 23
55 ± 13
96 ± 8
89 ± 7
71 ± 10
51 ± 11
93 ± 3
87 ± 9
74 ± 11
61 ± 5
#
4.7%
#
18%
#
#
#
28%
#
#
3.5%
32%
#
#
20%
32%
43 ± 8
32 ± 3
19 ± 3
8.1 ± 0.8
32 ± 3
23 ± 3
18 ± 2
11 ± 1
48 ± 11
32 ± 7
21 ± 5
9.7 ± 2.0
62 ± 3
51 ± 3
29 ± 5
17 ± 1
53%
65%
77%
89%
65%
72%
79%
85%
46%
57%
72%
87%
33%
41%
69%
81%
# indicates that the positive control recoveries were less than the decontamination test recoveries.
ND -Actual concentration of VHP™ was not determined for this test.
3.12 Discussion of Effect of mVHP®
Decontamination Diffuser
Configuration on Test Results
and Additional Positive Controls
During the initial mVHP® tests (PC 5 Test 17 and 18
and PC 6), uneven evaporation of agent drops and
positive control test results indicated a location bias
within the chamber. The bias was suspected to be a result
of the diffuser configuration (design and orientation).
A series of studies was conducted and modifications
were made to the diffuser configuration (Section 2.6.5).
The effect of the diffuser configuration on the test
was characterized by applying distilled water droplets
to the surface of a stainless steel pan to simulate the
evaporation of agent droplets. Three modifications to
the diffuser configuration were evaluated and a final
modification was incorporated for the remaining tests.
All mVHP® tests but Test 17 (diffuser Configuration 2)
and Test 18 (Configuration 3) were conducted with the
final diffuser configuration (Configuration 4). Tests 17
-------
and 18 consisted of HD 250 ppmv challenges at
10% flow. The corresponding tests (at 100 % flow at
250 ppmv - Tests 19 and 20) and (10% flow at 350
ppmv - Tests 25 and 26) have a degree of uncertainty
regarding the potential effect of the different diffuser
configurations. The absolute effect of the different
diffuser configurations is not clear.
As stated above in Section 2.5.6, additional positive
control experiments were run with the mVHP® generator
operating with DI water instead of hydrogen peroxide
solution where either 100% of the generator flow was
pushed through the test chamber (Tests 6, 10 for HD
and Test 7 for VX) and or 10% of the flow (Tests 5,
9 for HD and Test 8 for VX) was pushed through the
test chamber. PC-5 and PC-6 were run with the first
diffuser configuration and both of these tests had lower
observed average temperatures and relative humidities,
as shown in Tables D. 1, D.2, D.7, and D.8, than the
repeated positive controls (PC-9 and PC-10, respectively,
also shown in the same Tables). These lower average
temperatures and relative humidities may have been due
to uneven distribution of both the conditioning input air
as well as the fumigant. A larger percent recovery of
HD was observed for most of the positive controls run
with the earlier diffuser configuration for the reduced
flow tests (PC-5), as seen in Tables D.I, D.2, D.7, and
D.8, indicating that evaporative loss was likely occurring
during the positive control test run with the final diffuser
configurations (PC-9 and PC-10). In addition, the
positive control samples run at the full flow condition
in the final diffuser configuration (PC-9 and PC-10)
experienced higher average temperatures than the
corresponding ambient positive control test samples as
shown in Tables D-4 and D5. This diffuser configuration
was used during the HD decontamination testing and
similar average temperatures were observed during
this testing. Therefore, the full flow positive controls
indicate that some of the reduction in the HD agent mass
on the coupons during the full flow tests was likely due
to evaporation.
The VX positive control and decontamination tests
were all run with the final diffuser configuration and
the results from these tests are shown in Tables D-13
through D-18. The positive control samples run at
the full flow condition experienced higher average
temperatures than the corresponding ambient positive
control test samples as shown in Tables D-16 and D17.
The VX recoveries on all four building materials for
the full flow positive control samples also appear to be
lower than the VX recoveries observed for the ambient
positive control samples indicating that there were some
losses due to evaporation during the full flow positive
control tests. The positive controls were run at the
same diffuser configuration as the VX decontamination
tests and had a similar average temperature during the
decontamination phase, indicating that some evaporative
losses may have occurred during the VX mVHP®full
flow decontamination tests.
As discussed in Section 4.9, procedural blank coupons
(data shown in Tables 3.10-1 through 3.10-4) placed in
the chamber along with CWA-contaminated coupons
showed that HD was found on the CA and CT coupons
for many of the exposure periods. This demonstrates that
the chemical CWA vapor cross-contaminated the "clean"
samples. Non-porous procedural blank extracts (GM and
DL) contained less agent mass than the extracts of the
porous materials (CAand CT).
3.13 CWAs in Chamber Vapor
Samples Collected during mVHP®
Decontamination
During the mVHP® tests, vapor from the test chamber
was collected onto sorbent tubes to determine whether
vaporized CWA was present in the atmosphere of the
chamber during decontamination. Vapor samples were
collected onto solid sorbent tubes with a metered flow
rate for a known period of time. The sampling covered
the initial time period during the decontamination or
corresponding positive control test. Samples were
collected for a shorter period of time (60 versus 120 min)
at the higher air exchange rates (100% versus 10%) in
the test chamber. The contents of the sorbent tubes were
extracted with ethyl acetate and the resulting extract
analyzed using GC/MS.
The initial vapor study (Section 2.4.10) was conducted
using thermal desorption of the sorbent tubes prior to
GC/MS analysis. During the actual tests, the vapor
adsorbent tubes were extracted with 2 mL of ethyl
acetate to allow reanalysis or dilution of the samples
(it is not possible to reanalyze sorbent tubes following
thermal desorption), as needed.
-------
3.13.1 HD Vapor Sample Results
HD was detected in all the vapor samples but the mass
collected (0.001 to 0.625 mg) was relatively low in
comparison to the total mass of agent applied to the
coupons (80 to 100 mg). Vapor concentrations based
on the volume of vapor sampled over the periods of
time are presented in Table 3.13-1. The vapor sample
concentrations taken during the positive control tests
(PC-9 and PC-10) were higher than the test vapor sample
concentrations; however, a proper statistical analysis
of these data, to determine if this difference was due
to the absence of the fumigant, could not be completed
because the type of IBMs in the test chamber during the
positive control and test runs was different. There was
a significant difference (t- test of unequal variances,
95% confidence level, n=3, t , 4.3> t 3.5) between
' ' cnt stat '
the vapor samples collected during Tests 27 and 28
(100% flow rate, target 350 ppmv VHP®) indicating a
difference in the amount of agent available in the gas
phase for the two different groups of samples. The Test
28 samples taken with GM and DL in the test chamber
showed higher values of agent than the Test 27 samples
taken with CT and CA in the test chamber indicating the
agent was evaporating at a slower rate from the porous
materials. There were no significant differences between
Tests 25 and 26 indicating that the material type did not
affect the amount of available agent for this lower flow
rate. The two flow rates could not be compared because
their sampling times were different.
3.73.2 VX Vapor Sample Results
VX was not detected in any of the vapor samples. The
VX may have been present but the actual detection
limit for VX was much higher (0.036 mg/m3) than the
detection limit for HD because the VX sorbent tubes
required extraction with a solvent creating a significant
dilution factor (2 mL = 2000 x dilution factor). Lower
vapor concentrations were expected, at least for the
positive control vapor samples, because VX has a lower
vapor pressure than HD and as observed in the positive
controls, VX is more persistent on these IBMs than HD.
3.14 mVHP® - IBM Physical
Compatibility
No observable changes were noted for the physical
properties of the IBM coupons exposed to mVHP®. The
galvanized metal ductwork exhibited a white residue
at the agent application sites. There did not appear to
be any discernible differences in the appearance of the
coupons with different mVHP® fumigant flow rates.
Table 3.13-1 - mVHP® HD Vapor Results
CUBRC Test
Agent
STERIS
Output
Sampling Mean Concentration
period, and Standard Number of
min Deviation, mg/m3 replicates
PC-9
T-18
T-26
T-25
PC-10
T-19
T-20
T-28
T-27
GMDL
CACT
GMDL
GMDL
CACT
GMDL
CACT
GMDL
CA, CT
GMDL
CACT
HD
HD
HD
HD
HD
HD
HD
HD
HD
0
250
350
350
0
250
250
350
350
10%
10%
10%
10%
100%
100%
100%
100%
100%
0-120
0-300
0-120
0-120
0-60
0-60
0-60
0-60
0-60
13.2 ± 2.3
8.4 ± ND
7.4 ± 0.3
6.8 ± 0.3
5.4 ± 0.8
2.4 ± ND
2.5 ± 0.3
3.6 ± 4.E-03
2.4 ± 0.6
(n=3)
(n=l)
(n=3)
(n=3)
(n=3)
(n=l)
(n=3)
(n=3)
(n=3)
-------
4.1 ISO 9001 Audit
There were four findings during the ISO 9001 audit. All
findings were classified as minor and none were found to
affect the quality of the data.
4.2 Deviations from the QAPP
The mVHP® test matrix was modified as an additional
VHP® test concentration was added (350 ppmv) and GB
and TGD were not tested. The mVHP® test flow output
4.0
Quality Assurance
was changed from 425.0 L/min and 42.5 L/min to 340 L/
min and 34 L/min.
4.3 Quality Assurance Indicators
Table 4.3-1 contains data quality indicators that were
monitored in accordance with the QAPP. No findings
required corrective actions.
Table 4.3-1 - Measurements and Data Quality Indicators for Decontamination and Persistence Testing
Measurement
Parameter Method Data Quality Indicators
Temperature
Relative humidity
Air exchange rate in
chamber
Hydrogen peroxide
concentration
Ammonia
concentration
Agent on Positive
Control
Agent on Laboratory
Blank
Agent on Procedural
Blank
NIST-traceable
thermometer
NIST-traceable
hygrometer
Mass flow controller
mVHP®
electrochemical sensor
NH3 electrochemical
sensor
Extraction/GC
Extraction/GC
Extraction/GC
Compare against calibrated thermometer before and after
experiment, agree ± 10%
Compare against calibrated hygrometer before and after
experiment, agree ± 10%
Compare to second NIST-traceable calibrated flow meter before
and after experiment, agree ± 10%
Compare to AATCC Test Method 102-2007 (AATCC, 1957),
agree ± 10%
Run NH3 calibration gas, agree ± 10%
The relative standard deviation of the percent recoveries at each
time point should be < ±25%.
Laboratory blanks should have less than 1% of the amount of
analyte compared to that found on positive controls.
Procedural blanks should have less than 10% of the amount
(recovery corrected) compared to that found on positive controls.
4.4 Temperature and Relative Humidity
During the ambient positive control tests the RH
and temperature were measured with an internally
chamber-mounted calibrated VWR® hygrometer/
temperature probe (35519-020 VWR®, Bridgeport, NJ
08014). During the steam shakedown tests a bare wire
thermocouple ( CHAL-020 Type K, Omega, Stamford,
CT 06907) was inserted into the chamber verifying that
the temperature was 100 °C. During the remainder
of the testing the temperature was monitored on the
exterior of the chamber between the heating blankets
and the chamber exterior wall. This temperature was
maintained at > 100 °C throughout the testing. The
VWR® hydrometer (VWR.com, DH-011 - Cal dates
1/09 - 1/10) was also used to monitor the laboratory
temperature and humidity.
During the mVHP® experiments (decontamination
tests and positive controls), the RH and temperature
were monitored by the vendor-provided sensors. The
STERIS unit was plumbed into the glovebox along with
a STERIS remote sensor box containing the temperature,
humidity, hydrogen peroxide and ammonia sensors.
The first STERIS sensor bundle did not work properly
and STERIS had to reconfigure the unit. The STERIS
reconfigured unit's sensors had the following calibration
dates: 5/07-5/08 and 6/07-6/08. A replacement sensor
unit that was within calibration did not arrive in time for
incorporation into the testing (>50 % of the test program
was complete). Neither the hydrometer nor the STERIS
Humidity and Temperature Sensor were compared to a
-------
NIST-calibrated thermometer and hygrometer.
4.5 Air Exchange Rate
4.5.1 Steam Air Exchange
The steam air exchange rate was 0.016 exchanges/min.
This air exchange rate was dictated by the feed rate of
the steam into the test chamber.
4.5.2 mVHP®Air Exchange
There were two mVHP®air exchange rates; one for
the high flow (100% - 340 L/min) = 2.3 air exchanges
per min and one for the low flow (10% - 34 L/min) =
0.23 air exchanges per minute. The flow output was
controlled by a Venturi Flow Meter (STERIS ARD 1000,
STERIS, Mentor OH 44060 )with a differential pressure
sensor. STERIS reported that the unit was calibrated on
July 27, 2008. This flow meter was not compared to a
NIST-calibrated flow meter.
The ammonia output was controlled by an AALBORG
0-100 mL/minflow controller supplied by STERIS
(calibrated June 26, 2008). This flow meter was not
compared to a NIST-calibrated flow meter.
4.5.3 Hydrogen Peroxide Concentrations
The stock hydrogen peroxide solutions used to generate
VHP® were titrated following AATCC Test Method 102-
2007s The intention of this QA indicator was to verify
the sensor-determined VHP® vapor concentrations by
pulling an air sample from the test chamber through an
impinger filled with acidified potassium permanganate
and then titrating this solution. Because the hydrogen
peroxide stock solutions used for generation of the
VHP® were titrated instead, these titration results along
with the test duration, air flow rate and mean rate of
delivery of the titrated solution through the STERIS
unit (information obtainable from the data file generated
by the STERIS unit) were used to calculate a mean
theoretical hydrogen peroxide vapor concentration for
the duration of a test. In addition, condensation was not
observed in any of the tests and the RH and temperatures
achieved in the test chamber also did not indicate that
condensation occurred during any of the testing. No
losses of the vapor to the condensation phase should
therefore be occurring. The calculated concentrations
were compared to the concentrations reported by the
STERIS unit sensor (used in the feedback loop). In Table
4.5-2 the calculated concentrations and the average
sensor concentrations are shown. The sensors were
biased high at the higher flow tests, likely due to the
increased temperature during these runs.
Table 4.5-2 - VHP™ Concentrations and Ammonia Concentrations for Each of the Test Runs
Average Delta
VHP8 Temperature
STERIS Target sensor Average (from
output - VHP® - reading - Average Temperature ambient
Average
Calculated
RH - %
25 °C) - °C concentration
17
18
19
20
21
22
23 R
24
25
26
27
28
HD
HD
HD
HD
VX
VX
VX
VX
HD
HD
HD
HD
34
34
340
340
340
340
34
34
34
34
340
340
250
250
250
250
250
250
250
250
350
350
350
350
268
259
257
247
254
250
264
262
353
355
347
340
47
60
36
43
38
37
50
55
54
62
37
41
28
26
37
37
37
36
26
26
29
30
42
40
4
2
13
13
13
12
2
2
5
6
18
16
210
246
157
158
X
154
261
272
394
316
219
215
28%
5%
64%
56%
X
63%
1%
-4%
-10%
12%
59%
58%
X - Peroxide solution not titrated.
-------
4.6 Ammonia Concentration
4.6.7 Ammonia Concentration Control
During testing the ammonia sensor was observed to
be saturated even at the 16 ppmv concentration. This
saturation did not allow the conditioning phase of the
fumigant generation to be completed successfully.
Software control of the flow controller was therefore
overridden and manual control was implemented by
connecting the flow controller to a Hastings flow control
box (Model 400, Teledyne Hastings Instruments,
Hampton, VA 23669) for all remaining tests (Table 4.6-1
- manual control tests are highlighted in yellow).
Two distinct levels of operation were programmed
into the unit: 250 and 350 ppmv hydrogen peroxide
concentration. The generator-selected ammonia levels
were 6% of the target hydrogen peroxide concentration.
All of the above parameters were set up and tested at the
340 L/min output flow. The ammonia levels recorded in
full flow checkout runs were then used to establish the
flow rates programmed into the external flow controller.
The following ammonia flow rates were established for
each test type:
100% flow = ammonia flow rate of 5.5 g/min
10% flow = ammonia flow rate 4.1 g/min.
The STERIS unit was pre-programmed to set
the ammonia concentration during the ammonia
characterization cycle prior to the conditioning cycle
which starts the introduction of hydrogen peroxide. The
ammonia levels established by the STERIS unit's logic
always came in above the 6% target value (typically
-10%). In order to maintain some order of consistency
in the testing, the levels targeted for the 350 ppmv runs
were adjusted to the actual higher ratios (-10%) defined
in the 250 ppmv test and not the software-targeted value
of 6%.
These values closely approximate the average values
generated on previous runs with valid "ammonia
characterization" phase operation.
4.6.2 Verification of Ammonia Concentration
The results from the Drager tube measurements, taken
during Tests 17, 18 and 19, are shown in Table 4.6-2.
The Drager measurements consistently indicated a lower
ammonia concentration (40-50%) than the 1000-ARD
sensor. The presence of the VHP® may have affected the
Drager measurement, but from the offline measurements
by the Drager tubes the concentration measured
during the experiments was relatively consistent from
measurement to measurement and test to test. Drager
sampling was therefore discontinued after Test 19.
Table 4.6-1 - Ammonia Generation Data for the
Various Tests
CUBRC Ammonia Ammonia STERIS
Test ID Date Agent ppmv g/min output
PC-5
17
18
PC-6
19
20
PC-7
21
22
PC-8
23
23 R
24
25
26
PC-9
PC-10
27
28
05/19/09
05/21/09
05/22/09
05/25/09
05/26/09
05/27/09
05/28/09
05/29/09
06/01/09
06/02/09
06/03/09
06/04/09
06/05/09
06/08/09
06/09/09
06/10/09
06/11/09
06/12/09
06/12/09
HD
HD
HD
HD
HD
HD
VX
VX
VX
VX
VX
VX
VX
HD
HD
HD
HD
HD
HD
na
28
21
na
23
43
na
43
37
na
na
4.18
3.27
na
4.18
6.18
na
6.2
5.29
na
ABORTED
41
33
39
40
na
na
41
38
5.54
4.18
4.63
4.54
na
na
4.71
4.41
10%
10%
10%
100%
100%
100%
100%
100%
100%
10%
10%
10%
10%
10%
10%
10%
100%
100%
100%
na = Indicates that these tests are the positive controls. The fumigant was not
introduced into the chamber during these tests.
Yellow highlight indicates manual control of NH3 flow controller.
Table 4.6-2 - Drager Tube Ammonia Measurements
Test 19
Tube 2
0-100
ARZH-0551
<5
12
4.7 Agent on Positive Controls
The data quality indicator for the positive controls is
that the relative standard deviation of the triplicate set
of positive control percent recoveries at each time point
should be < 25%. This criterion was met for most of
the triplicate positive control sets except for the GB on
carpet and the TGD on galvanized metal, decorative
laminate, and carpet. The relative standard deviations
appear to be highest for the GB and TGD.
-------
Table 4.8-1 - Relative Standard Deviation for
Ambient Positive Controls
Relative Standard Deviation for Ambient Positive
Controls
HD
GB
VX
TGD
GM
DL
CA
CT
CA
CT
GM
DL
CA
CT
GM
DL
CA
CT
1.2
3.1
3.4
27
5.5
1.7
7.2
9.4
4.5
15
6.7
17
2.7
2.9
1.6
6.0
9.4
5.1
2.2
3.1
7.3
3.7
48
17
14
5.4
4.7
4.5
5.9
13
8.2
2.2
2.3
7.5
10
76
22
31
6.5
4.2
5.9
2.0
20
26
6.2
1.9
2.1
2.9
14
18
23
17
17
10
7.6
16
12
7.6
2.6
7.6
4.2
44
18
3.9
were analyzed every 10 samples. The daily check of the
calibration curve at one midpoint level had to be within
± 25% of the actual concentration. The method detection
limit for the CWAs is typically 5 ng depending upon the
substrate being analyzed.
Table 4.10-1 - Equipment Calibration Schedule
Responsible
Equipment Group Frequency
RH probe
Gas Sensors
Volumetric
Flow Controller
Thermocouples
Manufacturer
Manufacturer
Teledyne Hastings
CUBRC Personnel
Prior to
testing
Prior to
testing
Prior to
testing
One-time
two point
calibration
4.8 Agent on Laboratory Blanks
Agent was not detected in any of the mVHP® or steam
laboratory blank extracts.
4.9 Agent on Procedural Blanks
Procedural blank coupons placed in the chamber along
with CWA-contaminated coupons showed that in
only a few cases (HD on CA and CT) the CWA vapor
cross-contaminated the "clean" samples. Non-porous
procedural blank extracts (GM and DL) contained less
agent mass than the extracts of the porous materials (CA
and CT).
4.10 Equipment Calibration
The instrumentation and QA/QC procedures used to
determine chemical agents are identified in Section
2.5.6. The analytical equipment needed for the analytical
methods was maintained and operated according to
the quality requirements and documentation of the
Ashford Test Facility. All equipment was calibrated with
the appropriate standards. Table 4.10-1 contains the
equipment calibration schedule.
The Agilent GC/MS used for the analysis of the coupon
extracts and vapor samples collected on sorbents was
checked prior to beginning the analysis of each batch of
samples using a minimum of five calibration reference
standards. The GC/MS was recalibrated if the square
of the correlation coefficient (r2) from the regression
analysis of these standards was O.99. With each batch
of samples continuing calibration verification standards
-------
5.0
Conclusions
This research program addressed the following specific
questions:
• What is the decontamination efficacy of steam for
removal of CWAs on IBMs as compared to controls
at ambient environmental conditions?
• What is the decontamination efficacy of modified
Vaporous Hydrogen Peroxide(VHP®) (mVHP®),
modified in that ammonia is added to VHP®,
for removal of CWAs on IBMs under various
environmental and operational conditions as
compared to controls at ambient environmental
conditions?
• What are the physical effects of the decontaminants
on the IBMs?
Prior to addressing these questions, persistence studies
were completed. The persistence studies, completed for
HD, indicate that the agent remained on the coupons for
time periods equivalent to the selected decontamination
exposure times; but HD did not persist beyond the
4-hour time point on the non-porous surfaces (DL and
GM). The HD persistence was further studied in the
ambient positive controls, which were run under the
identical environmental conditions. These controls
indicate that HD persisted for 400 min on all of the
IBMs.
Persistence studies for GB deposited onto GM and DL
demonstrate that GB falls below detectable limits within
30 minutes. These results led to these material-agent
combinations being dropped from the decontamination
efficacy and corresponding ambient positive control test
matrices.
The steam results indicate a decontamination efficacy
of greater than 99% for HD, GB and TGD within 60
minutes for all IBM types at both feed rates (1.5 and 3
kg/hr). The efficacy of steam decontamination against
VX was >99% after 180 min exposure time for all
materials except CA. Increased steam output appears
to lead to higher decontamination efficacy for the VX
on the carpet (98% efficacy at 180 min for 1.5 kg/hour
output versus >99% at 180 min for 3 kg/hr). Due to
normalization of the recovery masses to the single dose
confirmation mass, statistical treatments of these data to
discern if this difference is significant was not possible.
Detectable concentrations of VX, GB, and GD were
found in the condensate, indicating that these agents are
not completely hydrolyzing during application of the
steam and that the IBM decontamination process is a
combination of physical removal and hydrolysis. Finally,
the steam fumigant did impact both porous materials
(CA and CT) causing permanent damage to the CT
The mVHP® results appear to indicate that increasing
fumigant concentration slightly improved the HD
decontamination efficacy for most of the material-
exposure time combinations in the test matrix. The best
efficacies were observed for the full flow conditions,
with efficacies of 99% or better for all materials, except
for CT, at the 350 ppmv target concentration conditions.
The most significant findings of this study were related
to the effect of the generator output flow. This effect was
seen for all HD-material combinations. For example,
under mVHP® 250 ppmv full flow conditions, the DL
was decontaminated to > 99% after 120 min whereas at
the 10% output, an efficacy of only 88% was observed
after 400 min. This effect was also seen for porous
surfaces where 99% efficacy was observed at 120 min
for CT, at the 350 ppmv full flow condition, while only
34% efficacy was observed at 120 min for the 10% flow
condition. This same effect was also seen for VX where
the decontamination efficacies for the 34 L/min flow
were all less than or equal to 32% at 400 min while the
efficacies for the full flow condition were 81-89%.
These test results, along with the corresponding
environmental data (shown in Appendix D), indicate that
two factors significantly affect test results - output flow
and temperature. The individual effect of these variables
could not be ascertained from this testing. The resulting
chamber temperature differed with each of the output
flows. The 10% flow test chamber temperature was
typically around 24 °C while the chamber temperature
was around 40 °C during the full flow tests. Agent
evaporation is enhanced by elevated temperatures and
increased flow resulting in reduction of the amount of
agent on the coupon surface. Evaporation of HD was
confirmed by the presence of the agent in all of the
sorbent tube samples taken during decontamination
testing.
The various diffuser configurations also complicated the
analysis of the mVHP® data because these configurations
may have altered the amount of fumigant each of the
individual coupons was exposed to during a test. The
standard deviations in the percent recoveries in the test
samples were not higher than the standard deviations in
the percent recoveries in the ambient positive controls
indicating that these deviations were due to the dosing
and extraction of the agent from the IBMs. These results
-------
show that there was at least some homogeneity (within
the area occupied by the 5 replicate coupons) in the
fumigant-agent interactions and possibly the fumigant
concentration.
An additional variation in the mVHP® test parameters
was the gas concentration data generated by the vendor-
provided sensors that were operated in a feedback loop
with the fumigant generator. These sensors did not
allow the VHP® or ammonia concentrations to reach
and maintain their target concentrations. The VHP®
sensor exhibited a positive bias during the tests where
the test chamber temperature was elevated (the full flow
tests) and the ammonia sensors saturated at their target
concentration levels. During the reduced flow test runs,
the VHP® concentrations were within 28% of the target
concentrations yielding at least one set of test data that
was generated using the vendor-recommended VHP®
concentrations. The target ammonia concentration
during all of the mVHP® tests exceeded the vendor-
recommended concentration (10% of the target VHP®
concentration versus the vendor-recommended 6%)
but the generator typically operates at this higher
concentration. This excess of ammonia should not affect
the HD and VX test results because the purpose of the
ammonia is to make the fumigant reactive towards the
G-agents. Finally, the mVHP® fumigant did not impact
the materials causing only a white residue to form where
the agent droplet was applied onto the GM.
Scaling of these fumigation technologies was the largest
technical challenge encountered during implementation
of this testing program. For the interior of buildings, the
air exchange rate is a significant factor in the reduction
of contaminants. Any efficacy testing must consider
the impact of air exchange (created by the generator
output flow) on the results and this testing attempted to
mimic the air exchange rate that is typical in a building
or building section undergoing decontamination. As
the testing progressed, the limitations of trying to adapt
a system designed for use in large spaces to a small
volume test chamber became increasingly apparent.
Considerable effort was expended in trying to mitigate
artifacts from testing discussed above (design changes
to VHP® diffuser and STERIS sensor interface). The
STERIS peroxide and ammonia flow controllers and
software were not designed for small areas. Several
"patches" were made over time to resolve issues as they
were identified. The test schedule also impacted the
program as several delays occurred, and the STERIS unit
had limited availability which did not allow an extension
to the test schedule. As a result, high priority was given
to running tests within a short period of time resulting
in delayed review of test data and test conditions. The
consequence of this accelerated test schedule was that
any potential test issues that negatively impacted the
results could not be addressed before further testing
proceeded.
Conversely, scaling-up of the steam fumigation
technology could be problematic. The presence of
the GB, TGD and VX in the condensate indicated that
if the steam fumigation were used to decontaminate
the interior of a facility or section of a facility there
would likely be re-deposition of these agents on other
surfaces. However, the non-detectable levels of the
agents on the procedural blanks directly adjacent to the
test coupons indicate that the steam might be a suitable
decontamination method if it were used in small areas
where the condensate could be collected, such as a steam
cleaner.
There are many different research questions that remain
unanswered related to these two fumigants. First, the
efficacies of these fumigants against the agents in vapor
form are likely different from the efficacies observed
for the liquid forms as determined under this effort.
Future testing with the mVHP® and steam should include
efficacy testing against vapor-contaminated samples. In
addition, the effects of agent dwell time, especially for
the persistent agents like VX, on the decontamination
efficacy for these fumigants should be studied. Any
future mVHP® testing should also incorporate the final
diffuser configuration as well as adjustment of the VHP®
target levels in the software to allow the concentrations
to reach the vendor-recommended levels. Lastly,
additional testing with the mVHP® fumigant should also
include longer decontamination exposure times and
additional materials.
-------
6.0
Appendices
Appendix A - Determination of Extraction Efficiency of CWAs
From IBMs
Table A. 1 - Results of Solvent Extraction of VX from IBMs
Table A.2 - Results of Solvent Extraction of TGD from IBMs
Table A.3 - Results of Solvent Extraction of HD from IBMs
Table A.4 - Results of Solvent Extraction of GB from IBMs
Table A.5 - GB Persistence Time Study
-------
Table A.1 - Results of Solvent Extraction Study for Extracting VX from IBMs
Sample ID
Reference in 10 mL
solvent, Rl
Reference in 10 mL
solvent, R2
Reference in 10 mL
solvent, R3
Mean
Blank decorative laminate
Decorative laminate, Rl
Decorative laminate, R2
Decorative laminate, R3
Mean
Blank metal ductwork
Metal ductwork, Rl
Metal ductwork, R2
Metal ductwork, R3
Mean
Blank industrial grade
carpet
Industrial grade carpet,
Rl
Industrial grade carpet,
R2
Industrial grade carpet,
R3
Mean
Blank ceiling tile
Ceiling tile, Rl
Ceiling tile, R2
Ceiling tile, R3
Mean
nexane j^myi Acetate
mg/ %
sample recovery
1.84 91
1.95 97
1.69 84
91
0.00 0
2.18 108
2.16 107
2.14 106
107
0.00 0
1.86 92
2.14 106
1.94 96
98
0.00 0
2.06 102
2.19 109
2.18 108
106
0.00 0
1.76 87
1.51 75
1.53 76
79
mg/ %
sample recovery
1.94 96
2.05 101
2.21 110
102
-
-
-
0.00 0
2.11 105
2.18 108
2.54 126
113
0.00 0
2.17 108
1.94 96
1.89 94
99
Methylene Chloride Hexane:Acetone
mg/ %
sample recovery
2.08 103
2.14 106
1.98 98
102
-
-
-
0.00 0
1.98 98
2.12 105
2.15 106
103
0.00 0
1.67 83
1.47 73
1.43 71
75
mg/ %
sample recovery
2.08 103
2.18 108
1.85 92
101
-
-
-
0.00 0
2.09 104
2.43 121
2.30 114
0.00 0
2.15 107
1.87 93
1.96 97
99
Agent was applied as four 0.5 uL droplets on each sample to result in 2.016 mg VX per coupon.
Laminate, ductwork and carpet samples were extracted in 10 mL of solvent. Ceiling tile was extracted in 20 mL solvent.
Coupons were extracted in solvent by ultrasonication for 10 minutes.
-------
Table A.2 - Results of Solvent Extraction Study for Extracting TGD from IBMs
Sample ID
Reference in 10 mL
solvent, Rl
Reference in 10 mL
solvent, R2
Reference in 10 mL
solvent, R3
Mean
Blank decorative
laminate
Decorative
laminate, Rl
Decorative
laminate, R2
Decorative
laminate, R3
Mean
Blank metal
ductwork
Metal ductwork,
Rl
Metal ductwork,
R2
Metal ductwork,
R3
Mean
Blank industrial
grade carpet
Industrial grade
carpet, Rl
Industrial grade
carpet, R2
Industrial grade
carpet, R3
Mean
Blank ceiling tile
Ceiling tile, Rl
Ceiling tile, R2
Ceiling tile, R3
Mean
Hexane Ethyl Acetate Methylene Chloride 1:1 Hexane: Acetone
mg/ %
sample recovery
1.78 87
1.85 91
1.84 90
89
0.00 0
1.74 85
1.84 90
1.87 91
89
0.00 0
1.54 75
1.99 98
2.00 98
90
0.00 0
1.79 88
1.75 86
1.84 90
88
0.00 0
1.22 60
1.52 75
1.40 69
68
mg/ %
sample recovery
1.61 79
1.82 89
1.89 93
87
_
_
_
_
-
_
_
_
_
0.00 0
1.96 96
1.68 82
1.92 94
91
0.00 0
1.89 93
1.87 92
2.05 100
95
mg/ %
sample recovery
1.94 95
1.79 87
1.80 88
90
_
_
_
_
-
_
_
_
_
0.00 0
1.49 73
1.88 92
1.81 89
85
0.00 0
1.65 80
1.96 96
1.64 80
86
mg/ %
sample recovery
1.99 98
1.76 86
2.17 106
97
_
_
_
_
-
_
_
_
_
0.00 0
1.83 90
1.82 89
1.71 84
87
0.00 0
1.74 85
1.98 97
1.85 91
91
Agent was applied as four 0.5 uL droplets on each sample to result in 2.04 mg GD per coupon.
Laminate, ductwork and carpet samples were extracted in lOmL of solvent. Ceiling tile was extracted in 20 mL solvent.
Coupons were extracted in solvent by ultrasonication for 10 minutes.
-------
Table A.3 - Results of Solvent Extraction Study for Extracting HD from IBMs
Sample ID
Reference in 10 mL
solvent, Rl
Reference in 10 mL
solvent, R2
Reference in 10 mL
solvent, R3
Mean
Reference in 20 mL
solvent, Rl
Reference in 20 mL
solvent, R2
Reference in 20 mL
solvent, R3
Mean
Blank decorative laminate
Decorative laminate, Rl
Decorative laminate, R2
Decorative laminate, R3
Mean
Blank metal ductwork
Metal ductwork, Rl
Metal ductwork, R2
Metal ductwork, R3
Mean
Blank industrial grade
carpet
Industrial grade carpet, Rl
Industrial grade carpet, R2
Industrial grade carpet, R3
Mean
Blank ceiling tile
Ceiling tile, Rl
Ceiling tile, R2
Ceiling tile, R3
Mean
nexane
mg/ %
sample recovery
2.63 104
2.67 105
2.65 104
104
2.68 106
2.60 103
2.62 103
104
0.00 0
2.83 112
2.77 109
2.73 107
109
0.00 0
2.77 109
2.72 107
2.69 106
107
0.00 0
2.62 103
2.47 97
2.79 110
103
0.00 0
2.71 107
2.78 109
2.75 108
108
• MJMRTJ
mg/ %
sample recovery
2.82 111
2.71 107
2.77 109
109
-
-
-
0.00 0
2.82 111
2.58 102
2.59 102
105
0.00 0
2.78 109
2.78 109
2.75 108
109
Methylene
Chloride
mg/ %
sample recovery
2.88 113
99
2.97 117
110
-
-
-
0.00 0
2.52 99
2.64 104
2.66 105
103
0.00 0
2.83 111
2.64 104
2.77 109
108
•M
m££jjj2£j2£2liifl
mg/ %
sample recovery
2.62 103
a/ a/
3.00 118
111
-
-
-
0.00 0
2.15 108
2.59 102
2.51 99
103
0.00 0
2.96 117
2.77 109
2.60 102
109
a/Sample Lost
Agent was applied as four 0.5 uL droplets on each sample to result in 2.54 mg HD per coupon.
Laminate, ductwork and carpet samples were extracted in 10 mL of solvent. Ceiling tile was extracted in 20 mL solvent.
Coupons were extracted in solvent by ultrasonication for 10 minutes
-------
Table A.4 - Results of Solvent Extraction Study for Extracting GB from IBMs
Sample ID
Reference in 10 mL
solvent, Rl
Reference in 10 mL
solvent, R2
Reference in 10 mL
solvent, R3
Mean
Blank decorative
laminate
Decorative laminate,
Rl
Decorative laminate,
R2
Decorative laminate,
R3
Mean
Blank metal ductwork
Metal ductwork, Rl
Metal ductwork, R2
Metal ductwork, R3
Mean
Blank industrial grade
carpet
Industrial grade carpet,
Rl
Industrial grade carpet,
R2
Industrial grade carpet,
R3
Mean
Blank ceiling tile
Ceiling tile, Rl
Ceiling tile, R2
Ceiling tile, R3
Mean
Hexane
mg/ %
sample recovery
1.73 79
1.78 81
1.61 73
77
0.00 0
1.59 72
1.44 65
1.76 80
72
0.00 0
1.91 87
1.86 84
1.82 83
84
0.00 0
1.99 90
1.73 79
1.70 77
82
0.00 0
1.46 66
1.39 63
1.26 57
62
Ethyl Acetate *
mg/ %
sample recovery
1.91 86
1.84 84
1.21 55
75
-
-
-
-
-
-
-
-
0.00 0
2.25 102
2.77 126
2.72 124
117
0.00 0
2.74 124
2.76 125
2.78 126
125
Methylene 1:1
Chloride Hexane:Acetone
mg/ %
sample recovery
1.73 79
1.70 77
1.68 76
77
-
-
-
-
-
-
-
-
0.00 0
2.02 92
1.83 83
1.88 85
87
0.00 0
1.69 76
1.69 77
1.84 84
79
mg/ %
sample recovery
1.90 86
1.62 73
1.27 58
73
-
-
-
-
-
-
-
-
0.00 0
1.77 80
1.80 82
1.84 83
82
0.00 0
1.57 71
1.56 71
1.44 65
69
* Chromatography of GB in ethyl acetate is not suitable for testing due to extreme tailing.
Agent was applied as four 0.5 uL droplets on each sample to result in 2.18 mg GB per coupon.
Laminate, ductwork and carpet samples were extracted in 10 mL of solvent. Ceiling tile was extracted in 20 mL solvent.
Coupons were extracted in solvent by ultrasonication for 10 minutes
-------
Table A.5 - Results of GB Persistence Study on Glass
GB Persistence Time Study
t
a
c
I
l
i
i
1
i
1
L
1
T=120s, R3
T=120s, R2
T=120s. R1
T=60s, R3 .JS
T=60s, R2 8
Ml
T=6Qs, R1
*-
2
•a
.2
"E.
ft
a
1/3
"a
M
a
1
a
^
a
JE
pfi
•a
.0
ft
I
**
•^
.a
j^*
S
a
a
S
"a
1
c*
OS
O
ne whether
i
d>
Q
Purpose
solvent.
"a
.a
S
(S
•a
_
•
0
IM
e
.2
"*^
s
*-i
0>
t/3
a
ft
E
o
>-t
•a
_3
tfi
tfi
la
H
S
a
a
••
a
s2
•a
a
0
}
a
S
o
H
ft
ft
a
a
•t
I
0
M
a
•a
B
a
x
«
JS
M
_B
*C
Q
Hi
_§
*•
recoverj
^
E
90
ri
i/
ft
i
vi
l
•a
a
u
•a
_a
fi
"a
1
•g
o>
^
«
"S
a
•a
a
O
"S
"S.
0
•-
•a
-J
d
1-
1
a
•a
1
a
a
M
S
a
a
S
o
!••
S
a
•a
"S
u
'a
0
a
0
!t
"ft
a
If!
-------
Appendix B - Determination of Method Detection Limits
(MDLs)
Table B.I - Determination of Method Detection Limit (MDL) - Decorative Laminate Coupons
Table B.2 - Determination of Method Detection Limit (MDL) - Galvanized Steel Coupons
Table B.3 - Determination of Method Detection Limit (MDL) - Ceiling Tile Coupons
Table B.4 - Determination of Method Detection Limit (MDL) - Carpet Coupons
-------
Table B.I - Determination of Method Detection Limit
(MDL) - Laminate Coupons
Table B.3 - Determination of Method Detection Limit
(MDL) - Carpet Coupons
GB,
Sample Description ng
Laminate Rep 1
Laminate Rep 2
Laminate Rep 3
Laminate Rep 4
Laminate Rep 5
Laminate Rep 6
Laminate Rep 7
STD Dev.
GC/MS MDL
(STDDEVx3.143)
0.68
0.81
0.81
0.80
0.80
0.99
0.81
0.09
0.28
GD,
ng
0.77
0.76
0.76
0.45
0.46
0.50
0.52
0.15
0.48
HD,
ng
0.83
0.93
0.89
0.80
0.76
0.79
0.83
0.06
0.18
Coupon DL, jug
based upon 1 0 mL
extract 2.8 4.8 1.8
VX,
ng
0.61
0.59
0.55
0.49
0.53
0.58
0.47
0.05
0.16
1.6
Table B.2 - Determination of Method Detection Limit
(MDL) - Galvanized Steel Coupons
Sample GB, GD, HD, VX,
Description ng ng ng ng
Galvanized Steel
Repl
Galvanized Steel
Rep 2
Galvanized Steel
Rep 3
Galvanized Steel
Rep 4
Galvanized Steel
Rep 5
Galvanized Steel
Rep 6
Galvanized Steel
Rep 7
STD Dev.
MDL (STDDEV x
3.143)
0.86
0.90
0.78
0.91
0.70
0.74
1.12
0.14
0.44
0.46
0.53
0.44
0.45
0.45
0.46
0.47
0.03
0.10
0.78
0.86
0.66
0.70
0.68
0.83
0.81
0.08
0.25
Coupon DL, jug
based upon 1 0 mL
extract 4.4 1.0 2.5
0.40
0.49
0.50
0.57
0.53
0.60
0.55
0.06
0.20
2.0
GB,
Sample Description ng
Carpet Rep 1
Carpet Rep 2
Carpet Rep 3
Carpet Rep 4
Carpet Rep 5
Carpet Rep 6
Carpet Rep 7
STD Dev.
MDL(STDDEVx
3.143)
0.71
0.85
0.70
0.79
0.94
0.88
0.83
0.09
0.28
GD,
ng
0.63
0.62
0.60
0.59
0.68
0.61
0.68
0.03
0.11
HD,
ng
0.67
0.58
0.51
0.67
0.60
0.76
0.72
0.09
0.27
Coupon DL, jug based
upon 10 mL extract 2.8 1.1 2.7
VX,
ng
0.75
1.02
0.96
1.02
0.92
1.07
1.19
0.14
0.43
4.3
Table B.4 - Determination of Method Detection Limit
(MDL) - Ceiling Tile Coupons
Sample GB, GD***, HD, VX***,
Description ng ng ng ng
Ceiling Tile
Rep 1
Ceiling Tile
Rep 2
Ceiling Tile
Rep 3
Ceiling Tile
Rep 4
Ceiling Tile
Rep 5
Ceiling Tile
Rep 6
Ceiling Tile
Rep 7
STD Dev.
MDL
(STDDEV x
3.143)
0.91
0.80
0.86
0.71
0.75
0.67
0.67
0.09
0.29
0.64
0.68
0.57
0.52
0.59
0.61
0.45
0.08
0.24
1.19
1.01
0.88
0.56
0.59
0.71
0.68
0.23
0.73
Coupon DL, jug
based upon 20
mL extract 5.9 4.9 14.6
0.33
0.27
0.35
0.30
0.39
0.45
0.31
0.06
0.19
3.8
*** 50% hexane/acetone solvent system was used for these
extractions
-------
Appendix C - Steam Results Graphs
Figure C.I Bar graph of jig of VX recovered per sample for the 3 kg/hr steam tests.
Figure C.2 Bar graph of jig of VX recovered per sample for the 1.5 kg/hr steam tests.
Figure C.3 Bar graph of jig of FID recovered per sample for the 1.5 kg/hr steam tests.
-------
Steam 3 kg/hr VX
• s'ij--:ij-i«.
•wuiiM-njiyAY;,
• IS [31 !«Hm 3 la* «V6
Figure C.I - Bar graph of fig of VX recovered per
sample for the 3 kg/hr steam tests.
Steam 1.5 kg/hr VX
unuirv 1*. 1W»
•
- I: .
Figure C.2 - Bar graph of fig of VX recovered per
sample for the 1.5 kg/hr steam tests.
Steam 1.5 kg/hr TGD
F KDi-j^t v 11 7WXI
•WBiiKa» 1-5 r*l< *«
Figure C.3 - Bar graph of fig of HD recovered per
sample for the 1.5 kg/hr steam tests.
-------
Appendix D - mVHP® Recovery Graphs
Samples, Positive Controls, Procedural Blanks and Environmental Conditions
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
Table
D.I
D.2
D.3
D.4
D.5
D.6
D.7
D.8
D.9
D.10
D.ll
D.12
D.13
D.14
D.15
D.16
D.17
D.18
HD, GM/DL.
HD, CA/CT,
Temperature,
HD, GM/DL
HD, CA/CT,
Temperature,
HD, GM/DL.
HD, CA/CT,
Temperature,
HD, GM/DL
HD, CA/CT,
Temperature,
VX, GM/DL.
VX, CA/CT,
Temperature,
VX, GM/DL
VX, CA/CT,
Temperature,
, 10%,250ppmv
10%, 250 ppmv
RH and Flow - HD,
, 100%, 250 ppmv
100%, 250 ppmv
RH and Flow - HD,
, 10%, 350 ppmv
10%, 350 ppmv
RH and Flow - HD,
, 100%, 350 ppmv
100%, 350 ppmv
RH and Flow - HD,
, 10%, 250 ppmv
10%, 250 ppmv
RH and Flow - VX,
, 100%, 250 ppmv
100%, 250 ppmv
RH and Flow - VX,
10%, 250 ppmv
100%, 250 ppmv
10%, 350 ppmv
100%, 350 ppmv
10%, 250 ppmv
100%, 250 ppmv
-------
Table D.I - HD, GM/DL, 10%, 250 ppmv
Environmental Conditions & HD Test Results - Galvanized Metal and Decorative Laminate
10% Output With a Peroxide Level of 250 ppmv
Agent Recovery
Galvanized Metal Ductwork
Decorative Laminate
GM sample (n=5) T-18
GM procedure blk(n=2) T-18
GM ambient PC (n=3) PC-1
-GM l/10th PC(n=3) PC-5
-»-GM l/10th PC(n=3) PC-9
i so.o
I 40'°
30.0
20.0
10.0
.. DL ambient PC(n=3) PC-1
--*-• DL l/10th PC (n=3) PC-5
- X - DL l/10th PC (n=3) PC-9
60 120 150 180 240 400
Diffuser Configuration (DC): T-18 = DC 3, PC-5 = DC 1, PC-9 = DC 4, PCI =
i Relative Humidity (%) HTemperature (C)
0.24 air exchanges/min
D.017 air
exchanges/min
Positive Controls
Table D.2 - HD, CA/CT, 10%, 250 ppmv
Environmental Conditions & HD Test Results - Industrial Grade Carpet and Ceiling Tile
10% Output With a Peroxide Level of 250 ppmv
Agent Recovery
t
§
1
Industrial Grade Carpet
— • — CA sample (n=5) T-17
*""*•.
— * — CA procedure blk(n=2) T-17
_^^2SS^
«7 * ^^^^^''a *+
** ^^tm~- ...&... CA ambient PC (n=3) PC-1
S ^-- . Nt»
-»- ^j
»
60 120 150 180 240 400
£.
i
1
» 4°'°
Ceiling Tile
— • — CT sample (n=5) T-17
— B — CT procedure blk (n=2) T-17
'•x.
CT ambient PC (n=3) PC-1
"^ -•-..
~-~^ x
3 ^^^[ _ rT,/,^r,r/
60 120 150 180 240 400
Diffuser Configuration (DC): T-17 = DC 2, PC-5 = DC 1, PC-9 = DC 4, PCI = NA
I Relative Humidity (%) yiemperature (C)
0.24 air exchanges/min
0.017 air
exchanges/min
T-17
Test Samples
PC-9 PC-1 (ambient)
Positive Controls
-------
Table D.3 - Temperature, RH and Flow - HD, 10%, 250 ppmv
Temperature. RH and Flow Results for Testing of mVHF® Fumigant for HP
10% Flow With a Peroxide Level of 250 ppmv
-GM sample (n=5) T-18
-DL sample (n=5) T-18
CA sample (n=5) T-17
-CT sample (n=5) T-17
60 120 150 180 240 400
Diffuser Configuration (DC): T-18 = DC 3, T-17 = DC 2
y Relative Humidity (%) U Temperature (C)
Table D.4 - HD, GM/DL, 100%, 250 ppmv
Environmental Conditions & HDTest Results - Galvanized Metal and Decorative Laminate
100% Output With a Peroxide Level of 250 ppmv
£•
1
1
a 4°'°
Galvanized Metal Ductwork
— • — GM sample (n=5) T-19
""•o.. o0'°
""•••^ • GM procedure blk(n=2) T-19 1 >•
°""*\ 1 °J
•"•O—GM ambient PC (n=3) PC-1 §
°-
\\
•^^^
60 120 150 180 240 400
m'n""S
Decorative Laminate
n_ — * — DL sample (n=5) T-19
""*•.., 1 DL procedureblk(n=2) T-19
'"•n.
""O —«••• DL ambient PC (n-3) PC-1
* "• ~»--DL 100%PC(n=31 PC-6
^\
"•>.
60 120 150 180 240 400
Diffuser Configuration (DC): T-19 = DC 4, PC-6 = DC 3, PC-10 = DC 4, PCI = NA
d Relative Humidity (%) y Temperature (C)
2.4air exchanges/min
0.017 air
exchanges/min
Positive Controls
-------
Table D.5 - HD, CA/CT, 100%, 250 ppmv
Environmental Conditions & HP Test Results - Industrial Grade Carpet and Ceiling Tile
100% Output With a Peroxide Level of 250 ppmv
Asent Recovery
t
1
1
„
Industrial Grade Carpet
— • — CA sample (n=5) T-20
"*•&,
— * — CA procedure blk(n=2) T-20
*"a«.
••••&••• CA ambient PC (n=3) PC-1
<^>^
60 120 150 180 240 400
minutes
t
1
Ceiling Tile
— •— CT sample (n=5) T-20
X
— H — CT procedure blk (n=2) T-20
'*
CT ambient PC(n~3} PC-1
fr--~ X
60 120 150 180 240 400
minutes
Diffuser Configuration (DC): T-20 = DC 4, PC-6 = DC 3, PC-10 = DC 4, PCI = NA
1 Relative Humidity (%) B Temperature (C)
Positive Controls
Table D.6 - Temperature, RH and Flow - HD, 100%, 250 ppmv
Temperature. RH and Flow Results for Testing of inVHP8 Fumigant for HD
100% Output With a Peroxide Level of 250 ppmv
1 ,.
GM sample (n=5) T-19
DL sample (n=5) T-19
CA sample (n=5) T-20
-CT sample (n=S) T-20
60 120 ISO 180 24C
minutes
Diffuser Configuration (DC): T-19 - DC 4, T-20 - DC 4
• Relative Humidity (%) UTemperature (C)
-------
Table D.7 - HD, GM/DL, 10%, 350 ppmv
Environmental Conditions & HP Test Results - Galvanized Metal and Decorative Laminate
10% Output With a Peroxide Level of 350 ppmv
Agent Recovery
E-
1
1
a.
Galvanized Metal Ductwork
-------
Table D.9 - Temperature, RH and Flow - HD, 10%, 350 ppmv
Temperature. RH and Flow Results for Testing of mVHI^Fiimigant for HP
10% Output With a Peroxide Level of 350 DDIIIV
GM sample (n=5) T-26
CA sample (n=5) T-25
CT sample (n = 5) T-25
Diffuser Configuration (DC): T-26 = DC4, T-25 = DC 4
I Relative Humidity (%) U Temperature (C)
Table D.10 - HD, GM/DL, 100%, 350 ppmv
Environmental Conditions & HD Test Results - Galvanized Metal and Decorative Laminate
100% Output With a Peroxide Level of 350 ppmv
Agent Recovery
£-
§
1
8.
Galvanized Metal Ductwork
— • — GM sample (n=5) T-28
"•o...
""••.^ • GM procedure blkfn=2) T-28
""'KI
— -0— GM ambient PC (n=3) PC-1
\
A* V — X — GM 100%PCfn=3) PC-6
\\
^^_ — »— GM 100%PCfn=3) PC-10
60 120 150 180 240 400
£•
§
1
a
Decorative Laminate
^ * PL sample (n=5) T-28
"**•.. ^ 1 DL procedureblkfn = 2) T-28
"*'O.
""-,3 ....Q... DL ambient PC (n=3) PC-1
— • — DL 100%PC(n=3) PC-6
\
^_ - * — DL 100% PC fn-3) PC-10
60 120 150 180 240 400
minutes
Diffuser Configuration (DC): T-28 = DC 4, PC-6 = DC 3, PC-10 = DC 4, PCI = NA
i Relative Humidity (%) y Temperature (C)
2.4 air exchanges/min
0.017 air
exchanges/min
Positive Controls
-------
Table D.ll - HD, CA/CT, 100%, 350 ppmv
Environmental Conditions & HP Test Results - Industrial Grade Carpet and Ceiling Tile
100% Output With a Peroxide Level of 350 ppmv
Agent Recovery
£•
§
1
s.
Industrial Grade Carpet
— • — CA sample fn=5) T-27
"'a.
...A— CA ambient PC fn=3) PC-1
h— CA 100%PC(n = 3) PC-6
"N^*^— — — — 4- -»-CA 100%PCfn = 3) PC-10
60 120 150 180 240 400
minutes
=r
§
S
a 4°'°
Ceiling Tile
— 9 — CT 100%PC(n=3) PC-6
X
\ — •— CT sample (n=5) T-27
'x
"••„. — H — CT procedure blk(n=2) T-27
"'•..f ..,-x— CT ambient PC (n=3) PC-1
"X
- • — CT 100%PC(n=3) PC-10
60 120 150 180 240 400
minutes
Diffuser Configuration (DC): T-27 = DC 4, PC-6 = DC 3, PC-10 = DC 4, PCI = NA
• Relative Humidity (%) HTemperature (C)
2.4 air exchanges/min
0.017 air
exchanges/min
Test Samples
Positive Controls
Table D.12 - Temperature, RH and Flow - HD, 100%, 350 ppmv
Temperature. RH and Flow Results for Testing of mVHP8Fumigant for HD
100% Output With a Peroxide Level of 350 ppmv
Diffuser Configuration (DC): T-28=DC4, T-27 = DC 4
I Relative Humidity (%) UTemperature (C)
-------
Table D.13 - VX, GM/DL, 10%, 250 ppmv
Environmental Conditions & VX Test Results - Galvanized Metal and Decorative Laminate
10% Output With a Peroxide Level of 250 ppmv
Agent Recovery
& "u
1 60
1
1
Galvanized Metal Ductwork
*•-... ..~ y*~*m^^^~~>K
GM procedure blk(n=2) T-24
( }
60 120 180 240 300 400
1
* 40
Decorative Metal
^>C" "^^_ — * — DL Sample (n=5) T-24
* •"-.?.—--*k--.
•O"'"-E>..\. *^»
DL procedure blk(n=2) T-24
X
60 120 180 240 300 400
Diffuser Configuration (DC): T-24 = DC 4, PC-8 = DC 4, PC3 = NA
j Relative Humidity (%) HTemperature (C)
0.24air exchanges/min
0.017 air
exchanges/min
Positive Controls
Table D.14 - VX, CA/CT, 10%, 250 ppmv
Environmental Conditions & VX Test Results - Industrial Grade Carpet and Ceiling Tile
10% Output With a Peroxide Level of 250 ppmv
Agent Recovery
s
- 60
3
Si
Industrial Grade Carpet
— • — CA Sample (n=5) T-23R
* £-CS£± — 1---I
^^^^ — A — CA procedure blk(n=2) T-23R
^.
....&... (^ smbient PC (n=3) PC-3
1 — CA l/10th PC (n=3) PC-8
60 120 180 240 300 400
minutes
§
- 60
S
Si
Ceiling Tile
«
** ^V — •— CT Sample (n=5) T-23R
*^^r"- *
^^**^**^ — a — CT procedureblk(n=2) T-23R
...$$... CT ambient PC (n=3) PC-3
— »— CT l/10th PC (n=3) PC-8
60 120 180 240 300 400
minutes
Diffuser Configuration (DC): T-23R = DC 4, PC-8 = DC 4, PC3 = NA
t Relative Humidity (%) y Temperature (C)
0.24 air exchanges/min
0.017 air
exchanges/min
Positive Controls
-------
Table D.15 - Temperature, RH and Flow - VX, 10%, 250 ppmv
Temperature. RH and Flow Results for Testing of mVHP Fumigant for VX
One-tenth Output With a Peroxide Level of 250 ppmv
" 60
£
a
-GM Sample (n=5) T-24
-DL Sample (n=5) T-24
CA Sample (n=5) T-23R
CT Sample (n=5) T-23R
60 120 180 240 300 400
minutes
Diffuser Configuration (DC): T-24 = DC 4, T-23R = DC 4
1 Relative Humidity (%} HTemperature (C)
Table D.16 - VX, GM/DL, 100%, 250 ppmv
Environmental Conditions & VXTest Results - Galvanized Metal and Decorative Laminate
100% Output With a Peroxide Level of 250 ppmv
Asent Recovery
Galvanized Metal Ductwork
-GM Sample (n=5) T-21
Decorative Laminate
•••O— GM ambient PC (n=3) PC-3
"*"GM 100% PC (n=3J
- DL Sample (n=5) T-21
- DL procedure blk(n=2) T-21
>• DL ambient PC (n=3) PC-3
"•" DL 100% PC ("=3J
Diffuser Configuration (DC): T-21 = DC 4, PC-7 = DC 4, PCS = NA
I Relative Humidity (%) H Temperature (C)
2.4 air exchanges/min
Positive Controls
-------
Table D.17 - VX, CA/CT, 100%, 250 ppmv
Environmental Conditions & VX Test Results - Industrial Grade Carpet and Ceiling Tile
100% Output With a Peroxide Level of 250 ppmv
Agent Recovery
Industrial Grade Carpet
-CA Sample (n=5) T-22
-CA procedure blk(n=2) T-22
—•A— CA ambient PC (n=3) PC-3
60 120 180 240 300 400
Ceiling Tile
-CT Sample (n=5) T-22
-CT procedure blk(n=2) T-22
.... CT ambient PC (n=3) PC-3
Diffuser Configuration (DC): T-22 = DC 4, PC-7 = DC 4, PC3 = NA
i Relative Humidity (%) y Temperature (C)
2.4 air exchanges/min
0.017 air
exchanges/min
37 36
38 37
T-22
Test Samples
PC-3 (ambient)
Positive Controls
Table D.18 - Temperature, RH and Flow - VX, 100%, 250 ppmv
Temperature. RH and Flow Results for Testing of mVHF*8 Fumigant for VX
100% Output With a Peroxide Level of 250 DDIIIV
-GM Sample (n=5) T-21
-DL Sample (n=5) T-21
-CA Sample (n=5) T-22
CT Sample (n=5) T-22
60 120 ISO 240 300 400
Diffuser Configuration (DC): T-21 = DC 4, T-22 = DC 4
J Relative Humidity (%) UTemperature (C)
-------
7.0
References
i VHP™ 1000-ARD Biodecontamination Unit
Modified for mVHP® Chem/Bio Decontamination
Supplemental Operating Procedure, STERIS, July
2008.
2 Ryan, S.; Snyder, E. "Systematic Decontamination
of Chemical Warfare Agents and Toxic Industrial
Chemicals", Response to Chemical Emergencies
Including Terrorist Attacks and Industrial
Accidents, St. Petersburg, Russia, September 20,
2007.
3 Persistence of Toxic Industrial Chemical
Warfare Agents on Building Materials under
Conventional Environmental Conditions, Battelle,
EPA/600/R-08/075, July 2008.
4 Systematic Evaluation of Developmental and
Commercially Available Methods for Chemical
Agent Decontamination. Quality Assurance Project
Plan, CUBRC/ERG, December 23, 2008.
5 Zamejc, E.R. ; Mezey, E.J. ; Hayes, L. et al.,
Development of Novel Decontamination
Techniques for Chemical Agents (GB, VX HD)
Contaminated Facilities AMXTII-TE-TR-85012,
June 1985.
6 Schwartz L,; McVey I. Large Scale Tests of Vaporous
Hydrogen Peroxide (VHP™) for Chemical and
Biological Weapons Decontamination. Scientific
Conference on Chemical and Biological Defense
Research, Edgewood, MD, November 15, 2004.
7 Lelain, T; Brickhouse, M. Chemical-Warfare Agent
Decontamination Efficacy Testing: Large-Scale
Chamber mVHP® Decontamination System
Evaluation Part 1: Comparison to the ORD. ECBC-
TR-MS-3163. August 15, 2006.
8 Determination of Hydrogen Peroxide by Potassium
Permanganate Titration. AATCC Test Method 102-
2007, AATCC Committee RA34, 1997.
9 Environmental Protection Agency. 40 CFR Appendix
B to Part 136 - Definition and Procedure for the
Determination of the Method Detection Limit
- Revision 1.11 in Code of Federal Regulations -
Title 40: Protection of Environment, Washington,
D.C., 2005.
-------
-------
United States
Environmental Protection
Agency
PRESORTED STANDARD
POSTAGES FEES PAID
EPA
PERMIT NO. G-35
Office of Research and Development (8101R)
Washington, DC 20460
Official Business
Penalty for Private Use
$300
------- |