EPA/600/R-19/023 | May 2019
www.epa.gov/homeland-security-research
United States
Environmental Protection
Agency
&EPA
Feasibility Study for Reuse of
Activated Carbon for Capture
of Methyl Bromide
Office of Research and Development
Homeland Security Research Program
-------
This page left intentionally blank
-------
EPA 600/R-19/023
Feasibility Study for Reuse of Activated Carbon for the
Capture of Methyl Bromide Used for Decontamination
Authors:
Stella McDonald
Jacobs Technology, Inc.
Joseph Wood
National Homeland Security Research Center
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
-------
Disclaimer
The U.S. Environmental Protection Agency (EPA), through its Office of Research and
Development's (ORD) National Homeland Security Research Center (NHSRC), directed, funded
and managed this investigation through contract EP-C-15-008 with Jacobs Technology, Inc.
This report has been peer and administratively reviewed and has been approved for publication
as an EPA document. It does not necessarily reflect the views of the Agency. No official
endorsement should be inferred. EPA does not endorse the purchase or sale of any commercial
products or services.
Questions concerning this document, or its application, should be addressed to:
Joseph Wood
Decontamination and Consequence Management Division
National Homeland Security Research Center
U.S. Environmental Protection Agency (MD-E343-06)
Office of Research and Development
109 T.W. Alexander Drive
Research Triangle Park, NC 27711
Phone: 919-541-5029
E-mail: wood.ioe@epa.gov
-------
Acknowledgments
The principal investigator from the U.S. Environmental Protection Agency (EPA), through its
Office of Research and Development's National Homeland Security Research Center (NHSRC),
directed this effort with intellectual support of a project team from across EPA. The input and
contributions of the individuals listed below have been a valued asset throughout this effort.
EPA Project Team
Joseph Wood, Principal Investigator, NHSRC/Decontamination and Consequence Management
Division (DCMD)
Leroy Mickelsen, Office of Land and Emergency Management (OLEM)/Consequence
Management Advisory Division (CMAD)
Shannon Serre, OLEM/CMAD
EPA Quality Assurance
Eletha Brady-Roberts, NHSRC
Jacobs Technology, Inc.
Stella McDonald
Abderrahmane Touati
External Reviewers of Report
Alden Adrion, PhD
Warfighter Directorate
Aberdeen Test Center
Jasper Hardesty, PhD
Sandia National Laboratories
-------
Executive Summary
Methyl bromide (MeBr) fumigation has been shown to be effective in decontaminating buildings
contaminated with Bacillus anthracis (Ba) spores. Previous laboratory and field tests have
shown that activated carbon (AC) is effective in capturing the MeBr following such fumigation, to
prevent its release to the atmosphere. In the event of a large urban Ba spore release, large
quantities of MeBr would be needed and hence potentially large quantities of AC as well. The
purpose of this study was to evaluate the feasibility of processing the AC to allow it to be reused
for further capture of MeBr.
In the present study, we quantified how well an AC sample maintained its adsorption capacity
for MeBr over several adsorption/desorption cycles, using three different gas conditions. During
the adsorption phase, the AC was exposed to a feed gas of 5.3% MeBr until saturation of the
AC sample was achieved. Subsequently, the MeBr was desorbed from the saturated AC by
exposing the carbon bed to dry or ambient air heated to 100 °C.
Overall, the results for the adsorption tests showed relatively high levels of adsorption and
ranged from approximately 0.43 to 0.58 g MeBr per gram AC. These higher-than-expected
adsorption capacities may be due to the preconditioned AC samples we used, as well as the
low relative humidity (RH) levels in the challenge gases (applicable to the first two test series).
The differences in adsorption capacity as a function of challenge gas and/or desorption gas
characteristics were generally minor, although in some cases the differences were statistically
significant. Tests to determine adsorption of moisture at high RH (75%), without the presence of
MeBr, suggest that the adsorption capacity for moisture was approximately 10% of the AC
capacity for MeBr.
The adsorption capacity of the AC was not affected (did not diminish) through the five
adsorption/desorption cycle series that each AC sample was subjected to. The process of
desorbing the MeBr using a 100 °C temperature gas was effective in maintaining relatively high
and stable adsorption levels of the AC samples for at least five cycles; the adsorption capacity
did not diminish over the course of the five cycles but remained rather stable. However, we
caveat that the tests used only five cycles and may not yield sufficient data to assess the effect
of numerous repeated adsorption/desorption cycles.
While we have demonstrated in this study the ability to efficiently remove MeBr from AC and in
the process, allow the AC to be reused (i.e., regenerate the AC without losing its adsorption
capacity for MeBr over several cycles), further research is recommended related to the reuse of
MeBr on a wide scale following a Ba incident, including development of methods to allow reuse
of the MeBr desorbed from AC.
-------
Table of Contents
Disclaimer ii
Executive Summary iv
List of Figures vi
List of Tables vi
Acronyms and Abbreviations viii
1 Introduction 1
2 Materials and Methods 2
2.1 Test Matrix and Study Description 2
2.2 Activated Carbon and MeBr 3
2.3 Test facility 4
2.4 AC Adsorption/Desorption Test Apparatus 4
2.5 Gas Flow Rate 8
2.6 Relative Humidity and Temperature 8
2.7 Sample Preparation and Conditioning 8
2.8 AC Moisture Content 9
2.9 Methyl Bromide Concentration 10
2.10 Adsorption Tests 10
2.11 Desorption Tests 12
2.12 Gravimetric Method to Determine Adsorption Capacity 13
2.13 Integration Method to Determine Adsorption Capacity 13
2.14 Adsorption Capacity Characterization 15
2.15 Control Test with Elevated RH 15
2.16 Statistical Analysis 15
3 Quality Assurance/Quality Control 16
3.1 Sampling, Monitoring, and Equipment Calibration 16
3.1.1. VIG FID Model 20 Concentration Measurement 16
3.1.2. Temperature and RH Measurements 16
3.1.3. Mass Flow Rate 17
3.1.4. Mass Measurements 18
3.1.5. Equipment Calibrations 18
3.2 Acceptance Criteria for Critical Measurements 19
4 Results and Discussion 23
4.1 Stability of AC Moisture Content and Related Measures 23
4.2. Activated Carbon Regeneration Tests 24
4.2.1 Adsorption and Desorption Results for Each Test Cycle 24
v
-------
4.2.2 Effect of Challenge Gas Conditions on Adsorption Capacity 29
4.2.3 The Effect of Reuse on Adsorption Capacity 31
4.3. Control Tests with Elevated RH Only 33
5 Summary and Conclusions 34
6 References 35
Appendix A: Test Conditions 1
Appendix B: Carbon tube mass stability preliminary tests 1
Appendix C: Preliminary Activated Carbon Adsorption Tests 1
List of Figures
Figure 2-2. Sample column prepared for installation 9
Figure 2-3. Sealed sample tube 9
Figure 2-4. Typical Adsorption Curve 11
Figure 2-5. Typical Desorption Curve 12
Figure 4-1. Comparison of MeBr desorbed vs MeBr adsorbed using gravimetric analysis 28
Figure 4-2. Comparison of MeBr desorbed vs MeBr adsorbed using the integration method FID 29
Figure 4-3. Average adsorption capacity for each test series (±SD) 30
Figure 4-4. Average adsorption capacities for Test 1 cycles (±SD) 31
Figure 4-5. Average adsorption capacities for Test 2 cycles (±SD) 32
Figure 4-6. Average adsorption capacities for Test 3 cycles (±SD) 32
Figure 4-7. Adsorbed and desorbed moisture with an exposure stream of 75% RH in ambient air 33
List of Tables
Table 2-1. Test Matrix 2
Table 2-2. AC Manufacturer Specifications 3
Table 3-1. FID Operating Specifications 16
Table 3-2. Relative Humidity Sensor and Temperature Probe Operating Specifications 17
Table 3-3. Mass Flow Controller Specifications 17
Table 3-4. Flow Calibrator Specifications 18
Table 3-5. Balance Operating Specifications 18
Table 3-6. Summary Sampling and Monitoring Equipment QA/QC Checks 19
vi
-------
Table 3-7. Critical Measurement Acceptance Criteria 20
Table 3-8. Accuracy (% Error) of Critical Measurements 21
Table 3-9. Data Precision (SD) of Critical Measurements 21
Table 3-10. FID Drift for Each Test Duration 22
Table 4-1. Moisture Content (%) of Bulk Carbon During Evaluation Period 23
Table 4-2. Amount of Moisture Removed from Fresh Carbon Bed (g/g carbon) 24
Table 4-3. Adsorbed and Desorbed MeBr Using Gravimetric and Integration Analysis 26
Methods 26
Table 4-4. Adsorption Capacity Comparison for Each Test Series 30
Table A-1. Adsorption Phase Test Conditions A-2
Table A-2. Desorption Phase Test Conditions A-4
Table A-3. RH only Test Conditions A-6
vii
-------
Acronyms and Abbreviations
AC activated carbon
ACS activated carbon system
ASTM American Society for Testing and Materials
ANOVA analysis of variance
Ba Bacillus anthracis
cc cubic centimeter
CCM cubic centimeters per minute
CCU carbon tetrachloride
CMAD Consequence Management Advisory Division (EPA)
DAS data acquisition system
DCMD Decontamination and Consequence Management Division (EPA)
EPA U.S. Environmental Protection Agency
ft feet
FID flame ionization detector/detection
g gram(s)
GC gas chromatograph/chromatography
KOH potassium hydroxide
L liter
Lpm liter per minute
MeBr methyl bromide
MFC mass flow controller
mg milligram(s)
min minute(s)
NA not applicable, not available
N2 nitrogen
NHSRC National Homeland Security Research Center (EPA)
NIST National Institute of Standards and Technology
OLEM Office of Land and Emergency Management (EPA)
ORD EPA Office of Research and Development
PDAQ personal data acquisition
ppm part(s) per million
psi pound(s) per square inch
QA quality assurance
QC quality control
RH relative humidity
SD Standard deviation
slpm standard liter(s) per minute
UHP ultra high purity
WACOR work assignment contracting officer's representative
viii
-------
1 Introduction
Methyl bromide (MeBr) has been demonstrated in the laboratory (Wood et al., 2016;
Juergensmeyer et al., 2007) as well as in full-scale field testing (Serre et al., 2016) to be an
effective decontaminant for the inactivation of Bacillus anthracis (Ba) spores on a wide
range of materials. Other advantages of using MeBr as a decontaminant include many
personnel trained in its use (as an agricultural/commodity fumigant), ease of penetration of
materials, and relative compatibility with most materials.
To prevent release of MeBr to the atmosphere to avoid human exposure (MeBr is toxic,
with a Permissible Exposure Limit of 20 parts per million; The National Institute for
Occupational Safety and Health, 2018), air quality impact, and depletion of stratospheric
ozone, laboratory studies have demonstrated the capture of MeBr onto activated carbon
(AC) (Leesch et al., 2000; Snyder and Leesch, 2001; Gan et al., 2001). Additionally, a full-
scale study was conducted to evaluate an activated carbon system (ACS) employed for the
capture of MeBr under the conditions that would be used for decontaminating a building
structure contaminated with Ba spores (Wood et al., 2015).
In the event of a large-scale Ba spore release, large quantities of MeBr would potentially be
needed for decontamination and ideally, the AC used to capture the MeBr could be
regenerated and deployed for reuse. (The reuse of the captured MeBr is also desired, but
that is a topic for future research.) Indeed, once the MeBr is adsorbed onto the AC
("adsorption" refers to when a gas such as MeBr adheres to the microscopic surfaces within
the pores of the carbon granules), there are processes that can be used to remove it from
the AC and render the MeBr into a less hazardous chemical (Yang et al., 2015; Joyce and
Bielski, 2010). Further, in the temperature swing adsorption process, hot air is used to
desorb (remove) the MeBr from the AC, and in the process, the hot air also regenerates the
AC so that it can be reused (Value Recovery, 2018). This technology is used at a few
quarantine and pre-shipment locations in the U.S, where MeBr is used to fumigate
agricultural products.
In the present study, we evaluated a similar temperature swing adsorption process, utilizing
gas streams with different temperatures and relative humidity levels for adsorption and
desorption of the AC. More specifically, we quantified how well an AC sample maintained
its adsorption capacity for MeBr over several adsorption/desorption cycles, using three
different adsorption/purge gas conditions. Additionally, a few tests were also conducted to
determine adsorption of water vapor from air without MeBr. Lastly, preliminary scoping test
results are provided in Appendix C that demonstrate the effect of various operational
parameters (e.g., gas temperature, MeBr concentration, relative humidity) on the adsorption
capacity of AC.
1
-------
2 Materials and Methods
2.1 Test Matrix and Study Description
Three tests were conducted in triplicate, with each test replicate consisting of a series of five
complete cycles of adsorption and desorption phases. Each test replicate began with a fresh 1 g
sample of AC. During the adsorption phase of testing, 53,000 parts per million (ppm) MeBr
(5.3%) in nitrogen gas (N2) was used as the challenge gas at a flow rate of 0.5 liters per minute
(Lpm) at ambient temperature (uncontrolled but averaged 22 °C throughout the study). (Note,
although chloropicrin is sometimes added to cylinders of MeBr as an odorant to detect leaks,
none was used in these tests.) The RH of the MeBr gas stream in Tests 1 and 2 was relatively
low (~ 3%), and the MeBr gas was at the moisture content of the gas cylinder. The RH was
elevated to 75% for the Test 3 evaluations. The desorption cycles were conducted with a gas
flow rate of 1 Lpm and a temperature of 100 °C. The desorption gas stream was heated to the
target temperature prior to entering the column. Dry compressed air was used as the desorption
gas for Test 1 and ambient air was used for Tests 2 and 3. Table 2-1 details the test parameters
evaluated for this study.
Table 2-1. Test Matrix
Test ID
Cycle Phase
Challenge Gas
Gas
Temperature
(°C)
RH
(%)
1
Adsorption
5.3% MeBr
in N2
Ambient
As received (typically 3%)
Desorption
Dry air from the laboratory
air compressor system
("house air")
100
Typically, 1.0% @ 100°C
2
Adsorption
5.3% MeBr in N2
Ambient
As received (typically 3%)
Desorption
Ambient air
100
Ambient (typically 5%)
3
Adsorption
5.3% MeBr in N2, with moisture
added to raise RH
Ambient
75
Desorption
Ambient air
100
Ambient (typically 5%)
Control
test with
air and
elevated
RH-no
MeBr
Adsorption
Ambient air with moisture
added to raise RH - no MeBr
Ambient
75
Desorption
Ambient air
100
Ambient (typically 5%)
Ancillary tests were also conducted outside the primary test matrix as follows:
2
-------
Tests to determine mass adsorption of water vapor only. (Control test listed in Table 2-
1.) This test was done to compare the mass of water adsorbed to the total mass
adsorbed in Test 3 (mass gain on AC attributed to both MeBr and water vapor).
• A mass stability analysis of the carbon tube over an observation period representative of
a regeneration test series, to ensure that variations in the mass of the tube did not affect
the measurements of the mass of carbon.
• Temporal evaluations of the bulk AC moisture content.
• Scoping tests to assess the effect of operational parameters such as gas temperature,
MeBr concentration, relative humidity, and AC moisture content on the adsorption
capacity of the AC.
2.2 Activated Carbon and MeBr
Premium coconut shell-derived granular AC (General Carbon Corp., Paterson, N.J., part
number (p/n) 30100) was used for this study because of its absorptivity for MeBr compared to
other ACs (Snyder and Leesch, 2001). The manufacturer's specifications for the AC are detailed
in Table 2-2.
Table 2-2. AC Manufacturer Specifications
Parameter
Value
Mesh Size -4x8,%
90
Less than No. 4, %
5
Greater than No. 8, %
5
CCI4a Activity, %
70
Iodine No., milligrams (mg)/gram (g)
1200
Hardness No., %
98
Ash Content, %
5
Moisture Content, % (as packaged)
5
Typical Density, g/cubic centimeter (cc)
0.47-0.50
PH
6-8
'Carbon tetrachloride. CC14 activity is used as a relative measure of pore volume.
The AC was stored indoors throughout the duration of the study in double-bagged vinyl plastic,
to maintain dryness.
The challenge gas used for this study was 53,000 ppm (209 mg/liter (L)) MeBr gas mixed in
nitrogen (N2) (Custom Gas Solutions; Durham, NC). A comparable concentration was previously
used in a full-scale study to evaluate an ACS for the capture of MeBr at conditions that would be
used for decontaminating a building structure contaminated with Ba spores (Wood et al., 2015).
3
-------
2.3 Test facility
All work for this effort was conducted in the Air and Energy Management Division's test facility
located at the U.S. EPA, Research Triangle Park High Bay Building room 226. Because of the
hazardous nature of MeBr, the test facility was designated as chemical safety level 4. The MeBr
gases were stored in an external gas closet located approximately 20 feet from the facility and
piped into the test facility using connected %-inch stainless steel tubing. A fixed gas regulator
located in the test facility served as the supply point for the gases located in the gas closet. A
portable gas cabinet located in the facility housed the dilution N2 and hydrogen/helium (hh/He)
fuel for the MeBr analyzer. A flexible duct connected the internal gas cabinet to the central
exhaust system of the building, effectively isolating the gas cabinet from the work area in the
event of a gas leak.
A bench top oven (1300U, VWR, Radnor, PA) used for drying the AC (for moisture content
measurement) was maintained at 150 °C (302 °F) using an Omega CSC32-K bench top
temperature controller (Omega Engineering, Inc., Stamford, CT). A bench top desiccator (model
1340; Boekel Scientific; Philadelphia, PA) containing silica was also used for moisture content
evaluations. A HOBO® temperature and humidity logger (UX100, Onset Computer Corp., Cape
Cod, Massachusetts) placed inside the desiccator indicated that the internal RH was 15%
throughout the duration of testing. An ENTIRS 124-1S top loading balance (Sartorius,
Gottingen, Germany) with a range of 120 g and resolution of 0.0001 g was used for all mass
measurements. A custom plexiglass enclosure was used as a secondary barrier against moving
air that could adversely affect the accuracy of the mass readings.
2.4 AC Adsorption/Desorption Test Apparatus
The MeBr experiments were carried out in a bench-scale system custom-built to maintain and/or
monitor the prescribed test conditions. The AC sample was contained in a custom borosilicate
glass AC tube. The bench-scale system was assembled under a chemical hood located in the
test facility as a safety precaution in the event of an unplanned MeBr gas release. The system
consisted of:
• Approximately 36 feet (ft) of 316 stainless steel tubing (McMaster Carr, Elmhurst, IL)
• Flame ionization detector (FID) (VIG Industries, Anaheim, CA)
• 4 Sierra Smart-Trak mass flow controllers (MFCs) (Sierra Instruments, Monterey,
CA)
• 1 Vaisala HMP50 temperature and relative humidity probe (Vaisala, Helsinki,
Finland)
• 3 Omega K-Type thermocouples (Omega Engineering, Stamford, CT)
• Gas humidity bottle (Fuel Cell Technologies, Inc., Albuquerque, New Mexico)
• MeBr gases: 500 ppm (2 mg/L) MeBr, 13,000 ppm (51 mg/L) MeBr, and 53,000 ppm
(209 mg/L) (Custom Gas Solutions, Durham, NC)
• Ultra-high purity (UHP) nitrogen (Airgas, Durham, NC)
• FID fuel: 40% hydrogen/60% helium UHP gas mixture (Airgas, Durham, NC)
• Dry compressed air (Airgas, Durham, NC)
4
-------
• Pump for ambient air desorption (Cole-Parmer L-792000-30)
* XX
M3 X
M2 -X+
Ml -><-~
VI V2 V3 V4
External Gas Closet
High Bay
Building
room 226
Adsorption gas flow
Desorption gas flow
Adsorption & Desorption gas flow
Fuel
Data
Figure 2-1. Flow diagram of test apparatus
5
-------
Table 2-3. Identification of Test Equipment
Fig. 2-1
I.D.
ACT
Equipment
Description
Ambient collection tube
Open to the atmosphere; under constant vacuum.
Prevents significant spikes and drops in system
pressure.
AF
Sample filter
Protects FID by filtering particles from sample gas.
CT
Carbon trap
Removes MeBr from bypass streams before being
vented into the hood
C1
Coiled heating tube
Increases resonance time of the exposure gas stream to
allow adequate heating and mixing ahead of the carbon
tube.
C2
Coiled mixing tube
Mixes dilution nitrogen with the exposure gas ahead of
the FID
DAQ
Data acquisition system
Digitally records MeBr concentration, RH, temperatures
and flow rates
FID
Flame ionizing detector
VIG Industries Model 20 total hydrocarbon analyzer
FM Flow meter
20 Lpm rotameter.
G1
Nitrogen gas cylinder
Zero gas used for FID calibrations
G2
0.05% MeBr cylinder
Span gas used for FID calibrations
G3
1.3% MeBr cylinder
Span gas used for FID calibrations
G4
5.3% MeBr cylinder
Exposure gas used for adsorption
G5
Nitrogen gas cylinder
Dilution gas for sample
G6
H2 in He gas cylinder
40% hydrogen in helium. Used to fuel FID
HB
Gas humidity bottle
Adds water vapor to exposure gas stream
HT
Humidity transmitter
Vaisala, used to measure the relative humidity of the
exposure gas stream
M1
Mass Flow Controller
10 Lpm, used to regulate zero and span gases during
FID calibration
M2
Mass Flow Controller
1 Lpm, used to regulate 5.3% MeBr gas flow during
adsorption phase
M3
Mass Flow Controller
5 Lpm, used to regulate ambient air flow during
desorption phase
M4
Mass Flow Controller
15 Lpm, used to regulate nitrogen dilution flow
OV
GC oven
Heats coiled tubing to 100°C
P1
Thomas pump
Generates ambient airflow during desorption phase
6
-------
P2
Chemical pump
Moves contents of ambient collection tube through the
carbon trap
ST
Carbon tube
Borosilicate glass chamber, 3" effective length, 5/8" I.D.
to 5/16" tube studs (Contains the AC.)
TC1
Thermocouple probe
Located at inlet port to measure the temperature of inlet
gas stream
Affixed to the outer surface of the carbon tube adjacent
to the carbon sample to monitor temperature changes
during testing
TC2
Surface thermocouple
TC3
Thermocouple probe
Located at outlet port to measure the temperature of
outlet gas stream
V1-4
On/Off Valve
Starts and stops gas flow of respective calibration gases.
Located on gas manifold in the H-240 gas cabinet
V5
Regulator Valve
Dilution nitrogen regulator
V6
Regulator Valve
FID fuel cylinder regulator
V7
Regulator Valve
Gas regulator located in H-226 connected to H-240 gas
manifold
V8
Flow adjustment valve
Located between the inlet and outlet of the ambient air
pump. Used to set the gas pressure for the air MFC (M3)
V10-12
On/Off Valve
Starts and stops gas flow through manifold in the H-226
chemical hood
V13
Directional Valve
Directs exposure gas through the test stand or bypasses
the test stand and sends gas to the ambient collection
tube
V14
Flow adjustment valve
Used in conjunction with V15 to direct a portion of the
exposure gas through the gas humidity bottle
V15
On/Off Valve
Opened to allow flow through gas humidity bottle, closed
to bypass the gas humidity bottle
V16
Directional Valve
Directs flow through or around the GC oven or around it
V17
Directional Valve
Directs flow through or around the sample port inlet
V18
Directional Valve
Directs flow from the outlet sample port or the bypass to
the mixing coil (C2)
V19
On/Off Valve
Opens/closes access to the ambient collection tube
V20
Directional Valve
Directs flow from either the hood (ambient air) or the
sample lines
V21
Flow adjustment valve
Located between the inlet and outlet of the chemical
pump. Allows the flow rate to be reduced to 10 Lpm
without deadheading the 20 Lpm pump
7
-------
2.5 Gas Flow Rate
Velocity measurements of each gas stream (exposure, dilution, and bulk) were controlled using
Sierra Smart-Trak MFCs (Sierra Instruments; Monterey, CA). During flow rate verifications, each
gas was introduced to the system in the same manner and location as during a test scenario.
MeBr and desorption gases entered the system at a gas manifold located in the test facility and
were measured at a point located immediately in front of the FID. Nitrogen gas was used to
dilute the MeBr gas stream leaving the carbon tube prior to being measured by the FID.
Concentrations of MeBr that exceeded 1.3% were diluted with ISMo maintain the accuracy of the
FID (Section 2.9). A series of five flow rate measurements was collected before and after each
phase (adsorption and desorption) using a calibrated Gilibrator-2 Standard Air Flow Calibrator
(Sensidyne LP, St. Petersburg, FL). The average of the series was used as the measurement.
2.6 Relative Humidity and Temperature
The RH of the challenge gas stream during both the adsorption and desorption phases was
measured at the inlet of the carbon bed using a temperature and RH transmitter that had been
calibrated within seven days of the test dates. This "HOBO" was calibrated prior to each test by
comparing its RH data with known RH values that were generated in the sealed headspace
above the individual saturated solutions of various salt compounds. The transmitter sensors
were replaced if the calibration criteria could not be met.
The temperature was measured with Omega K-type thermocouples in three locations: (1) at the
inlet of the column, (2) on the external surface of the column at the carbon bed, and (3) at the
outlet of the column. The glass tube would not accommodate the direct insertion of a
thermocouple probe into the carbon bed; therefore, a K-Type surface thermocouple was affixed
to the outer surface of the AC tube at the carbon bed to monitor the progression of the
exothermic reaction driven by the MeBr adsorption onto the AC. The thermocouples were
calibrated annually by comparisons to the temperature sourced by a Fluke 744 Documenting
Process Calibrator. The temperature and RH measurements data were logged continuously
using a personal data acquisition (PDAQ) system (Measurement Computing Corporation;
Norton, MA.
2.7 Sample Preparation and Conditioning
The AC sample was housed in a custom borosilicate glass AC tube with an inside diameter of
5/8 inch (in) and a length of 3 in that reduced to 1/4 in tube ports on both ends. An approximate
1 g sample was collected from the center of the bulk container of the AC and transferred to the
glass AC tube. Leco® fine quartz wool (Saint Joseph, Ml) was packed into the column on both
sides of the carbon bed to hold the sample in the approximate center of the column. The initial
weight of the tube assembly with AC was measured gravimetrically using the procedure detailed
in Section 2.12. Figure 2-1 shows the column containing the AC sample prepared for installation
into the test stand.
8
-------
Figure 2-2. Sample column prepared for installation
Prior to starting the series of five cycles of adsorption/desorption, the carbon bed was
conditioned to ensure that the physical state of the AC sample prior to initiating the first cycle
was consistent with the physical state of the sample for each subsequent cycle. During the
conditioning phase, the carbon bed was exposed to an air (dry compressed or ambient) flow of
1 liter per minute (Lpm) at 100 °C. These conditions were maintained overnight for
approximately 18 hours. Upon completion, the sample column was removed from the test stand,
the tube ends were sealed with silicone rubber plugs (McMaster-Carr, Atlanta, GA; p/n
9277K37), and the column then was transferred to a desiccator that was maintained at 15% RH
and room temperature to cool for 1 hour. The sample mass was subsequently analyzed
gravimetrically (Section 2.12). Figure 2-2 shows a sample column with both tube ends plugged,
which is the typical sample column configuration when it is not installed in the test stand.
Figure 2-3. Sealed sample tube
2.8 AC Moisture Content
The moisture content of the bulk AC was monitored over the duration of the project. A daily
calibration check was performed on the balance prior to use. (Briefly, the initial weight of each of
three empty Pyrex dishes was measured. A sample of fresh AC (approximately 10 g) was
collected from the bulk container and transferred to each dish. The initial mass for each sample
replicate was measured and recorded (AC + dish). The samples were placed in the oven (150 °C)
for 6 hours (±15 min), then allowed to cool in a desiccator containing silica indicating desiccant
9
-------
(p/n S162-500, Fisher Scientific, Hampton, NH) for 1 hour. The samples (AC + dish) were weighed
individually for the final mass measurement. The moisture content of each AC sample was
calculated as follows:
B - c
Moisture content (%) = x 100
B -A
where:
A was the weight of the dish, g
B was the weight of fresh AC + dish, g
C was the weight of dry AC + dish, g
2.9 Methyl Bromide Concentration
A dual-channel VIG Industries (Anaheim, CA) model 20/2 flame ionization detector (FID) was
used to continuously and simultaneously monitor MeBr levels at two of the three sample
locations. Hydrogen gas was supplied to the instrument from a pressurized gas cylinder for the
flame source. MeBr data were collected, logged, and stored using a data acquisition system
(lotech Corporation). The FID calibration and the MeBr dilution were checked before and after
each cycle using a bias span (calibration gases traveled through the sample line prior to
detection). Calibration gases were obtained from Custom Gas Solutions (Durham, NC), and
included 0.05% MeBr in N2, 1.3% MeBr in N2, and 5.3% MeBr in N2, to span the range of MeBr
concentrations that would be expected during adsorption and desorption tests. The FID was
zeroed using ultra-high-purity N2 (Airgas, Inc., Fort Lauderdale, FL). The channel 1 detector of
the FID served as the high-level MeBr monitor and was calibrated using the 1.3% MeBr gas,
while channel 2 was the lower level detector and was calibrated using the 0.05% gas.
Previous method development work identified issues with the FID response factor at
concentrations exceeding 1.3%. Therefore, a dilution system consisting of ultra-zero N2 and an
MFC was used to dilute the gas exiting the carbon tube prior to analysis by the FID.
The total mass of MeBr adsorbed onto the AC was determined using two methods, gravimetric
analysis and integration using the gas measurements of MeBr.
2.10 Adsorption Tests
The FID instrument was used during the adsorption tests to continuously measure the MeBr
concentration in the gas at the outlet of the AC bed. Immediately before and after each
adsorption test, the MeBr concentration in the bulk stream was confirmed via bypassing the
carbon tube and sending the gas directly to the FID as follows. A nitrogen dilution system was
used to adjust the bulk flow rate to 4 liters per minute (Lpm), approximately twice the required
flow for the FID analyzer. The FID continuously sampled the bulk gas while the excess portion
or the overflow stream was directed to the ambient carbon trap to scrub the MeBr before being
released into the chemical hood. During Test 3, the RH of the MeBr exposure stream was
controlled using a gas humidity bottle (model LF-HBA, Fuel Cell Technologies, Inc.,
Albuquerque, NM) and two inlet feed streams as shown in Figure B-1. Valves V14 and V15
10
-------
were adjusted to achieve the desired prescribed RH condition. The flow rate of the 5.3% MeBr
challenge gas was controlled at 0.5 Lpm using a calibrated 0-1 Lpm MFC. The flow rates of all
streams were verified with a calibrated Gilian Gilibrator-2 (Sensidyne, LP, St. Petersburg, FL)
calibration system before and after each test.
After verifying MeBr concentration, flow rates and the target RH were reached, the challenge
gas was directed through the carbon bed, initiating the MeBr adsorption cycle. See Figure 2-4
for a typical adsorption breakthrough curve. (The curve below shows the MeBr concentrations
after dilution with N2. The initial portion of the curve is the verification of the MeBr concentration,
which was approximately 7,200 ppm after dilution, and remained steady for 5 minutes.
Following the verification, the MeBr gas enters the carbon tube, with breakthrough occurring
rapidly.) The adsorption phase continued until the bulk sample concentration at the outlet of the
AC tubes was stable, indicating that the carbon bed was saturated. A stable concentration was
marked when the FID reading showed less than 1% change within 2 min. Upon completion of
adsorption, the MeBr gas bypassed the carbon tube, and the sample column was removed from
the test stand, the carbon tube ends were plugged, and the column allowed to cool for 1 hour.
The mass of MeBr adsorbed onto the carbon sample was analyzed gravimetrically (Section 2-
12). Adsorption capacity was assessed by the mass of MeBr adsorbed per gram of conditioned
AC. After the AC tube was weighed, it was placed back into the test stand, and the desorption
cycle commenced as discussed below.
9000
8000
-p 7000
6000
5000
4000
3000
CO
2000
1000
0
12:57
13:19
13:40
Time hr:min
14:02
14:24
Figure 2-4. Typical Adsorption Curve
11
-------
2.11 Desorption Tests
The FID was used during the desorption cycles to continuously measure the MeBr
concentration in the gas at the outlet of the sorbent bed. Immediately before and after each
desorption test, a bump test was performed on the FID with 1.3% MeBr certified gas. A nitrogen
dilution system was used to adjust the bulk flow rate to 4 Lpm, approximately twice the required
flow for the FID analyzer. The FID sampled the bulk sample stream continuously while the
excess portion or the overflow stream was directed to the ambient carbon trap to scrub the
MeBr before the flowing gas was released into the chemical hood. The flow rate of the air (zero
and ambient) gas was controlled at 1 Lpm using a calibrated 0-5 Lpm MFC. The flow rates of all
streams were verified with a calibrated Gilian Gilibrator-2 (Sensidyne, LP, St. Petersburg, FL)
calibration system before and after each test.
After verifying flow rates, the desorption gas was directed through the saturated carbon bed,
initiating the desorption phase. The desorption phase continued until the bulk sample outlet
concentration was stable, indicating the end of the desorption cycle. A stable concentration was
marked when the FID reading showed less than 1% change within 2 minutes (min), which
typically occurred within two hours, although desorption was allowed to proceed overnight.
Refer to Figure 2-5 for a typical desorption curve. Upon completion, 1.3% MeBr gas was
directed through the bypass, and the sample column was removed from the test stand, the
carbon tube ends were then plugged, and the column was cooled for 1 hour. The mass of MeBr
adsorbed onto the carbon sample was analyzed gravimetrically. Desorption efficiency was
assessed by the mass of MeBr desorbed per mass adsorbed. After the AC tube was weighed, it
was placed back into the test stand to conduct another adsorption cycle. Five
adsorption/desorption cycles were conducted for each AC sample.
12000
10000
(
i
8000
c
i
»
c_
Q-
~ 6000
CO
CD
4000
>
k
1
V
2000
n
V
u
15
59 16:24 16:50 17:16 17:42
Time hr:min
Figure 2-5. Typical Desorption Curve
12
-------
2.12 Gravimetric Method to Determine Adsorption Capacity
At the start of each sampling day, the balance was verified with Class-S calibration weights.
Calibration data were recorded in the laboratory record book associated with the balance.
Gravimetric analyses of the carbon samples were performed with the AC in the tube assembly,
which included borosilicate glass tube, quartz wool (model 502-177; Leco; St. Joseph, Ml),
Keck® joint clamp (Cole Parmer; Chicago, IL), and two plugs, on a custom weighing stand.
When weighing the carbon tube, the balance reading was allowed 20 min to stabilize before
recording the reading. This procedural step was based on preliminary test findings that are
detailed in Appendix C.
Each fresh AC sample was weighed in a tared weigh boat (model YBW 01L; Gottingen,
Germany) before transferring to the tube assembly, henceforth carbon tube.
The mass of MeBr adsorbed by the AC sample during a cycle was determined as the difference
of the masses of the carbon tube with conditioned (or desorbed) AC and the mass of the carbon
tube with saturated AC from the subsequent adsorption cycle. The adsorption capacity of the
AC sample was determined by dividing the mass of the adsorbed MeBr by the mass of the
conditioned AC sample.
The amount of desorbed MeBr was determined by the difference in the masses of the carbon
tube with saturated AC from the adsorption phase and the mass of the carbon tube with
desorbed AC from the following desorption cycle. The amount of MeBr desorbed was
normalized by dividing by the mass of the conditioned AC.
All gravimetric measurements were recorded to the fourth decimal place.
2.13 Integration Method to Determine Adsorption Capacity
Real-time gas flowrates (Lpm), gas temperatures (°C), and MeBr concentration (ppm) data were
used for the integration method to determine mass of MeBr mass adsorbed and desorbed and
were digitally recorded (ten-second averages) during testing for post-test processing. For the
integration method, the total mass of MeBr adsorbed onto the AC was calculated based on the
difference between the carbon tube challenge concentration (target of 53,000 ppm) and outlet
concentration of MeBr in the gas stream, and then converted to mass (as described below), for
each time step, for the duration of the adsorption phase.
Inlet MeBr Mass (adsorption)
Prior to the start of adsorption testing, the inlet (challenge) MeBr gas concentration was
measured by bypassing the carbon tube. Avogadro's Law was used to determine the number of
moles detected for the time interval (moles (mol)/min) (using the recorded gas temperature and
assuming the pressure was close to atmospheric).
Nmiet PV (1 atm)(y)
1 RT (0-082057sS)
-------
where ISLet is the total number of moles of gas at the inlet, t is a unit of time (e.g., minutes), V is
the volumetric flow rate of the inlet gas (L/minute), P is pressure (1 atmosphere), R is the ideal
gas law constant, and T is the temperature (K) of the inlet gas.
The moles of gas at the inlet for each time interval were multiplied by the elapsed time for
number of moles.
^ in} pt
Ninlet=^fi>< At
where ISLet is the number of moles of gas at the inlet, t is a unit of time, and At is the time
interval.
The moles of MeBr were calculated using the concentration of MeBr measured by the FID (CwieBr
[ppm]) at the time interval:
CMeBr(ppm) Af
MeBr ~ 1,000,000 inlet
where NwieBr is the number moles of MeBr at the inlet.
The molecular weight of methyl bromide (94.94 g/mol) was used to determine the mass (g)
adsorbed at the time interval:
mMeBr inlet ~ MWMeBr * ^VjvfeBr
where mMeBr iniet is the mass of MeBr entering the AC tube for a specific time step, and MWwieBr is
the molecular weight of MeBr
Outlet MeBr mass (desorption)
A similar method was used to calculate the mass of MeBr desorbed at the outlet of the carbon
tube. Instead of using the MeBr concentration measurement at the inlet, real-time MeBr
measurements at the outlet were used to calculate the MeBr mass exiting the carbon tube.
The total mass of MeBr desorbed was determined by subtracting the calculated MeBr mass at
the outlet from the inlet mass (FID reading when sampling the desorption air adjusted by the
baseline reading) for each time interval, then adding the amounts over the duration of the
desorption cycle. The point at which the system reached steady state marked the time the test
cycle was complete.
14
-------
M = If-i f, xQxAtx 1/VM x Mw)
Zji_1 Vl,000,000 v ' M )
where M is the total mass of MeBr (g), Ci is the concentration of MeBr (ppm) at time interval I
(minute), Q is the volumetric flow rate of gas (liters/minute), At is the time interval (minutes), Vm
is the molar volume of ideal gas at the temperature of the gas (liters), and MW is the molecular
weight of MeBr (94.94 g/mol).
2.14 Adsorption Capacity Characterization
Each adsorption cycle was conducted for approximately 33 minutes to ensure that the
concentration of MeBr at the outlet of the carbon tube had stabilized (indicating saturation or
maximum adsorption capacity of the AC). The outlet level was considered stable when the
difference between each successive 10-min block average of MeBr concentration was less than
0.3%.
The mass of MeBr adsorbed and desorbed for each cycle was determined using the two
methods: a gravimetric approach and an integration approach.
For the integration method (Section 2.13), the total mass of MeBr adsorbed onto the AC was
calculated based on the difference between the carbon tube challenge concentration (target of
53,000 ppm) and outlet concentration of MeBr in the gas stream, and then converted to mass
(as discussed in Section 2.13), for each time step. Initially, the mass adsorbed was calculated
using a time step of six seconds, but for the last two tests (3.2 and 3.3), a time step of one
second was used to identify the moment of breakthrough more accurately. The total mass
adsorbed was then accumulated for the total time the adsorption cycle took place, i.e., until the
stable MeBr level was reached. This approach to determining mass adsorbed is like the method
used by Li et al. (2003).
The adsorption capacity was also determined gravimetrically as the mass of MeBr adsorbed
divided by the conditioned mass of AC tested. AC samples were weighed with a gravimetric
balance before the adsorption and desorption phases of each regeneration cycle (Section 2.12).
2.15 Control Test with Elevated RH
Control tests were conducted using a challenge gas with elevated RH (75%) but no MeBr, to
assess the mass of moisture adsorbed onto the AC bed compared to the total mass gain when
exposed to a challenge gas with an RH of 75% and MeBr (Test 3). These control tests were
carried out in the same manner as Test 3 except ambient air at 75% RH was used as the
challenge gas instead of 5.3% MeBr in N2, at 75% RH.
2.16 Statistical Analysis
Adsorption capacities as well as other measurements associated with the study parameters
were compared using one-way analysis of variance (ANOVA) in MS Excel. The p-value from
two-sided (non-directional) tests was used to test for significance (a = 0.05).
15
-------
3 Quality Assurance/Quality Control
Quality assurance (QA)/quality control (QC) procedures and results are summarized below.
3.1 Sampling, Monitoring, and Equipment Calibration
3.1.1. VIG FID Model 20 Concentration Measurement
The Model 20 is a microprocessor-based oven-heated total hydrocarbon gas analyzer designed
for high accuracy, sensitivity, and stability. The Model 20 uses an FID for continuous
measurement. All components that contact the sample throughout analysis are maintained in a
temperature-controlled oven to prevent condensation and to provide repeatable, reliable
performance in the analysis. Ultra-zero N2 was used for zero calibrations. Certified
concentrations of 0.05% and 1.3% MeBr (balanced in nitrogen) were used for span calibrations.
The specifications for the instrument are listed in Table 3-1.
Table 3-1. FID Operating Specifications
Parameter
Specifications
Zero and Span Noise
Less than 0.2% of full scale
Zero and Span Drift
±1% full scale per 24 hours
Range3
0-100, 0-1000, 0-10000, 0-100000 ppm
Linearity
Within 1% of full scale through all ranges
Repeatability
Within 1% of full scale through all ranges
Stability
Within 1% of full scale through all ranges
Sample Flow Rate
4 L/min
Avg. Accuracy (0.05% MeBr)b
13.3% (± 12.5% RSD)
Avg. Accuracy (1.3% MeBr)b
1.5% (± 1.8% RSD)
Precision as Calibrated (%RSD)C
± 1.818%
aFour ranges per amplifier
b Expressed as percent error. Assessed in-situ with certified MeBr (0.0500% ± 0.001% or 1.3% ±
0.0026%) prior to each test over the duration of the study
cAssessed with certified MeBr (1.3% ± 0.0026%) over the duration of the study
3.1.2. Temperature and RH Measurements
The real-time temperature and RH measurements were collected using a K-type thermocouple
and Vaisala (Vantaa, Finland) HMD53 temperature and humidity probe, respectively. The
specifications for each device are shown in Table 3-2.
16
-------
Table 3-2. Relative Humidity Sensor and Temperature Probe Operating Specifications
Parameter
K-Type Thermocouple
Vaisala
RH range
NAa
0 to 98%
RH accuracy: 0-90 %
NA
± 3%
RH accuracy: 90-98 %
NA
± 5%
RH resolution
NA
0.001 %b
Temperature range
-200 to 1200 °C
-10 to 60 °C
Temperature accuracy
± 1.2°C@25 °C
±0.6 °C @20 °C
Temperature resolution
0.01 °C
0.001 °Cb
Uncertainty as Calibrated0
± 0.15 °C
1.5%
Precision as Calibrated11 (%RSD)
N/A
± 1.7%
aNot applicable.
bVaisala resolution
cAssessed by the Metrology Laboratory prior to testing
dAssessed in-situ with Vaisala NaCI calibration cell periodically over the duration of the study
3.1.3. Mass Flow Rate
SmartTrak® 100 (Sierra Instruments, Inc., Monterey, CA) mass flow controllers were used to
control and collect real-time measurements of each gas introduced to the system. Table 3-3
shows the specifications of each MFC used in this investigation.
Table 3-3. Mass Flow Controller Specifications
Parameter
Range
0 - 1 slpm
0-5 slpm
0-15 slpm
Accuracy3
±1.0% full scale
±1.0% full scale
±1.0% full scale
Linearity
± 0.05% full scale
± 0.05% full scale
± 0.05% full scale
Repeatability
± 0.2% full scale
± 0.2% full scale
± 0.2% full scale
Uncertainty as Calibrated15
± 2.9 SCCM
± 0.07 % of corrected
reading and ± 0.01
LPM
± 0.03 % of corrected
reading and ± 0.10
LPM
Precision as Calibrated0
(%RSD)
± 1.19%
± 0.692%
± 1.27%
aAs received from the manufacturer
bAssessed by the Metrology Laboratory prior to testing
°Assessed in-situ at test conditions with the Gilian flow calibrator over the duration of the study
Prior to each test, flow rates were verified using the Gilian Gilibrator-2 Flow Calibrator
(Sensidyne, LP, St. Petersburg, FL). An average of five readings was used as the accepted
measurement. For each sequence of readings, no individual reading varied more than 10
standard cubic centimeters per minute (CCM) from the average. The specifications of the flow
calibrator are shown in Table 3-4.
17
-------
Table 3-4. Flow Calibrator Specifications
Parameter
Specification
Flow Cell Type
Standard
Flow range
20 CCM to 6 Lpm
Accuracy3 (wet cell)
Better than 1%
Uncertainty as Calibrated15
± 0.04% of corrected reading and ± 25
SCCM
aAs received from the manufacturer
bAssessed by the Metrology Laboratory prior to testing
3.1.4. Mass Measurements
Mass measurements of the carbon and associated tube/column apparatus were performed
using an Entris 124-1S analytical balance (Sartorius, LLC, Columbus, OH). The weighing range
was up to 210 g with a readability of 0.0001 g. Prior to each use, the calibration was verified
using Troemner-certified class 2 weights (ASTM International (ASTM) Certification 782762A).
Table 3-5 shows the balance specifications.
Table 3-5. Balance Operating Specifications
Parameter
Specification
Capacity
120 g
Readability
0.0001 g
Repeatability
0.0001 g
Linearity
0.0002 g
Accuracy as Calibrated3 b
± 0.001 Og
Precisionb
-------
Table 3-6. Summary Sampling and Monitoring Equipment QA/QC Checks
Equipment
Matrix
Measurement
Calibration
Frequency
Calibration
Method
Acceptance
Criteria
Corrective
Action
RH probe
Challenge
gas
Inlet RH
Weekly
Saturated salt
cells
± 5%
Replace
Vaisala
sensor.
FID
Outlet gas
MeBr in outlet
stream
Before each
test
Zero and
certified span
gases
± 5% target
Leak check
system and
repeat
calibration.
Thermocouple
Challenge
gas
Inlet
temperature
Annually
Compared to
N ISP-
traceable
thermometer
± 1.1 °C
Replace
thermocouple.
Balance
Activated
carbon
sample
MeBr gained/
lost
Before each
use
Comparison
to class 2
weights
± 0.1% target
Recalibrate
balance.
MFC
Feed gases
Gas flow rate
Annually
Comparison
to a Gilibrator
calibration
system
± 1% of full
scale
Check line for
leaks.
Recalibrate
MFC, if
necessary.
N ISP-
traceable
timer
Exposure
duration
Time
Annually
Compared to
NIST-
calibrated
timer
± 1 minute
per hour
Return to
manufacturer
for
recalibration.
aNational Institute of Standards and Technology
3.2 Acceptance Criteria for Critical Measurements
Equipment detection limit values were provided by the manufacturer in the product literature.
Failure to provide a measurement method or device that meets these goals resulted in a
rejection of results derived from the critical measurement. For instance, some data points can
be missing for the real-time test MeBr concentration, but the concentration can be calculated for
whatever time interval data are available. Failure to collect the sorbent weight before the test,
however, completely invalidated the test, and the test was repeated. Table 3-7 lists the
quantitative acceptance criteria for critical measurements. Table 3-8 lists the critical
measurements, their target values and, the accuracy of the actual values with respect to the
target values (expressed as percent error) for each test. Table 3-9 lists the precision (as
assessed by standard deviation) of the critical measurements.
19
-------
Table 3-7. Critical Measurement Acceptance Criteria
Critical
Measurement
Accuracy
Precision
Detection
Target
Corrective
Measurement
Device
Limit
Action
MeBr test
FID
Unknown
±5%
250 ppm
NA
Check
concentration
sample lines
for leaks.
Recalibrate
FID.
MeBr
FID
± 5% of
±5%
250 ppm
13,000
Check
challenge
calibration
ppm ± 1 %
sample lines
concentration
gas
for leaks.
Recalibrate
FID.
RH of MeBr
Vaisala
±5%
±5%
NA
75%a
Replace
challenge
Vaisala
stream
sensor.
Temperature
of air stream
K-type
thermocouple
± 1.2 °C @
100 °C
NA
0.01 °C
100±1 °C
Adjust GC
oven set
(desorption)
point.
MeBr flow rate
Gilibrator
±5%
±5%
.02 Lpm
0.47 Lpm
Check
(adsorption)
sample lines
for leaks.
Air flow rate
Gilibrator
±5%
±5%
.02 Lpm
1.1 Lpm
Check
(desorption)
sample lines
for leaks.
Carbon bed
Sartorius
± 1%
±0.001 g
0.001 g
1.0 ±0.01
Add/remove
mass
Entris 124-1S
g
sample (virgin
only).
NA = not applicable. GC = Gas chromatograph.
aTest 3 series only
20
-------
Table 3-8. Accuracy (% Error) of Critical Measurements
Test ID
AC
Sample
Mass
MeBr
Challenge Gas
Flow Rate
RHa
Desorption
Air Flow Rate
Inlet
Temperature15
Target Value
1 g
0.5 LPM
75%
1 LPM
100°C
1.1°
3.4
28.5
-
-
-
1.2
0.4
20.7
-
12.6
1.4
1.3
3.5
26.0
-
11.9
0.9
1.4
1.9
26.0
-
11.9
1.4
2.1
4.8
26.0
-
9.6
0.8
2.2
4.9
26.0
-
9.0
0.4
2.3
2.5
26.0
-
9.7
0.6
3.1c
2.8
-
-
-
-
3.1b
3.2
4.0
3.4
10.3
0.6
3.2
4.4
0.86
1.8
9.4
1.2
3.3
5.8
0.87
2.0
8.3
2.1
Values in unshaded cells are in units of % error, i.e., the % difference between the target
value and the actual value
aRH was a critical measurement only for the Test 3 series.
bTemperature was monitored during adsorption and controlled during desorption.
Test 3.1 was aborted due to a technical issue, incomplete data. A new test was run which
is referred to as 3.1b
Table 3-9. Data Precision (SD) of Critical Measurements
Test ID
MeBr Flow Rate
(Lpm)
RHa
(%)
Air Flow Rate
(Lpm)
Inlet Tempb
(°C)
1.1°
-
-
-
-
1.2
0.058581
0.6
0.00447
2.4
1.3
0.000015
0.7
0.00028
1.8
1.4
0.000022
0.7
0.00028
1.7
2.1
0.000082
0.52
0.00038
1.42
2.2
0.000010
0.50
0.0051
0.84
2.3
0.000038
0.61
0.0087
1.84
3.1c
-
-
-
-
3.1b
0.044721
2.06
0.0074
1.04
3.2
0.000034
1.7
0.0047
1.2
3.3
0.000020
1.4
0.015
1.4
aRH was a critical measurement only for the Test 3 series.
bTemperature was a critical measurement only for the desorption phase.
Test aborted; incomplete data. A new Test 3.1b was conducted in place of
Test 3.1
21
-------
While the MeBr concentration was a critical measurement, the accuracy of the FID could not be
assessed because it is not specific to MeBr. That is, there was the small possibility that trace
amounts of hydrocarbons or gases capable of producing a signal on the FID (other than MeBr)
were present in the gas stream, especially from the desorption cycle. Nevertheless, the FID was
zeroed and spanned on certified MeBr calibration gases before and after each test.
A post-test span bias check on the FID was performed with 1.3% MeBr after each test,
repeating the procedure performed for the pre-test span. Both points were used to assess the
analyzer drift over the test duration (5 cycles) of five to seven days. The FID maintained less
than 5% drift for each test. Table 3-10 shows the FID drift (calculated as the percent difference
in instrument readings taken before and after each test, when using the 1.3% MeBr calibration
gas).
Table 3-10. FID Drift for Each Test Duration
Test ID
Pre-test 1.3% MeBr
Span
(PPm)
Post-test 1.3% MeBr
Span
(PPm)
MeBr Analyzer
Drift
(%)
1.1a
-
-
-
1.2
13004
13174
1.31
1.3
12808
12694
-0.89
1.4
12551
13116
4.50
2.1
13072
13459
2.97
2.2
13256
13517
1.97
2.3
12677
12939
2.07
3.1a
-
-
-
3.1b
13090
13658
4.34
3.2
13019
12643
-2.89
3.3
12824
12653
-1.34
aTest aborted; incomplete data. A new Test 3.1b was conducted in the place of
Test 3.1
22
-------
4 Results and Discussion
The results from the test matrix (see Table 2-1) and some additional ancillary/preliminary tests
are presented in this section. Test results to assess the stability of the carbon tube mass
measurements over time are presented in Appendix B, and scoping results related to
operational factors affecting adsorption capacity are presented in Appendix C.
4.1 Stability of AC Moisture Content and Related Measures
Each AC sample was collected from the bulk container stored in the test facility (indoors in the
laboratory). Upon arrival, the bulk container contained approximately 50 pounds (lb) of AC. The
moisture content of the AC was monitored over the duration of the investigation, since moisture
levels of the AC could potentially impact adsorption capacity.
The moisture content observation period totaled 693 days during which the measurements
ranged from the minimum (2.46% [±0.03 SD]) to maximum (3.53% [±0.03 SD]). Table 4-1
shows that the moisture content of the bulk AC under storage conditions was generally stable.
Table 4-1. Moisture Content (%) of Bulk Carbon During Evaluation Period
Date
Time Under
Observation (Days)
Ave. Moisture Content3
(%, ±SD)
1/12/16
0
2.46 ±0.03
5/11/16
120
2.55 ±0.03
2/1/17
386
2.70 ±0.0
9/22/17
619
3.38 ±0.05
10/17/17
644
3.53 ±0.03
11/9/17
667
3.50 ±0.17
11/24/17
682
3.38 ±0.12
12/5/17
693
3.44 ±0.07
a average of 3 samples
On average, 0.02 g (±0.005 SD) of moisture were removed from AC samples during the
conditioning phase portion of testing. Conditioned AC samples ranged from 1.0267 g (Test 2.1)
and 1.1237 g (Test 3.2); the precision within the group of AC test samples (Test Series 1-3) was
0.0335 (SD). Table 4-2 details the amount of moisture removed from fresh AC samples at
completion of the conditioning phase, adjusted per gram of carbon.
23
-------
Table 4-2. Amount of Moisture Removed from Fresh Carbon Bed (gig carbon)
Test ID
Unconditioned
Conditioned
Moisture
Carbon (g)
Carbon
Removed (g
(g)
water/g carbon)*
1.2
1.0959
1.0647
0.0284
1.3
1.0616
1.0267
0.0329
1.4
1.0787
1.0516
0.0251
2.1
1.0471
1.0276
0.0186
2.2
1.0466
1.0276
0.0181
2.3
1.0730
1.0560
0.0158
3.1b
1.0653
1.0418
0.0220
3.2
1.1480
1.1237
0.0211
3.3
1.0360
1.0133
0.0220
"calculated by dividing the difference in mass by the unconditioned carbon mass
4.2. Activated Carbon Regeneration Tests
Actual measured test conditions for regeneration (adsorption/desorption) tests are detailed in
Appendix A. The prescribed challenge gas conditions for each test series were as follows:
• Test 1
- Adsorption (challenge) gas: dry 5.3% MeBr in N2 at ambient temperature (22 ±
0.03 °C)
- Desorption: dry, laboratory air at 100 °C
• Test 2
- Adsorption (challenge) gas: dry 5.3% MeBr in N2 at ambient temperature (22 ±
0.03 °C)
- Desorption: ambient air at 100 °C
• Test 3
- Adsorption (challenge) gas: 5.3% MeBr in N2 at 75% RH at ambient temperature
(22 ± 0.03 °C)
- Desorption: ambient air at 100 °C
4.2.1 Adsorption and Desorption Results for Each Test Cycle
Table 4-3 shows the MeBr adsorption and desorption results, determined using both the
gravimetric and integration methods, for every test cycle (n=45 for total of these tests). Note: the
MeBr scrubber system with potassium hydroxide (KOH) solution was replaced with an MeBr
scrubbing system containing carbon (see Appendix A) between Tests 2.2 and Test 2.3. This
24
-------
was done because during Test 2.2, KOH scrubber solution inadvertently back-flowed into the
FID due to a change in pressure in the line. The FID had to be repaired, and to prevent potential
recurrence of having KOH back-flow into the FID again, the KOH scrubber was replaced with a
dry carbon trap.
We note also that about a third of the test replicate results from the "integration" method
showed that the mass desorbed was somewhat greater than the mass adsorbed. We believe
this is an artifact of the method, and indicative of the inherent variability in such measurements
using the FID; we do not believe more mass is being desorbed than adsorbed in these cases.
We note also that when we average the replicate results for each test, there was always less
mass being desorbed than adsorbed.
25
-------
Table 4-3. Adsorbed and Desorbed MeBr Using Gravimetric and Integration Analysis
Methods
Gravimetric
Integration
Test ID
Cycle
MeBr
MeBr
MeBr
MeBr
Adsorbed,
Desorbed,
Adsorbed,
Desorbed,
g/g AC
g/g AC
g/g AC
g/g AC
1.2
1
0.5132
0.5079
0.4399
0.475
2
0.5003
0.4999
0.3396
0.445
3
0.5026
0.5013
0.4189
0.492
4
0.5028
0.5043
0.5061
0.446
5
0.5065
0.5052
0.4480
0.534
1.3
1
0.4884
0.4874
0.3876
0.484
2
0.4744
0.4738
0.5551
0.494
3
0.4716
0.4722
0.5499
0.470
4
0.4805
0.4788
0.5982
0.480
5
0.4856
0.4853
0.6100
0.508
1.4
1
0.4520
0.4505
0.4959
0.471
2
0.4507
0.4499
0.5517
0.459
3
0.4618
0.4611
0.3558
0.412
4
0.4421
0.4416
0.4998
0.384
5
0.4487
0.4482
0.3663
0.425
Average ± SD
0.4787±0.023
0.4778±0.023
0.4749±0.089
0.4653±0.038
2.1
1
0.4612
0.4591
0.6138
0.500
2
0.4742
0.4752
0.7378
0.540
3
0.4754
0.4747
0.4670
0.519
4
0.4746
0.4741
0.6706
0.539
5
0.4813
0.4802
0.6577
0.519
2.2
1
0.4709
0.4711
0.6283
0.503
2
0.4628
0.4609
0.6123
0.521
3
0.4643
0.4649
0.6088
0.501
4
0.4665
0.4637
0.5788
0.521
5
0.4592
0.4585
0.6595
0.493
2.3
1
0.4310
0.4283
0.5613
0.489
2
0.4382
0.4392
0.4695
0.495
3
0.4294
0.4284
0.4924
0.496
4
0.4362
0.4360
0.5242
0.478
5
0.4326
0.4309
0.5454
0.497
Average ± SD
0.45719±0.018
0.45634±0.019
0.5885±0.079
0.5074±0.018
26
-------
Gravimetric
Integration
Test ID
Cycle
MeBr
MeBr
MeBr
MeBr
Adsorbed,
Desorbed,
Adsorbed,
Desorbed,
gig AC
gig AC
gig AC
gig AC
1
0.4515
0.4459
0.4631
0.470
2
0.4504
0.4476
0.4971
0.457
a3.1b
3
0.4474
0.4490
0.4244
0.457
4
0.4564
0.4533
0.4966
0.488
5
0.4532
0.4590
0.4501
0.487
1
0.4598
0.4580
0.4466
0.507
2
0.4553
0.4552
0.4831
0.468
3.2
3
0.4602
0.4581
0.4786
0.467
4
0.4571
0.4543
0.4921
0.473
5
0.4646
0.4637
0.4301
0.477
1
0.4457
0.4494
0.4284
0.488
2
0.4590
0.4543
0.4916
0.473
3.3
3
0.4467
0.4467
0.4993
0.471
4
0.4532
0.4576
0.4766
0.510
5
0.4599
0.4574
0.5401
0.484
Average ± SD
0.4547±0.006
0.4540±0.005
0.4732±0.03
0.478±0.016
a Due to a technical issue, Test 3.1 was not completed, and so a new Test 3.1b was conducted in its
place.
As Table 4-3 shows, the results for the adsorption and desorption tests range from
approximately 0.43 to 0.58 g MeBr per gram AC. The results were generally similar for the two
methods used to determine adsorption capacity, except in Test 2. In Test 2, results determined
via the "integration" method were ~ 25% higher compared to the gravimetric method. It is
unclear why the difference in the results, but the difference may be related to having the FID
repaired during the Test 2 series, as discussed above. No other changes in methods or
equipment occurred.
The correlation between the adsorption and desorption cycles (mass of MeBr) is further
discussed below; the effects of the challenge gas characteristics and the number of reuse
cycles on adsorption capacity are also discussed below.
27
-------
Figure 4-1 shows good correlation between the adsorbed and desorbed MeBr gravimetric
measurements, for all 45 cycles conducted in the study. The high linear correlation between the
two phases indicates low variability in the MeBr mass data as well as effective desorption of the
MeBr.
0.5200
0.5100
— 0.5000
<
CuO 0.4900 mr''*
"5s .®*
0.4800 ^ y = 0.9876X + 0.0049
S „ R2 = 0.9886
¦£ 0.4700 ,W
v/
,»c
a) 0.4600
Q
m 0.4500
-------
0.600
0.550
0.500
0.450
u
<
W)
• * .*
- * £•*
• V * "• *
# f y = 0.1995x +0.3816
~a
-------
Overall, the results show comparable adsorption capacities for the three test conditions,
regardless of the RH of the adsorption gas. Table 4-4 provides a comparison of the adsorption
capacity (gravimetric method) for each of the three tests.
The adsorption capacity results obtained in this study (average of approximately 0.45 g MeBr/g
AC) are much higher than adsorption capacity results that were obtained in a field study (0.05 g
MeBr/g AC; Wood et al., 2015). This difference may be because in the field study, the challenge
concentration diminished over time as the fumigated building was aerated. The adsorption
capacity results for the present study are also somewhat higher than what was obtained in
method development tests (~ 0.30 g MeBr/g AC; refer to Appendix D). This difference in results
may be due to the lower RH of the present study compared to what was used in method
development (-40%) tests; the present study also incorporated preconditioning (essentially
drying) of the AC, whereas the method development tests did not.
0.60
T1 T2 T3
Test Series
Figure 4-3. Average adsorption capacity for each test series (±SD)
Table 4-4. Adsorption Capacity Comparison for Each Test Series
Test Series
Ave. Carbon Bed
Inlet RH
(%, ±SD)
Ave. Desorption
AirRH @100°C
(%, ±SD)
Ave. MeBr Adsorption
Capacity (g/g AC)
(±SD)
P-Value
(a = 0.05)
T1
3.8 ±0.6
1.0 ±0.2
0.4787 ±0.0236
0.00016
T2
3.1 ±2.6
4.6 ±0.3
0.4572 ±0.0184
0.14
T3
74.4 ±2.1
0.9 ±0.2
0.4547 ±0.0056
0.04
30
-------
4.2.3 The Effect of Reuse on Adsorption Capacity
The results show stable adsorption capacities over the course of the five adsorption/desorption
cycles, for each of the three conditions (with each test having three replicates) tested. As seen
in Figures 4-4 through 4-6, average adsorption capacity (gravimetric method) is essentially
unchanged after the fifth cycle.
¦ Avg. MeBr Adsorbed
¦ Avg. MeBr Desorbed
Figure 4-4. Average adsorption capacities for Test 1 cycles (±SD)
31
-------
¦ Avg. MeBr Adsorbed
¦ Avg. MeBr Desorbed
Figure 4-5. Average adsorption capacities for Test 2 cycles (±SD)
¦ Ave. MeBr Adsorbed
¦ Ave. MeBr Desorbed
Figure 4-6. Average adsorption capacities for Test 3 cycles (±SD)
32
-------
4.3. Control Tests with Elevated RH Only
The purpose of these control tests using air only (no MeBr) with elevated RH was to gauge the
amount of water vapor that could be adsorbed onto the AC and compare this value to the mass
adsorbed for the Test 3 evaluations that used the same elevated RH but with MeBr.
The average moisture adsorption capacity was 0.0442 g/g AC (± 0.0003 SD), which was less
than 10% of the average adsorption capacity determined for Test 3 (0.45 g/g AC). Figure 4-7
shows the moisture loaded versus the moisture removed on the AC sample. The desorption
portion for Test 1 was inadvertently not completed.
0.0800
0.0700
0.0600
^ 0.0500
ao
0.0400
&_
CD
| 0.0300
0.0200
0.0100
0.0000
I Water Adsorbed
I Water Desorbed
Test 1
Test 2
Test No.
Figure 4-7. Adsorbed and desorbed moisture with an exposure stream of 75% RH in ambient air
33
-------
5 Summary and Conclusions
Overall, the results for the adsorption tests showed relatively high levels of MeBr adsorption and
ranged from approximately 0.43 to 0.58 g MeBr per gram AC. These high adsorption capacity
levels may be due to the preconditioned (dried) AC samples we used, as well as the low RH
levels in the challenge gases (applicable to the first two test series).
The differences in adsorption capacity as a function of challenge gas and/or desorption gas
humidity levels were generally minor, although in some cases the differences were statistically
significant. Tests to determine adsorption of moisture at high RH (75%), without the presence of
MeBr, suggest that the adsorption capacity for moisture was approximately 10% of the AC
capacity for MeBr.
The adsorption capacity of the AC samples was not affected (did not diminish) in the five
adsorption/desorption cycle series to which each AC sample was subjected. That is, the
process of desorbing the MeBr using a 100 °C temperature gas was effective in maintaining
relatively high adsorption levels of the AC samples for at least five cycles; the adsorption
capacity did not diminish over the course of the five cycles but remained rather stable. However,
we caveat that using only five cycles may not yield sufficient data to assess the effect of
numerous repeated adsorption/desorption cycles.
While we have demonstrated in this study the ability of AC to be reused (i.e., regenerated)
without losing its adsorption capacity for MeBr over several cycles, further research remains
related to the use of MeBr on a wide scale following a B. anthracis incident. Specifically,
research related to the reuse of the MeBr desorbed from AC is recommended. Anecdotal
evidence suggests that the MeBr gas desorbed from AC may contain impurities (i.e., other
chemical species produced from the chemical interactions between the MeBr and AC during the
adsorption/desorption process) that could make the gas unsuitable for reuse as a
decontaminant. Further investigation into this potential is recommended.
34
-------
6 References
Gan, J, Megonnell NE, Yates SR (2001) Adsorption and catalytic decomposition of methyl
bromide and methyl iodide on activated carbons. Atmos. Environ. 35: 941-947.
doi: 10.1016/S1352-2310(00)00339-3
Joyce PJ, Bielski R (2010) Method and system for removing alkyl halides from gases. United
States Patent Application US 2010/0101412 A1. April 29, 2010.
Juergensmeyer MA, Gingras BA, Scheffrahn RH, Weinberg MJ (2007) Methyl bromide fumigant
lethal to Bacillus anthracis spores. J. Environ. Health 69:24-26.
Li YH, Lee, CW, Gullet, BK, Importance of activated carbon's oxygen surface functional groups
on elemental mercury adsorption; Fuel 2003, 82, 451-457. https://doi.org/10.1016/S0016-
2361(02)00307-1.
Leesch JG, Knapp GF, Mackey BE (2000) Methyl bromide adsorption on activated carbon to
control emissions from commodity fumigations. J. Stored Prod. Res. 36:65-74.
doi:10.1016/S0022-474X(99)00028-4
National Institute for Occupational Safety and Exposure (2018). NIOSH Pocket Guide to
Chemical Hazards, Methyl bromide, https://www.cdc.qov/niosh/npq/npqd0400.html: Accessed
11/7/18.
Serre S, Mickelsen L, Calfee MW, Wood JP, Gray Jr. MS, Scheffrahn, RH, Perez, R, Kern Jr.
WH, Daniell N (2016) Whole-building decontamination of Bacillus anthracis Sterne spores by
methyl bromide fumigation. J. Appl. Microbiol, doi: 10.1111/jam.12974
Snyder, JD, Leesch, JG (2001) Methyl bromide recovery on activated carbon with repeated
adsorption and electrothermal regeneration. Ind. Eng. Chem. Res. 2001, 40, 2925-2933.
doi: 10.1021/ie000395v
Value Recovery Inc. (2018) Fumigation emissions controls.
http://www.valuerecoverv.net/Technoloqy.html (Accessed October, 2018).
Wood JP, Clayton MJ, McArthurT, Serre SD, Mickelsen L, Touati A (2015) Capture of methyl
bromide emissions with activated carbon following the fumigation of a small building
contaminated with a Bacillus anthracis spore simulant, Journal of the Air & Waste Management
Association, 65:2, 145-153, doi: 10.1080/10962247.2014.980017
Wood JP, Wendling M, RichterW, Lastivka A, Mickelsen L (2016) Evaluation of the efficacy of
methyl bromide in the decontamination of building and interior materials contaminated with B.
anthracis spores. Appl. Environ. Microbiol.. April 2016; 82:7 2003-2011; Accepted manuscript
posted online Jan 22, 2016. doi:10.1128/AEM.03445-15
35
-------
Yang Y, Yuanqing L, Walse SS, Mitch WA (2015) Destruction of methyl bromide sorbed to
activated carbon by thiosulfate or electrolysis. Environ. Sci. Technol. 49: 4515-4521.
36
-------
Appendix A: Test Conditions
A-1
-------
Table A-1. Adsorption Phase Test Conditions
Test ID
Cycle
Ave. MeBr
Ave. Dilution N2
Ave. Inlet
Ave. Inlet
Ave. Outlet
Cycle
Flow Rate
Flow Rate
RH
Temp
Temp
Duration minutes
Lpm (± SD)
Lpm (± SD)
% (± SD)
°C (± SD)
°C (± SD)
1.2
1
0.5 ±0.0001
3.56 ±0.0010
4.4 ±0.4
20.3 ±0.1
21.7 ±0.9
31.8
2
0.37 ± 0.00008
3.07 ±0.0006
3.2 ±0.3
20.7 ±0.2
23.0 ±0.4
31.2
3
0.37 ± 0.00008
3.58 ±0.0007
4.1 ±0.4
21.6 ±0.4
23.0 ±0.4
30.8
4
0.37 ± 0.00008
3.48 ±0.0007
3.0 ±0.3
20.9 ±0.1
22.5 ±0.4
29.8
5
0.37 ± 0.00008
3.53 ±0.0008
3.6 ±0.4
22.6 ±0.2
22.1 ±0.5
29.3
1.3
1
0.37 ± 0.000077
3.12 ±0.0007
2.8 ±0.3
20.0 ±0.1
21.4 ±0.8
30.7
2
0.37 ± 0.000069
3.53 ±0.0008
4.1 ±0.4
21.1 ±0.2
21.9 ±0.4
30.8
3
0.37 ± 0.000068
3.46 ±0.0008
3.5 ±0.3
20.6 ±0.2
22.2 ±0.4
32.8
4
0.37 ± 0.000076
3.55 ±0.0007
4.3 ±0.4
22.2 ±0.5
21.8 ±0.4
36.3
5
0.37 ± 0.000068
3.55 ±0.0019
4.3 ±0.4
20.5 ±0.3
21.1 ±0.1
36.5
1.4
1
0.37 ± 0.000073
3.55 ±0.0008
3.3 ±0.4
21.1 ±0.1
23.0 ±0.1
32.7
2
0.37 ± 0.000070
3.55 ±0.0008
4.4 ±0.4
23.2 ±0.4
22.8 ±0.3
32.3
3
0.37 ± 0.000069
3.55 ±0.0007
4.6 ±0.4
21.0 ±0.2
21.8 ±0.1
34.8
4
0.37 ± 0.000078
3.55 ±0.0021
4.6 ±0.3
24.6 ±0.7
24.1 ±0.3
36.0
5
0.37 ± 0.000078
3.48 ±0.0006
3.5 ±0.3
20.8 ±0.1
23.1 ±0.2
32.7
2.1
1
0.37 ± 0.000096
3.48 ±0.0006
4.8 ±0.4
22.5 ±0.4
25.2 ±1.02
33.0
2
0.37 ± 0.000071
3.55 ±0.0007
4.7 ±0.3
21.7 ±0.1
23.0 ±0.1
33.2
3
0.37 ± 0.000072
3.53 ±0.0015
4.5 ±0.4
21.3 ±0.2
23.2 ±0.2
33.5
4
0.37 ± 0.000075
3.56 ±0.0007
5.5 ±0.3
20.7 ±0.04
22.7 ±0.1
33.0
5
0.37 ± 0.000075
3.56 ±0.0008
4.6 ±0.5
21.0 ±0.2
22.9 ±0.03
33.7
2.2
1
0.37 ± 0.000070
3.56 ±0.0012
5.2 ±0.2
21.1 ±0.3
22.8 ±0.1
33.5
2
0.37 ± 0.000067
3.46 ±0.0007
5.2 ±0.2
20.7 ±0.04
23.2 ±0.3
33.5
3
0.37 ± 0.000065
3.53 ±0.0020
4.6 ±0.4
20.9 ±0.1
23.1 ±0.3
33.4
4
0.37 ± 0.000082
3.5 ±0.00080
4.4 ±0.5
20.8 ±0.04
22.8 ±0.04
33.5
A-2
-------
0.37 ± 0.000075
3.52 ±0.0007
4.5 ±0.4
20.9 ±0.1
23.5 ±0.4
33.0
Test ID
Cycle
Ave. MeBr
Flow Rate
Lpm (± SD)
Ave. Dilution N2
Flow Rate
Lpm (± SD)
Ave. Inlet
RH
% (± SD)
Ave. Inlet
Temp
°C (± SD)
Ave. Outlet
Temp
°C (± SD)
Cycle
Duration minutes
2.3
0.37 ± 0.000079
3.32 ±0.0009
0.3 ±0.1
21.8 ±0.3
25.3 ±0.4
34.2
0.37 ± 0.000072
3.47 ±0.0007
0.3 ±0.3
21.5 ±0.2
24.1 ±0.4
34.0
0.37 ± 0.000069
3.5 ±0.00080
-0.8 ±0.2
22.4 ±0.3
25.3 ±0.5
33.3
0.37 ± 0.000074
3.5 ±0.00070
-0.9 ±0.3
22.0 ±0.2
24.2 ±0.3
33.3
0.37 ± 0.000078
3.5 ±0.00080
-0.8 ±0.1
22.2 ±0.3
24.5 ±0.4
32.2
3.1b
0.40 ± 5.8E-15
3.45 ±0.0026
71.8 ±0.9
21.0 ±0.2
23.6 ±0.3
33.6
0.5 ±0.00026
3.51 ±0.0026
72.2 ±1.3
21.0 ±0.1
23.4 ±0.4
33.4
0.50 ± 0.00026
3.51 ±0.0026
71.4 ±0.8
21.3 ±0.1
23.7 ±0.4
33.3
0.50 ± 0.000081
3.49 ±0.0008
72.9 ±8.0
20.3 ±0.1
22.9 ±0.1
33.2
0.50 ± 0.00026
3.48 ±0.0026
74.4 ±0.9
21.3 ±0.2
23.2 ±0.2
34.0
3.2
0.50 ± 0.00026
3.50 ±0.00250
75.3 ±1.9
21.1 ±0.1
23.8 ±0.2
33.7
0.50 ± 0.00026
3.51 ±0.0026
73.8 ±0.6
21.1 ±0.1
23.5 ±0.6
33.9
0.50 ± 0.00027
3.51 ±0.0028
73.3 ±1
20.6 ±0.1
22.8 ±0.3
33.7
0.50 ± 0.00026
3.49 ±0.0027
71.9 ±0.7
21.0 ±0.2
23.1 ±0.1
33.6
0.50 ± 0.00009
3.5 ±0.00090
76.5 ±1.1
20.5 ±0.1
22.9 ±0.1
33.5
3.3
0.50 ± 0.00027
3.51 ±0.0026
77.6 ±1.6
21.3 ±0.2
23.6 ±0.6
33.5
0.50 ± 0.00023
3.53 ±0.0026
75.6 ±1.3
21.0 ±0.2
23.4 ±0.3
33.8
0.50 ± 0.00026
3.45 ±0.0026
75.7 ±0.3
21.6 ±0.1
24.2 ±0.4
33.4
0.50 ± 0.00026
3.43 ±0.0026
77.1 ±1.1
21.3 ±0.1
23.4 ±0.2
33.6
0.50 ± 0.000083
3.50 ±0.0015
76.3 ±0.9
21.8 ±0.2
23.6 ±0.2
33.7
Note: the MeBr scrubber system with
carbon between Tests 2.2 and Test 2
potassium hydroxide (KOH) solution was replaced with an M
.3.
eBr scrubbing system containing
A-3
-------
Table A-2. Desorption Phase Test Conditions
Ave. MeBr
Ave. Dilution N2
Ave. Inlet
Ave. Inlet
Ave. Outlet
Cycle
Test ID
Cycle
Flow Rate
Flow Rate
RH
Temp
Temp
Duration
Lpm (± SD)
Lpm (± SD)
% (± SD)
°C (± SD)
°C (± SD)
min
1.2
1
0.87 ±0.00026
3.07 ±0.0018
1.3 ±0.08
100.3 ±2.3
20.3 ±0.1
998.9
2
0.87 ±0.00025
3.07 ±0.0008
1.3 ±0.10
100.0 ±0.9
20.7 ±0.2
1105.8
3
0.87 ±0.00026
3.07 ±0.0016
1.2 ±0.08
103.2 ±1.3
21.6 ±0.04
1058.9
4
0.87 ±0.00027
3.07 ±0.0006
1.3 ±0.09
104.0 ±2.4
20.9 ±0.1
1105.7
5
0.88 ±0.00028
3.12 ±0.0016
1.2 ±0.08
100.1 ±0.6
22.6 ±0.2
1120.9
1.3
1
0.88 ±0.00026
3.11 ±0.0009
0.9 ±0.13
99.4 ±1.9
22.9 ±1.4
1040.2
2
0.88 ±0.00027
3.11 ±0.0009
0.9 ±0.12
101.7 ±1.2
24.0 ±1.3
1145.8
3
0.88 ±0.00027
3.11 ±0.0008
0.9 ±0.13
100.4 ±0.7
23.1 ±1.3
1157.7
4
0.88 ±0.00026
3.11 ±0.0008
1.0 ±0.16
100.0 ±0.8
23.0 ±1.0
1102.0
5
0.88 ±0.00027
3.11 ±0.0017
0.9 ±0.15
102.9 ±1.6
22.8 ±1.1
1061.2
1.4
1
0.88 ±0.00026
3.11 ±0.0008
0.9 ±0.10
100.2 ±0.9
24.1 ±0.9
1119.4
2
0.88 ±0.00027
3.11 ±0.0008
0.9 ±0.15
102.3 ±1.3
24.1 ±0.9
1152.8
3
0.88 ±0.00027
3.11 ±0.0008
0.8 ±0.14
101.8 ±2.3
24.0 ±0.7
1130.5
4
0.88 ±0.00026
3.11 ±0.0008
0.9 ±0.17
101.9 ±1.7
24.0 ±0.7
1234.4
5
0.88 ±0.00027
3.07 ±0.0007
0.9 ±0.15
100.8 ±0.6
24.2 ±0.7
1135.6
2.1
1
1.10 ±0.00037
2.97 ±0.0007
4.5 ±0.61
102.9 ±1.3
24.2 ±0.7
1154.4
2
1.10 ±0.00037
3.07 ±0.0009
4.6 ±0.41
100.4 ±1.2
25.0 ±0.9
1079.2
3
1.10 ±0.00037
3.07 ±0.0016
4.4 ±0.52
100.0 ±0.9
24.1 ±0.4
1142.6
4
1.10 ±0.00037
3.07 ±0.0012
4.9 ±0.53
100.3 ±0.4
24.4 ±0.6
1091.3
5
1.10 ±0.00039
3.07 ±0.0008
4.4 ±0.33
100.5 ±0.8
24.6 ±0.8
1086.5
2.2
1
1.10 ±0.00039
3.56 ±0.0012
4.4 ±0.47
100.6 ±0.9
24.8 ±0.6
555.2
2
1.10 ±0.0004
3.03 ±0.0008
4.7 ±0.49
100.2 ±0.5
23.8 ±0.9
1154.7
3
1.09 ±0.00038
3.05 ±0.0019
4.9 ±0.46
100.6 ±1.3
24.5 ±0.7
1280.9
4
1.09 ±0.00037
3.04 ±0.0018
5.0 ±0.49
100.2 ±0.5
24.1 ±0.7
988.2
5
1.09 ±0.0004
3.05 ±0.0007
4.8 ±0.64
100.6 ±0.7
24.6 ±0.6
1087.7
A-4
-------
Test ID
Cycle
Ave. MeBr
Flow Rate
Lpm (± SD)
Ave. Dilution N2
Flow Rate
Lpm (± SD)
Ave. Inlet
RH
% (± SD)
Ave. Inlet
Temp
°C (± SD)
Ave. Outlet
Temp
°C (± SD)
Cycle
Duration
min
2.3
1
1.09 ±0.00038
2.93 ±0.0017
5.1 ±0.47
101.1 ±0.9
27.1 ±1.3
1036.0
2
1.09 ±0.00039
2.96 ±0.0007
4.4 ±0.77
103.0 ±2.5
26.6 ±1.2
1085.0
3
1.11 ±0.0004
3.01 ±0.0008
4.3 ±0.32
99.9 ±0.9
26.5 ±0.9
1088.5
4
1.11 ±0.00043
2.97 ±0.0008
4.1 ±0.38
100.2 ±1.0
26.5 ±1.0
1137.8
5a
1.10 ±0.0004
2.45 ±0.5105
5.0 ±0.54
99.7 ±1.0
26.4 ±0.5
1048.5
3.1b
1a
1.11 ±0.0004
2.29 ±0.7647
0.0 ±0.18
100.1 ±1.0
27.0 ±1.5
1073.4
2 a
1.11 ±0.00041
2.59 ±0.6730
-0.5 ±0.17
100.5 ±0.7
26.3 ±1.4
1094.7
3
1.11 ±0.00041
2.93 ±0.2493
-0.2 ±0.17
100.7 ±0.4
27.1 ±1.4
1178.8
4a
1.11 ±0.00041
2.03 ±0.7412
2.4 ±0.13
100.1 ±0.6
26.5 ±1.2
1045.3
5a
1.09 ±0.00043
2.43 ±0.7352
-0.2 ±0.22
102.2 ±0.8
27.3 ±1.5
1135.4
3.2
1 a
1.09 ±0.0004
2.75 ±0.5718
0.5 ±0.20
100.4 ±0.9
26.2 ±1.8
1089.2
2
1.09 ±0.00053
3.00 ±0.0026
-0.4 ±0.10
100.8 ±0.8
24.6 ±0.7
1172.1
3
1.09 ±0.00048
3.00 ±0.0023
0.2 ±0.09
101.1 ±0.7
24.8 ±1.3
1207.6
4
1.10 ±0.0005
2.99 ±0.0025
1.7 ±0.14
102.1 ±1.3
25.2 ±1.5
1512.8
5
1.10 ±0.00052
2.90 ±0.4444
1.5 ±0.22
100.9 ±1.3
25.2 ±1.3
1031.5
3.3
1
1.10 ±0.00051
3.02 ±0.0025
-0.4 ±0.10
101.2 ±0.9
25.3 ±1.1
1196.8
2
1.10 ±0.00056
3.02 ±0.0027
-2.8 ±0.09
103.4 ±1.8
25.1 ±1.1
1087.7
3
1.09 ±0.00056
2.99 ±0.0026
4.6 ±0.19
102.3 ±1.1
26.0 ±1.4
1179.4
4
1.06 ±0.00048
2.94 ±0.0028
4.6 ±0.16
101.6 ±1.4
25.2 ±0.8
1079.1
5
1.07 ±0.00052
2.99 ±0.0067
2.5 ±0.09
102.1 ±0.7
26.3 ±1.1
1196.6
aThe dilution flow setting was reduced to conserve gas after the FID reading was less than 100 ppm.
Note: the MeBr scrubber system with potassium hydroxide (KOH) solution was replaced with an MeBr scrubbing system containing
carbon between Tests 2.2 and Test 2.3.
A-5
-------
Table A-3. RH only Test Conditions
Test ID
Phase
Challenge Gasa
Ave. Flow Rate
Lpm (±SD)
Ave. Inlet
RH
% (±SD)
Ave. Inlet
Temp
°C (±SD)
Ave. Outlet
Temp
°C (±SD)
Cycle
Duration
min
1b
Adsorption
0.47 ±0.025
75.8 ±0.4
22.3 ±0.1
24.8 ±0.3
33.2
Desorption
-
-
-
-
-
2
Adsorption
0.48 ±0.0001
75 ±0.3
22.6 ±0.2
24.9 ±0.5
33.1
Desorption
1.08 ±0.0004
4.1 ±0.4
101 ±0.5
28.1 ±0.6
1067.0
aUltra-high-purity nitrogen was used for the adsorption gas and ambient air for the desorption gas.
Resorption phase was not completed for Test 1.
A-6
-------
Appendix B: Carbon tube mass stability preliminary tests
B-1
-------
Mass Stability Observations
The following three reactor tube configurations were used:
1. Empty tube: The mass of an empty reactor tube with ends sealed with Parafilm was
monitored to assess the stability of the materials and balance at ambient temperature
and RH over a period of three days.
2. Two identical tubes, A and B, were empty and secured with a Keck® joint clamp. Tube A
was sealed with Parafilm while Tube B remained open on both ends. The masses of
both tubes were monitored over 20 minutes (measurements taken at 1-minute intervals)
each day, for a period of five days.
3. Two identical tubes, A and B, were assembled with 1 g of carbon, quartz wool, and a
Keck® joint clamp. This configuration was consistent with the test configuration. The
tube ends of tube A were plugged with chemical resistant stoppers while Tube B
remained open on both ends. The masses of both tubes were monitored at 20 min over
a duration of 20 minutes (measurements taken at 1-minute intervals) each day, over a
period of four days.
The agreement in the measurements was assessed as standard deviation. An ANOVA analysis
was performed to assess the statistical significance of the variability in the data.
Empty - Sealed Tube
Mass measurements were recorded every second for a 15-minute duration using a calibrated
balance. This procedure was performed for three consecutive days. The data showed excellent
agreement within each set of readings, suggesting the reactor tube was not a likely source of
measurement variability. Although the data were precise, there was a period of time lasting
approximately 10 minutes when the mass appeared to be stabilizing. As a result, the procedural
step was taken during the AC regeneration tests to allow the tube 20 minutes (double the
observed time) to stabilize prior to recording the mass reading.
A comparative analysis performed on the data over the three-day periods suggests a significant
change occurred (p-value =1.67516E-56). The change was likely the result of temperature and
RH fluctuations in the facility which were also likely present during AC regeneration testing. A
subsequent test was performed to determine if the fluctuations were caused or exacerbated by
sealing the tube ends. The SD in the data collected over the 3-day period was 0.002719. Figure
4-1 shows the mass reading for the reactor tube over the 15-minute observation period. Table
4-1 provides descriptive statistics for each of the data sets.
B-2
-------
111.9
111.89
111.88
111.87
111.86
¦Day 1
¦Day 2
¦Day 3
111.85
5 10 15
elapsed time (min)
20
Figure 1. Mass stability of an empty tube with sealed ends over 15 min interval
Table 1. Descriptive Statistics for Mass Stability Observation of an Empty Tube
A Time
(Days)
Min
Max
Average
SD
(each day)
SD
(3-day)
P-value
(a=0.05)
1
111.8644
111.8651
111.8649
0.0002566
0.002719
< 0.0001
2
111.8699
111.8703
111.8702
0.0001412
3
111.8709
111.8710
111.8710
4.4722E-05
Sealed vs. Unsealed Tube - Empty
As previously mentioned, this observation was conducted to determine the effect of sealing the
tube ends on the mass stability of the reactor tube. As before, the data for the sealed tube
showed excellent agreement within each set of measurements. The period of time required for
the mass to stabilize lasted for nearly 20 min at times (Day 4 and Day 5).
As before, comparative analysis performed on the data over the three-day periods suggests a
significant change occurred (p-value = 2.2552E-116). The change may be a result of
temperature and RH fluctuations in the facility which were also likely present during AC
regeneration testing. The SD in the data collected over the three-day period was 0.002220.
Figure 4-2 shows the mass reading for the reactor tube over the 15-minute observation period.
Table 4-2 provides descriptive statistics for each of the data sets.
B-3
-------
119.48
119.475
119.47
119.465
_ 119.46
aa
% 119.455 .
CD _
E 119.45
119.445
119.44
119.435
119.43
0
Figure 2. Mass readings empty, sealed tube over 20 min intervals for 5 days
Table 2. Descriptive Statistics for Mass Stability Observation of an Empty, Sealed Tube
Over 5 Days
A Time
(Days)
Min
Max
Average
SD
(each day)
SD
(5-day)
P-value
(a=0.05)
1
119.4586
119.4594
119.4591
0.0002598
0.002220
< 0.0001
2
119.4546
119.4547
119.4546
0.00002182
3
119.4538
119.4542
119.4539
0.00008536
4
119.4563
119.4564
119.4563
0.00002182
5
119.4525
119.4531
119.4528
0.0001989
The unsealed tube mass measurements showed improved correlation as demonstrated by the
significantly lower standard deviations. Additionally, the open tube ends appear to eliminate the
need to wait for the tube to stabilize; the readings were stable from the onset of the observation
period. Although there is excellent correlation in each set of data, there is considerably more
variation in the data over the entire observation period compared to the sealed tube data
(standard deviations are 0.01082 and 0.002220, respectively). ANOVA analysis indicates that
there is a significant change in the data during the observation period (p-value < 0.0001). These
findings show that while an unsealed tube has superior correlation in the data for individual days
and does not require time for stabilization, there was significantly higher variation in the
collective data for the five-day period. Again, this is likely due to temperature and RH changes.
However, a follow-up test was performed to determine if preparing the reactor tube in the same
manner as a regeneration test (i.e., with AC and packed with quartz wool) will affect these
10 15
6t (min)
20
25
¦Day 1
•Day 2
¦Day 3
¦Day 4
¦Day 5
B-4
-------
findings. Figure 4-3 shows the mass reading for the reactor tube over the 20-minute observation
period. Table 4-3 provides descriptive statistics for each of the data sets.
117.54
117.535
117.53
117.525
— 117.52
8 117.515
E 117.51
117.505
117.5
117.495
117.49
¦Day 1
¦Day 2
¦Day 3
¦Day 4
¦Day 5
10 15
6t (min)
20
25
Figure 3. Mass reading for empty, unsealed tube (Tube B) 20 min intervals for five days
Table 3. Descriptive statistics for mass stability measurements of an empty, unsealed
tube (Tube B) over five Days
A Time
(Days)
Min
Max
Average
SD
(each day)
SD
(5-day)
P-value
(a=0.05)
1
117.5269
117.5272
117.5271
0.0001044
0.01082
< 0.0001
2
117.5186
117.5187
117.5187
0.00002182
3
117.5136
117.514
117.5136
0.00009258
4
117.5132
117.5133
117.5132
0.00004976
5
117.4942
117.4944
117.4943
0.00006796
Sealed vs. Unsealed Tube with AC
As previously mentioned, this observation was conducted to determine if preparing a reactor
containing AC and quartz wool and AC (i.e., no air pocket) would yield results like those
discussed in Section 4.2.2. As before, the data for the sealed tube showed excellent agreement
within each set of measurements.
Compared with the sealed empty tube, variability in the data over the four-day evaluation was
comparable (SD values are 0.001017 and 0.002220, respectively).
As before, comparative analysis performed on the data over the five-day periods suggests a
significant change occurred (p-value < 0.0001). The change may be a result of temperature and
RH fluctuations in the facility, which were also likely present during AC regeneration testing. The
B-5
-------
SD in the data collected over the three-day period was 0.002220. Figure 4-4 shows the mass
reading for the reactor tube over the 15-minute observation period. Table 4-4 provides
descriptive statistics for each of the data sets.
96.3700
96.3650
96.3600
96.3550
— 96.3500
3B
8 96.3450
ro
E 96.3400
96.3350
96.3300
96.3250
96.3200
0 5 10 15 20 25
6t (min)
Figure 4. Mass readings for tube prepared with plugged ends (Tube A) over 20 min
intervals for four days
Table 4. Descriptive statistics for mass stability observation of sealed prepared tube
(Tube A) over four days
A Time
(Days)
Min
Max
Average
SD
(each day)
SD
(5-day)
P-value
(a=0.05)
1
96.3454
96.346
96.3457
0.00021657
0.001017
< 0.0001
2
96.3428
96.3429
96.3429
0.00004976
3
96.3444
96.3445
96.3444
0.00002182
4
96.3446
96.3447
96.3447
0.00005071
The unsealed tube mass measurements showed comparable correlation with the sealed
counterpart. Although there is excellent correlation in each set of data, there is slightly more
variation in the data over the four-day observation period compared to the sealed tube data
(standard deviations are 0.001017 and 0.002514, respectively). An ANOVA analysis indicates
there was a significant change in the data at a point during the four-day observation period (p-
value < 0.0001). Figure 4-5 shows the mass reading for an unsealed tube over the 20-minute
observation period, and Table 4-5 provides descriptive statistics for each of the data sets.
4/13/2017
4/14/2017
4/17/2017
4/18/2017
B-6
-------
96.0200
96.0100
96.0000
4/13/2017
8 95.9900
| 4/14/2017
95.9800 4/17/2017
4/18/2017
95.9700
95.9600
0 5 10 15 20 25
6t (min)
Figure 5. Mass readings for prepared tube with open ends (Tube B) over 20 min intervals
for four days
Table 5. Descriptive statistics for mass stability measurements of an empty, unsealed
tube (Tube B) over four days
A Time
(Days)
Min
Max
Average
SD
(each day)
SD
(5-day)
P-value
(a=0.05)
1
95.992
95.9921
95.9921
0.00004024
0.002514
< 0.0001
2
95.9861
95.9862
95.9862
0.00002182
3
95.9862
95.9863
95.9863
0.00004830
4
95.9865
95.9866
95.9866
0.00004364
The findings in this section were used to develop the procedures implemented for preparing and
weighing the reactor tubes during AC regeneration testing. While in this orientation, the sample
tube proved stable during the weighing process. Additionally, an acceptable level of variation
over the test cycle was observed. The tube ends were plugged during the weighing procedure
and for sample storage between regeneration cycles.
B-7
-------
Appendix C: Preliminary Activated Carbon Adsorption Tests
C-1
-------
Figure C-1: Adsorption capacity of activated carbon at 25 °C as a function of RH and
MeBr concentration
25 degrees C
40% RH
75% RH
.
l
18A
i
19A
1.3%
2.7%
Concentration MeBr in N,
5.3%
C-2
-------
Figure C-2: Adsorption capacity of activated carbon at 37 °C as a function of RH and
MeBr concentration
0.3
37 degrees C
0.2 -
40% RH
75% RH
0.1 -
T
T
1.3%
2.7%
MeBr concentration in N,
5.3%
C-3
-------
Figure C-3: Adsorption capacity of MeBr as a function of carbon moisture content
0.30
0.25
+-»
5 0.20
o
u
a, 0.15
o 0.10
0.05
0.00
«x
~ Test 7 Modified TIGG
¦ Test 3 Modified TIGG
Test 3 Jar 1
XTest 3, Jar 3
0.00 0.05 0.10 0.15 0.20 0.25
Sorbent capacity
Figure C-4: Carbon moisture content adjusted using different RH levels in purge air
X
cc
a>
t)J0
100.00
90.00
80.00
70.00
60.00
50.00
40.00
30.00
20.00
10.00
0.00 +
~ Seriesl
0.00% 10.00% 20.00% 30.00% 40.00% 50.00%
C-4
-------
vvEPA
United States
Environmental Protection
Agency
PRESORTED STANDARD
POSTAGE & FEES PAID
EPA
PERMIT NO. G-35
Office of Research and Development (8101R)
Washington, DC 20460
Official Business
Penalty for Private Use
$300
------- |