vvEPA
Protection EPA-600/R-03/101 a
Agency November 2003
Source Sampling Fine
Particulate Matter:
Stationary Source
Characterization
Testing of a Smelt Tank
at a Pulp and Paper
Facility: Volume 1,
Report
-------
-------
EPA-600/R-03/101 a
November 2003
Source Sampling Fine Particulate Matter:
Stationary Source Characterization Testing
of a Smelt Tank at a Pulp and Paper
Facility: Volume 1, Report
by
Joan T. Bursey and Dave-Paul Dayton
Eastern Research Group, Inc.
1600 Perimeter Park Drive
Morrisville, NC 27560
Contract No. 68-D7-0001
EPA Project Officer: N. Dean Smith
Air Pollution Prevention and Control Division
National Risk Management and Research Laboratory
Research Triangle Park, NC 27711
Prepared for:
U.S. Environmental Protection Agency
Office of Research and Development
Washington, DC 20460
-------
Abstract
Fine particulate matter (PM) of aerodynamic diameter 2.5 |im or less (PM2 5) has been
implicated in adverse health effects, and a National Ambient Air Quality Standard for PM2 5 has
been promulgated (July 1977) by the U.S. Environmental Protection Agency. A national
network of ambient monitoring stations has been established to assist States in determining
areas which do not meet the ambient standard for PM2 5. For such areas, it is important to
determine the major sources of the PM2 5 so States can devise and institute a control strategy to
attain the ambient concentrations set by the standard.
One of the tools often used by States in apportioning ambient PM2 5 to the sources is a
source-receptor model. Such a model requires a knowledge of the PM2 5 chemical composition
emitted from each of the major sources contributing to the ambient PM2 5 as well as the
chemical composition of the PM2 5 collected at the receptor (ambient monitoring) sites. This
report provides such a profile for a smelt tank at a pulp and paper facility. Along with the
PM2 5 emission profile, data are also provided for gas-phase emissions of several organic
compounds. Data are provided in a format suitable to be included in the EPA source profile
database, SPECIATE.
li
-------
Foreword
The U.S. Environmental Protection Agency (EPA) is charged by Congress with protecting the
Nation's land, air, and water resources. Under a mandate of national environmental laws, the
Agency strives to formulate and implement actions leading to a compatible balance between
human activities and the ability of natural systems to support and nurture life. To meet this
mandate, EPA's research program is providing data and technical support for solving
environmental problems today and building a science knowledge base necessary to manage our
ecological resources wisely, understand how pollutants affect our health, and prevent or reduce
environmental risks in the future.
The National Risk Management Research Laboratory (NRMRL) is the Agency's center for
investigation of technological and management approaches for preventing and reducing risks from
pollution that threaten human health and the environment. The focus of the Laboratory's research
program is on methods and their cost-effectiveness for prevention and control of pollution to air,
land, water, and subsurface resources; protection of water quality in public water systems;
remediation of contaminated sites, sediments and ground water; prevention and control of indoor
air pollution; and restoration of ecosystems. NRMRL collaborates with both public and private
sector partners to foster technologies that reduce the cost of compliance and to anticipate
emerging problems. NRMRL's research provides solutions to environmental problems by:
developing and promoting technologies that protect and improve the environment; advancing
scientific and engineering information to support regulatory and policy decisions; and providing
the technical support and information transfer to ensure implementation of environmental
regulations and strategies at the national, state, and community levels.
This publication has been produced as part of the Laboratory's strategic long-term research
plan. It is published and made available by EPA's Office of Research and Development to assist
the user community and to link researchers with their clients.
Lawrence W. Reiter, Acting Director.
National Risk Management Research Laboratory
111
-------
EPA Review Notice
This report has been peer and administratively reviewed by the U.S. Environmental Protection
Agency and approved for publication. Mention of trade names or commercial products does not
constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Information Service,
Springfield, Virginia 22161.
iv
-------
Table of Contents
Volume 1, Report
Section Page
Abstract ii
List of Tables viii
List of Figures x
Nomenclature xi
Acknowledgments xii
Introduction 1
Characterization of a Smelt Tank at a Pulp and Paper Facility 2
Organization of Report 3
Conclusions 5
Methods and Materials 7
Description of Test Equipment 8
Dilution Sampling System 9
Dilution Sampling System Control Instrumentation 12
Process Description/Site Operation 14
Pre-Test Survey 18
Experimental Procedures 21
Preparation for Test Setup 21
Traverse Point Determination Using EPA Method 1 22
Volumetric Flow Rate Determination Using EPA Method 2 23
Pitot Tube Calibration 23
Calculation of Average Flue Gas Velocity 23
Nozzle Size Determination 24
Measurement of 02 and C02 Concentrations 25
Stationary Gas Distribution (as Percent Volume) 25
Dry Molecular Weight of Flue Gas 25
Wet Molecular Weight of Flue Gas 25
Determination of Average Moisture Using EPA Method 4 26
Volume of Dry Gas Sampled at Standard Conditions (dscf) 26
Volume of Water Vapor at Standard Conditions (dscf) 27
v
-------
Table of Contents (continued)
Section Page
Calculation of Moisture/Water Content (as Percent Volume) 27
Calculation of Dry Mole Fraction of Flue Gas 28
Setup of the Dilution Sampling System 28
Pre-Test Leak Check 31
Orifice Flow Check 33
Determination of Test Duration 33
Canister/Veriflow Blanks 33
Determination of Flow Rates 34
Sample Collection Arrays 34
Use of the ELPI Particle Size Distribution Analyzer 37
Operation of the Dilution Sampling System with Sample Collection Arrays 40
Dilution System Sample Collection Arrays Recovery 49
Laboratory Experimental Methodology 51
PM2 5 Mass 51
Elemental Analysis 51
Water-Soluble Inorganic Ions 51
Elemental Carbon/Organic Carbon 52
Organic Compounds 52
Carbonyl Compounds 53
Canister Analysis: Air Toxics and Speciated Nonmethane Organic
Compounds 55
Particle Size Distribution Data 60
Results and Discussion 61
Calculated Emission Factors for PM Mass, Carbonyl Compounds, and
Nonmethane Organic Compounds 62
Gas-Phase Carbonyl Compounds 63
Gas-Phase Air Toxics Whole Air Samples 66
Gas-Phase Speciated Nonmethane Organic Compounds 67
PM2 5 Elemental Carbon/Organic Carbon, Inorganic Ions, and Element Profile 72
Semivolatile Organic Compounds 73
Particle Size Distribution Data 74
Quality Assurance/Quality Control 79
Field Sampling 79
vi
-------
Table of Contents (continued)
Section Page
Carbonyl Compound Analysis 81
Concurrent Air Toxics/Speciated Nonmethane Organic Compound Analysis 83
PM Mass Measurements, Elemental Analysis, Water-Soluble Ion Analysis, and
GC/MS Analysis 85
References 89
Volume 2, Appendices
A Table of Unit Conversions A-l
B Smelt Tank Vent, Chain of Custody Documentation B-l
C Example Calculations: NMOC, Carbonyl, and PM25 Mass Emission Factors .... C-l
D Data Tables for Individual PM25 Mass Measurements D-l
E Data Tables for Individual Carbonyl Samples E-l
F Data Tables for Individual NMOC Samples F-l
G Data Tables for Individual Air Toxics Samples G-l
H Data Tables for Individual PM2 5 Elemental Samples H-l
I Data Tables for Individual PM2 5 EC/OC Samples 1-1
J Data Tables for Individual PM2 5 Inorganic Ion Samples J-l
K Data Tables for Individual Semivolatile Organic Compounds K-l
L List of ERG SOPs and EPA MOPs by Title I.-l
vii
-------
List of Tables
Table Page
3-1 Sampling Medium Used for Collection of Samples, Analysis Performed,
Analytical Method, and Responsible Laboratory 7
3-2 Fuel Use During the Smelt Tank Vent Test Period 14
3-3 Black Liquor, Black Liquor Solids, and #2 Oil Process Data for Testing Days 15
3-4 Analysis of Black Liquor 15
4-1 EPA Method 1-Traverse Point Locations for the Circular Smelt Tank Vent 22
4-2 Average Flue Gas Velocity for Each Traverse Point 24
4-3 Moisture Recovery for Method 4 (Measured on 11/01/01) 26
4-4 Blank Values for Veriflows and Canisters 33
4-5 Denuder Sampling Scheme 38
4-6 Run Time Summary Information, Test Run 1 (12/14/01) 40
4-7 Run Time Summary Information, Test Run 2 (12/15/01) 42
4-8 Run Time Summary Information, Test Run 3 (12/16/01) 43
4-9 Carbonyl Compounds Analyzed by High Performance Liquid Chromatography:
Method Detection Limits 54
4-10 Detection Limits for Air Toxics Compounds (Analytical Method TO-15) 56
4-11 Detection Limits for Speciated Nonmethane Organic Compounds 58
5-1 Fine Particle, Carbonyl, and Nonmethane Organic Compound Emission Factors
from a Smelt Tank Vent at a Pulp and Paper Facility 63
5-2 Carbonyl Compounds, Smelt Tank Vent: Carbonyl Compounds Collected in
Dilution Air Subtracted from Carbonyl Compounds Collected in
Residence Chamber, Mean Fraction, Mean and Uncertainty 64
5-3 Summarized Analytical Results for Air Toxics Compounds Observed at the
Smelt Tank Vent on Each of the Three Test Days 66
5-4 Speciated and (Speciated Plus Unspeciated) NMOC Data for all Three Test Days .... 68
5-5 Fine Particle Chemical Composition of Emissions from a Smelt Tank Vent
at a Pulp and Paper Facility 72
5-6 Trace Semivolatile Organic Compounds Found Above Background Levels in
the Particle-Phase Emissions from the Smelt Tank Vent 74
viii
-------
List of Tables (continued)
Table Page
5-7 Particle Size Distribution Data 75
6-1 Field Sampling Equipment Quality Control Measures 80
6-2 Carbonyl Analysis: Quality Control Criteria 81
6-3 Quality Control Procedures for the Concurrent Analysis for Air Toxics and
SNMOC 84
6-4 PM Mass Measurements: Quality Control Criteria 86
6-5 Elemental Analysis: Quality Control Criteria 86
6-6 Water-Soluble Ion Analysis: Quality Control Criteria 87
6-7 Quality Control Procedures for Gas Chromatography-Mass Spectrometry
Analysis of Semivolatile Organic Compounds 88
IX
-------
List of Figures
Figure Page
3-1 Diagram of the Dilution Sampler and Dilution Air Conditioning System 10
3-2 Instrumentation for Control and Analysis of the Dilution Sampler 13
3-3 Schematic Diagram of the Sampling Location (Top View) 17
3-4 Sampling Port at the Smelt Tank Vent Location 18
4-1 Schematic Diagram of the Sampling Location (Top View) 28
4-2 Dilution Sampling System Sampling Probe Installed in 6 in. id. Flanged Port 29
4-3 Dilution System Control Module Positioned at the Sampling Location 30
4-4 Dekati ELPI Positioned at the Sampling Location 30
4-5 Dilution System with Sample Collection Arrays and Instruments Attached 31
4-6 Recovery Area for Dilution Sampling System Sample Collection Arrays 32
4-7 Sample Collection Arrays Used for Testing at the Smelt Tank Vent 35
4-8 Blower Flow. Day 1 (12/14/01) 45
4-9 Dilution Flow. Day 1 (12/14/01) 45
4-10 Venturi Flow, Day 1 (12/14/01) 46
4-11 Blower Flow, Day 2(12/15/01) 46
4-12 Dilution Flow. Day 2(12/15/01) 47
4-13 Venturi Flow. Day 2(12/15/01) 47
4-14 Blower Flow. Day 3 (2/16/01) 48
4-15 Dilution Flow. Day 3(12/16/01) 48
4-16 Venturi Flow, Day 3 (12/16/01) 49
5-1 Plots of Particle Counts vs. Size, Particle Mass vs. Size, and Particle Mass per
Stage for Test Day 1 (12/14/01) 76
5-2 Plots of Particle Counts vs. Size, Particle Mass vs. Size, and Particle Mass per
Stage for Test Day 2 (12/15/01) 77
5-3 Plots of Particle Counts vs. Size, Particle Mass vs. Size, and Particle Mass per
Stage for Test Day 3 (12/16/01) 78
x
-------
Term
Nomenclature
Definition
CMB
chemical mass balance
DNPH
2,4-dinitrophenylhydrazine
EC
elemental carbon
ELPI
Electrical Low Pressure Impactor
EPA
U.S. Environmental Protection Agency
ERG
Eastern Research Group
FID
flame ionization detector
GC/MS
gas chromatography/mass spectrometry
GRAV
gravimetric analytical method
HEPA
high efficiency particulate arresting
HPLC
high performance liquid chromatography
IC
ion chromatography
MDLs
method detection limits
MSD
mass selective detector
NAAQS
National Ambient Air Quality Standards
NaOH
sodium hydroxide
Na2C03
sodium carbonate
Na2S
sodium sulfide
Na2S04
sodium sulfate
nh3
ammonia
NMOCs
nonmethane organic compounds
NOx
nitrogen oxides
oc
organic carbon
PM
particulate matter
PM10
PM of aerodynamic diameter 10 |_im or less
pm25
PM of aerodynamic diameter 2.5 |_im or less
PUF
polyurethane foam
RH
relative humidity
SIPs
State Implementation Plans
SOx
sulfur oxides
SNMOCs
speciated nonmethane organic compounds
TOE
thermal-optical evolution
TRS
total reduced sulfur
TSP
total suspended particulate
VOCs
volatile organic compounds
XRF
X-ray fluorescence
XI
-------
Acknowledgments
Dave-Paul Dayton, Mark Owens, and Robert Martz of Eastern Research Group, Inc.
(ERG) were responsible for conducting sampling at the test site and for preparing collected
samples for transport to the analytical laboratories. Amy Frame, Donna Tedder, and Randy
Bower of ERG were responsible for the carbonyl and volatile organic compound analyses. Joan
Bursey and Raymond Merrill of ERG provided calculations, data analysis, and sections of the
report pertaining to the ERG work on this project. Wendy Morgan of ERG prepared the
typewritten manuscript.
Michael Hays, Kara Linna, and Jimmy Pau of the EPA, NRMRL-RTP, were responsible
for the analysis of organic compounds, elements, and ionic species. Yuanji Dong, John Lee,
David Proffitt, and Tomasz Balicki of ARCADIS, Geraghty & Miller, Inc., provided technical
support in preparing the dilution sampling system and sampling substrates, in performing the
elemental/organic carbon analyses, and in extracting organic compounds from the various
sampling substrates. N. Dean Smith was the EPA Project Officer responsible for overall project
performance.
Xll
-------
Introduction
In July 1997, the U.S. Environmental Protection Agency (EPA) promulgated new National
Ambient Air Quality Standards (NAAQS) for ambient particulate matter (PM) of aerodynamic
diameter 2.5 |im or less (PM2 5) and revised the existing standard for ambient particles of
aerodynamic diameter 10 |im or less (PM10). In 1999, a national network of ambient monitoring
stations was started under the overall guidance of the EPA's Office of Air Quality Planning and
Standards to assist the States in determining regulatory non-attainment areas and to develop
State Implementation Plans (SIPs) to bring those areas into compliance with the PM2 5 and
revised PM10 regulations. One component of the monitoring network is a number of regional air
sheds in which intensive coordinated PM-related research will be carried out to better
understand the linkages between source emissions and actual human dosages of fine PM.
The mission of the Emissions Characterization and Prevention Branch of the National Risk
Management Research Laboratory's Air Pollution Prevention and Control Division is to
characterize source emissions and to develop and evaluate ways to prevent those emissions.
Source characterization as defined here includes the measurement of PM mass emission
factors, source PM profiles (PM chemical composition and associated chemical mass emission
factors), and emission factors for ambient aerosol precursors such as sulfur oxides (SOx),
nitrogen oxides (NOx), and ammonia (NH3).
PM mass emission factors are used in emission inventories and as inputs to atmospheric
dispersion models that yield estimates of ambient PM concentrations from considerations of
atmospheric transport and transformation of emitted particles. Emissions composition data are
used in receptor models to enable apportionment of ambient concentrations of PM to the
various sources that emitted the particles. EPA has interest and investments in source
apportionment, ambient monitoring, and regulatory matters related to fine PM. For example,
States rely on source-receptor and dispersion models to target major sources of PM25 and to
devise cost-effective strategies for achieving compliance with the standard. EPA has a
longstanding effort to produce the models for use by the States and EPA. An example of a
source-receptor model is the Chemical Mass Balance (CMB) model, which requires chemical
composition data from both ambient and source samples as input. The field test reported here
focused on the collection of fine particles emitted by a smelt tank at a pulp and paper facility.
1
-------
Data were collected to evaluate new measurement techniques and to update and improve source
emission profiles and emission factors for PM2 5.
Characterization of a Smelt Tank at a Pulp and Paper Facility
During the last quarter of 2001 a sampling campaign was conducted at a large pulp and
paper mill to measure emissions from three of the mill's largest process sources of atmospheric
emissions. These sources were:
1. A recovery boiler firing concentrated black liquor;
2. An auxiliary boiler co-firing a mixture of wood bark (hogged wood waste) and
bituminous coal; and
3. A vent from the smelt tank.
This test report presents results from the emissions testing at the smelt tank vent. The
primary aim of these tests is to determine the amount and nature at the fine PM emissions from
the sources. Fine PM is here defined as PM having an aerodynamic diameter less than 2.5 |im.
Much of the PM sampling conducted at stationary sources in the past focused on total
suspended particulate (TSP) or PM10, and PM2 5 source emissions data have been limited.
Emissions data, in general, are very limited for pulp mill for smelt tank vents.
Most of the previous work to measure PM10 or total PM emissions followed the protocols in
EPA Method 51 and/or EPA Method 2022. A number of potential biases have been identified
with the use of these methods including both a negative mass bias when filterable PM is
collected in a hot exhaust stream without first cooling and diluting the exhaust and a positive
mass artifact of condensible PM when the hot exhaust is quenched by passing it through a
series of cold impinger solutions. To minimize these sampling artifacts, the present test
campaign employed a state-of-the-art dilution sampling system designed to dilute and cool a
hot exhaust gas to near ambient conditions prior to collection of the PM. Also, sufficient time
was provided after cooling and dilution of the sampled gas prior to collection of the PM to
enable semivolatile organic compounds to distribute between the gas and particle phases as
they would in the ambient air downstream of the stack. Sampling in this way should yield more
accurate, artifact-free PM mass emission factors and PM samples whose composition is the
same as that in the ambient air some distance from the source.
In mills utilizing the Kraft pulping process, the first step involves digestion of the wood
chips in an aqueous solution of sodium sulfide and sodium hydroxide at elevated temperature
and pressure. This process extracts the cellulose from the wood by dissolution of the lignin that
binds the cellulose fibers together. Spent digestion liquor ("black liquor") is combined with
2
-------
water used to wash the resulting pulp and undergoes a sequence of evaporation steps to yield a
concentrated black liquor with a solids content of 50% to 65%. Once the black liquor is
concentrated to this level, it is supplied as fuel to the recovery boiler used to generate heat for
process steam in the plant. Combustion of the black liquor results in a molten smelt, which is
comprised mostly of inorganic chemicals, at the bottom of the recovery furnace.
The smelt is collected and dissolved in water to form "green liquor," which is subsequently
transferred to a tank where quicklime (calcium oxide) is added to regenerate the digestion
reactants, sodium sulfide and sodium hydroxide. The recovered digestion solution, or "white
liquor," is then recycled to the initial digestion process. A calcium hydroxide sludge
precipitates from the smelt tank and is returned to a lime kiln to regenerate quicklime.
Headspace gases in the smelt tank at this mill were exhausted to the atmosphere from two vents
using powered ventilation fans. Total reduced sulfur gases and PM emissions in the vented
exhaust were controlled by a dedicated alkaline scrubber system.
Although fine PM was the focus of this particular test campaign, gas-phase organic
emissions were also collected concurrently and analyzed. Reduced sulfur gases such as
hydrogen sulfide, methyl mercaptan, and dimethyl sulfide, which are primarily responsible for
the characteristic odor associated with Kraft mills, were not sampled in these tests. This report
presents the results from three replicate test runs on the smelt tank vent that were conducted on
three consecutive days in December 2001. The report describes the nature of the source, the
method of sampling, analysis methods used to determine the composition of the PM and gas-
phase organic emissions, and the analysis results—both in the form of mass emission factors
(mass of emitted species per mass of fuel consumed) and as mass fraction compositions. Since
the amount of recovery boiler bottoms (smelt) fed to the tank were not measured directly, mass
emission factors reported here are related to the black liquor fuel flow rate to the corresponding
recovery boiler. Results presented as mass emission factors are useful for compiling emission
inventories. The composition of PM and gas-phase organic emissions expressed as mass
fractions can be used as source profiles for input to source-receptor models. Such models are
used to apportion ambient PM and gases to the various contributing sources.
Organization of the Report
This report is organized into five additional sections plus references and appendices.
Section two provides the conclusions derived from the study results, and section three describes
the process operation and the test site. Section four outlines the experimental procedures used
in the research, and section five presents and discusses the study results. Section six presents
3
-------
the quality control/quality assurance procedures used in the research to ensure generation of
high quality data.
4
-------
Conclusions
Results of the three replicate test runs at the smelt tank vent are summarized below.
PM2 5 mass emission factors ranged from 4.28 to 8.08 mg/kg fuel and averaged 6.16 mg/kg
fuel over the three test days. Therefore, the smelt tank vent was the lowest emitter of fine PM
of the three sources tested at the facility(23.3 mg/kg fuel for the recovery boiler; 50.0 mg/kg
fuel for the hogged waste boiler). However, as discussed in the Results an Discussion section of
this report, the high moisture content of the smelt tank vent gas coupled with a high relative
humidity of the ambient air used to dilute the sampled vent gas on Test Day 1 resulted in some
water condensation in the sampler for at least that test day. This event likely resulted in a
negative bias in the particle mass emission factors and possibly for emissions of water-soluble
species, especially for the first test day. The PM2 5 from the smelt tank vent had a mean
aerodynamic diameter of 0.47 |im, and particle size was unimodally distributed about the mean.
Approximately 51% of the fine PM mass from the smelt tank vent was identified and
quantified. Sodium and potassium sulfates, thiosulfates, and chlorides comprised the bulk of
the PM (46.58% of the 51% characterized). The smelt tank vent was the only one of the three
tested sources for which thiosulfate was found in quantifiable amounts, and the amount of
thiosulfate was nearly the same as the sulfate in the PM. Thiosulfate ion is formed by oxidation
of sulfide ion and could be formed during the evaporative concentration of the black liquor and
possibly in the smelt tank itself where air contacts the sulfide-containing liquids. Because
thiosulfate was found in significant quantities in the fine PM from only the smelt tank vent, it
may be considered a good PM marker species for this source.
The carbon content of the fine PM was only 4.0% on average, and nearly all of the carbon
was organic rather than elemental. However, the fact that less than 4.0% of the fine PM was
comprised of organic carbon coupled with the relatively low overall fine PM mass emitted
from the vent made accurate quantitation of any individual organic compounds in the PM
difficult. As a result, a number of specific organic compounds were identified and are reported
but were not quantified.
Of the three sources tested at the mill, the smelt tank emissions contained the greatest
concentration of volatile organic compounds (VOCs). Total VOC emissions averaged 136.5
|ig/m3 and ranged from 110.0 to 169.3 |ig/m3. Total VOCs were comprised of a large variety of
5
-------
low molecular weight alkanes, alkenes, and aromatic organic compounds with no individual
species predominating. Ambient air sampled at the site on the first day of the smelt tank test
campaign was found to contain 249.8 |ig/m3 of VOCs with //-hexane, methylcyclopentane, and
a-pinene present in the largest amounts. Carbonyl analyses showed high emission levels of
acetone, especially on Test Day 1, which was not observed in the background sample, and
unusually high levels of acetaldehyde; formaldehyde levels were at least a factor of 20 lower
than the acetaldehyde levels. Anomalously high levels of methylene chloride, acetone, and n-
hexane on Test Day 1 were attributed to contamination arising from the denuder solvent; these
compounds were eliminated for Test Day 1 in subsequent calculations involving air toxics,
nonmethane organic compounds (NMOCs), and carbonyl compounds.
6
-------
Methods and Materials
A field test was conducted (December 14-16, 2001) on a smelt tank vent at a pulp and paper
facility to obtain source emissions measurements of high and known quality. The objectives of
the testing activities were to evaluate the sampling equipment and to characterize the fine
particulate and volatile organic emissions from a smelt tank. To simulate the behavior of fine
particles as they enter the ambient atmosphere from an emissions source, dilution sampling was
performed to cool, dilute, and collect gaseous and fine particulate samples from the smelt tank
exhaust. Gaseous and fine particulate samples were collected and chemically characterized.
ERG coordinated all field test activities; laboratory testing activities were divided between
EPA and ERG according to the breakdown shown in Table 3-1.
Table 3-1. Sampling Medium Used for Collection of Samples, Analysis
Performed, Analytical Method, and Responsible Laboratory
Sampling Medium Analysis Method Laboratory
Teflon Filter
PM2 5 Mass
Gravimetric (GRAV)
EPA
Teflon Filter
Elemental Analysis
X-Ray Fluorescence (XRF)
EPA
Teflon Filter
Inorganic Ions
Ion Chromatography (IC)
EPA
Quartz Filter
Elemental Carbon/Organic
Carbon
Thermal-Optical Evolution
(TOE)
EPA
Quartz Filter
XAD-4 Denuder
PUF
Semivolatile Organic Species
Gas Chromatography/ Mass
Spectrometry (GC/MS)
EPA
DNPH-Impregnated
Silica Gel Tubes
Carbonyl Compounds
High Performance Liquid
Chromatography (HPLC)
ERG
SUMMA Canisters
Air Toxics
Speciated NMOCs
Method TO-15 (GC/MS)
ERG Concurrent Analysis
ERG
Particle Size Analyzer
Particle sizes
Electrical Low Pressure
Impactor (ELPI)
ERG
ERG performed source sampling to collect artifact-free, size-resolved particulate matter in a
quantity and form sufficient to identify and quantify trace elements and organic compounds and
to distinguish gas-phase and particle-phase organic compounds. Total PM mass in the diluted
7
-------
and cooled emissions gas was size resolved at the PM10 and PM2 5 cut points with the PM2 5
fraction further continuously resolved down to 30 nm diameter using an Electrical Low
Pressure Impactor (ELPI). Fine particle emission profiles can be used in molecular marker-
based source apportionment models, which have been shown to be powerful tools to study the
source contributions to atmospheric fine particulate matter.
To assist in the characterization of the smelt tank stationary source emissions and to obtain
chemical composition data representative of particle emissions after cooling and mixing with
the atmosphere, ERG performed the following activities at the test site:
• Performed preliminary measurements using EPA Methods 1 through 43"5 to evaluate
source operating conditions and parameters;
Installed the precleaned dilution sampling system, sample collection trains, and ancillary
equipment at the field site without introduction of contaminants;
Calibrated flow meters before and after sampling, monitoring, and adjusting gas flows
(as necessary) throughout the tests;
Acquired process data for the test periods, including temperatures, pressures, flows, fuel
consumption, etc.;
• Determined the type of combustion fuel and the rate of consumption during the source
testing;
Collected three sets of stationary source samples as prescribed in the Site-Specific Test
Plan, including one set of field blanks; and
• Recovered the dilution sampling unit and sample collection arrays for analysis for
specific parameters and return of the dilution sampling unit to EPA.
ERG transported the dilution sampling system to the test site to collect integrated samples,
performed whole air analysis of volatile organic compounds collected in SUMMA polished
stainless steel canisters and gas-phase carbonyl compounds collected on silica gel cartridges
impregnated with 2,4-dinitrophenylhydrazine (DNPH), and evaluated particle size distribution
data. EPA was responsible for pre-test cleaning of the dilution sampling system, for analysis of
semivolatile organic compounds collected by XAD-4 denuders and polyurethane foam (PUF)
sampling modules resulting from the test efforts, and for characterization of the particle-phase
emissions and mass loading on quartz and Teflon filters.
Description of Test Equipment
The test equipment consisted of a dilution sampling system and its instrumentation.
8
-------
Dilution Sampling System
The dilution sampling system used in the source test was based on an original design by
L.M. Hildemann6, modified to incorporate more secure closure fittings and electronic controls.
Automatic flow control and data acquisition capabilities were added to the dilution sampler to
improve the ease of operation of the unit. A touch screen interface connected to the main
controller was used to monitor current conditions and allow set points to be entered into the
system readily. A laptop computer was used for continuous monitoring of operating parameters
and logging of the sampler operation.
The dilution sampling system dilutes hot exhaust emissions with clean air to simulate
atmospheric mixing and particle formation. Control of residence time, temperature, and
pressure allows condensible organic compounds to adsorb onto fine particles as they might in
ambient air. The sampler is also designed and fabricated to minimize any contamination of
samples, especially organic compound contamination, and to minimize particle losses to the
sampler walls.
Figure 3-1 shows a schematic diagram of the dilution sampling system and dilution air
cleaning and conditioning system. As shown, the dilution air cleaning system provides high
efficiency particulate arresting (HEPA) and activated carbon-filtered air. Acid gases (if present)
will not be removed completely by the dilution air cleaning system, but the presence of acid
gases can be monitored in the dilution tunnel immediately downstream of the dilution air inlet.
The dilution air cleaning system can be modified to add a heater, cooler, and dehumidifier as
needed. Cleaned dilution air enters the main body of the sampling system downstream of the
dilution air orifice meter.
The key zones of the dilution sampling system and their function are described below.
Sample Inlet Zone-
Stack Emissions Inlet: designed to allow stationary source exhaust gas to be sampled
through an inlet cyclone separator to remove particles with nominal aerodynamic
diameters greater than 10 |im. The PM10 cyclone prevents large particles from entering
the sampler to plug or damage the equipment. Three ports are dedicated to sampling the
dilution air before it mixes with the source gas.
Heated Inlet Line: 3/4 in. heated stainless steel sampling probe draws source gas
through a venturi meter into the main body of the sampler. Sample flow rate can be
adjusted from 15 to 50 L/min (typically 30 L/min).
9
-------
Turbulent Mixing Chamber
Re =10,000
RESIDENCE
TIME
CHAMBER
ACCESS
PORTS \
O LU
il
STACK
EMISSIONS
10 jjm
CYCLONE
2.1 jjm
CYCLONES
FILTER
VENTURI
HEATED INLET LINE
STACK
EMISSIONS INLET
SAMPLE
ARRAYS
D1—C]
D2—[]
D3—C]
Blower
SAMPLE
ARRAYS
ACTIVATED
CARBON
BED
HEPA
FILTER
COOLING
UNIT
BLOWER
DILUTION
AIR
DILUTION AIR
INLET
VACUUM
PUMPS
Figure 3-1. Diagram of the Dilution Sampler and Dilution Air Conditioning System.
-------
Venturi Meter-
Constructed of low carbon, very highly corrosion-resistant stainless steel; equipped
for temperature and pressure measurement. Wrapped with heating coils and
insulated to maintain the same isothermal temperature as the inlet cyclone and inlet
line.
Turbulent Mixing Chamber-
The mixing chamber incorporates an entrance zone, U-bend, and exit zone. The
inside diameter is 6 in., yielding a Reynolds number of approximately 10,000 at a
flow rate of 1000 L/min. Dilution air enters the mixing chamber in a direction
parallel to the flow of source gas. Hot source emission gas enters the chamber
perpendicular to the dilution air flow, 4.5 in. downstream of the dilution air inlet.
The combined gas flow travels 38 in. before entering the U-bend. After the
residence chamber transfer line, the mixing chamber continues for 18 in., then
expands to an in-line, high-volume sampler filter holder. Collected particulate
material has not experienced time to equilibrate with the gas phase in the diluted
condition. Sampling and instrumentation ports are installed on the turbulent mixing
chamber at various locations, as shown in Figure 3-1.
Residence Time Chamber-
The inlet line to the residence time chamber expands from a 2 in. line (sized to
provide a quasi-isokinetic transfer of sample gas from the turbulent mixing chamber
to the residence time chamber at a flow rate of approximately 100 L/min) within the
mixing chamber to a 7-in. line at the wall of the residence chamber. The flow rate is
controlled by the total sample withdrawal from the bottom of the residence time
chamber and provides a 60-sec residence time in the chamber. Twelve ports are
installed at the base of the residence time chamber: nine ports for sample
withdrawal, three ports for instrumentation.
Sample Collection Zone-
Samples collected from the sampling ports at the base of the residence time
chamber have experienced adequate residence time for the semivolatile organic
compounds to re-partition between the gas phase and the particle phase.
Since it is very difficult to maintain both isokinetic sampling and a fixed cyclone size
cut during most stack sampling operations, the inlet cyclone may be operated to provide a
rough PM10 cut while maintaining near-isokinetic sampling. The rough inlet size cut has
minimal impact on sampling operations since the dilution sampling system is used mainly
to collect fine particulate matter from combustion sources, and the critical fine particle size
11
-------
cut is made at the end of the residence time chamber. Typically, the calculated total time
the sample spends in the dilution sampling system ranges from 58-75 s: 2-3 s for the
turbulent mixing chamber and 56-72 s for the residence chamber.
Dilution Sampling System Control Instrumentation
Instrumentation for control and analysis of the dilution sampling system is shown in
Figure 3-2. Differential pressure measurements made across the venturi and orifice meters
are used to determine the dilution air flow rate, the sample gas flow rate, and the exhaust
gas flow rate. Since flow equations used for determination of the flow across venturi and
orifice meters correct for flowing temperature and pressure, the flowing temperature and
pressure of the venturi and orifice meters must be recorded during sampling operations.
Thermocouples for monitoring temperature are placed at each flow meter as well as at the
inlet PM10 cyclone, at various points on the sample inlet line, at the inlet to the mixing
chamber U-bend, and at the outlet of the residence time chamber. An electronic relative
humidity probe is used to determine the relative humidity of the sample gas. The dilution
sampling system is equipped with automated data logging capabilities to better monitor
source gas testing operations and to minimize manpower requirements during sampling
operations. Dilution sampling system flows and temperatures are monitored and controlled
automatically at setpoints established by the operator using a QSI Corporation QTERM-
K56 electronic touchscreen interface. The dilution sampling system was operated by three
testing staff members during the test at the smelt tank at the pulp and paper facility.
In operation, the source sample flow, the dilution air flow, and the total air flow (not
including the sample collection arrays) were each measured by separate flow meters and
pressure transducers. A venturi meter measured the source sample flow and orifices were
used for the dilution and total flows. A ring compressor was used to push the dilution air
through a HEPA filter, a carbon adsorber, and a final filter into the turbulent mixing
chamber. The compressor motor was modulated by a variable frequency drive to match the
desired dilution flow based on a set point entry. A separate blower (connected to a speed
controller adjusted to achieve the desired sample flow based on a set point entry) at the end
of the dilution sampling system pulled the source sample flow through the venturi. Flow
through this blower consisted of the dilution air flow plus the source sample flow, not
including the flow exiting through the sample collection arrays.
The main controller modulated the power used to heat the sample probe (64 in. long,
two heated zones). The controller switched solid state relays on and off as needed to
maintain the probe temperature entered by the operator.
12
-------
Residence
Time
Chamber
PM 10
Cyclone
PT-104
PT-102
HE PA Filter
Carbon Bed
Dilution Air
Blower
Exhaust
Blower
TE-108
PT-101
TE-104
TE-105
PT-103
TE-106
TE-107
TE-101
TE-102
TE-103
Ambient
TE-109
RH-1
Key:
TE = Temperature Indicator
PT = Pressure Indicator
RH = Relative Humidity Indicator
Figure 3-2. Instrumentation for Control and Analysis of the Dilution Sampler.
-------
Virtually any ambient sampling equipment—including filters of various types (quartz,
Teflon, Nylon), denuders, PUF modules, DNPH-impregnated silica gel sampling
cartridges, SUMMA polished canisters, cyclones, particle size distribution measurement
instrumentation—can be employed with the dilution sampling system.
Process Description/Site Operation
The first step in the Kraft process involves digestion of the wood chips in an aqueous
solution of sodium sulfide (Na2S) and sodium hydroxide (NaOH) at elevated temperature
and pressure. This process extracts the cellulose from the wood by dissolution of the lignin
that binds the cellulose fibers together. Spent digestion fluid, or black liquor, is combined
with water used to wash the resulting pulp, which combination forms weak black liquor.
Weak black liquor is a dilute solution (approximately 12% to 15% solids) of wood
lignin, organic materials, and oxidized inorganic compounds. This weak black liquor is
first directed through a series of evaporators to increase the solids content to 50—65%. The
concentrated black liquor is then sent to a recovery boiler—Recovery Boiler No. 5 in this
research—where it is burned to produce heat for process steam and to allow recovery of
inorganic solids contained in the black liquor, which collect at the bottom of the boiler as
smelt. Recovery Boiler No. 5, one of several recovery boilers that function independently,
is a noncontact, low-odor boiler with a rated capacity of 520 gal/min of black liquor
(density 11.4 lb/gal), or a mass flow rate of 121 tons/h of black liquor solids. In addition,
Recovery Boiler No. 5 is equipped to fire fuel oil during startup, shutdown, and
malfunction. The heat from the recovery boiler is used to produce 35% of the steam and
electricity required by the facility.
Table 3-2 summarizes the fuel use for Recovery Boiler No. 5 during the testing period.
Table 3-2. Fuel Use During the Smelt Tank Vent Test Period
Test
No.
Fuel
Type
Feed Rate
(gal/min)
Test
Duration
(min)
Total Volume
Used
(gal)
Combined Overall
Total Volume Used
(gal)
1
Black Liquor
491.43
400
196,572
196,572
1
#2 Oil
None
400
None
2
Black Liquor
356.93
479.4
171,112.2
180,906.3
2
#2 Oil
20.43
479.4
9794.1
3
Black Liquor
454.54
479.58
217,988.3
217,988.3
3
#2 Oil
None
479.58
None
14
-------
A more detailed description of the fuel fired during each specific test is provided in
Table 3-3.
Table 3-3. Black Liquor, Black Liquor Solids, and #2 Oil Process Data for
Testing Days
Test
No.
Test
Date
2001
Start
Time
End
Time
Black
Liquor
Firing
Rate
(gal/min)
Black
Black
Total Black
#2 Oil
Liquor
Liquor
Liquor Solids
Firing
Solids
Solids
Fired
Rate
(%)
(tons/h)
(tons)
(gal/min)
67.0
112.61
900.85
0
67.51
82.41
659.28
20.43
66.82
103.87
803.99
0
: 8 x 10 quartz filter was
changed approximately
1A3
IB3
2
3
12/14
12/15
12/16
0745
1130
0715
0715
1118
1436
1515
1515
491.43
356.93
454.54
halfway through the test due to moisture.
A sample of the black liquor was collected for analysis by facility environmental staff.
The black liquor sample was collected at the inlet to the fuel injector during the first test
day; analytical results are shown in Table 3-2.
Table 3-4. Analysis of Black Liquor
Parameter Value
Sulfur
2.73%
Ash
23.18%
Carbon
23.86%
Hydrogen
6.93%
Nitrogen
0.07%
Oxygen
43.23%
Chlorine
870.6 ppm
Heating Value
4208 Btu/lb
Flue gases leaving the boiler are routed to an electrostatic precipitator for PM control.
The recovered PM, which is primarily sodium sulfate (Na2S04) and sodium carbonate
(Na2C03), is added to the concentrated black liquor, and the clean flue gas exits through
15
-------
the stack. Approximately 95% of the Na2S04 in the flue gas is recovered. The Na2S04 is
reduced to Na2S in the reducing zone (lower section) of the boiler and, together with
Na2C03 and other inorganic chemicals, drains as molten (1038-1149 °C, 1900-2100 °F)
smelt from the boiler bottom through a char bed filter into smelt dissolving tanks, which
are large, covered vessels. As the smelt falls several feet into the tank, it is sprayed by high
pressure steam or recirculating green liquor in order to break the molten smelt into small
droplets and cool it.
This molten smelt and process steam are the main products from the combustion of
black liquor; the smelt is refined and reused as solvent in the first process at the mill,
digestion of the wood chips. Refining the smelt begins in smelt dissolving tanks where the
filtrate—mainly calcium oxide (CaO) and water—from a process later in the sequence is
added to dissolve the smelt and form green liquor, which is primarily Na2S, Na2C03, and
NaOH. The green liquor level in the smelt dissolving tank is maintained such that the tank
remains about half full. The tank is constantly agitated to prevent the formation of hot
spots on the surface of the liquor and the accumulation of solids on the bottom of the tank.
Surface hot spots can contribute to the formation of explosive hydrogen gas from the
dissociation of water reacting with the hot smelt. A calcium hydroxide sludge precipitates
from the smelt tank and is returned to a lime kiln to regenerate CaO.
The vapor space above the liquid level, or headspace, is filled with large volumes of
steam and particulate matter from the quenching of the molten smelt. An induced-draft fan
constantly draws this headspace gas, containing vapor and entrained PM, through a wet
scrubber. The smelt tank scrubber is a Ducon venturi-rod dynamic scrubber. The gas
stream from the vertical vent pipe of the smelt dissolving tank is withdrawn into a
horizontal duct leading to the rod deck section of the Ducon scrubber. Prior to entering the
rod deck section, the gas stream passes through a set of louvers and a presaturator spray
chamber. The gas is then accelerated through the openings between a set of hardened rods,
and a wash fluid of dilute green liquor is used for scrubbing in this section. After passing
through the rod deck, the gas stream passes through a 45 in.x31 V2 in. duct that enters the
lower chamber of the main scrubber vessel. Droplets are removed cyclonically in this
lower chamber. Gas from the lower chamber is withdrawn into a duct that leads to the 250-
hp radial-blade centrifugal fan. A single spray nozzle, injecting a wash liquid of dilute
green liquor is used for supplemental particle removal. The fan discharge gas stream
tangentially enters the upper chamber used for droplet removal. The treated gas stream is
then injected into the smelt tank vent stack.
16
-------
The green liquor is sent to clarifiers that remove insoluble impurities (dregs), which are
concentrated for disposal at the facility's permitted landfill. The clarified green liquor is
subsequently transferred to a tank where CaO is added to regenerate the digestion
reactants, Na2S and NaOH. The recovered digestion solution, or white liquor, is then
recycled to the initial digestion process.
The clarifiers, storage tank, and dregs surge tank have no forced air flow during normal
operations and vent to the atmosphere. The dregs filter is also open to the atmosphere.
Although the filter operates under a vacuum, the hood over the filter drum pulls air across
the dregs and will, therefore, collect some minor air pollutant emissions.
The smelt tank vent is located on the roof of the smelt tank building, with the sampling
port installed at a point that meets EPA Method 1 requirements for length of straight run
and for orientation of the port with respect to the plane of bends in the duct work. The
sampling port was located downstream of the ventilation fan and approximately 300 feet
above ground level. A schematic diagram depicting the orientation of the sampling port
and equipment is shown in Figure 3-3. The area around the sampling port is an open
unobstructed rooftop that did not impede sampling activities.
Top View
Smelt Tank Vent Stack Location Layout
Ofcftbn
Um
f
£iP!
i *>«*» Suppuft
Stock
Probe
*
Flanged
Part
. V
M. W-
i '
AltMfH
tsfcxMe
Certrot
Unit
SmeN
Hoot Arm
{- 300 k«rt at»
-------
The sampling location could be accessed by freight elevator or stairs. The sampling
equipment was loaded into the freight elevator and transported to the rooftop level, where
it was positioned as shown in Figure 3-3. The dilution unit was located outside on the
rooftop, with the control unit located inside the elevator equipment building and connected
to the dilution unit using flexible hose and electrical wiring. Supplemental equipment such
as the ELPI was transported to the sampling site by freight elevator. There was no space in
the vicinity of the sampling port for an enclosed area in which to prepare sampling
components or recover the sample collection arrays. Therefore, an area at ground level was
identified for preparation of sampling components and for recovery of the sample
collection arrays and preparation for transport to the laboratories.
Pre-Test Survey
A thorough survey of the test site was conducted in order to determine that the test
equipment could gain access to the test location and that the dilution sampling system and
the control module would fit in the test location, to identify and gain access to the utilities
needed to operate the dilution sampling system and its ancillary equipment, to arrange for
the installation of a sample collection port (Figure 3-4), and to determine and evaluate the
Figure 3-4. Sampling Port at the Smelt Tank Vent Location.
18
-------
means of positioning the dilution sampling system at the desired location. The pre-test
survey also considered access to utilities and personnel, legal, and safety requirements.
Source data were obtained on parameters—such as exhaust gas flow rate, velocity,
temperature, water vapor content, and approximate particulate matter concentration—that
are useful for estimating appropriate dilution ratios and duration of sample collection. A
second pre-test survey was made to verify that the sampling port had been installed
correctly and that all arrangements necessary to access utilities were complete. Electrical
power (250V, single phase, 40 A) was provided and installed by the facility.
19
-------
Blank Page
20
-------
Experimental Procedures
For sampling undiluted hot exhaust gas streams, the EPA/ECPB dilution sampling
system (schematic diagram in Figure 3-1), sample collection arrays, sample substrates, and
dilution air cleaning system were used. ERG transported the dilution sampling system and
ancillary equipment to and from the sampling site. To minimize introduction of
contaminants, EPA pre-cleaned and preassembled the dilution sampling system and sample
collection arrays in a clean environment prior to transport to the test site. The dilution
sampling system and dilution air cleaning system were assembled on separate portable
aluminum frames equipped with wheels and tie-down for transport to and from the test
site. ERG maintained the dilution sampling system and sample collection arrays in a
contaminant-free condition prior to collection of stationary source samples and field
blanks.
A sampling system blank test was performed prior to transporting the dilution sampling
system to the test site to ensure that the system had been cleaned properly and was leak
free. The blank test was performed in the laboratory by completely assembling the dilution
sampling system, including the sample collection arrays connected to the residence time
chamber, and all instrumentation. The blank test was conducted for approximately 8 hours,
which is a time period consistent with the expected duration of the source tests. Following
the blank test, the dilution sampling system was shut down in reverse order from startup,
and all substrates were unloaded, preserved as appropriate, and analyzed to verify the
absence of contamination in the dilution sampling system.
Preparation for Test Setup
Prior to the deployment of the dilution sampling system at the test site and initiation of
sampling with the dilution sampling system and associated sample collection arrays, EPA
Methods 1 through 4 were conducted to establish key experimental parameters for test
conditions.
21
-------
Traverse Point Determination Using EPA Method 1
EPA Method l3 was used to establish the number and location of sampling traverse
points necessary for isokinetic and flow sampling. These parameters are based on how
much duct distance separates the sampling ports from the closest downstream and
upstream flow disturbances. The smelt tank vent sampling site was located on the vertical
wall of a vent, with the sampling port at a location that meets EPA Method 1 requirements
for length of straight run and for orientation of the port with respect to the plane of bends
in the duct work. Sampling at the test site was performed at the point determined by
Method 1 to represent the average exhaust velocity at the smelt tank vent.
The following duct dimensions were measured:
Inside of far wall to outside of nipple (distance A): 72 in.
Inside of near wall to outside of nipple (distance B): 7 in.
Inside stack dimensions: 65 in.
Traverse point locations for the circular smelt tank vent are listed in Table 4-1. A table
of metric unit conversions is shown in Appendix A.
Table 4-1. EPA Method 1-Traverse Point Locations for the Circular Smelt
Tank Vent
Traverse
Fraction of Inside Stack
Distance from Stack
Traverse Point
Point
Dimension3
Wall
Location
Number
(%)
(in.)
(in.)
1
2.6
1 3/4
8 %
2
8.7
5 %
12%
3
14.6
9 »/2
16 »/2
4
22.6
14 %
21 3/4
5
34.2
22 %
29 %
6
65.8
42 %
49 %
7
77.4
50 %
57 %
8
85.4
55 »/2
62 »/2
9
91.8
59 %
66 %
10
97.4
63 %
70 %
a Inside stack depth: 65 in. Distance from lip of flange to inside stack wall: 7 in.
22
-------
The absolute pressure of the flue gas (in inches of mercury) was calculated by
(4-1)
Where:
PS = absolute gas pressure, inches of mercury
Pbar = barometric pressure, inches of mercury (29.89 in.)
Pg = gauge pressure, inches of water (static pressure) (-0.11 in.)
The value 13.6 represents the specific gravity of mercury (1 in. of mercury = 13.6 in. of
water). For the stack tested, the absolute gas pressure under the test conditions was 29.882
in. of mercury.
Volumetric Flow Rate Determination Using EPA Method 2
Volumetric flow rate was measured according to EPA Method 2.4 A Type K
thermocouple and S-type Pitot tube were used to measure flue gas temperature and
velocity, respectively. All of the isokinetically sample methods that were used
incorporated EPA Method 2.
Pitot Tube Calibration
The EPA has specified guidelines concerning the construction and geometry of an
acceptable S-type Pitot tube. If the specified design and construction guidelines are met, a
pitot tube coefficient of 0.84 is used. Information pertaining to the design and construction
of the S-type Pitot tube is presented in detail in Section 3.1.1 of report EPA 600/4-77-
027b, and only S-type Pitot tubes meeting the required EPA specifications were used. Pitot
tubes were inspected and documented as meeting EPA specifications prior to field testing.
Calculation of Average Flue Gas Velocity
The average flue gas velocity for each traverse point is calculated by
(4-2)
Where:
Vs = average flue gas velocity, ft/s
23
-------
Kp = Pitot constant (85.49)
Cp = Pitot coefficient (dimensionless), typically 0.84 for Type S
APmg = average flue gas velocity head, inches of water
460 = 0 °F, expressed as degrees Rankin
Ts = flue gas temperature, °F
Ps = absolute stack pressure (barometric pressure at measurement site plus stack
static pressure), inches of mercury (29.882 in.)
Ms = wet molecular weight, pounds per pound-mole (24.9 lb/lb-mole).
The flue gas velocity calculated for each traverse point and the average velocity are
shown in Table 4-2.
Table 4-2. Average Flue Gas Velocity for Each Traverse Point
Traverse Pont
Velocity
(Calculated in Table 4-1)
(ft/min)
1
775.8
2
947.9
3
991.2
4
1122.0
5
1033.8
6
947.9
7
1164.0
8
1164.0
9
1032.0
10
1032.0
Average Velocity
1021.1
The point of average velocity has the closest relationship to Traverse Point #5.
Nozzle Size Determination
It is desirable to sample at or near isokinetic velocities at the probe inlet nozzle. The
nozzle size is based on the required sample flow rate. Prior to using an Excel macro to
perform nozzle size calculations according to the procedures of EPA Method 51, the
velocity in the stack (feet per minute) must be determined from the Pitot traverses prior to
the start of the test run. The additional input required by the macro is sampling rate
(liters/minute). The nozzle selected for use at the smelt tank vent was 0.441 in. inner
diameter (id), nozzle size 12.
24
-------
Measurement of 02 and C02 Concentrations
The 02 and C02 concentrations were determined by use of a Fyrite bulb during the
traverse.
Stationary Gas Distribution (as Percent Volume)
The following values were measured by ERG (02, C02) and by the facility (CO).
Measured 02 = 20.0%V
Measured C02 = 0.0%V
Measured CO = 0.0%V
The percentage of nitrogen (N2) was calculated by
N2 %V = 100- (02 %V + C02 %V + CO%v) = 80.00%V (4-3)
Dry Molecular Weight of Flue Gas
The dry molecular weight of the flue gas (Md) was calculated by
Md = (CO%V x 0.44) + (o %V x 0.32) + [(cO%V + N2 %v) x 0.28
Md = 28.8 lb I lb - mole
Where:
Md = molecular weight of flue gas, dry basis (lb/lb-mole)
C02%V = percent C02 by volume (20.0)
02%V = percent 02 by volume (0.0)
CO%V= percent CO by volume (0.0)
N2%V = percent N2 by volume (80.0)
0.44 = molecular weight of C02, divided by 100
0.32 = molecular weight of 02, divided by 100
0.28 = molecular weight of N2 or CO, divided by 100
Wet Molecular Weight of Flue Gas
The wet molecular weight of the flue gas (Ms) was calculated by
(4-4)
Ms = ('Md X Mfd) + (o.l8 X H20%V) = 24.9 wet lb / lb - mole (4.5)
Where:
Ms = wet molecular weight of flue gas, wet lb/lb-mole
25
-------
Md = molecular weight of flue gas, dry basis (28.8 lb/lb-mole)
Mfd = dry mole fraction of effluent gas, calculated as [1 -H2O%V/100] (0.64)
0.18 = molecular weight of H20, divided by 100
H20%V = percent H20, by volume (35.81)
Determination of Average Moisture Using EPA Method 4
EPA Method 45 was used to determine the average moisture content of the duct gas. A
gas sample was extracted from the vent duct, moisture was removed from the sample
stream, and the moisture content was determined gravimetrically. Before sampling, the
initial weight of the impingers was recorded. When sampling was completed, the final
weights of the impingers were recorded and the weight gain was calculated. The weight
gain and the volume of gas sampled were used to calculate by computer the average
moisture content (percent) of the duct gas. Method 4 was incorporated into the techniques
used for all of the manual sampling methods that were used during the test.
The measurements shown in Table 4-3 were made on November 1, 2001, using Method
4 to determine moisture recovery.
Table 4-3. Moisture Recovery for Method 4 (Measured on 11/01/01)
Weight of
Impinger
Tip
Configuration
Impinger Weight
Impinger
Number
Impinger
Solution
Impinger
Contents
(g)
Final
(g)
Initial
(g)
Weight
Gain
(g)
1
Water
100
Standard
827.4
608.9
218.5
2
Water
100
Standard
761.4
521.2
240.2
3
Empty
"
Standard
499.9
483.4
16.50
4
Silica Gel
300
Standard
773.7
758.3
15.40
Total Weight Gain (g)
490.6
Volume of Dry Gas Sampled at Standard Conditions (dscf)
The volume of dry gas sampled under standard conditions was calculated by
A H
= 17-64 x Vm x PbaT + 13+6 = 41.45 dscf (4-6)
m
26
-------
Where:
Vm(std) = volume of dry gas sampled at standard conditions, dry standard cubic feet
(dscf)
Vm = volume of gas metered, cubic feet, dry (43.89)
Pbar = barometric pressure at measurement site, inches of mercury (29.89)
AH = sampling rate, measured as differential pressure at the meter orifice, inches
of water (1.816)
Tm = dry gas meter temperature, degrees Fahrenheit (89.6 °F).
The constant 17.64 was used for conversion to standard conditions (84.7 °F + 460
°R)/30.24 in. Hg; 460 °R is 0 °F. Using measured values from the field data sheet, the
volume of dry gas sampled at standard conditions is calculated to be 41.45 dscf.
Volume of Water Vapor at Standard Conditions (dscf)
The volume of water vapor under standard conditions was calculated by
K<,m = 004707 x K = 23.1 dscf (4-7)
Where:
Vw(std) = volume of water vapor at standard conditions, dry standard cubic feet (dscf)
Vlc = volume of liquid catch, grams (490.6)
The constant 0.04707 is the standard cubic feet per gram (or milliliter) of water at
standard conditions. Using the total weight gain for water determined using Method 4
(Table 4-3, above), the volume of water vapor at standard conditions is calculated to be
23.1 dscf.
Calculation of Moisture/Water Content (as Percent Volume)
The moisture content of the gaseous stack emissions is calculated by
y
H 0%V =100 ' -—^ =35.81%T (4-8)
w(std) m(std)
Using values measured using EPA Method 4 and values calculated previously, the
moisture content was calculated to be 35.81%V.
27
-------
Calculation of Dry Mole Fraction of Flue Gas
The dry mole fraction of flue gas is calculated by
HJ)%V
(4-9)
Where:
Mfd = dry mole fraction of effluent gas
Using the percent moisture determined above, the dry mole fraction of effluent gas is
calculated as 0.64.
The smelt tank vent sampling site is located on the roof of the smelt tank, with the
sampling port installed at a point that meets EPA Method 1 requirements for length of
straight run and for orientation of the port with respect to the plane of bends in the duct
work. The sampling port is located downstream of the ventilation fan, and the sampling
location is elevated approximately 300 feet above ground level. A schematic diagram
presenting the orientation of the sampling port and equipment is presented in Figure 4-
l.The area around the sampling port is an open unobstructed rooftop and does not impede
sampling activities.
Setup of the Dilution Sampling System
Top View
Smelt Tank Vent Stack Location Layout
Stev-3®! SUpfKrt
PAjlpfl kiPt
imoor : ? t
h'V.'tjWf
flared
Port
Contra*
Rgftf Arts*
i - J00 feet above srsKtef
Senator
Figure 4-1. Schematic Diagram of the Sampling
Location (Top View).
28
-------
Figure 4-2 shows the sampling probe installed in the 6 in. flanged port used for
sampling.
Figure 4-2. Dilution Sampling System Sampling Probe
Installed in 6 in. id Flanged Port.
Access to this location is by freight elevator or stairs. The sampling equipment was
loaded into the freight elevator, transported to the rooftop level, and then positioned as
shown in Figure 4-1. The dilution unit was located outside on the rooftop, and the control
module (Figure 4-3) was located inside the elevator equipment building and connected to
the dilution sampling unit using flexible hose and electrical wiring.
An ELPI, manufactured by Dekati (Figure 4-4), with associated laptop computer, was
transported by freight elevator and stairs to reach the sampling site. The ELPI was also
connected to the sampling module together with other sample collection arrays; one
sample collection array is visible in the background.
The dilution system sampling module with all sample collection arrays and instruments
attached is shown in Figure 4-5. The ELPI is in the foreground and the various sample
collection arrays (the white filter holders are readily visible) are attached to the various
ports of the dilution system sampling module.
29
-------
Figure 4-3. Dilution System Control Module
Positioned at the Sampling Location.
Figure 4-4. Dekati ELPI Positioned at the Sampling
Location.
30
-------
Figure 4-5. Dilution System with Sample Collection
Arrays and Instruments Attached.
Because of the lack of available space in the immediate vicinity of the sampling
location, sample recovery was conducted inside the ERG mobile laboratory located at
ground level on the host facility property (Figure 4-6). Sample collection arrays were
transported intact to the Recovery Area in the ERG mobile laboratory (note assembled
collection arrays on the table in front of the analyst) and disassembled. Samples were then
labeled, packaged for transport, and placed in a chest-style freezer. Copies of the chain of
custody documentation are shown in Appendix B
Pre-Test Leak Check
To perform a pre-test leak check on the assembled dilution sampling system in the
field, the end of the probe was plugged with a Swagelok fitting. Solvent-cleaned blank-off
plates were inserted in place of the orifice plates at the orifice meter ran flanges using
gaskets on each side. A new, tared 8 in.xlO in. quartz filter was inserted into the filter
holder, and the fittings were carefully sealed. A vacuum pump was attached to the
residence chamber, and a Magnehelic gauge was attached to an available port. The valve
between the pump and the chamber was opened, and the vacuum was read as the pump
was turned on. As the reading passed 27 in., the stopwatch was started, and the valve
between the pump and the chamber was closed. The leak rate was timed between 25 to 20
31
-------
Figure 4-6. Recovery Area for Dilution Sampling System
Sample Collection Arrays.
in. and again from 20 to 15 in., and the two times were averaged. Using the recorded data,
the leakage rate in cubic feet per minute was calculated according to Equation 4-10.
leakage rate =
A P
A T
¦x Fx CF
(4-10)
Where:
leakage rate = rate of leakage (ftVmin)
AP = change in pressure (in. water)
A71 = time increment (sec)
V = volume of the dilution sampling system (15.3 ft3)
CF = unit conversion factors
• 60 sec/min
• 1 atm/406.8 in. water
An acceptable leak rate is less than or equal to 0.1 ftVmin, or at least 1 min 53 sec for a
pressure change of 5 in. U20. For this test, the average time for a 5-in. H20 pressure
change to occur was 2 min 29 sec. The resulting leak rate was 0.076 ftVmin, which
satisfies the criteria for acceptability.
32
-------
Orifice Flow Check
Critical orifice flows on the sampling pumps were checked without sample collection
arrays in place, using a rotameter to verify that the channels on sampling array pumps were
at the specified flow rate of 16.7 L/min. Rotameters were calibrated with a National
Institute of Standards and Technology (NIST)-traceable electronic bubble flow meter, and
the readings were converted to flow (L/min) using a spreadsheet.
Determination of Test Duration
To ensure the best possible collection of PM, sampling was conducted for 8 hours, the
maximum amount of time permitted by the facility.
Canister/Verifiow Blanks
Prior to deployment in the field, SUMMA polished canisters and Veriflow canister
filling units were cleaned and blank analysis was performed in the laboratory. All units
met the cleanliness criterion of less than 10 parts per billion Carbon (ppbC, Table 4-4).
Table 4-4. Blank Values for Veriflows and Canisters
Unit
Blank Value
lEEbCi
Veriflows
EPA Unit #418 (Source Veriflow)
ERG-1 Ambient Veriflow
EPA Unit #315 (Dilution Veriflow)
0.0
1.9
0.0
Canisters
1442
1441
4040
3506
3957
3947
3503
4006
0.0
6.0
0.0
0.0
0.0
0.0
1.0
10.0
33
-------
Determination of Flow Rates
A Visual Basic macro was written to process raw data files of flow rate information and
convert this information to actual flow based on temperature, pressure, and calibration
data. For venturi flows, the macro converted differential pressure into a reported flow rate.
The square root of the differential pressure was then multiplied by a previously determined
calibration factor based on the flowing temperature, and the resulting value was converted
to standard liters per minute (sL/min) using Ideal Gas Law relationships (1 atm, 70 °F).
Calibration data for the venturi were generated by placing a dry gas meter at the inlet to
the sample probe. The flows reported by the data acquisition system were corrected to
actual liters per minute (aL/min) and compared to those produced by the dry gas meter
corrected to the venturi conditions. An Excel macro automatically selected a correct
calibration value to be applied based on the flowing temperature.
Since the actual venturi flow was dependent upon the operating conditions, the set
point value displayed and entered on the viewing screens needed to be adjusted to achieve
the desired flow. Information to be entered included the desired flow, flow temperature,
flow pressure, and barometric pressure; the Excel macro automatically selected the correct
value to be applied based on the flow temperature.
Sample Collection Arrays
Prior to actual testing (Test Run 1, 12/14/01; Test Run 2, 12/15/01; Test Run 3,
12/16/01), sample collection arrays were attached to various ports on the dilution sampling
system, as shown in Figure 4-7. Up to ten sampling ports were available attached to either
the dilution chamber or the residence chamber (available sampling ports are shown in
Figure 3-1). The following sample collection arrays were used for Tests 1, 2, and 3:
Dilution Chamber Sample Collection Arrays
Samples of the dilution air were collected to evaluate the analyte background in
the dilution air.
- Dilution Chamber Collection Array Dl, Port Dl
Sample Collection Array Dl collected gas-phase semivolatile organic
compounds, particle-bound organic materials, and metals. The array consists of
a cyclone separator to remove particulate matter with an aerodynamic diameter
greater than 2.5 |im. One leg contains a quartz filter followed by two PUF
sampling plugs in series. The other leg of Array Dl consists of a Teflon filter.
34
-------
Dilution chamber
Port #D1
Port #D2 Port #D3
B QF B TF
Cyclone
Residence chamber
Port #R2 Port#R3
B TF B TF
Port #R5 Port #R6
B TF B TF
Port #R4
B qf gQF
Cyclone
Port #R8
B QF g qf
Cyclone
Field Blanks
QF
1
TF
1
PUF
1 Pair
SUMMA
1
DNPH
1
Total
Substrates
QF
8
TF
6
PUF
9 Pair
Denuder
2
SUMMA
4
DNPH
7
Port#R10
: puf
BQF E~3 c
Leaend
QF
= Quartz Filter
KOH QF
KOH Quartz Filter
TF-0.5
= Teflon© Filter - 0.5
TF
= Teflon© Filter
PUF
= Polyurethane Foam Sampling Module
Denuder
= Denuder
SUMMA
= SUMMA®Canister
DNPH
= 2,4-Dinitrophenylhydrazine
-impregnated silica gel cartridge
Figure 4-7. Sample Collection Arrays Used for Testing at the Smelt Tank
Vent.
- \ilution Chamber Collection Array D2, Port D2
35
-------
Sample Collection Array D2 collected fine particulate matter and gas-phase
organic compounds. This array consisted of a single filter unit followed by a
SUMMA polished stainless steel canister.
- Dilution Chamber Collection Array D3, Port D3
Sample Collection Array D3 collected carbonyl compounds using three DNPH-
impregnated silica gel sampling cartridges in series and a pump.
Residence Chamber Sample Collection Arrays
Samples of the air were collected from the Residence Chamber to evaluate the
analyte presence in diluted stationary source air.
- Residence Chamber Sample Collection Array R2, Port R2
Sample Collection Array R2 collected fine particulate matter. The array consists
of a 2.5 |im cyclone followed by two identical legs containing Teflon filters.
- Residence Chamber Sample Collection Array R3, Port R3
Sample Collection Array R3 collected fine particulate matter and carbonyl
compounds. This array consists of three carbonyl collection cartridges in series,
with a pump.
- Residence Chamber Sample Collection Array R4, Port R4
Sample Collection Array R4 collected fine particulate matter on paired quartz
filters for total carbon and elemental analysis, as well as semivolatile organic
compounds using two PUF sampling modules in series. This sampling array
consists of a 2.5 |im cyclone with two quartz filters in parallel, with one of the
quartz filters followed by two PUF sampling modules in series.
- Residence Chamber Sample Collection Array R5, Port R5
Sample Collection Array R5 collected fine particulate matter and gas-phase
organic compounds. This array consists of a single Teflon filter unit followed
by a SUMMA polished stainless steel canister.
- Residence Chamber Sample Collection Array R6, Port R6
Sample Collection Array R6 collected fine particulate matter. This array
consists of a 2.5-|im cyclone followed by two identical legs containing Teflon
filters.
- Residence Chamber Sample Collection Array R8, Port R8
Sample Collection Array R8 collected fine particulate matter on paired quartz
filters for total carbon and elemental carbon analysis, as well as semivolatile
organic compounds using two PUF sampling modules in series. This sample
collection array consists of a 2.5 |im cyclone with two quartz filters in parallel,
with one of the quartz filters followed by two PUF sampling modules in series.
36
-------
- Residence Chamber Sample Collection Array RIO, Port RIO
Sample Collection Array RIO collected fine particulate matter on two quartz
filters in parallel, semivolatile organic compounds on two XAD-4-coated
denuders in series as well as two PUF sampling modules in series. This sample
collection array consists of a 2.5 |im cyclone immediately prior to two XAD-4-
coated annular denuders in series, followed by two identical legs each
containing a quartz filter, with one of the quartz filters followed by two PUF
sampling modules in series.
In the field, denuders were used in series as pairs on Residence Chamber Port 10. On
Test Day 1, the paired denuders were changed and extracted every half hour across the
duration of testing, as shown in Table 4-5. The paired denuders were removed from the
sample collection array and separated. Each denuder was rinsed with a mixture of
methylene chloride, acetone, and hexane in a volume ratio of 2:3:5. The solvent mixture
was added to the denuder, and the denuder tube was capped and shaken (4 times); an
internal standard was added to the first extraction. The rinses were combined in a pre-
cleaned glass jar for paired denuders, the jar was labeled, sealed with Teflon tape, chain of
custody documentation was initiated for the extract, and the jar was stored over ice.
Denuder extracts for each half hour sampling episode were combined, but each half-hour
sampling episode generated a separate sample (i.e., 13 denuder extract samples generated).
After extraction, the denuders and caps were dried using high purity nitrogen and capped
until ready for re-use. On Test Days 2 and 3, a different sampling scheme was used for the
paired denuders. The first pair of denuders collected sample for a half hour; the second set
of paired denuders collected sample for 1 hour, the third pair for 2 hours, and the fourth
pair for 4 hours, as shown in Table 4-5.
Use of the ELPI Particle Size Distribution Analyzer
In addition to the sample collection arrays, an ELPI continuous particle size analyzer
was used on the residence chamber to collect data on particle size distribution in the
diluted source sample. The ELPI measures airborne particle size distribution in the size
range 30-1000 nm (0.03 - 10 |im) with 12 channels. The principle of operation is based on
charging, inertial classification, and electrical detection of the aerosol particles. The
instrument consists primarily of a corona charger, low pressure cascade impactor, and
multichannel electrometer.
In operation, the sample first passed through a unipolar positive polarity charger where
particles in the sample were charged electrically by small ions produced in a corona
37
-------
Table 4-5. Denuder Sampling Scheme
Test
Pair
Duration
Number
Number
(min)
1
1
30
2
30
3
30
4
28
5
29
6
30
7
26
8
30
9
29
10
30
11
30
12
30
13
30
Total
13 samples
382 min
2
1
30
2
60
3
120
4
240
Total
4 samples
450 min
3
1
30
2
60
3
120
4
240
Total
4 samples
450 min
discharge. After the charger, the charged particles were size-classified on a low pressure
impactor. The impactor is an inertial device classifying the particles according to their
aerodynamic diameter, not charge. The stages of the impactor are insulated electrically,
and each stage is connected individually to an electrometer current amplifier. The charged
particles collected in a specific impactor stage produce an electrical current, which is
recorded by the respective electrometer channel. A larger charge correlates to a higher
particle population. The current value of each channel is proportional to the number of
particles collected and, thus, to the particle concentration in the particular size range. The
38
-------
current values are converted to aerodynamic size distribution using particle size-dependent
relationships describing the properties of the charger and the impactor stages.
The precalibrated instrument was transported to the field and placed in the vicinity of
the sample collection arrays on a sturdy table. Twenty minutes prior to the start of the test
run, the ELPI was turned on to warm up and equilibrate. The computer was turned on, and
the sample acquisition program was initiated in the flush mode. On the ELPI, the sample
flow was manually adjusted to the manufacturer's specifications (vacuum setting to 100±1
mbar). The ELPI was set to monitor the range of 0.03 - 10 |im (30 - 1000 nm) midpoint
particle diameter to provide an indication of particle size distribution in the range below 10
|im, as well as the concentration distribution of the particles within this size range. The
data system was initially set up to collect data for particles ranging from 0.03 to 10 |im; the
particles were collected over the duration of each test. The stages of the ELPI impactor are
electrically insulated, and each stage was connected individually to an electrometer current
amplifier. The charged particles collected in a specific impactor stage produce an electrical
current which is recorded by the respective electrometer channel. A larger charge
correlates to a higher particle population. Values are converted to aerodynamic size
distribution using particle size-dependent relations describing the properties of the charges
and the impactor stages.
Shortly before the initiation of the test run, the data system was programmed to collect
particulate data that encompassed the expected duration of the test run. The instrument
completed a charging process whereby particles in the sample were electrically charged by
ions produced in a corona discharge. After the charger, the particles are size-classified in a
low pressure impactor, producing particle size and concentration data for the selected scan
range. The ELPI was the last piece of equipment connected to the residence chamber.
When the test run was started, the inlet line was attached to the port, and "Flush Off' was
initiated on the data system. Data were continually saved on the computer hard drive, and a
real-time display on the computer screen showed the particle distribution. Graphical
presentations of the data were prepared off-line.
For each run, impactor stages were covered with tared aluminum foils. After test runs,
the foils were recovered and individually weighed for additional mass data. Collected
materials on each foil were also extracted to determine organic composition of each stage.
39
-------
Operation of the Dilution Sampling System with Sample
Collection Arrays
After completion of the pre-test run to establish experimental parameters for the test,
the dilution sampling system was prepared for a full test run. Sampling probe temperature
setpoints were set equal to or slightly above the measured stack temperature, and the
system was equilibrated at temperature. Sample collection arrays were loaded with
appropriate collection media, and flow/leak checks were performed with each array to
ensure that the entire system would be leak free in operation. Sampler flows were set just
before initiation of the test to prevent heat loss from the heated probe. The blower and the
ring compressor were started to achieve a slightly positive pressure; then the blower flow
was adjusted to cause the stack gas to flow into the dilution sampling system after the
probe was inserted into the duct. Sample collection array pumps were started, and valves
for the SUMMA canisters were opened to initiate canister air sample collection. The
sampling process was carefully monitored by the sampling team based on the pressure
change in the canister to ensure that the filters were not overloaded in the course of
sampling. Start time and other pertinent data were recorded. At the end of the 8-hour
sampling interval, the sampling process was stopped by stopping the pumps for the sample
collection arrays and closing the valves on the SUMMA canisters. The probe was
withdrawn from the stack, the blower and ring compressor were turned off, and heaters
were turned off and allowed to cool. Each sample collection array was leak checked at the
end of the sampling period, and ending flow rates were documented. Experimental
parameters for Tests 1, 2, and 2 are shown in Tables 4-6 to 4-8; blower flow, dilution flow,
and venturi flow for Tests 1, 2, and 3 are shown graphically in Figures 4-8 through 4-16.
Table 4-6. Run Time Summary Information, Test Run 1 (12/14/01)
Run Parameter Value
Start Time 7:45:06 AM
End Time 11:18:56 AM
Start Time3 11:30:06 AM
End Time 2:36:16 PM
Run Time 400.00 min
Barometric Pressure 29.32 in. Hg
Nozzle Size #12 (162 °C, 1021.1 fit/min)
Canister Flow Dilution canister, 9.75 cm3/min; residence chamber canister,
9.75 cm3/min; ambient canister, 10.75 cm3/min
(continued)
40
-------
Table 4-6. (continued)
Measurement Parameter Average
Venturi Flow
PT-101d
TE-104f
Dilution Flow
PT-102
TE-108
Blower Flow
PT-103
TE-105
Dilution Ratio
TE-101
TE-102
TE-103
Sample Flow Rates
29.53 aL/minb (22.69 sL/minc)
-0.74 in. WC
101.78 °C
882.83 aL/min (841.25 sL/min)
-1.22 in. WC
26.91°C
959.51 aL/min (869.25 sL/min)
-22.34 in. WC
25.56 °C
39.06
69.28 °C
109.06 °C
59.84 °C
Sample
Location
Port
Start/
End
Flow
Average Flow
(sL/min) (aL/min) (sL/min)
pm25
dilution air
-
start
17.13
17.35
pm25
dilution air
-
end
16.98
17.20
pm25
residence chamber
10
start
15.94
16.15
pm25
residence chamber
10
end
15.64
15.85
pm25
residence chamber
8
start
16.98
17.20
pm25
residence chamber
8
end
16.83
17.05
pm25
residence chamber
6
start
17.13
17.35
pm25
residence chamber
6
end
17.13
17.35
pm25
residence chamber
4
start
16.98
17.20
pm25
residence chamber
4
end
17.13
17.35
pm25
residence chamber
2
start
17.13
17.35
pm25
residence chamber
2
end
16.98
17.20
DNPH
residence chamber
3
start
0.78
0.79
DNPH
residence chamber
3
end
0.68
0.69
DNPH
dilution chamber
3
start
0.82
0.83
DNPH
residence chamber
3
end
0.73
0.74
17.06
15.79
16.91
17.13
17.06
17.06
0.73
0.78
(continued)
41
-------
Table 4-6. (continued)
a Due to the high moisture content of the smelt tank exhaust gas, it was necessary to change the 8 in. x 10 in. quartz
filter approximately halfway through the test due to a reduction system flow. Consequently, two start and stop times
apply to this test.
b aL/min = actual liters per minute.
c sL/min = standard liters per minute
<1 px = Pressure Transducer
e WC = Water Column
f TE = Thermocouple
Table 4-7. Run Time Summary Information, Test Run 2 (12/15/01)
Run Parameter
Value
Start Time
End Time
Run Time
Barometric Pressure
Nozzle Size
Canister Flow
7:15:36 AM
3:14:59 PM
479.38 min
29.35 in. Hg
#12(162 °C, 1021.1 ft/min)
Dilution canister, 8.14 cm3/min, residence chamber canister
8.14 cm3/min
Measurement Parameter
Average
Venturi Flow
PT-101C
TE-104e
Dilution Flow
PT-102
TE-108
Blower Flow
PT-103
TE-105
Dilution Ratio
TE-101
TE-102
TE-103
30.09 aL/min3 (23.20 sL/minb)
-0.77 in. WCd
100.69 °C
894.93 aL/min
851.97 sL/min
-1.41 in. WC
27.53 °C
766.37 aL/min
705.99 sL/min
-14.79 in. WC
27.88 °C
37.92
64.90 °C
100.09 °C
50.32 °C
(continued)
42
-------
Table 4-7. (continued)
Sample Flow Rates
Sample Location
Port
Start/
End
Flow
Average Flow
(sL/min) (aL/min) (sL/min)
pm25
dilution air
-
start
16.93
17.25
pm25
dilution air
-
end
16.93
17.25
pm25
residence chamber
10
start
15.46
15.75
pm25
residence chamber
10
end
15.75
16.04
pm25
residence chamber
8
start
16.93
17.25
pm25
residence chamber
8
end
16.93
17.25
pm25
residence chamber
6
start
17.08
17.40
pm25
residence chamber
6
end
17.08
17.40
pm25
residence chamber
4
start
16.93
17.25
pm25
residence chamber
4
end
17.08
17.40
pm25
residence chamber
2
start
17.08
17.40
pm25
residence chamber
2
end
17.08
17.40
DNPH
residence chamber
3
start
0.82
0.84
DNPH
residence chamber
3
end
0.80
0.81
DNPH
dilution chamber
3
start
0.82
0.84
DNPH
dilution chamber
3
end
0.84
0.86
16.93
15.60
16.93
17.08
17.01
17.08
0.81
0.83
a aL/min = actual liters per minute.
b sL/min = standard liters per minute
c PT = Pressure Transducer
d WC = Water Column
e TE = Thermocouple
Table 4-8. Run Time Summary Information, Test Run 3 (12/16/01)
Run Parameter
Value
Start Time
End Time
Run Time
Barometric Pressure
Nozzle Size
Canister Flow
7:15:53 AM
3:15:28 PM
479.58 min
29.77 in. Hg
#12(162 °C, 1021.1 ft/min)
Dilution canister, 9.38 cm3/min, residence chamber canister
8.97 cm3/min
(continued)
43
-------
Table 4-8. (continued)
Measurement Parameter
Venturi Flow
30.06 aL/min3 (22.91 sL/minb)
PT-101C
-0.80 in. WCd
TE-104e
110.66 °C
Dilution Flow
868.12 aL/min (856.03 sL/min)
PT-102
-1.34 in. WC
TE-108
21.42 °C
Blower Flow
877.71 aL/min (829.52 sL/min)
PT-103
-17.25 in. WC
TE-105
22.18 °C
Dilution Ratio
38.42
TE-101
69.08 °C
TE-102
128.98 °C
TE-103
49.82 °C
Sample Flow Rates
Location
Start/
Average Flow
Sample
Port End
(sL/min) (aL/min)
(sL/min)
pm25
pm25
dilution air
dilution air
- start
- end
17.40
17.40
17.08
17.08
17.40
pm25
pm25
residence chamber
residence chamber
10 start
10 end
15.74
15.74
15.46
17.40
15.74
pm25
pm25
residence chamber
residence chamber
8 start
8 end
17.25
17.25
16.93
16.93
17.25
pm25
pm25
residence chamber
residence chamber
6 start
6 end
17.40
17.25
17.08
16.93
17.32
pm25
pm25
residence chamber
residence chamber
4 start
4 end
17.10
17.40
16.78
17.08
17.25
pm25
pm25
residence chamber
residence chamber
2 start
2 end
17.25
17.10
16.93
16.78
17.17
DNPH
residence chamber
3 start
0.86
0.84
0.84
DNPH
residence chamber
3 end
0.82
0.80
DNPH
dilution chamber
3 start
0.84
0.82
0.84
DNPH
dilution chamber
3 end
0.84
0.82
a aL/min = actual liters per minute.
b sL/min = standard liters per minute
c PT = Pressure Transducer
d WC = Water Column
e TE = Thermocouple
44
-------
1200
1000
800
600
400
200
Actual
Actual
Standard
-200
Standard
Standard
Actual
6:00:00 7:00:00 8:00:00 9:00:00 10:00:00 11:00:00 12:00:00 13:00:00 14:00:00 15:00:00 16:00:00
Time
Figure 4-8. Blower Flow, Day 1 (12/14/01).
1200
Actual
Actual
Standard
Standard
C 600
Standard
Actua
6:00:00
1:00:00 9:00:00 10:00:00 11:00:0* 12:00:00 13:00:00 14:00:00 #5:00.00 16:00:00
Time
-200 -
Figure 4-9. Dilution Flow, Day 1 (12/14/01).
45
-------
50 T
Actual
Actua
Standard
Standard
Actual
Standard
6:00:00 7:00:00 8:00:00 9:00:00 10:00:00 11:00:00 12:00:00 13:00:00 14:00:00 15:00:00 16:00:00
Time
Figure 4-10. Venturi Flow, Day 1 (12/14/01).
1200
1000 -
800
600
400-
Actual
Actual
Standard
Actual
Standard
Standard
200 -
Standard | /Actual
6:00:00 7:00:00 8:00:00 9:00:00 10:00:00 11:00:00 12:00:00 13:00:00 14:00:00 15:00:00 15:00:00
Time
-200 -
Figure 4-11. Blower Flow, Day 2 (12/15/01).
46
-------
1400 -
Actual
Standard
F SOO
Actual
Standard
400
200
6:0(1:00 7:00:00 8:00:00 9:00:00 10:00:00 11:00:00 12:00:00 13:00:00 14:00:00 15:00:00 16:00:00
Time
-200
Figure 4-12. Dilution Flow, Day 2 (12/15/01).
50-r
Actual
Actual
Standard
Standard
Actual.
J<4
Standard
6:00:00 7:00:00 8:00:00 9:00:00 10:00:00 11:00:00 12:00:00 13:00:00 14:00:00 15:00:00 16:00:00
Time
Figure 4-13. Venturi Flow, Day 2 (12/15/01).
47
-------
i
'
Actual
Standard
¦
5
Standard
..-•Actual
6:00:00 7:00:00 8:00:00 9:00:00 10:00:00 11:00:00 12:00:00 13:00:00 14:00:00 15:00:00 16:00:00
Time
Figure 4-14. Blower Flow, Day 3 (12/16/01).
1200
1000
800
c
£
600
400
200
-200
Standard
Actual
Actual -
Standard
Actual
Standard
rr—1 1 1 1 ¦ ¦ 1—
6:0:0C ?!<»«0O 8:00:00 9:00:00 1 0:00:00 11:00:00 1 2:00:00 13:00:00 1 4:00:00 1 5:00:50 1 6:00:00
Time
Figure 4-15. Dilution Flow, Day 3 (12/16/01).
48
-------
45
40
Actual
_l
Standard
Actual
Standard
6:00:00 7:00:00 8:00:00 9:00:00 10:00:00 11:00:00 12:00:00 13:00:00 14:00:00 15:00:00 16:00:00
Time
Figure 4-16. Venturi Flow, Day 3 (12/16/01).
Dilution System Sample Collection Arrays Recovery
At the end of the sampling period, the pumps on the dilution system were turned off,
and recovery of the dilution sampling system consisted of removing the sample collection
arrays and turning off the particle size analyzer. Sample collection arrays were then carried
to the recovery area, where they were disassembled, the parts carefully labeled, and the
components of the sample collection arrays carefully packaged for transport back to the
laboratories. Denuder extractions were performed in the field according to the scheme
described above.
The sample collection arrays were removed sequentially at the cyclone connection.
Each individual collection array was removed, and the ends of the assembly were covered
with aluminum foil. As each sample collection array was removed from the dilution
sampling system, the sampling aperture was covered to avoid introduction of any
contaminants into the dilution sampling system. The ends of the sample collection array
were capped and the array placed in a secure container for transport to the sample recovery
area.
49
-------
In the sample recovery area, the sample collection arrays were disassembled into the
following components:
• Polyurethane foam (PUF) modules were disassembled from the sample collection
array as a module. Both ends of the PUF sampling module were capped, the module
placed in a sealable plastic bag, the bag appropriately labeled, and chain of custody
documentation initiated.
• Filters were positioned in specific filter holder assemblies as part of several of the
sample collection arrays. In the sample recovery area, the filter holder assemblies
were disassembled, and the filter was removed with Teflon-tipped tweezers and
placed in a prenumbered custom filter container with a locking lid. The appropriate
label was affixed to the filter container, and chain of custody documentation was
initiated. The filter holder assembly was reassembled without the filter, placed in a
sealable plastic bag, and labeled.
• Denuders were disassembled, the ends of the sorbent tube closed with Teflon caps
and sealed with Teflon tape, the sealed denuder tubes placed in a sealable plastic
bag, labeled, and chain of custody documentation initiated.
• Carbonyl sampling tube assemblies (three carbonyl sampling tubes in series) were
disassembled. The ends of the individual tubes were sealed with plastic caps, and the
sealed tubes placed in an aluminum foil packet, labeled to preserve the front/back
order from the sample collection array, placed in a plastic bag, labeled, and chain of
custody documentation initiated.
• Canister sampling was terminated by closing the valve on the canister at the end of
the sampling period. The canister with closed valve was disconnected from the
dilution sampling system and capped; chain of custody documentation was initiated.
Extraction was performed on-site for the denuders, according to the scheme described
above. The denuders were rinsed with a mixture of methylene chloride:acetone:hexane in a
volume ratio of 2:3:5. The solvent mixture was added to the denuder, and the denuder tube
was capped and shaken 4 times. An internal standard was added to the first extraction. The
rinses were combined in a precleaned glass jar for paired denuders; the jar was labeled and
sealed with Teflon tape. Chain of custody documentation was initiated for the extract, and
the jar was stored over ice. After extraction, the denuders and caps were dried using high
purity nitrogen and capped until ready for re-use.
Denuders, PUF modules, and filters were all bagged and stored over ice. Canisters and
carbonyl tubes were transported to the ERG laboratory for analysis, and the filters, PUF
modules, and denuder extracts were transported to the EPA laboratory for analysis.
50
-------
Laboratory Experimental Methodology
Components of the sample collection arrays, filters, DNPH-impregnated silica gel tubes
used to sample carbonyl compounds, and canisters used to sample volatile organic com-
pounds were returned for analysis to EPA and ERG laboratories (see Table 3-1 for respon-
sible laboratory). The analyses described in the following sections were performed with
the analytical methodology used by the respective laboratories summarized in Table 3-1.
PM2 5 Mass
Teflon membrane (Gelman Teflo) filters of 2 |im pore diameter were used to collect
fine PM samples for mass determinations. Filters before and after sample collection were
maintained at 20-23 °C and a relative humidity of 30-40% for a minimum of 24 hours
prior to weighing on a micro-balance. Sample mass was determined by gravimetric
analysis before and after sample collection.
Elemental Analysis
Individual elements above atomic number 9 (fluorine) were measured using a Philips
Model 2404, wavelength-dispersive, X-ray fluorescence (XRF) spectrometer running the
UniQuant program. This program gives qualitative and quantitative information on the
elements present on a Teflon membrane filter. The filter to be analyzed was covered with a
0.4 |im thick Prolene film which was attached using glue. The glue was only on the outer
rim of the filter and did not interfere with the analysis. Only elements which gave amounts
greater than one standard error above the detection limit were reported.
Water-Soluble Inorganic Ions
Teflon filter samples were analyzed for major inorganic anions and cations using a
Dionex DX-120 ion chromatograph equipped with a 25 |iL sample loop and a conductivity
detector. Major ions determined were chloride, nitrate, sulfate, calcium, magnesium,
potassium, and ammonium. Prior to extraction, the filters were wetted with 350-500 |iL of
ethanol. Two sequential extractions with HPLC-grade water were performed using mild
sonication of the filters followed by filtration of the extracts. The two extracts were
combined for analysis.
Anions were separated using an Ion Pac AS14 (4x250 mm) column with an alkyl
quaternary ammonium stationary phase and a carbonate-bicarbonate mobile phase. Cations
were separated using an Ion Pac CS12 (4x250 mm) column with an 8 |im
poly(ethylvinylbenzene-divinylbenzene) macroporous substrate resin functionalized with a
51
-------
relatively weak carboxylic acid stationary phase and a sulfuric acid mobile phase. Ion
concentrations were determined from four-point calibration curves using an external
standard method. All samples were extracted and analyzed in duplicate or triplicate.
Elemental Carbon/Organic Carbon
Elemental carbon (EC) and organic carbon (OC) content of PM samples collected on
pre-fired quartz filters was determined by NIOSH Method 50408 using a Sunset
Laboratory thermal evolution instrument. In this method, a 1.0x1.5 cm punch of the quartz
filter sample is placed in the instrument, and organic and carbonate carbon are evolved in a
helium atmosphere as the temperature is raised to 850 °C. Evolved carbon is catalytically
oxidized to C02 in a bed of granular Mn02, then reduced to methane in a methanator.
Methane is subsequently quantified by a flame ionization detector (FID). In a second stage,
the sample oven temperature is reduced, an oxygen-helium mixture is introduced, and the
temperature is increased to 940 °C. With the introduction of oxygen, pyrolytically
generated carbon is oxidized, and the transmittance of a laser light beam through the filter
increases. The point at which the filter transmittance reaches its initial value is defined as
the split between OC and EC. Carbon evolved prior to the split is considered OC
(including carbonate), and carbon volatilized after the split is considered elemental (EC).
Elemental carbon evolved is similarly oxidized to C02 and subsequently reduced to
methane to be measured by the FID.
Organic Compounds
Individual organic compounds present in the fine PM collected on prefired quartz
filters were determined by extracting the filters with hexane (two extractions) followed
with a 2:1 mixture by volume of benzene and isopropanol (three extractions). Prior to
extraction, the filters were composited as necessary to achieve a total of approximately 0.5
mg of OC and spiked with a mixture of deuterated internal recovery standards. These
standards were selected to represent the range of expected solubilities, stabilities,
chromatographic retention times, and volatilities of organic compounds present in the
samples. All extracts from the five extraction steps were combined and concentrated using
an automated nitrogen blow-down apparatus.
An aliquot of the combined extract was derivatized with diazomethane to yield methyl
esters of any fatty acids which might be present. After the methylation reaction was
complete, the methylated extract aliquot was reconcentrated by nitrogen blow down. A
separate portion of the methylated extract was derivatized a second time using bis
(trimethylsilyl) trifluoroacetamide-N,0-bis (trimethylsilyl) acetamide (Sylon BFT) reagent
52
-------
to convert compounds such as levoglucosan and cholesterol to their trimethylsilyl (TMS)
derivatives. Both derivatizations were performed in order to allow the compounds to be
separated and eluted from a gas chromatograph column. Since the TMS derivatives are
somewhat unstable over time, the silylation was carried out just prior to analysis.
Gas chromatography coupled with a GC/MS was used to identify and quantify the
individual organic compounds present in the extracts. A Hewlett-Packard 6890 GC
equipped with an HP 5973 mass spectrometer detector was used. A 5MS column (30 m,
0.25 mm diameter, 0.25 |im film thickness) was employed along with an injector
temperature of 65 °C and a GC/MS interface temperature of 300 °C. The initial GC oven
temperature was set at 65 °C with an initial hold time of 10 min. The oven temperature was
then ramped upward at 10 °C/min to 300 °C and held at the upper temperature for an
additional 41.5 minutes. Helium was used as the carrier gas (1 mL/min) and the GC was
operated in the split/splitless mode.
Positive identification of target compounds was obtained by comparing mass spectra of
the analytes with those obtained from over 100 authentic compound standards. Iso- and
anteiso-alkanes were identified using secondary standards derived from paraffin candle
wax. Additional compounds were identified as "probable" based on a comparison of the
GC retention times and mass spectra with commercially available spectral libraries.
Quantification of the individual compounds involved referencing each compound against
one or more of the deuterated internal standards spiked into the sample to correct for losses
of the analytes which may have occurred in the compositing, extracting, concentrating, and
derivatizing steps. An extensive set of standards of target compounds at known
concentrations, which also included the deuterated internal standard compounds, was used
to establish 3-point or 5-point calibration curves from which the concentrations of the
analytes were determined.
Carbonyl Compounds
Sep-Pak chromatographic-grade silica gel cartridges impregnated with DNPH were
used in series for carbonyl sample collection. The tubes were used in series to check for
compound breakthrough. Following sample collection in the field, the cartridges and
accompanying chain of custody documentation were transported to the ERG laboratory,
where they were logged into the laboratory sample tracking system. The cartridges were
extracted and analyzed for the compounds listed in Table 4-9 using EPA Compendium
Method TO-11 A,9 "Determination of Formaldehyde in Ambient Air Using Adsorbent
Cartridge Followed by High Performance Liquid Chromatography (HPLC)." The
53
-------
analytical instalment was a Varian 5000 HPLC with a multiwavelength detector operated
at 360 nm. The HPLC was configured with a 25 cm, 4.6 mm id, C18 silica analytical
column with a 5-|im particle size. An automatic sample injector was used to inject 25 |iL
aliquots into the HPLC.
Table 4-9. Carbonyl Compounds Analyzed by High Performance Liquid
Chromatography: Method Detection Limits
Compound
CAS No.
Method Detection Limits
(M-g)
Formaldehyde
50-00-0
0.0838
Acetaldehyde
75-07-0
0.0916
Acetone
67-64-1
0.0428
Propionaldehyde
123-38-6
0.0934
Crotonaldehyde
4170-30-3
0.1283
Butyraldehyde
123-72-8
0.0956
Benzaldehyde
100-52-7
0.0959
Isovaleraldehyde
590-86-3
0.1076
Valeraldehyde
110-62-3
0.1758
o-Tolualdehyde
529-20-4
0.1439
m-Tolualdchydc
620-23-5
0.1439
/;-Tolualdehvde
104-87-0
0.1439
Hexaldehyde
66-25-1
0.1377
2,5 -Dimethylbenzaldehy de
5779-94-2
0.1337*
Diacetyl
432-03-8
0.0154*
Methacrolein
78-85-3
0.0125*
2-Butanone
78-93-3
0.0125*
Glyoxal
107-22-2
0.0412*
Acetophenone
98-86-2
0.0250*
Methylglyoxal
78-98-8
0.0244*
Octanal
124-13-0
0.0100*
Nonanal
124-19-6
0.0182*
*Estimated value
The chromatography data acquisition system was used to retrieve data from the HPLC.
The data were processed, and peak identifications were made using retention times and
relative retention times determined by analysis of analytical standards. After peak
identifications were made, the concentration of each target analyte was determined using
individual response factors for the carbonyl compounds.
54
-------
Daily calibration checks were performed to ensure that the analytical procedures were
in control. Daily quality control checks were performed after every 10 samples on the days
that samples were analyzed, with compound responses within ±15% relative to the
responses from the current calibration curve. Compound retention time drifts were also
measured from the analysis of the quality control check sample and tracked to ensure that
the HPLC was operating within acceptable parameters.
As part of the daily quality control check, a second injection of the quality control
standard was performed if the analysis of the daily quality control sample was not
acceptable. If the second quality control check did not meet acceptance criteria or if more
than one daily quality control check did not meet acceptance criteria, a new calibration
curve (at five concentration levels) was produced. All samples analyzed with the
unacceptable quality control checks would be re-analyzed.
An acetonitrile system blank was analyzed after the daily calibration check and before
sample analysis. The system was considered in control if target analyte concentrations
were less than the current method detection limits.
Canister Analyses: Air Toxics and Speciated Nonmethane
Organic Compounds
The combined analysis for gas-phase air toxics and speciated nonmethane organic
compounds was performed on a gas chromatograph (GC)/flame ionization
detector(FID)/mass selective detector (MSD). A Hewlett-Packard 5971 MSD and a
Hewlett-Packard 5890 Series II GC with a 60-m by 0.32-mm id and a 1 |im film thickness
J&W DB-1 capillary column followed by a 2:1 splitter was used to send the larger portion
of the column effluent to the MSD and the smaller fraction to the FID. The chromatograph
oven containing the DB-1 capillary column was cooled to -50 °C with liquid nitrogen at
the beginning of the sample injection. This temperature was held for 5 min and then
increased at the rate of 15 °C/min to 0 °C. The oven temperature was then ramped at 6
°C/min to 150 °C, then ramped at 20 °C/min to 225 °C and held for 8 min. The gas eluting
from the DB-1 capillary column passed through the 2:1 fixed splitter, to divide the flow
between the MSD and the FID.
The air toxics analysis was performed according to the procedures of EPA
Compendium Method TO-15,9 "Determination of Volatile Organic Compounds (VOCs) in
Air Collected in Specially-Prepared Canister and Analyzed by Gas Chromatography/Mass
Spectrometry (GC/MS)." The analysis of speciated nonmethane organic compounds was
55
-------
performed according to the procedures of "Technical Assistance Document for Sampling
and Analysis of Ozone Precursors."10 Method detection limits11 (MDLs) for air toxics are
shown in Table 4-10 and for the speciated nonmethane organic compounds in Table 4-11.
Table 4-10. Detection Limits for Air Toxics Compounds (Analytical Method
TO-15)
Target Compounds*
CAS No.
Method Detection Limit
(Hg/m3)
Acetylene
74-86-2
0.24
Propylene
115-07-1
0.17
Dichlorodifluoromethane
75-71-8
0.40
Chloromethane
74-87-3
0.24
Dichlorotetrafluoroethane
1320-37-2
0.70
Vinyl chloride
75-01-4
0.31
1,3-Butadiene
106-99-0
0.31
Bromomethane
74-83-9
0.70
Chloroethane
75-00-3
0.42
Acetonitrile
75-05-8
0.84
Acetone
67-64-1
1.23
Trichlorofluoromethane
75-69-4
0.45
Acrylonitrile
107-13-1
0.91
1,1 -Dichloroethene
75-35-4
0.79
Methylene chloride
75-09-2
0.42
Trichlorotrifluoroethane
26523-64-8
1.07
trans-1,2-Dichloroethylene
56-60-5
0.47
1,1 -Dichloroethane
75-34-3
0.65
Methyl tert-butyl ether
1634-04-1
1.29
Methyl ethyl ketone
78-93-3
0.88
Chloroprene
126-99-8
0.73
cis-1,3 -Dichloroethylene
156-59-2
0.79
Bromochloromethane
74-97-5
1.26
Chloroform
67-66-3
0.49
Ethyl tert-butyl ether
637-92-3
1.25
1,2-Dichloroethane
107-06-2
0.48
1,1,1 -Trichloroethane
71-55-6
0.65
Benzene
71-43-2
0.25
Carbon tetrachloride
56-23-5
1.01
tert-Amyl methyl ether
994-05-8
1.00
(continued)
56
-------
Table 4-10. (Continued)
Target Compounds*
CAS No.
Method Detection Limit11
(ng/m3)
1,2-Dichloropropane
78-87-5
0.65
Ethyl acrylate
140-88-5
1.31
Bromodichloromethane
75-27-4
0.80
T richloroethylene
79-01-6
0.75
Methyl methacrylate
80-62-6
1.47
cis-1,2-Dichloropropene
10061-01-5
0.82
Methyl isobutyl ketone
108-10-1
1.36
trans-1,2-Dichloropropene
10061-02-6
1.00
1,1,2-Trichloroethane
79-00-5
0.65
Toluene
108-88-3
0.45
Dibromochloromethane
124-48-1
1.36
1,2-Dibromoethane
106-93-4
1.23
«-Octanc
111-65-9
0.56
T etrachloroethylene
127-18-4
0.81
Chlorobenzene
108-90-7
0.55
Ethylbenzene
100-41-4
0.35
m-, /^-Xylene
108-38-3/106-42-3
0.87
Bromoform
75-25-2
1.65
Styrene
100-42-5
0.59
1,1,2,2-Tetrachloroethane
79-34-5
0.82
o-Xylene
95-47-6
0.43
1,3,5 -T rimethy lbenzene
108-67-8
0.69
1,2,4-Trimethylbenzene
95-63-6
0.69
m-Dichlorobcnzcnc
541-73-1
0.60
Chloromethylbenzene
100-44-7
0.72
/;-Dichlorobenzene
106-46-7
1.08
o-Dichlorobenzene
95-50-1
0.72
1,2,4-Trichlorobenzene
120-82-1
0.89
Hexachloro-1,3-butadiene
87-68-3
1.28
a MDLs are instrument detection limits based on Fed. Reg., 1984. MDLs reported here are based on nominal
injection volume of 200 mL of gas.
57
-------
Table 4-11. Detection Limits for Speciated Nonmethane Organic
Compounds
Compound
CAS No.
Method Detection Limits
(l^g/m3)
Ethylene
Acetylene
Ethane
Propylene
Propane
Propyne
Isobutane
Isobutene/1 -butene
1,3-Butadiene
Butane
/ram-2-Butcnc
C7.V-2-Butene
3 -Methyl-1 -butene
Isopentane
1-Pentene
2-Methyl-1 -butene
/7-Pentane
Isoprene
/ram-2-Pentene
c7.v-2-Pcntene
2-Methyl-2-butene
2.2-Dimethylbutane
Cyclopentene
4-Methyl-1 -pentene
Cyclopentane
2.3-Dimethylbutane
2-Methylpentane
3-Methylpentane
2-Methyl-1 -pentene
1-Hexene
2-Ethyl-l-butene
/7-Hexane
fra«s-2-Hexene
c7.v-2-Hcxene
74-85-1
74-86-2
74-84-0
115-07-1
74-98-6
74-99-7
75-28-5
115-11-7/106-98-0
106-99-0
106-97-8
624-64-6
590-18-1
563-45-1
78-78-4
109-67-1
563-46-2
109-66-0
78-79-4
646-04-8
627-20-3
513-35-9
75-83-2
142-29-0
691-37-2
287-92-3
79-29-8
107-83-5
96-14-0
763-29-1
592-41-6
760-21-4
110-54-3
4050-45-7
7688-21-3
0.50
0.47
0.54
0.44
0.46
0.42
0.43
0.21
0.40
0.43
0.42
0.42
0.32
0.33
0.32
0.45
0.33
0.31
0.33
0.33
0.32
0.46
0.31
0.45
0.32
0.46
0.46
0.46
0.46
0.46
0.45
0.46
0.46
0.46
(continued)
58
-------
Table 4-11. (continued)
Compound
CAS No.
Method Detection Limits
Milll
Methylcyclopentane
2,4-Dimethylpentane
Benzene
Cyclohexane
2-Methylhexane
2,3 -Dimethylpentane
3-Methylhexane
1-Heptene
2,2,4-Trimethylpentane
^-Heptane
Methylcyclohexane
2.2.3-Trimethylpentane
2.3.4-Trimethylpentane
Toluene
2-Methylheptane
3-Methylheptane
1-Octene
/7-Octane
Ethylbenzene
m-, /^-Xylene
Styrene
0-Xylene
1-Nonene
«-Nonane
Isopropylbenzene
a-Pincnc
/7-Propylbenzene
m-Ethyltolucnc
/;-Ethyltoluene
1,3,5 -T rimethylbenzene
0-Ethyltoluene
P-Pinene
1,2,4-Trimethylbenzene
1-Decene
«-Decane
96-37-7
108-08-7
71-43-2
110-82-7
591-76-4
565-59-3
589-34-4
592-76-7
540-84-1
142-82-5
108-87-2
564-02-3
565-75-3
108-88-3
592-27-8
589-81-1
111-66-0
111-65-9
100-41-4
108-38-3/106-42-3
100-42-5
95-47-6
124-11-8
111-84-2
98-82-8
80-56-8
103-65-1
620-14-4
622-96-8
108-67-8
611-14-3
127-91-3
95-63-6
872-05-9
124-18-5
0.45
0.35
0.42
0.45
0.40
0.40
0.40
0.39
0.51
0.40
0.39
0.51
0.51
0.37
0.51
0.51
0.50
0.51
0.52
0.47
0.46
0.47
0.40
0.41
0.38
0.39
0.38
0.38
0.38
0.38
0.38
0.39
0.38
0.33
0.33
(continued)
59
-------
Table 4-11. (continued)
Compound
CAS No.
Method Detection Limits
(ng/m3)
1,2,3 -Trimethylbenzene
526-73-8
0.38
m-Dicthylbcnzcnc
141-93-5
0.32
/;-Dicthylbcnzcnc
105-05-5
0.32
1-Undecene
821-95-4
0.49
/7-Undccanc
1120-21-4
0.50
1-Dodecene
112-41-4
0.49
/7-Dodccanc
112-40-3
0.50
1-Tridecene
2437-56-1
0.49
/7-Tridccanc
629-50-5
0.50
Particle Size Distribution Data
The ELPI was operated and collected data during all three test days. Data were
reduced using the Dekati software package.
60
-------
Results and Discussion
Analyses were performed in either EPA or ERG laboratories according to the scheme
shown in Table 3-1, using the analytical procedures described in Section 4. Results of
these analyses are discussed in this section.
Sampling the smelt tank vent presented a special challenge because the moisture
content of the vented gas was very high (36%). With the vent gas temperature at 156 °F
(69 °C), the vent gas was super-saturated with water. Under these high moisture
conditions, it is feasible that water could condense in the sampling system as the vent gas
is cooled and diluted. The likelihood of condensation occurring in the sampler is further
enhanced as the temperature of the ambient dilution air decreases and the relative humidity
(RH) of the ambient dilution air increases.
Were moisture condensation to occur in the sampler, the condensation process would
tend to remove some water-soluble species along with some particles and would result in a
negative bias in the measured emission factors and concentrations. This condensation
effect could also vary from test to test depending especially on the temperature and RH of
the air used to dilute the sampled stack gas. For a vent gas of 36% moisture, psychometric
calculations indicate that the dew point temperature of the vent gas, when diluted by a
factor of 40 (as was done in this study), would vary from 10 to 30 °C as the dilution air
temperature and RH ranged from 15 to 30 °C and 20% to 80% RH, respectively.
Rain occurred during the first test day of the smelt tank vent sampling campaign, and
the temperature and RH of the dilution air were 18.2 °C and 100%, respectively—condi-
tions at which the dew point of the diluted vent gas was exceeded—and condensation of
moisture in the sampler did occur. During the second and third test days,.the temperature
and RH of the dilution air were 21.4 °C, 39% RH and 14.5 °C, 45.4% RH, respectively.
Therefore, condensation should not have occurred on Test Day 2, and the possibility of
condensation occurring was marginal on Test Day 3. Stripping of any moisture that may
have collected in the activated carbon bed of the dilution air-conditioning system during
Test Day 1 could also have increase the possibility of some condensation occurring in the
sampler during Test Days 2 and 3.
61
-------
That condensation in the sampler occurred on Test Day 1 was confirmed by water
blinding the 8x 10 in. filter located at the exit of the dilution sampler during the test run.
This test was interrupted during the planned sampling duration in order to replace the
filter. This difficulty occurred only during the first of the three days of testing. Because of
the wet sampling conditions on Test Day 1, it is likely that the PM mass emission factor
measured for that day is biased low. In fact, Test Day 1 at the smelt tank vent gave the
lowest PM2 5 mass emission factor for all three tests.
Because of the high moisture level of the ambient air on Test Day 1, it was also a
concern that the activated carbon bed used to scrub organic contaminants from the ambient
dilution air would become saturated with water, thereby decreasing its capacity. Had that
saturation of the carbon bed occurred, a higher level of organic contaminants in the
collected dilution air samples would be expected. However, higher levels of organic
compounds (carbonyls, air toxies, and nonmethane organics) were not observed in the
cleaned dilution air for any of the tests, indicating that the capacity of the carbon bed had
not been exceeded and that saturation of the carbon bed was not a problem.
Calculated Emission Factors for PM Mass, Carbonyl
Compounds, and Nonmethane Organic Compounds
Emissions of PM mass, carbonyl compounds, and nonmethane organic compounds are
reported in Table 5-1. Emission factors reported for the smelt tank vent are based on the
amount of black liquor and distillate #2 oil fuels fed to the recovery boiler that generated
the smelt. Results reported in Table 5-1 show the following:
• Approximately 47% of the PM2 5 mass was chemically characterized on average.
Most of the PM25 mass was inorganic ions (i.e., 14.3% sodium, 13.4% thiosulfate,
12.2%) sulfate, 4.9%> potassium, and 1.8% chloride). Organic carbon averaged about
4% and elemental carbon only about 0.2% by weight of the PM2 5 mass.
• The average emission factor for total carbonyl compounds was skewed to a higher
value due to a particularly large concentration of acetone observed in the emissions
on the first test day. Acetone levels for the second and third test days were only 10%
and 5%>, respectively, of the level found on the first test day. This high level of
acetone on Test Day 1 has been attributed to contamination arising from the solvents
used in the recovery of the denuders; emission factors have been recalculated with
the deletion of this artifactually high value.
• Emission factors for speciated and total nonmethane organic compounds are very
nearly the same, indicating that little unidentified hydrocarbon material was
observed on the test days. The high //-hexane artifact on the first test day, due to
62
-------
contamination from the denuder solvents, has been deleted, and values have been
recalculated.
Example calculations for the emission factors are shown in Appendix C. Supporting
data for emission factor calculation for PM2 5, carbonyl compounds, and nonmethane
organic compounds are shown in Appendices D, E, and F, respectively.
Table 5-1. Fine Particle, Carbonyl, and Nonmethane Organic Compound
Emission Factors from a Smelt Tank Vent at a Pulp and Paper Facility
12/14/01 12/15/01 12/16/01 Mean Uncertainty
PM2 5 Mass Emission Factor (mg/kg fuel
4.28
8.08
6.13
6.16
1.90
burned)
Speciated Carbonyl Compounds Mass
0.13
0.51
0.19
0.28
0.20
Emission Factor (mg/kg fuel burned)
Total (Speciated + Unspeciated)
0.17
0.58
0.21
0.32
0.23
Carbonyl Compounds Mass Emission
Factor (mg/kg fuel burned)
Speciated NMOC Mass Emission Factor
0.37
0.81
0.32
0.50
0.27
(mg/kg fuel burned)
Total (Speciated + Unspeciated) NMOC
0.88
1.42
0.76
1.02
0.35
Mass Emission Factor (mg/kg fuel
burned)
Gas-Phase Carbonyl Compounds
Analytical results for the carbonyl field samples for each of the three test days are
shown in Table 5-2. Results of the analysis are reported for the difference of the combined
total of three DNPH-impregnated silica gel tubes for the residence chamber (RC) and the
combined total of three DNPH-impregnated silica gel tubes for the dilution air (DA). The
"Total Unspeciated" entry at the bottom of the table is the total mass (all three tubes com-
bined) of the compounds characterized as carbonyl compounds but not identified as a
specific compound because no analytical standard was available. The "Total (Speciated +
Unspeciated)" entry includes the total mass (all three tubes combined) of both specifically
identified carbonyl compounds and unspeciated carbonyl compounds.
The significant portion of the reported results for carbonyl compounds is the speciated
compounds. Mass fractions reported are calculated by dividing the mass of an individual
species by the total mass of speciated plus unspeciated carbonyl compounds. Concentra-
tions of the unspeciated carbonyl compounds are based on the calibration factor for
formaldehyde; the concentrations of the unspeciated carbonyls for this sequence of tests
63
-------
Table 5-2. Carbonyl Compounds, Smelt Tank Vent: Carbonyl Compounds Collected in Dilution Air Subtracted
from Carbonyl Compounds Collected in Residence Chamber, Mass Fraction, Mean and Uncertainty
Field
Carbonyls
RC-DA
12/14/01
(M-g)
Mass
Carbonyls
RC-DA
12/15/01
0*g)
Mass
Carbonyls
RC-DA
12/16/01
(ng)
Mass
Mean
Compound
CAS No.
Blank
(H-g)
Fraction
(M-g)
Fraction
(Hg)
Fraction
(Hg)
Mass
Fraction
Uncertainty
formaldehyde
50-00-0
0.0230
0.0500
0.0089
0.1910
0.0095
0.0490
0.0053
0.0079
0.0023
acetaldehyde
75-07-0
0.0210
2.2730
0.4053
5.8270
0.2908
2.7380
0.2966
0.3309
0.0645
acetone
67-64-1
0.2280
ND
ND
6.8190
0.3404
3.3250
0.3602
0.2335
0.2025
propionaldehyde
123-38-6
ND
0.0770
0.0137
0.3400
0.0170
0.1170
0.0127
0.0145
0.0022
crotonaldehyde
4170-30-0
ND
ND
ND
0.0110
0.0005
ND
ND
0.0002
0.0003
butyr/isobutyraldehyde
123-72-8
0.0840
0.2620
0.0467
0.6580
0.0328
0.2120
0.0230
0.0342
0.0119
benzaldehyde
100-52-7
ND
0.0080
0.0014
0.0860
0.0043
0.0030
0.0003
0.0020
0.0020
isovaleraldehyde
590-86-3
ND
ND
ND
ND
ND
ND
ND
ND
ND
valeraldehyde
110-62-3
ND
0.0450
0.0080
0.0920
0.0046
0.0520
0.0056
0.0061
0.0018
o-tolualdehyde
529-20-4
ND
ND
ND
ND
ND
0.0090
0.0010
0.0003
0.0006
/w-tolualdehyde
620-23-5
ND
ND
ND
ND
ND
ND
ND
ND
ND
/Moliialdchvdc
104-87-0
ND
0.0130
0.0023
0.0880
0.0044
0.0110
0.0012
0.0026
0.0016
hexaldehyde
66-25-1
0.0280
0.0220
0.0039
0.0510
0.0025
0.0340
0.0037
0.0034
0.0007
2,5-dimethylbenzaldehyde
5779-94-2
ND
ND
ND
0.0210
0.0010
ND
ND
0.0003
0.0006
diacetyl
431-03-8
ND
ND
ND
ND
ND
ND
ND
ND
ND
methacrolein
78-85-3
ND
ND
ND
ND
ND
ND
ND
ND
ND
2-butanone
78-93-3
0.0050
1.5030
0.2.680
3.3960
0.1695
1.7620
0.1909
0.2095
0.0518
glyoxal
107-22-2
0.0820
0.0750
0.0134
0.0060
0.0003
ND
ND
0.0046
0.0076
acetophenone
98-86-2
ND
0.0300
0.0053
0.0410
0.0020
ND
ND
0.0025
0.0027
methylglyoxal
78-98-8
0.0490
0.0010
0.0002
ND
ND
0.0140
0.0015
0.0006
0.0008
octanal
124-13-0
ND
0.0040
0.0007
0.0290
0.0014
ND
ND
0.0007
0.0007
nonanal
124-19-6
0.1070
0.0060
0.0011
0.0620
0.0031
0.0200
0.0022
0.0021
0.0010
(Continued)
-------
Table 5-2. (Continued)
Field
Carbonyls
RC-DA
37239
(M-g)
Mass
Carbonyls
RC-DA
37240
(Hg)
Mass
Carbonyls
RC-DA
37241
(ng)
Mass
Mean
Compound CAS No.
Blank
(H-g)
Fraction
(M-g)
Fraction
(Hg)
Fraction
(Hg)
Mass
Fraction
Uncertainty
Sum, Speciated
0.6270
4.3690
0.7791
17.7180
0.8844
8.3460
0.9041
Sum, Unspeciated
0.8850
1.2385
0.2209
2.3170
0.1156
0.8850
0.0959
Total (Speciated + Unspeciated)
1.5120
5.6075
20.0350
9.2310
Mean
Uncertainty
Mass Emission Factor, mg/kg fuel burned
(Speciated)
0.1287
0.5108
0.1922
0.28
0.20
Mass Emission Factor, mg/kg fuel burned
(Total)
0.1652
0.5776
0.2126
0.32
0.23
-------
averages 10% or less of the concentration of speciated carbonyl compounds. On Test
Day 1, the relatively high value for concentration of carbonyl compounds is driven by the
concentration of acetone in the emissions. Acetone is not observed at any significant level in
the blank or in the dilution air. Since acetone is one of the solvents used for desorption of
the denuders, the presence of acetone was attributed to contamination from the denuders.
Supporting data showing results for each individual carbonyl sampling tube are included in
Appendix E.
Gas-Phase Air Toxics Whole Air Samples
Analytical results for the air toxics canister samples are shown in Table 5-3. The ERG
concurrent analysis produces analytical results for both air toxics and speciated/non-
speciated nonmethane organic compounds; these results are presented separately.
Table 5-3. Summarized Analytical Results for Air Toxics Compounds
Observed at the Smelt Tank Vent on Each of the Three Test Days
Compounds CAS No. Ambient Air Toxics Air Toxics Air Toxics
12/14/01 RC-DA RC-DA RC-DA
(|ig/m3) 12/14/01 12/15/01 12/16/01
(Hg/m3) (ng/m3) (ng/m3)
acetylene
74-86-2
0.49
ND
1.81
ND
dichlorodifluoromethane
75-71-8
2.94
ND
ND
ND
chloromethane
74-87-3
1.43
ND
ND
ND
trichlorofluoromethane
75-69-4
1.40
ND
ND
ND
methylene chloride
75-09-2
159.71
153.51
9.50
ND
trichlorotrifluoroethane
26523-64-8
0.30
ND
ND
ND
methyl ethyl ketone
78-93-3
ND
12.67
ND
ND
benzene
71-43-2
1.10
0.80
7.56
ND
carbon tetrachloride
56-23-5
0.21
ND
ND
ND
toluene
108-88-3
1.28
3.46
6.85
2.68
ethylbenzene
100-41-4
0.25
ND
ND
ND
m-, />-xylene
108-38-3/106-42-3
0.60
ND
ND
ND
o-xylene
95-47-6
0.19
ND
ND
ND
1,2,4-trimethylbenzene
95-63-6
0.44
ND
ND
ND
Method detection limits for the air toxics compounds are shown in Table 4-10, with
values typically ranging from 1 |ig/m3 and lower. Most of the observed values in the field
66
-------
samples are at the lower end of the calibration curve for this analysis, with a small number
of air toxics compounds actually observed; analytical results are shown in Table 5-3.
Analytical results for an ambient canister taken at the test location, on Test Day 1 only, are
included for reference. More of the air toxics compounds are observed in the ambient
sample than are observed in the smelt tank vent emissions. For nearly all of the air toxics
compounds, the values observed in the ambient air are higher than the values observed in
the stack emissions. The single compound observed in highest concentration in the ambient
sample is methylene chloride, at a concentration of approximately 160 |ig/m3. On Test Day
1, the same day that the ambient sample was taken at the test location, the concentration of
methylene chloride in the smelt tank vent emissions is nearly the same as the concentration
of methylene chloride in the ambient sample. For the first test day, where the high value of
methylene chloride is observed in the emissions, the concentration of methylene chloride in
the dilution air is low (less than approximately 10 |ig/m3). On the second test day, the
concentration of methylene chloride in the emissions decreases (approximately 10 |ig/m3),
with the residence chamber value at approximately 30 |ig/m3 and the dilution air at
approximately 20 |ig/m3. On the third test day, methylene chloride was not detected in the
smelt tank vent emissions; a very slightly higher value (approximately 4 |ig/m3) for
methylene chloride was observed in the dilution air than in the residence chamber
(approximately 3 |ig/m3). The high value of methylene chloride in the ambient sample and
on Test Day 1 has been attributed to contamination arising from the solvents used to desorb
the denuders and has been deleted as an artifact. Supporting data for the air toxics analyses
are shown in Appendix G, with analytical results for each individual canister.
Gas-Phase Speciated Nonmethane Organic Compounds
Analysis of whole air samples of dilution air and residence chamber air using ERG's
concurrent analysis generated analytical data for speciated nonmethane organic compounds
(SNMOC), as well as unspeciated NMOC, and analytical results are reported in Table 5-4.
Mass emission factors for total SNMOC and total (speciated plus unspeciated) NMOC are
also provided. Results are reported as the difference between residence chamber samples
and dilution air. The dilution air is used to dilute the stationary source matrix; samples of the
dilution air have not been exposed to the stationary source matrix. The residence chamber
air represents the sample of the diluted stationary source matrix collected at the end of the
residence chamber. Concentrations of NMOC (both speciated and unspeciated) observed in
the source samples were negligible. Although «-hexane was observed at elevated levels in
the ambient air (approximately 140 |ig/m3), levels of //-hexane observed in the source
samples are negligible. Supporting data for the NMOC analysis are shown in Appendix F.
67
-------
Table 5-4. Speciated and (Speciated + Unspeciated) NMOC Data for all Three Test Days
SNMOC
SNMOC
SNMOC
Mean
Mass
Fraction
Compound
CAS No.
RC-DA
12/14/01
Mass
Fraction
RC-DA
12/15/01
Mass
Fraction
RC-DA
12/16/01
Mass
Fraction
Uncertainty
(us)
(us)
(US)
ethylene
74-85-1
0.0057
0.0139
0.0399
0.0768
0.0055
0.0154
0.0354
0.0359
acetylene
74-86-2
0.0008
0.0020
0.0097
0.0186
ND
ND
0.0069
0.0102
ethane
74-85-1
0.0065
0.0159
0.0363
0.0698
0.0068
0.0190
0.0349
0.0303
propylene
115-07-1
0.0040
0.0098
0.0154
0.0297
0.0043
0.0120
0.0172
0.0109
propane
74-98-6
0.0111
0.0273
0.0172
0.0331
0.0045
0.0125
0.0243
0.0106
propyne
74-99-7
ND
ND
ND
ND
ND
ND
ND
ND
isobutane
75-28-5
0.0016
0.0040
0.0034
0.0065
0.0012
0.0035
0.0047
0.0016
isobutene/1 -butene
115-11-7/106-98-0
0.0014
0.0034
0.0073
0.0140
0.0013
0.0036
0.0070
0.0061
1,3-butadiene
106-99-0
ND
ND
0.0023
0.0045
ND
ND
0.0015
0.0026
//-butane
106-97-8
0.0041
0.0102
0.0073
0.0140
0.0034
0.0095
0.0112
0.0025
trans-2-butene
624-64-6
0.0021
0.0052
0.0034
0.0065
0.0023
0.0064
0.0060
0.0008
c/.v-2-butcnc
590-18-1
0.0023
0.0058
0.0039
0.0075
0.0029
0.0079
0.0071
0.0012
3 -methyl-1 -butene
563-45-1
ND
ND
ND
ND
ND
ND
ND
ND
isopentane
78-78-4
0.0040
0.0099
0.0051
0.0098
0.0035
0.0098
0.0098
0.0001
1-pentene
109-67-1
0.0019
0.0047
0.0031
0.0059
0.0025
0.0070
0.0059
0.0012
2-methyl-1 -butene
563-46-2
ND
ND
ND
ND
ND
ND
ND
ND
//-pcntanc
109-66-0
0.0024
0.0059
0.0038
0.0073
0.0034
0.0096
0.0076
0.0018
isoprene
78-79-4
0.0029
0.0071
0.0030
0.0059
0.0031
0.0087
0.0072
0.0014
;ra«.v-2-pcntcne
646-04-8
ND
ND
ND
ND
ND
ND
ND
ND
c/.v-2-pcntcnc
627-20-3
0.0026
0.0063
0.0032
0.0062
0.0040
0.0112
0.0079
0.0028
2 -me thy 1-2 -butene
513-35-9
ND
ND
ND
ND
ND
ND
ND
ND
2,2-dimethylbutane
75-83-2
0.0039
0.0097
0.0048
0.0092
0.0051
0.0141
0.0110
0.0027
cyclopentene
142-29-0
ND
ND
ND
ND
ND
ND
ND
ND
(Continued)
-------
Table 5-4. (Continued)
Compound
CAS No.
SNMOC
RC-DA
12/14/01
(US)
4-methyl-1 -pentene
691-37-2
ND
cyclopentane
287-92-3
ND
2,3 -dimethylbutane
79-29-8
0.0055
2-methylpentane
107-83-5
0.0026
3-methylpentane
96-14-0
0.0072
2-methyl-1 -pentene
763-29-1
ND
1-hexene
592-41-6
0.0046
2-ethyl-l-butene
760-21-4
ND
«-hexane
110-54-3
0.1105
/ra/?.v-2-hcxcnc
4050-45-7
ND
67.v-2-hc.xcnc
7688-21-3
ND
methylcyclopentane
96-37-7
0.0193
2,4-dimethylpentane
108-08-7
0.0039
benzene
71-43-2
0.0054
cyclohexane
110-82-7
0.0046
2-methylhexane
591-76-4
ND
2,3 -dimethylpentane
565-59-3
0.0076
3-methylhexane
589-34-4
0.0022
1-heptene
592-76-7
0.0057
2,2,4-trimethylpentane
540-84-1
0.0029
//-heptane
142-82-5
0.0030
methylcyclohexane
108-87-2
0.0023
2,2,3 -trimethylpentane
564-02-3
ND
SNMOC
SNMOC
Uncertain
Mass
RC-DA
Mass
RC-DA
Mass
Mean
Fraction
12/15/01
Fraction
12/16/01
Fraction
Mass
(us)
(us)
Fraction
ND
ND
ND
ND
ND
ND
ND
ND
0.0041
0.0079
0.0013
0.0037
0.0039
0.0039
0.0136
0.0053
0.0102
0.0054
0.0150
0.0129
0.0025
0.0064
ND
ND
0.0041
0.0114
0.0059
0.0057
0.0139
0.0045
0.0087
0.0045
0.0125
0.0117
0.0027
ND
ND
ND
ND
ND
ND
ND
0.0112
0.0048
0.0092
0.0051
0.0142
0.0115
0.0025
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
0.0021
0.0057
0.0019
0.0033
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
0.0475
0.0018
0.0034
0.0022
0.0062
0.0190
0.0247
0.0097
0.0038
0.0074
0.0037
0.0103
0.0091
0.0016
0.0133
0.0300
0.0577
0.0026
0.0073
0.0261
0.0275
0.0112
0.0033
0.0064
0.0047
0.0131
0.0102
0.0035
ND
ND
ND
0.0043
0.0120
0.0040
0.0069
0.0186
0.0084
0.0161
0.0098
0.0273
0.0207
0.0059
0.0054
ND
ND
0.0013
0.0035
0.0026
0.0023
0.0109
ND
ND
ND
ND
0.0030
0.0027
0.0071
0.0031
0.0059
0.0029
0.0080
0.0070
0.0010
0.0075
0.0025
0.0048
0.0031
0.0087
0.0070
0.0020
0.0056
0.0034
0.0065
0.0027
0.0076
0.0065
0.0010
ND
ND
ND
ND
ND
ND
ND
(Continued)
-------
Table 5-4. (Continued)
Compound
CAS No.
SNMOC
RC-DA
12/14/01
(1X2)
2,3,4-trimethylpentane
565-75-3
0.0023
toluene
108-88-3
0.0146
2-methylheptane
592-27-8
0.0025
3-methylheptane
589-81-1
ND
1-octene
111-66-0
ND
«-octane
111-65-9
0.0030
ethylbenzene
100-41-4
ND
-xy 1 e nc/p-xy 1 e ne
108-38-3/106-42-3
ND
styrene
100-42-5
ND
o-xylene
95-47-6
0.0028
1-nonene
124-11-8
ND
rt-nonane
111-84-2
ND
isopropylbenzene
98-82-8
0.0024
alpha-pinene
80-56-8
0.0035
«-propyl benzene
103-65-1
ND
/w-cthvltolucnc
620-14-4
ND
p-cthyltolucnc
622-96-8
ND
1,3,5-trimethylbenzene
108-67-8
0.0026
o-ethyltoluene
611-14-3
0.0019
beta-pinene
127-91-3
ND
1,2,4-trimethylbenzene
95-63-6
0.0028
1-decene
872-05-9
ND
n-decane
124-18-5
ND
SNMOC
SNMOC
Mass
RC-DA
Mass
RC-DA
Mass
Mean
TT
Fraction
12/15/01
Fraction
12/16/01
Fraction
Mass
Uncertain
(UB)
(US)
Fraction
0.0056
0.0023
0.0045
0.0035
0.0097
0.0066
0.0027
0.0359
0.0277
0.0533
0.0122
0.0340
0.0411
0.0107
0.0062
0.0029
0.0056
0.0025
0.0071
0.0063
0.0008
ND
0.0017
0.0033
0.0019
0.0053
0.0029
0.0027
ND
ND
ND
ND
ND
ND
ND
0.0074
0.0021
0.0041
0.0026
0.0072
0.0062
0.0019
ND
ND
ND
0.0031
0.0085
0.0028
0.0049
ND
0.0055
0.0106
0.0028
0.0077
0.0061
0.0055
ND
0.0026
0.0050
ND
ND
0.0017
0.0029
0.0070
0.0023
0.0045
ND
ND
0.0038
0.0035
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
0.0059
0.0037
0.0072
0.0045
0.0124
0.0085
0.0034
0.0086
ND
ND
ND
ND
0.0029
0.0050
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
0.0041
0.0080
0.0023
0.0064
0.0048
0.0042
0.0064
ND
ND
ND
ND
0.0021
0.0037
0.0047
ND
ND
0.0017
0.0048
0.0032
0.0027
ND
ND
ND
ND
ND
ND
ND
0.0070
0.0037
0.0071
0.0025
0.0069
0.0070
0.0001
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
0.0024
0.0066
0.0022
0.0038
(Continued)
-------
Table 5-4. (Continued)
SNMOC
Compound CAS No. 12/14/01
lESL_
1,2,3-trimethylbenzene 526-73-8 ND
/w-diethylbenzene 141-93-5 ND
/j-dicthvlbcnzcnc 105-05-5 ND
1-undecene 821-95-4 ND
«-undecane 1120-21-4 0.0021
1-dodecene 112-41-4 ND
«-dodecane 112-40-3 0.0004
1-tridecene 2437-56-1 ND
«-tridecane 629-50-5 ND
Total Speciated 0.1751
Total Unspeciated 0.2317
Total (Speciated + Unspeciated) 0.4068
Mass Emission Factor, mg/kg fuel burned
Mass Emission Factor, mg/kg fuel burned
(Speciated)
(Total)
0.3717
0.8804
Mass
Fraction
SNMOC
RC-DA
12/15/01
(|lg)
Mass
Fraction
SNMOC
RC-DA
12/16/01
(|lg)
Mass
Fraction
Mean
Mass
Fraction
Uncertair
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
0.0053
0.0018
0.0035
ND
ND
0.0029
0.0027
ND
ND
ND
ND
ND
ND
ND
0.0010
ND
ND
ND
ND
0.0003
0.0006
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
0.4306 0.3045 0.5862 0.1559 0.4332
0.5696 0.2149 0.4137 0.2041 0.5669
0.5194 0.3600
Mean Uncertainty
0.8069
1.4237
0.3245
0.7633
0.5010 0.2659
1.0225 0.3524
-------
PM2 5 Elemental Carbon/Organic Carbon, Inorganic Ions,
and Element Profile
Emissions of EC/OC, major inorganic ions, and major elements are reported in Table 5-
5 as mass fraction of measured mass. Uncertainties in the reported mass fraction averages of
the PM2 5 compounds are the standard deviations of results from the three test days.
Table 5-5. Fine Particle Chemical Composition of Emissions from a Smelt
Tank Vent at a Pulp and Paper Facility
Constituent
12/14/01 12/15/01 12/16/01
PMj ; Composition (mass fraction)
OC
0.0233
1.0452
0.0508
0.0398
0.0145
EC
0.0018
0.0018
0.0023
0.0020
0.0003
Elemental Composition (mass fraction)
sodium (Na)
0.1374
0.1798
0.1113
0.1428
0.0346
sulfur (S)
0.0926
0.1120
0.0719
0.0922
0.0201
potassium (K)
0.0498
0.0520
0.0458
0.0492
0.0031
chlorine (CI)
0.0173
0.0168
0.0136
0.0159
0.0020
silicon (Si)
0.00069
0.00066
0.00054
0.0006
0.0001
calcium (Ca)
0.00056
0.00052
0.00057
0.0006
ND
magnesium (Mg)
0.00025
0.00032
0.00022
0.0003
0.0001
phosphorus(P)
0.00004
0.00006
0.00006
0.0001
ND
Major Water-Soluble Ions (mass fraction)
thiosulfate (H03S2)
0.1374
0.1071
0.1216
0.1220
0.0152
sulfate (S04=)
0.1150
0.1926
0.0944
0.1340
0.0518
potassium (K+)
0.0468
0.0450
0.0474
0.0464
0.0014
chloride (CI )
0.0193
0.0163
0.0179
0.0178
0.0015
Results reported in Table 5-5 show the following:
• Thiosulfate ion was present in relatively large amounts in the PM2 5 aerosol emitted
from the smelt tank vent. Among the three process sources tested at the facility, this
species was unique to the smelt tank vent PM2 5 composition and may, therefore, be a
good marker species for the vent emissions.
• Potassium is found in significant amounts in the PM emissions for the smelt tank vent
(as it was for the recovery boiler and the hogged fuel boiler), indicative of the biomass
72
-------
fuel contribution. Most of the potassium in the PM from the smelt tank vent is water
soluble.
• Sodium thiosulfate and sodium sulfate comprised the bulk of the characterized PM2 5
mass. Over half of the PM2 5 mass could not be characterized by the analytical methods
used in this study and may have consisted mostly of water, given the very wet nature
of the vent gas emissions.
Supporting data for PM2 5 elemental, EC/OC, and inorganic ions are shown in
Appendicies H, I, and J.
Semivolatile Organic Compounds
Since the smelt from the recovery boiler bottoms is composed largely of molten
inorganic salts and since the organic content of the black liquor fuel in combusted by the
boiler, the organic carbon in the PM emissions from the smelt tank vent should be minimal
as should any organic species. Indeed, as discussed above, the organic carbon content of the
PM2 5 emitted represented only 4% of the PM2 5 mass, and relatively small amounts of gas-
phase carbonyl and nonmethane organic compounds were emitted.
Likewise, only a few semivolatile organic compounds were positively identified and
found above the levels observed in the cleaned ambient air used to dilute the sampled vent
gas. All semivolatile organic species were below the quantitation limits for the analytical
method used. Table 5-6 provides a list of the semivolatile organic compounds positively
identified in particle-phase emissions of the smelt tank vent that were at levels above those
found in the cleaned ambient air.
The results shown in Table 5-6 were obtained from a solvent extract of the quartz filters
on the sampling array located on sampling port RIO of the dilution sampler residence
chamber. These quartz filters from all three test days were composited and extracted
together in order to achieve even the semi-quantitative results reported. XAD-coated annular
denuders were placed in front of these filters to remove gas-phase organic compounds that
could have adsorbed onto the quartz filters and/or the pm collected on the filters, resulting in
a positive PM mass artifact. Likewise, quartz filters located on the dilution air sampling port
D1 were composited for all three test days and extracted together. Quantities of compounds
found on the dilution air filters were subtracted from like compounds found in the residence
chamber filters to give emissions results corrected for background levels of the species.
73
-------
Quantitation limits for the semivolatile species were taken to be the lowest
concentrations of standards used to establish the GC/MS analysis calibration curves. The
concentration ranges of these standards are given in Appendix K.
Table 5-6. Trace Semivolatile Organic Compounds Found Above Background
Levels in the Particle-Phase Emissions from the Smelt Tank Vent
Semivolatile Organic Compounds
phytane squalane
fluoranthene pyrene
chrysene benzo(a)anthracene
benzo(k)fluoranthene benzo(b)fluoranthene, benzo(j)fluoranthene
benzo(a)pyrene indeno( 1,2,3 -cd)pyrene
(and other isomers of molecular weight 276)
dibenzo(a,h)anthracene, dibenzo(a,c)anthracene benzo(ghi)perylene
Particle Size Distribution Data
The ELPI system was operated in the "charged" mode on all three test days, collecting
data on particle size distribution in the range from approximately 30 to 10,000 nm. The
ELPI was run in continuous mode throughout all three of the analytical runs. When the
dilution sampling system was started and flow was initiated, the ELPI operational mode was
changed from "flush" mode to sampling mode. Sample was collected for the entire run of
slightly more than 8 hours.
Results of the individual runs are summarized in the following tables, diagrams, and
figures. Table 5-7 lists the collected mass in each of twelve stages for each test day. The
mean particle diameter (Di) of each stage is listed in increasing size order from 42.78 nm to
8328.12 nm. Also shown are the particle counts vs. size expressed as log plots dN/dlogDp,
1/cm3 and particle mass vs size expressed as log plots dM/dlogDp, mg/cm3. A simple bar
plot of particle mass by channel is also shown.
Foil stages of each impactor stage were recovered in the field for individual gravimetric
mass determinations. Foils were tared before shipment to the field, used for collection with
each sampling run, and individually recovered for determination of mass using a sensitive
electronic balance. After mass determination, the foils were also available for organic mass
determinations.
74
-------
Table 5-7. Particle Size Distribution Data
December 14, 2001
Stage3
1
2
3
4
5
6
7
8
9
10
11
12
Di,b nm
42.78
80.03
132.82
208.19
320.04
506.03
803.12
1276.71
2010.57
3157.47
5212.98
8328.12
dN/dlog(Dp),c 1/cm3
9.01xl04
3.15xl05
5.36xl05
4.25xl05
1.09xl05
3.69xl04
8.49xl03
1.39xl03
1.74xl02
3.01x10'
1.21x10'
1.30x10'
M,d mg/m3
0.0011
0.0199
0.1342
0.3741
0.3512
0.5276
0.4385
0.3223
0.1348
0.1042
0.2029
0.7113
dM/dlog(Dp), mg/m3
3.69 xlO"3
8.45 xlO"2
6.57x10"'
2.01
1.88
2.50
2.30
1.52
7.40x10"'
4.97x10"'
8.99x10"'
3.93
December 15, 2001
Stage
Di, nm
dN/dlog(Dp), 1/cm3
M, mg/m3
dM/dlog(Dp), mg/m3
1 2
42.78 80.03
9.38xl04 3.70xl05
0.0012 0.0234
3.85xlO"3 9.94xlO"2
3 4
132.82 208.19
5.36xl05 3.56xl05
0.1342 0.3133
6.57x10"' 1.68
5 6
320.04 506.03
1.38xl05 5.48xl04
0.4428 0.7837
2.37 3.72
7 8
803.12 1276.71
1.54xl04 3.45xl03
0.7962 0.7985
4.18 3.76
9 10
2010.57 3157.47
6.23xl02 1.03xl02
0.4832 0.3567
2.65 1.70
11 12
5212.98 8328.12
3.58x10' 3.25x10'
0.6000 1.7799
2.66 9.82
December 16, 2001
Stage 1 2 3 4 5 6 7 8 9 10 11 12
Di, nm 42.78 80.03 132.82 208.19 320.04 506.03 803.12 1276.71 2010.57 3157.47 5212.98 8328.12
dN/dlog(Dp), 1/cm3 1.79xl05 5.14xl05 5.86xl05 4.00xl05 1.07xl05 6.43xl04 3.05xl04 1.17xl04 4.47xl03 4.57xl02 1.25x10' 2.03x10'
M, mg/m3 0.0023 0.0325 0.1467 0.3524 0.3435 0.9199 1.5743 2.7027 3.4645 1.5800 0.2094 1.1131
dM/dlog(Dp), mg/m3 7.34xlQ-3 1.38xlQ-'7.19xlQ-' 1.89 1.84 4.36 8.27 1.27x10' 1.90x10' 7.53 9.28xlQ-' 6.14
a Stage shows the individual stages of the 12-stage ELPI.
b Di is the midpoint value used in the distribution calculations; Di is the geometric mean of the boundaries of each stage.
c Particle counts are expressed as log dN/dlogDp, 1/cm3 or as log dM/dlogDp, mg/cm3 and plotted vs. particle diameter (Dp).
d M (expressed in mg/m3) is the mass distribution, which gives the total mass of all particles in each size range. Mass distribution is calculated by multiplying the current
distribution by the conversion vector and by a vector formed from the masses of spheres having diameter equal to midpoint values (Di) of each stage.
-------
Plots of particle counts vs. size, particle mass vs. size, and particle mass vs. stage are
shown for each test day (Figures 5-1, 5-2, and 5-3). The mass of particles collected appears
to maximize at Stage 12 (8328.12 nm) for the first two test days, and at Stage 9 (2010.57) on
the third test day.
Particle Counts vs. Size
5
5
4X 10
5
2X 10
0
10
Dp, nm
Particle Mass vs. Size
4
3
2
1
o
10
Dp, nm
Particle Mass
0.80 -T-
"E 0.60
cb —
S 0.20
0.00
1 2 3 4 5 6 7 8 9 10 11 12
Stage
Figure 5-1. Plots of Particle Counts vs. Size, Particle Mass vs. Size, and
Particle Mass per Stage for Test Day 1 (12/14/01).
76
-------
Particle Counts vs. Size
5
5
4X 10
5
2X 10
0
10
Dp, nm
Particle Mass vs. Size
£
10
cb
E
8
CL
6
o
Sf)
4
o
T3
2
0
7f
1 1—I I I II I | 1 1—I I I lll|
10 102 103
Dp, nm
10^
Particle Mass
2.00 -i
*> 1.50
OJ
E 1.00
ui
in
£ 0.50
0.00
1 2 3 4 5 6 7 8 9 10 11 12
Stage
Figure 5-2. Plots of Particle Counts vs. Size, Particle Mass vs. Size, and
Particle Mass per Stage for Test Day 2 (12/15/01).
77
-------
Particle Counts vs. Size
*£ 8 X 10*
O 5
^ 6X10
g 4X105
o 2X105
p
¦o
0
-2X 10
51
10*
10
Dp, nm
Particle Mass vs. Size
CO
E
O)
E
CL
Q
CJ
o
¦p
¦o
Dp, nm
Particle Mass
4.00 -r
£ 3.00
E
in
iff
<4
2.00
1.00
0.00
I
1—F=l—t
1 2 3 4 5 6 7 8 9 10 11 12
Stage
Figure 5-3. Plots of Particle Counts vs. Size, Particle Mass vs. Size, and
Particle Mass per Stage for Test Day 3 (12/16/01).
78
-------
Quality Assurance/Quality Control
The sampling and analysis procedures performed for this study adhered to approved
EPA Quality Assurance Project Plans (QAPPs) QTRAK No. 9905112 and QTRAK No.
9900213, respectively. EPA method operating procedures (MOPs) and ERG standard
operating procedures (SOPs) prescribe the QC checks performed with each procedure. EPA
MOPs and ERG SOPs are listed by title in Appendix L; complete MOPs and SOPs are
available upon request. QAPPs, MOPs, SOPs, and files of raw data and QC supporting data
for the project were archived for future reference. Summaries of the QC measures
implemented for the field sampling activities and for various analytical methods are given in
Tables 6-1 through 6-8.
Field Sampling
In field sampling with the dilution sampling system, the following quality control
procedures14 were implemented:
• A leak check of the dilution sampling system with all sample collection arrays was
performed before field testing was initiated.
• Pitot tubes and meter boxes were calibrated before field testing was initiated.
• Analytical balance(s) were calibrated before field testing was initiated.
• Flow control collection devices for the canisters were calibrated using a primary flow
standard before field testing was initiated.
• Multipart forms recording field conditions and observations were used for canisters
and carbonyl samples.
• Strict chain of custody documentation for all field samples was maintained.
Field sampling equipment quality control requirements that were met in the course of
preparing for the field test and execution of testing activities are summarized in Table 6-1.
Strict chain of custody procedures were followed in collecting and transporting samples
and sampling media to and from the field sampling location. Sample substrates (filters,
denuders, PUF modules, DNPH cartridges) were prepared in advance in accordance with the
numbers and types of samples designated in the sampling matrix of the approved field test
plan. Clean SUMMA polished collection canisters and the DNPH-coated sampling
79
-------
Table 6-1. Field Sampling Equipment Quality Control Measures
Equipment
Effect
Acceptance
Criteria
Criteria
Achieved?
Orifice meters (volumetric gas
flow calibration)
Ensures the accuracy of flow
measurements for sample
collection
±1%
Yes
Venturi meters (volumetric
gas Flow calibration)
Ensures the accuracy of flow
measurements for sample
collection
±l%of
reading
Yes
Flow transmitter (Heise gauge
with differential pressure)
Ensures the accuracy of flow
measurements for sample
collection
±0.5% of
range
Yes
Analytical balances
Ensures control of bias for all
project weighing
Calibrated
with Class S
weights
Yes
Thermocouples
Ensures sampler temperature
control
±1.5 °C
Yes
Relative humidity probes
Ensures the accuracy of
moisture measurements in the
residence chamber
±2% relative
humidity
Yes
Sampling equipment leak
check and calibration (before
each sampling run)
Ensures accurate measurement
of sample volume
1%
Yes
Sampling equipment field
blanks
Ensures absence of
contamination in sampling
system
<5.0% of
sample
values
Yes
cartridges used to collect carbonyl compounds were prepared and supplied by ERG. The
PUF, XAD-4-coated denuders, and PM2 5 sampling substrates were prepared and supplied by
EPA. Chain of custody forms (Figure 6-1) were initiated when the sampling media were
prepared. Each sample substrate was assigned a unique identification number by the
laboratory supplying the substrates. Copies of the chain of custody forms are included in
Appendix B.
Sample identification numbers include a code to track
• Source type;
• Test date(s);
• Sampler type;
• Substrate type;
80
-------
• Sampler chamber (i.e., dilution chamber or residence chamber);
• Sampler port;
• Lane/leg;
• Position; and
• Holder number.
For samples to be analyzed in the EPA laboratories, whole sample collection arrays were
assembled by EPA, assigned a unique tracking number, and used for sample collection.
Sample collection arrays were recovered in the field as a complete unit and transferred to the
EPA laboratory for disassembly and analysis.
After collection, samples were transported to the analysis laboratories by ERG, with
careful documentation of sample collection and chain of custody records for the samples
being transported. Samples were stored in a secure area until they were transported to the
laboratories performing the analysis.
Carbonyl Compound Analysis
Quality control criteria for the carbonyl analysis performed by ERG are shown in
Table 6-2. Supporting analytical data are a part of the project file at ERG.
Table 6-2. Carbonyl Analysis: Quality Control Criteria
Parameter
Quality
Control Check
Frequency
Acceptance
Criteria
Corrective
Action
Criteria
Achieved?
HPLC
column
efficiency
Linearity
check
Analyze second
source QC
sample (SSQC).
Analyze 5-point
calibration curve
and SSQC in
triplicate.
At setup and 1
per sample
batch.
At setup or
when calibra-
tion check
does not meet
acceptance
criteria.
Resolution
between acetone
and propion-
aldehyde >1.0;
column efficiency repeat analysis.
>500 plates.
Eliminate dead
volume, back-
flush, or
replace column;
Correlation
coefficient
>0.999, relative
error for each
level against
calibration curve
at or within ±20%
relative error.
Check integra-
tion, reintegrate
or recalibrate.
Yes
Yes
(continued)
81
-------
Table 6-2. (Continued)
Parameter
Quality
Control Check
Frequency
Acceptance
Criteria
Corrective
Action
Criteria
Achieved?
Retention
time
Analyze
calibration
midpoint.
Once per 10
samples.
Calibration
check
Analyze midpoint Once per 10
standard. samples.
Calibration SSQC
accuracy
Once after
calibration in
triplicate.
Analyze 0.1 Once after
|ig/mL standard, calibration in
triplicate.
Intercept accep-
tance should be
< 10,000 area
counts per com-
pound; correlates
to 0.06 mg/mL.
Acetaldehyde,
benzaldehyde,
hexaldehyde
within retention
time window
established by
determining 3 a
or ±2% of the
mean calibration
and midpoint
standards, which-
ever is greater.
85%- 115%
recovery.
85%- 115%
recovery.
±25% difference.
Check integra-
tion, reintegrate
or recalibrate.
Check system
for plug,
regulate
column
temperature,
check gradient
and solvents.
Check integra-
tion, recalibrate
or reprepare
standard,
reanalyze
samples not
bracketed by
acceptable
standard.
Check integra-
tion; recalibrate
or reprepare
standard,
reanalyze
samples not
bracketed by
acceptable
standard.
Yes
Yes
Yes
Yes
(continued)
82
-------
Table 6-2. (continued)
Parameter
Quality
Control Check
Frequency
Acceptance
Criteria
Corrective
Action
Criteria
Achieved?
System blank Analyze
Bracket
Measured
Locate
Yes
acetonitrile.
sample batch,
1 at beginning
and 1 at end.
concentration
<5 times MDL.
contamination
and document
levels of
contamination
in file.
Duplicate
Duplicate
As collected.
±20% difference.
Check integra-
Yes
analyses
samples.
tion; check
instrument
function; re-
analyze
duplicate
samples.
Replicate
Replicate
Duplicate
< 10% RPD for
Check integra-
Yes
analyses
injections.
samples only.
concentrations
greater than 1.0
|ig/mL.
tion, check
instrument
function, re-
analyze
duplicate
samples.
Method
Analyze
One MS/MSD
80%- 120%
Check calibra-
Yes
spike/method MS/MSD.
per 20
recovery for all
tion, check
spike
samples.
compounds.
extraction
duplicate
procedures.
(MS/MSD)
Concurrent Air Toxics/Speciated Nonmethane Organic
Compound Analysis
The analytical system performing the concurrent analysis is calibrated monthly and
blanked daily prior to sample analysis. A quality control standard is analyzed daily prior to
sample analysis to ensure the validity of the current monthly response factor. Following the
daily quality control standard analysis and prior to the sample analysis, cleaned, dried air
from the canister cleaning system is humidified and then analyzed to determine the level of
organic compounds present in the analytical system. Upon achieving acceptable system
blank results—less than or equal to 20 ppbC—sample analysis begins. Ten percent of the
total number of samples received are analyzed in replicate to determine the precision of
83
-------
analysis for the program. After the chromatography has been reviewed, the sample canister
is returned to the canister cleaning laboratory to be prepared for subsequent sample
collection episodes or sent to another laboratory for further analysis. Quality control
procedures for the air toxics and SNMOC analyses are summarized in Table 6-3.
Table 6-3. Quality Control Procedures for the Concurrent Analysis for Air
Toxics and SNMOC
Quality Control
Check
Frequency
Acceptance
Criteria
Corrective Criteria
Action Achieved?
Air Toxics Analysis
BFB instrument tune
check
Five-point calibration
bracketing the expected
sample concentration
(0.25 - 15 ppbv)
Calibration check using
mid-point of calibration
range
System blank
Daily prior to
calibration check.
Following any major
change, repair, or
maintenance if daily
quality control check
is not acceptable.
Calibration is valid
for six weeks if
calibration check
criteria are met.
Daily. Response factor
<30% bias from
calibration curve
average response
factor.
Daily following tune 0.2 ppbv/analyte or
check and calibration MDL, whichever is
check.
Evaluation criteria in
data system software;
consistent with
Method TO-15.
RSD of response
factors <30%
relative retention
times (RRTs) for
target peaks ±0.06
units from mean
RRT.
Retune mass
spectrometer;
clean ion source
and quadrupoles.
Repeat individual
sample analysis;
repeat linearity
check; prepare
new calibration
standards and
repeat analysis.
Repeat
calibration
check; repeat
calibration curve.
Repeat analysis
with new blank;
check system for
greater.
Internal standard (IS) leaks,
area response ±40% contamination;
and retention time
±0.33 min of most
recent calibration
check.
re-analyze blank.
Laboratory Control Daily.
Standard (LCS)
Recovery limits
70% - 130%
IS Retention Time
±0.33 min of most
recent calibration.
Repeat analysis;
repeat calibration
curve.
Yes
Yes
Yes
Yes
Yes
(continued)
84
-------
Table 6.3 (continued)
Quality Control
Check
Frequency
Acceptance
Criteria
Corrective
Action
Criteria
Achieved?
Replicate analysis
Samples
All duplicate field
samples.
All samples.
<30% RPD for
compounds
>5xMDL.
IS RT ±0.33 min of
most recent
calibration.
Repeat sample
analysis.
Yes
Repeat analysis. Yes
SNMOC Analysis
System blank analysis
Multiple point
calibration (minimum
5); propane bracketing
the expected sample
concentration range
Calibration check:
midpoint of calibration
curve spanning the
carbon range (C2 - C10)
Replicate analysis
Daily, following
calibration check.
Prior to analysis and
monthly.
Daily.
All duplicate field
samples.
20 ppbC total.
Correlation
coefficient
(r2) >0.995.
Repeat analysis; Yes
check system for
leaks; clean
system with wet
air.
Repeat individual Yes
sample analysis;
repeat linearity
check; prepare
new calibration
standards and
repeat.
Response for selected Repeat Yes
hydrocarbons calibration
spanning the carbon check; repeat
range within ±30% calibration curve.
difference of
calibration curve
slope.
Total NMOC within Repeat sample Yes
±30% RSD. analysis.
PM Mass Measurements, Elemental Analysis, Water-
Soluble Ion Analysis, and GC/MS Analysis
Quality control criteria for EPA analyses (PM mass, elemental analyses, ion
chromatography analysis, and GC/MS analysis) are summarized in Tables 6-4 through 6-7;
supporting data are included in the project file in the EPA laboratory.
85
-------
Table 6-4. PM Mass Measurements: Quality Control Criteria
„ ^ Quality Control „ Acceptance Corrective Criteria
Parameter . Frequency „
Check Criteria Action Acheived I
Deposition on
Analyze laboratory
Bracket
Mass within
Adjust mass for
Yes
filter during
filter blank.
sample batch,
±15 mg of
deposition.
conditioning
1 at
previous
beginning
weight.
and 1 at end.
Laboratory
Analyze laboratory
Bracket
Mass within
Adjust mass to
Yes
stability
control filter.
sample batch,
±15 mg of
account for
1 at
previous
laboratory
beginning
weight.
difference.
and 1 at end.
Balance stability
Analyze standard
Bracket
Mass within ±3
Perform internal
Yes
weights.
sample batch,
mg of previous calibration of
1 at
weight.
balance, perform
beginning
external
and 1 at end.
calibration of
balance.
Table 6-5. Elemental Analysis: Quality Control Criteria
_ , Quality Control Acceptance Corrective Criteria
Parameter ' . Frequency „ ...
Check Criteria Action Achieved:
Performance Analyze monitor Once per <2% change in Recalibrate. Yes
evaluation check sample. month. each element from
previous
measurement
86
-------
Table 6-6. Water-Soluble Ion Analysis: Quality Control Criteria
„ ^ Quality Control „ Acceptance Corrective Criteria
Parameter , Frequency „ ...
Check Criteria Action Achieved I
Linearity
Analyze 4-point
At setup or
Correlation
Recalibrate.
Yes
check
calibration curve.
when
coefficient
calibration
>0.999.
check does not
meet
acceptance
criteria.
System dead
Analyze water.
Bracket sample
Within 5% of
Check system
Yes
volume
batch, 1 at
previous
temperature,
beginning and
analysis.
eluent, and
1 at end.
columns.
Retention
Analyze standard.
At setup.
Each ion within
Check system
Yes
time
±5% of standard
temperature and
retention time.
eluent.
Calibration
Analyze one
Once every
85%- 115%
Recalibrate or re-
Yes
check
standard.
4-10 samples.
recovery.
prepare standard,
reanalyze sample
not bracketed by
acceptable
standard.
System blank Analyze HPLC
Bracket sample
No quantifiable
Reanalyze.
Yes
grade water.
batch, 1 at
ions.
beginning and
1 at end.
Replicate
Replicate
Each sample.
< 10% RPD for
Check instrument
Yes
analyses
injections.
concentrations
function, re-
greater than 1.0
analyze samples.
ma/L.
87
-------
Table 6-7. Quality Control Procedures for Gas Chromatography-Mass
Spectrometry Analysis of Semivolatile Organic Compounds
Quality Control
Check
Frequency
Acceptance
Criteria
Corrective Criteria
Action Achieved?
Mass spectrometer Daily prior to
instrument tune check calibration check.
Mass assignments m/z
= 69, 219, 502 (±0.2).
Peak widths = 0.59 -
0.65.
Relative mass
abundances = 100%
(69); >30% (219);
> 1% (502).
Retune mass
spectrometer;
clean ion source.
Yes
Five-point calibration Following mainten-
bracketing the ance or repair of
expected concentration either gas
range chromatograph or
mass spectrometer or
when daily quality
control check is not
acceptable.
Calibration check
using midpoint of
calibration range
System blank
Daily.
As needed after
system maintenance
or repair.
Retention time check Daily
Correlation coefficient
of either quadratic or
linear regression
>0.999.
Check
integration, re-
integrate or
recalibrate
Yes
Compounds in a
representative organic
compound suite >80% curve,
are ±15% of individ-
ually certified values.
Values >20% are not
accepted.
Repeat analysis,
repeat calibration
Potential analytes
greater than or equal to check system
detection limit values, integrity.
Reanalyze blank
Verify that select Check inlet and
compounds are within column flows
±2% of established and the various
retention time window. GC/MS
temperature
zones.
Yes
Repeat analysis; Yes
Yes
88
-------
References
1. U. S. Government Printing Office, EPA Method 5, Determination of Particulate Matter
Emissions from Stationary Sources, in Code of Federal Regulations, Title 40, Part 60,
Appendix A, pp. 371-443, Washington, DC, 1989d.
2. U.S. Government Printing Office, EPA Method 202, Determination of Condensible
Particulate Emissions from Stationary Sources, in Code of Federal Regulations, Title
40, Part 51, Appendix M, pp. 330-338, Washington, DC, 2002.
3. U.S. Government Printing Office, EPA Method 1, Sample and Velocity Traverses for
Stationary Sources, in Code of Federal Regulations, Title 40, Part 60, Appendix A, pp.
181-206, Washington, DC, 1989a.
4. U.S. Government Printing Office, EPA Method 2, Velocity - S- Type Pilot, in Code of
Federal Regulations, Title 40, Part 60, Appendix A, pp. 214-253, Washington, DC,
1989b.
5. U.S. Government Printing Office, EPA Method 4, Moisture Content, in Code of Federal
Regulations, Title 40, Part 60, Appendix A, pp. 347-371, Washington, DC, 1989c.
6. Hildemann, L.M., G.R. Cass, and G.R. Markowski. A Dilution Stack Sampler for
Collection of Organic Aerosol Emissions: Design, Characterization and Field Test.
Aerosol Science and Technology 10:193-204, 1989.
7. U.S. EPA. National Emission Standards for Hazardous Air Pollutants (NESHAP): A
Plain English Description, U.S. Environmental Protection Agency, Office of Air Quality
Planning and Standards, Research Triangle Park, NC, EPA/456/R-01/003 [NTIS
PB2002-100518], September 2001. Also available from
http://www.epa.gov/ttnatw01/pulp/pulppg.html (Accessed September 2004).
8. NIOSH Method 5040, Elemental Carbon (DieselParticulate). National Institute for
Occupational Safety and Health (NIOSH) Manual of Analytical Methods (NMAM), 4th
Edition, Department of Health and Human Services (NIOSH) Publication 94-113,
August, 1994.
9. U.S. EPA. Compendium of Methods for the Determination of Toxic Organic Compounds
in Ambient Air, U.S. Environmental Protection Agency, Center for Environmental
Research Information, National Risk Management Research Laboratory, Office of
Research and Development, Cincinnati, Ohio, EPA/625/R-96/010b [NTIS PB99-
172355], January 1999. Also available at
http://www.epa.gov/ttnamtil/files/ambient/airtox/tocomp99.pdf (Accessed September
2004).
89
-------
10. U.S. EPA. Technical Assistance Document for Sampling and Analysis of Ozone
Precursors, U.S. Environmental Protection Agency, National Exposure Research
Laboratory, Office of Research and Development, Research Triangle Park, NC, EPA-
600/R-98-161 [NTIS PB99-124034], September 1998. Also available at
http://www.epa.gov/ttnamtil/files/ambient/pams/newtad.pdf (Accessed September
2004).
11. Federal Register, Volume 49, Number 209, Appendix B to Part 136—Definition and
Procedure for the Determination of the Method Detection Limit—Revision 1.11, pp.
198-199, October 26, 1984.
12. U.S. EPA. Quality Assurance Project Plan. Chemical Analysis of Fine Particulate
Matter, N. Dean Smith, QTRAKNo. 99051, U.S. Environmental Protection Agency,
National Risk Management Research Laboratory, Air Pollution Prevention and Control
Division, Research Triangle Park, NC 27711.
13. U.S. EPA. Quality Assurance Project Plan. Chemical Analysis of Fine Particulate
Matter, N. Dean Smith, QTRAKNo. 99002/III, Revision 4, U.S. Environmental
Protection Agency, National Risk Management Research Laboratory, Air Pollution
Prevention and Control Division, Research Triangle Park, NC 27711, August 2001.
14. von Lehmden, D.J., W.G. De Wees, and C. Nelson. Quality Assurance Handbook for Air
Pollution Measurement Systems. Volume III. Stationary Source Specific Methods,
EPA/600/4-77/027b [NTIS PB80-112303], U.S. Environmental Protection Agency,
Environmental Monitoring and Support Laboratory, Research Triangle Park, NC, May
1979.
90
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO. 2.
EPA-600/R-03/101 a
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Source Sampling Fine Particulate Matter: Stationary Source
Characterization Testing of a Smelt Tank at a Pulp and Paper
Facility Vnlump 1 Rpnnrt
5. REPORT DATE
November 2003
6. PERFORMING ORGANIZATION CODE
7. AUTHORS
Joan T. Bursey and Dave-Paul Dayton
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Eastern Research Group, Inc.
1600 Perimeter Park Drive
Morrisville, NC 27560
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
Contract No. 68-D7-0001
12. SPONSORING AGENCY NAME AND ADDRESS
U. S. EPA, Office of Research and Development
Air Pollution Prevention and Control Division
Research Triangle Park, North Carolina 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final; 02/05/01 - 05/30/03
14. SPONSORING AGENCY CODE
EPA/600/13
15. SUPPLEMENTARY NOTES
The EPA Project Officer is N. Dean Smith, mail drop E343-02, phone (919) 541-2708
16. ABSTRACT
The report provides a profile of the chemical composition of particulate matter (PM) with aerodynamic
diameter 2.5 |_im or less (PM25) emitted from a smelt tank at a pulp and paper facility using the Kraft pulping
process. The first step in the Kraft process is to form pulp by digesting wood chips in an aqueous solution of
sodium sulfide and sodium hydroxide at elevated temperature and pressure. This extracts the cellulose
from the wood by dissolving the lignin that binds the cellulose fibers together. The pulp is washed, and the
spent digestion liquor/wash water solution (black liquor) is separated from the cellulose (which forms the
paper) and is sent through a process to recover the digestion chemicals. The black liquor undergoes a
sequence of evaporation steps to yield a concentrated black liquor which is sent as fuel to a recovery boiler
used to generate heat for process steam in the plant. Combusting the black liquor results in a molten smelt,
which is composed primarily of inorganic chemicals, at the bottom of the recovery furnace. The smelt is
collected, dissolved in water, and transferred to a tank where quicklime is added to regenerate the digestion
reactants, sodium sulfide and sodium hydroxide. The data obtained during this research will assist States in
determining the major sources of PM2 5 so they can devise and institute a control strategy to attain the
ambient concentrations set by the National Ambient Air Quality Standard for PM2 5 that was promulgated in
July 1977 by the U.S. EPA. Along with the PM25 emission profile, data are also provided for gas-phase
emissions of several organic compounds. Data are provided in a format suitable to be included in the EPA
source profile database, SPECIATE.
17. KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Air Pollution
Paper Industry
Wood Pulp
Fine Particulate Matter
Chemical Composition
Organic Compounds
Volatility
Pollution Control
Stationary Sources
13B
11L
14G
07D
07C
20M
18. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
102
20. SECURITY CLASS (This Page)
Unclassified
22. PRICE
EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION IS OBSOLETE
forms/admin/techrpt.frm 7/8/99 pad
------- |