I&EFA
United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park NC 27711
ElvlB Report 83-CAT-12
August 1983
Air
Petroleum
Refineries -
Fluid Catalytic
Cracking Regenerators
Particulate Test
Method Evaluation
Emission Test Report
Exxon Company, USA
Baton Rouge, LA
-------
EMISSION TEST REPORT
METHOD DEVELOPMENT AND TESTING
FOR FCCU REGENERATORS
Exxon Company, U.S.A
Baton Rouge, Louisiana
EMB Report No. 82-CAT-12
ESED Project No. 82/04
by
PEDCo Environmental, Inc.
11499 Chester Road
P.O. Box 46100
Cincinnati, Ohio 45246-0100
Contract No. 68-02-3546
Work Assignment Nos. 14 and 20
PN: 3530-14 and 3530-20
EPA Task Manager
Mr. Winton Kelly
Emission Standards and Engineering Division
Emission Measurement Branch
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
April 1984
-------
DISCLAIMER
This report was furnished to the U.S. Environmental Protec-
tion Agency, Emission Measurement Branch, by PEDCo Environmental,
Inc., Cincinnati, Ohio, in fulfillment of Contract No. 68-02-3546,
Work Assignments 14 and 20. Its contents are reproduced herein
as received from PEDCo Environmental, Inc. The opinions, find-
ings, and conclusions expressed are those of the author and not
necessarily those of the Environmental Protection Agency. Men-
tion of company or product names does not constitute endorsement
or recommendation for use.
-------
CONTENTS
Figures iv
Tables v
Acknowledgment vii
Quality Assurance Element Finder viii
1. Introduction 1-1
2. Process Operation 2-1
3. Sampling and Analytical Plan 3-1
3.1 Sampling location 3-1
3.2 Sampling methods 3-1
3.3 Sample analysis 3-9
4. Summary and Discussion of Test Results 4-1
4.1 Sample data 4-1
4.2 Thermogravimetric analytical results 4-4
4.3 Water-soluble sulfate analytical data 4-21
4.4 Recommendations for sample and analytical
methodology 4-33
5. Quality Assurance 5-1
References R-l
Appendix A Computer printouts and example calculations A-l
Appendix B Raw field data B-l
Appendix C Raw laboratory data C-l
Appendix D Sampling and analytical procedures D-l
Appendix E Calibration procedures and results E-l
Appendix F Quality assurance summary F-l
Appendix G Project participants and sample log G-l
111
-------
FIGURES
Number Page
3-1 Sample Train Setup - Plan View 3-2
3-2 Sampling Location 3-3
4-1 Average Particulate Concentration for Run 7
at Indicated Sample Conditioning Temperature 4-12
5-1 Audit Report Dry Gas Meter (Meter Box FB-2) 5-5
5-2 Audit Report Dry Gas Meter (Meter Box FB-3) 5-6
5-3 Audit Report Dry Gas Meter (Meter Box FB-5) 5-7
5-4 Audit Report Dry Gas Meter (Meter Box FB-8) 5-8
5-5 Audit Report SO,, Analysis 5-11
IV
-------
TABLES
Number Page
3-1 Sample Matrix 3-5
3-2 Analytical Plan 3-10
4-1 Summary of Sample Conditions 4-2
4-2 Summary of Thermogravimetric Analytical
Results 4-5
4-3 Direct Comparison of Particulate Weights at
6- and 24-Hour Heat-Conditioning Periods 4-7
4-4 Comparison of Weight Losses Above 160°C 4-8
4-5 Comparison of Filterable Particulate Con-
centration After Conditioning at Tempera-
tures of 160°, 232°, and 315°C 4-10
4-6 Filterable Particulate Relative Percent
Weight Loss After Conditioning at Tempera-
tures 160°, 232°, and 315°C 4-11
4-7 Summary of H_SO. and SO_ Analytical Data 4-13
4-8 Statistical Data for Grouped Runs After Con-
ditioing at Indicated Temperatures 4-17
4-9 Summary of Precision Estimates After Condi-
tioning at Indicated Temperatures 4-18
4-10 Summary of Water-Soluble Sulfate Analytical
Results 4-22
4-11 Summary of Results for Residual Sulfate (SO ~)
in Within-Run Samples Conditioned at 315°C 4-25
4-12 Comparison of Within-Run Particulate Concen-
tration After Correction for Residual Sulfate
to the M5W Test Results 4-26
v
-------
TABLES (Continued)
Number Page
4-13 Summary of Results for Residual Sulfate (SO ~)
of Samples Conditioned at 315°C 4-28
4-14 Cations Found in Water Extraction by ICP 4-29
4-15 Soluble Sulfate Present in Sample Analyzed
by ICP 4-30
4-16 Charge Balance Results for Samples Analyzed
by ICP 4-31
5-1 Field Equipment Calibration 5-3
5-2 Example of a Thermogravimetric Analysis of
Filter and Acetone Blanks 5-9
5-3 Reagent Blank Analysis for IPA and H_0_ 5-13
L* £t
5-4 Ion Chromatography Checks 5-14
5-5 Non-Water-Soluble Sulfate Blank Analytical
Data 5-16
5-6 Ion Chromatography Blank Analytical Data 5-18
VI
-------
ACKNOWLEDGMENT
Mr. Winton Kelly, EPA Task Manager, provided overall project
coordination and guidance and observed the test program. Mr. Tim
Tucker represented the Exxon Company and provided assistance in
scheduling and process operation. Mr. Charles Bruffey was the
PEDCo Project Manager. Principal authors were Messrs. Charles
Bruffey and Thomas Wagner.
VII
-------
QUALITY ASSURANCE ELEMENT FINDER
Title page
Table of contents
Project description
QA objective for measurement of data in
terms of precision, accuracy, completeness,
representativeness, and comparability
Sampling procedures
Sample custody
Calibration procedures and frequency
Analytical procedures
Data reduction, validation, and
reporting
Internal quality control checks and
frequency
Performance and system audits and
frequency
Preventive maintenance procedures and
schedules
Specific routine procedures used to
assess data precision, accuracy, and
completeness of specific measurement
parameters involved
Corrective action
Quality assurance reports to management
Location
Section Page
Hi
1 l-l
Appendix F F-2
Appendix D D-l
Appendix C C-l
Appendix E E-l
Appendix D D-l
Section 5 5-1
Appendix F F-2
Section 5 5-1
Appendix F F-ll
Section 5 5-1
Appendix F F-3
Appendix F F-12
Appendix F F-4
Appendix F F-ll
Appendix F F-12
Vlll
-------
SECTION 1
INTRODUCTION
On March 8, 1974, the U.S. Environmental Protection Agency
(EPA) promulgated a New Source Performance Standard (NSPS)
governing particulate emissions from fluid catalytic cracking
unit (FCCU) regenerators. The testing procedures specified the
use of Method 5 for measurement of these emissions. The data to
support the NSPS were collected during 1971 and 1972 by use of
the Method 5 procedures. The facilities tested were conventional
regenerators equipped with electrostatic precipitators (ESP's)
and carbon monoxide (CO) boilers.
Since the promulgation, EPA has received several requests to
clarify the intent of the emission regulation. The EPA has
stated that the materials intended to be controlled were "cata-
lyst fines" or "mineral dust" and not condensible sulfates that
were in the gas phase at the operating temperature of the control
device.
In the public notice of proposed rulemaking for a revision
to the FCCU new source standard,* EPA stated that because Method
5 is capable of collecting condensible matter that is not con-
trollable by the best systems of emission reduction, a facility
employing such systems could be found in noncompliance if
*44 FR 60759 Monday, October 22, 1979.
1-1
-------
significant quantities of such condensibles were present as a
result of feed changes or process variations. Consequently, EPA
is evaluating sample and analytical parameters designed to mini-
mize the collection of condensible sulfate materials from these
sources.
Under contract to the Emission Measurement Branch of the
EPA, PEDCo Environmental, Inc., conducted the third in a series
of atmospheric emission test projects from March 2 through 9,
1983, at the Exxon Company U.S.A. refinery in Baton Rouge,
Louisiana. Testing was performed at the final exit stack of the
FCCU regenerator to provide data for the development of modifica-
tions to the existing test methodology to minimize the collection
of condensible sulfate materials during the measurement of par-
ticulate emissions from these sources.
All samples were collected by use of four single-sample
trains located at points of similar velocity and temperature in
the FCCU exit stack. Nine test runs were performed during the
test series. For evaluation of the effect of sample temperature
on sulfate collection, the temperatures of the probe and filter
box were varied for each run as follows:
Probe and filter box
Sample designation sample temperature
M5 121°C (250°F)
MSB 160°C (320°F)
M5-450 232°C (450°F)
M5W 121°C (250°F)
Paired trains were run at similar temperatures to allow within-
run data comparisons as well as comparisons between methods run
at different sample temperatures.
1-2
-------
Probe rinse and filter sample fractions were subjected to a
thermogravimetric analysis at predetermined temperatures to
assess sample weight loss as a function of drying temperature.
In addition, several samples (designated M5W) collected at 121°C
(250°F) were analyzed for total water-soluble sulfate and subse-
quent particulate mass determination by use of modified proce-
dures developed by the Texas Air Control Board (TACB).* This
method incorporates deionized water as the sample recovery sol-
vent and uses gravimetric analyses and an independent determina-
tion to measure water-soluble sulfates for subsequent derivation
of the mass of non-water-soluble particulate (matter that does
not contain any water-soluble sulfate). Ion chromatography was
used to measure water-soluble sulfate. Select sample fractions
also were analyzed for cation species to characterize the water-
soluble sulfate as other than sulfuric acid in the samples.
Each individual sample train was followed by a modified EPA
Method 8** impinger section to allow comparative analysis of
sulfates as sulfuric acid (H_SO.) and sulfur dioxide (S0_). Flue
gas temperature, moisture content, and composition [oxygen (0,,),
carbon dioxide (CO-), and carbon monoxide (CO)] were measured in
conjunction with the emission tests.
Mr. Winton Kelly, representing EPA, observed part of the
test program and provided overall project coordination and guid-
ance. Mr. Tim Tucker, representing Exxon, coordinated onsite
activities and unit operation.
*
TACB - Laboratory Division - Determination of Particulate in
Stack Gases Containing Sulfur Dioxide. December 1979.
**
40 CFR 60, Appendix A, Reference Method 8, July 1982.
1-3
-------
SECTION 2
PROCESS OPERATION
This methods development project was conducted on a fluid
catalytic cracking unit regenerator at Exxon's Baton Rouge refin-
ery. Exxon utilizes conventional regeneration techniques to
regenerate carbon-laden catalysts. Particulate and sulfur diox-
ide emissions are controlled by a jet-ejector venturi scrubber.
Carbon monoxide emissions are controlled by a CO boiler. A de-
tailed description of Exxon's process operation and control
equipment is not given in this report because of confidentiality
considerations.
2-1
-------
SECTION 3
SAMPLING LOCATION AND TEST AND ANALYTICAL METHODS USED
All samples were collected by means of four sampling trains,
each located at single points representing average velocities in
the FCCU final exit stack. The four-train sampling system was
used to conduct nine test runs for a total collection of 36
individual samples. Figure 3-1 presents a schematic of the
sampling site setup.
3.1 SAMPLING LOCATION
Testing was conducted at the FCCU scrubber exit stack as de-
picted in Figure 3-2. Four 7.0-cm (2.75-in.) i.d. sampling ports
were available at 90 degrees off-center. All four sampling ports
were used in this study. The sampling platform was approximately
61 meters (200 feet) above grade. The stainless steel sampling
port couplings were beveled inward by 0.64 cm (0.25 in.) to
produce an actual inside diameter of less than 7.0 cm (2.75 in.).
Inevitably, the tight fit caused some scraping of the sampling
port couplings during sampling train insertion and removal from
the stack.
3.2 SAMPLING METHODS
Flue gas samples were collected simultaneously from four
single points in the stack; each point represented similar
3-1
-------
SAMPLE
TRAIN D
CATWALK
SAMPLE
TRAIN C
SAMPLE
TRAIN A
SAMPLE
TRAIN B
Figure 3-1. Sampling train setup - plan view.
3-2
-------
CIRCULAR STAIRS-
CROSS-SECTION
.AMPLE POINTS
4.9-m (16-ft) i.d.
COUPLING LENGTH:30.5cm(12inO
4.9m(16ft)
i
6.1m(20ft) M.25 dd
24m(80ft) ^5 dd
SEPARATOR
•ELEVATOR
Figure 3-2. Sampling location,
3-3
-------
velocity pressures and temperatures. The desired sampling time
was 120 minutes, and readings of stack flue gas and sampling
train data were recorded at 10-minute intervals for each train.
In Run Nos. 4A-D, 6A-D, 1C, and 8A, the sampling time was reduced
because of problems with the sampling equipment or the weather.
Pitot tubes and thermocouples attached to each sampling probe
were used to set isokinetic sample rates for each train. Pro-
grammable calculators were used to determine sample rates. Prior
to sampling, velocity and temperature profiles were established
by use of procedures described in EPA Methods 1 and 2.* These
data were used to select the four sampling points.
Table 3-1 presents the sampling matrix performed during this
test series. The following are a brief descriptions of the
specific conditions for each train:
0 Method 5 - Designation M5
Filterable particulate was collected by use of a probe
and filter assembly heated to 121°C (250°F). Acetone
was used to rinse all sampling train components prior
to the filter.
0 Method SB - Designation M5B
Filterable particulate was collected by use of a probe
and filter assembly heated to 160°C (320°F). Acetone
was used to rinse all sampling train components prior
to the filter.
0 Method 5-450 - Designation M5-45Q
Filterable particulate was collected by use of a probe
and filter assembly heated to 232°C (450°F). Acetone
was used to rinse all sampling train components prior
to the filter.
*
40 CFR 60, Appendix A, Reference Methods 1 and 2, July 1983.
3-4
-------
TABLE 3-1. SAMPLING MATRIX
Run
No.
1
2
3
4
5
6
Sampling
Train No.
1A
IB
1C
ID
2A
2B
2C
2D
3A
3B
3C
3D
4A
4B
4C
4D
5A
5B
5C
5D
6A
6B
6C
6D
Sampling method3
M5
121°C (250°F)
X
X
X
X
X
X
MSB
160°C (320°F)
X
X
X
X
X
X
M5-450
232°C (450°F)
M5W
121°C (250°F)
X
X
X
X
X
X
X
X
X
X
X
X
(continued)
3-5
-------
TABLE 3-1 (continued)
Run
No.
7
8
9
Sampling
Train No.
7A
7B
7C
7D
8A
SB
8C
8D
9A
9B
9C
9D
Sampling method3
M5
121°C (250°F)
X
X
MSB
160°C (320°F)
X
X
X
X
X
X
M5-450
232°C (450°F)
X
X
X
X
M5W
121°C (250°F)
aM5 (Method 5) - Probe and filter heated to 121°C (250°F).
MSB (Method SB) - Probe and filter heated to 160°C (320°F).
M5 (Method 5) - Probe and filter heated to 232°C (450°F).
M5W (Method 5) - Probe and filter heated to 121°C (250°F); water rinse of
nozzle, probe, and front filter holder glassware.
3-6
-------
0 Method 5W - Designation M5W
Filterable particulate was collected by use of a probe
and filter assembly heated to 121°C (250°F). Deion-
ized, distilled water was used to rinse all sampling
train components prior to the filter.
In each train, the probe and filter temperatures were set at
a predetermined level and monitored during each test by means of
multiterminal digital indicators with thermocouple leads located
in each probe and immediately behind the Method 5 filter frits.
The back half of each sampling train consisted of a Modified
Method 8 assembly with either five or six impingers. An empty
impinger(s) was used to prevent back-half carryover resulting
from water condensation. An unheated Method 5 filter assembly
was inserted between the IPA and H-O- impingers to preclude any
sulfuric acid mist carryover. The contents of each impinger were
as follows:
Contents - Five Contents - Six
Impinger Impinger Runs Impinger Impinger Runs
1 Empty 1 Empty
2 150 ml 80% IPA 2 Empty
3 100 ml 10% H202 3 150 ml 80% IPA
4 100 ml 10% H202 4 100 ml 10% H2O2
5 300 grams silica gel 5 100 ml 10% H2O2
6 300 grams silica gel
All filters (Whatman Reeve Angel 934 AH) used in the Method
5 position were heated to 300°C prior to identification and tare
weighing.
Because the gas stream appeared saturated and, at times,
contained water droplets, two moisture determinations were made.
3-7
-------
The first determination involved calculations based on the water
collected in the sampling trains, and the second involved psy-
chrometric calculations. In each case, the lower value (satura-
tion at stack temperature) was used as the correct moisture
content in all calculations, as prescribed by EPA Method 4.*
Also, approximate measurements were made to determine the
degree of turbulent flow at the test location, as detailed in
Method 2 of the Federal Register.* Select traverse points were
checked by aligning the face openings of the pitot tube perpen-
dicular to the stack cross-sectional plane, designated "0 degree
reference." A null (zero) pitot reading obtained at 0 degree
reference indicates an acceptable flow condition at a given
point. A pitot tube angular notation of ±10 degrees is consid-
ered acceptable for achieving a null reading. At the scrubber
outlet test location, the degree of angular rotation was highly
variable and often greater than 10 degrees. Plume observation
also indicated cyclonic flow conditions in the stack. The rota-
tional, swirling flow would tend to cause larger particles and
particles entrained in water droplets to move toward the walls of
the stack. Since the four sampling trains were located approxi-
mately 1.2 m (4 ft) from the 4.9-m (16-ft) round stack wall at
each port, the spatial and temporal variations in velocity and
particulate concentration could have affected the within-run
precision and increased the variability in the reported results.
40 CFR 60, Appendix A, Reference Methods 2 and 4, July 1983.
3-8
-------
A flue gas grab sample collected during each of the first
five tests was also analyzed for 02, C02, and CO by means of an
Orsat analyzer. In the last four tests, a Fyrite analyzer was
used to measure the 0* and C02 content of the flue gas. The flue
gas composition was used to calculate stack gas molecular weight.
3.3 SAMPLE ANALYSIS
Table 3-2 lists all the samples collected during this test
program and their respective analyses.
3.3.1 Particulate Analysis
The filter particulate catch was placed in a tared glass
weighing dish, desiccated for 24 hours, and weighed until a
constant weight was achieved.* The probe rinse fraction was
transferred to a tared beaker, evaporated to dryness at ambient
temperature and pressure, desiccated for 24 hours, and weighed to
a constant weight.**
After this initial analysis, probe rinse and filter frac-
tions were heat-conditioned in an oven for 6 hours (except where
noted), according to the treatment sequence presented in Table
3-2. Each sample fraction was cooled and desiccated for 24 hours
after removal from the oven and weighed to a constant weight.**
Filter and acetone blanks were treated much that same as the
actual samples.
Previous data show that samples collected at 120°C will not
come to a constant weight. At least three separate weighings
were obtained, and the lowest weight achieved was reported as
the ambient weight.
* *
Criteria as specified in 40 CFR 60, Appendix A, EPA Reference
Method 5, July 1983.
3-9
-------
TABLE 3-2. ANALYTICAL PLAN
Run
No.
1
2
3
4
5
6
Sampling
Train No.
1A
IB
1C
ID
2A
2B
2C
2D
3A
3B
3C
3D
4A
4B
4C
4D
5A
5B
5C
5D
6A
6B
6C
6D
Sampling
method
M5
M5
M5W
M5W
M5W
M5W
M5B
MSB
M5W
M5W
M5
M5
M5W
M5W
MSB
MSB
MS
MS
M5W
M5W
M5W
MSB
MSB
M5W
Thermogravimetric analysis3
Ambient -»•
160° + 232°
-> 316°C
X (24)
X (24)
X
X (24)
Ambient -»•
232° +
316°C
X (24)
X (24)
X
X
Ambient
+ 316°C
X
X (24)
X (24)
X
Water-soluble
sulfate .
determination
X
X
X
X
X
X
X
X
X
X
X
X
Cations
by ICPC
X
X
X
X
X
X
X
X
(continued)
3-10
-------
TABLE 3-2 (continued)
Run
No.
7
8
9
Sampling
Train No.
7A
7B
7C
7D
8A
8B
8C
8D
9A
9B
9C
9D
Sampling
method
MSB
M5B
M5-450
M5-450
M5
M5
MSB
MSB
M5-450
MSB
MSB
M5-450
Thermogravimetric analysis3
Ambient ->
160° + 232°
+ 316°C
X
X
X (24)
X
X
X (24)
X
X (24)
Ambient -»•
232° +
316°C
Ambient
+ 316°C
X (24)
X
X (24)
X
Water-soluble
sulfate .
determination
Cations
by ICPC
Thermogravimetric conditioning of probe rinse and filter fractions at indicated
temperatures after initial desiccation and ambient weights were obtained. All
other samples were heat-conditioned for 6 hours, except where noted with (24).
These samples were conditioned for 24 hours.
In this procedure, the mass of total water-soluble sulfates in the sample was
determined and subtracted from the total sample mass.
cThese samples were analyzed for cations by ICP analytical techniques.
Note: All back halves represent a modified Method 8, with analysis for sulfates
as sulfuric acid and sulfur dioxide.
3-11
-------
3.3.2 Water-Soluble Sulfate Determination
This method is designed to determine the particulate catch
corrected for any water-soluble sulfate retained in the Method 5
sample fractions. As documented in previous studies, the con-
densible sulfate problem can be attributed to sulfuric acid.
This makes a direct gravimetric analysis difficult for two rea-
sons. First, sulfuric acid is a powerful desiccating agent
itself; therefore, if a significant amount of sulfuric acid is
present, the Method 5 criteria for constant weight of the partic-
ulate cannot be met. Second, the number of water molecules
associated with each sulfuric acid molecule is not consistent.
The water-soluble sulfate method developed by the Texas Air Board
was designed to overcome these problems. This method converts
any sulfuric acid present to a suitable form for accurate gravi-
metric analysis. Ammonium hydroxide is added to form ammonium
sulfate in the aqueous solutions. Ammonium hydroxide is used
because any excess reagent will evaporate. This procedure allows
the determination of the gross particulate (sulfate as ammonium
sulfate plus other particulate), the determination of sulfate as
ammonium sulfate from the Method 6 titration or ion chromatog-
raphy, and subsequently, the determination of non-water-soluble
sulfate particulate by subtraction of the sulfate (as ammonium
sulfate) from the gross particulate.
Each sample fraction plus blanks were handled and analyzed
as follows:
3-12
-------
0 Filter - The filter was cut into small pieces and
placed in a 125-ml Erlenmeyer flask with a standard
type joint equipped with an air condenser. The con-
tents of the shipping container were rinsed into the
flask. About 50 ml of distilled water was added and
the contents gently refluxed for 6 to 8 hours. The
solution was then cooled and diluted with water to
exactly 250 ml in a volumetric flask. This solution
was reserved for total soluble sulfate analysis, which
is described below.
0 Probe Rinse - The probe wash was poured into a 250-ml
volumetric flask. The sample bottle was rinsed with
distilled water and the rinsings were added to the
flask. The solution was then diluted to the mark with
distilled water (or, if greater than 250 ml, the volume
was measured). This solution was reserved for total
soluble sulfate analysis, which is described below.
Total Soluble Sulfate—
A 15-ml aliquot* was drawn from the settled samples (filter
and rinse) into separate sample containers with a clean, dry
pipet (only solution was transferred--no solid; if necessary, a
portion of the sample was centrifuged). The sulfate ion (SO ~)
concentration in each aliquot was determined by ion chromato-
graphy (1C). A syringe was used to inject 1 ml of the aliquot
into the 100-yl sample loop of the 1C. The conductivity response
of the sample was compared with the calibration curve to obtain
SO ~ concentration in ml/liter. Dilutions were prepared and
reanalyzed if the initial response was out of the linear cali-
bration range (0.1 to 15 mg/liter). Blank filter and water
samples were prepared and analyzed in the same manner as the
actual samples.
*
The pipet is not rinsed. This deviation from normal procedures
is necessary because the volume removed from the volumetric
flask is required in the calculations.
3-13
-------
Mass Determination—
Filter and Rinse Solution Preparation - The remaining con-
tents of each volumetric flask (235 ml) were poured into
separate tared 250-ml beakers, and the flask was rinsed with
distilled water to transfer all particulate matter. The
filter solution in Beaker A and the rinse solution in Beaker
B were evaporated to approximately 100 ml at 105°C and
allowed to cool before the next analysis was made.
Filter and Rinse Solution Analysis - Five drops of phenol-
phthalein indicator were added to all the tared beakers.
Concentrated NH.OH was then added drop by drop until the
solution turned pink. The samples were returned to the oven
and evaporated to dryness at 105°C, then cooled in a desic-
cator and weighed to a constant weight. Results were re-
ported to the nearest 0.1 mg. For this method, "constant
weight" means a difference of no more than 0.5 mg or 1
percent of the total weight less beaker and/or filter tare,
whichever is greater, between two consecutive weighings,
with no less than 6 hours of desiccation time between weigh-
ings.
Calculations—
Nomenclature—
FP = weight of particulate* on the filter in Beaker A,
mg
PRP = weight of probe rinse particulate* in Beaker B,
mg
NWSSP = weight of non-water-soluble sulfate particulate**,
mg
AS, = weight of ammonium sulfate in filter sample, mg
AS = weight of AS in probe rinse sample, mg
V = volume of solution evaporated in Beaker A (filter)
p or Beaker B (probe rinse), ml
C_o = concentration of sulfate in filter or probe rinse
4 solution aliquots, mg/liter
Particulate with H2SO4 converted to (NH4)2S04.
**
Particulate excluding water-soluble sulfates.
3-14
-------
Equations —
FP (mg) = gross weight Beaker A - tare weight (Eq. 1)
Beaker A - filter tare weight
PRP(mg) = gross weight Beaker B - tare weight (Eq. 2)
Beaker B
AS(mg) = Cso tag/liter) x V£vap (ml) x ml (Eq. 3)
Mass of Non-Water-Soluble Sulfate Particulate
The sum of the particulate* collected on the filter (FP) and
the particulate* collected in the probe rinse (PRP) is equal to
the sum of non-water-soluble sulfate particulate (NWSSP) and
ammonium sulfate (AS) in both samples:
FP + PRP = NWSSP + ASf + AS (Eq. 4)
The NWSSP can be found by rearranging the equation and
substituting appropriate values determined by Equations 1, 2,
and 3.
NWSSP = FP + PRP - AS. - AS (Eq. 5)
r pr
3.3.3 Sulfuric Acid Mist Analysis
The amount of sulfuric acid that passed through the particu-
late filter was initially determined through analysis of the IPA
solution recovered in the first impinger(s) and the backup fil-
ter. The volume of the sample solution was recorded. A 20-ml
aliquot of this solution was pipetted into a 250-ml Erlenmeyer
*
Particulate with H2S04 converted to (NHJ-SO-.
3-15
-------
flask with 2 to 3 drops of thorin indicator and titrated to a
pink end point by use of 0.0100 N barium perchlorate. Results
from the titrametric analysis were highly variable because the
amount of condensed water in the sample made it difficult to
distinguish the titrametric end point. Therefore, 10- to 20-ml
aliquots of each sample were diluted with 100 ml of deionized,
distilled water, and ion chromatography was used in the analyses
for sulfates (SO.~). A blank was titrated for each sample in the
same manner.
3.3.4 Sulfur Dioxide Analysis
A 20-ml aliquot of the hydrogen peroxide solution was
pipetted into a 250-ml Erlenmeyer flask with 80 ml of 100 percent
IPA and 2 to 3 drops of thorin indicator. The solution was then
titrated to a pink end point by use of 0.0100 N barium perchlo-
rate. Blanks were titrated in a similar manner. Several perox-
ide impingers were checked by 1C for comparative purposes. The
1C and titrametric analysis compared favorably.
3.3.5 Sulfate Analysis by Ion Chromatography
Samples were analyzed for total sulfates as SO. by use of
standard ion chromatography (1C) analytical techniques. The 1C
procedures were used to determine sulfate values obtained as part
of the water-soluble sulfate method. In addition, within-run
samples, heat-conditioned to 315°C, were extracted with distilled
water, as described in Method M5, and aliquots were analyzed for
total sulfates by 1C for comparative purposes.
3-16
-------
3.3.6 Cation Analysis by Inductively Coupled Plasma (ICP)
The extracts of selected filter and rinse particulate (see
Table 3-2) were analyzed for cations. Metallic ions and a cross-
check on sulfur were determined by ICP.
3-17
-------
SECTION 4
SUMMARY AND DISCUSSION OF TEST RESULTS
This section summarizes the results of the field sampling
program. The results are presented to allow both within-run and
between-run data comparisons with emphasis on the thermogravi-
metric and water-soluble sulfate test results. Appendix A
contains computer printouts and example calculations. Appendices
B and C contain the raw field and laboratory data sheets,
respectively. Appendix D details the sample and analytical
procedures used, and Appendix E addresses equipment calibration
guidelines and results.
4.1 SAMPLE DATA
Table 4-1 summarizes pertinent sample data. The actual
probe and filter temperatures, stack temperature, and moisture
content represent average values from each individual sampling
train. Isokinetic criteria defined in Reference Method 5* were
met in each case, with the exception of Sample Runs 2B, 3A, and
8A. Run No. 3A had a post-test leak rate three times greater
than the allowable limit because of a cracked piece of glassware
in the back half of the sampling train. During Run No. 8A, the
*
40 CFR 60, Appendix A, Reference Method 5, July 1983.
4-1
-------
TABLE 4-1. SUMMARY OF SAMPLE CONDITIONS
Test
No.
1
i
2
3
4
5
6
Date
(1983)
and
time
(24-h)
3/2
1056-
1257
3/3
0936-
1137
3/3
1414-
1615
3/4
1058-
1209
3/7
1024-
1225
3/7
1540-
1701
Tra i n
ID
1A
IB
1C
ID
2A
2B
2C
2D
3A
3B
3C
3D
4A
4B
4C
4D
5A
5B
5C
5D
6A
6B
6C
6D
Sample
type
M5
M5
M5W
M5W
M5W
M5W
M5B
M5B
M5W
M5W
M5
M5
M5W
M5W
M5B
M5B
M5
M5
M5W
M5W
M5W
M5B
M5B
M5W
Sample temperature, °C (°F)
Probe
Desired
121 (250)
121 (250)
121 (250)
121 (250)
121 (250)
121 (250)
160 (320)
160 (320)
121 (250)
121 (250)
121 (250)
121 (250)
121 (250)
121 (250)
160 (320)
160 (320)
121 (250)
121 (250)
121 (250)
121 (250)
121 (250)
160 (320)
160 (320)
121 (250)
Actual
131 (267)
138 (281)
120 (249)
128 (263)
124 (256)
129 (265)
161 (323)
163 (326)
127 (261)
125 (257)
126 (260)
122 (253)
126 (260)
125 (258)
164 (327)
165 (330)
124 (256)
129 (265)
130 (267)
128 (263)
125 (258)
168 (333)
168 (334)
124 (256)
Filter
Desired
121 (250)
121 (250)
121 (250)
121 (250)
121 (250)
121 (250)
160 (320)
160 (320)
121 (250)
121 (250)
121 (250)
121 (250)
121 (250)
121 (250)
160 (320)
160 (320)
121 (250)
121 (250)
121 (250)
121 (250)
121 (250)
160 (320)
160 (320)
121 (250)
Actual
134 (273)
134 (273)
129 (264)
128 (263)
129 (265)
130 (266)
167 (332)
162 (324)
128 (263)
126 (260)
125 (258)
127 (262)
126 (260)
127 (258)
170 (338)
165 (328)
131 (268)
133 (271)
129 (264)
128 (263)
122 (253)
157 (316)
156 (313)
122 (253)
Meter . .
volume,
dNm3 (dscf)
1.46 ( 51.66)
1.45 ( 51.28)
1.44 ( 50.71)
1.48 ( 52.14)
1.36 ( 48.06)
1.31 ( 46.24)
1.24 ( 43.93)
1.23 ( 43.33)
2.38 ( 83.87)
2.64 ( 93.35)
2.45 ( 86.67)
2.31 ( 81.43)
1.60 ( 56.46)
1.59 ( 56.00)
1.56 ( 55.21)
1.55 ( 54.65)
2.74 ( 96.86)
2.78 ( 98.21)
2.72 ( 96.14)
2.65 ( 93.73)
2.01 ( 71.13)
2.00 ( 70.90)
1.99 ( 70.20)
1.98 ( 70.02)
Average
stack
temper-
ature,
°C (°F)
70 / i co \
/ j \ loo )
74 (166)
73 (163)
71 (160)
77 (170)
71 (160)
Average
moisture
content, %
or o
C.D . J
25.6
25.6
26.6
24.3
27.5
(continued)
-------
TABLE 4-1 (continued)
Test
No.
7
8
Q
3
Date
(1983)
and
time
(24-h)
3/8
0934-
1135
3/8
1300-
1501
3/9
0945-
1146
Train
ID
7A
7B
7C
7D
8A
8B
8C
8D
9A
9B
9C
9D
Samplea
type
M5B
MSB
M5-450
M5-450
M5
M5
M5B
MSB
M5B
M5-450
M5B
M5-450
Sample temperature, °C (°F)
Probe
Desired
160 (320)
160 (320)
232 (450)
232 (450)
121 (250)
121 (250)
160 (320)
160 (320)
160 (320)
232 (450)
160 (320)
232 (450)
Actual
163 (325)
159 (319)
223 (434)
158 (317)
131 (268)
125 (258)
165 (329)
166 (330)
163 (325)
234 (455)
164 (326)
149 (300)
Filter
Desired
160 (320)
160 (320)
232 (450)
232 (450)
121 (250)
121 (250)
160 (320)
160 (320)
160 (320)
232 (450)
160 (320)
232 (450)
Actual
167 (333)
166 (331)
205 (403)
234 (453)
134 (273)
127 (261)
163 (325)
166 (330)
160 (320)
200 (393)
162 (324)
213 (416)
Meter ^
volume.
dNm3 (dscf)
2.84 (100.26)
2.87 (101.42)
2.18 ( 76.79)
2.83 (100.00)
2.38 ( 84.19)
2.74 ( 96.61)
2.60 ( 91.71)
2.71 ( 95.74)
2.86 (101.02)
2.85 (100.87)
2.71 ( 95.84)
2.88 (101.58)
Average
stack
temper-
ature,
°C (°F)
78 (172)
77 (170)
77 9 M 71 '
' / . f. \ it i ,
Average
moisturec
content, %
23.3
22.9
?A 0
m , o
I
U)
Designation:
M5 = Reference Method 5 - desired probe and filter temperature, 121°C (250°F).
M5B = Reference Method 5B - desired probe and filter temperature, 160°C (320°F).
M5-450 = Modified Method 5 - desired probe and filter temperature, 232°C (450°F).
M5W = Modified Method 5 - desired probe and filter temperature, 121°C (250°F) with a water rinse of
the probe and analysis for total water soluble sulfate and corresponding mass determination.
Meter volume measured.in dry normal cubic meters (dry standard cubic feet).
°Average of four sampling trains.
-------
vacuum suddenly dropped. Sampling was immediately discontinued.
An examination of the sampling train, revealed a broken glass
connector between the back-half filter holder and H-O- impinger.
Sample volumes for these runs were corrected for leakage, which
caused the nonisokinetic test condition for these runs. As noted
in Section 3, the desired sampling time was 120 minutes. Test
4A-D was run for 70 minutes, Test 6A-D for 80 minutes, Test 1C
for 90 minutes, and Test 8A for 103 minutes.
4.2 THERMOGRAVIMETRIC ANALYTICAL RESULTS
Table 4-2 presents the thermogravimetric analytical results.
The filterable particulate reported in Table 4-2 represents
material collected in the sampling probe and on the filter for
each sample type (M5, M5B, M5-450). Also reported are the ambi-
ent filter weights for samples designated M5W. All weights are
reported in milligrams (mg) and sample concentrations in milli-
grams per dry normal cubic meters (mg/dNm3).
As previously noted, samples were heat-conditioned at each
designated interval for 6 hours, except for select samples that
were heated for 24 hours for comparative purposes. Table 4-3
provides within-run comparisons on a total weight basis, and
Table 4-4 presents a comparison of weight loss in milligrams
above 160°C (320°F). Because of the variability of reported
results, an absolute assessment of sample conditioning time (6
hours versus 24 hours) is not possible. Data from Runs 2, 7, and
8 (Table 4-3) indicate no significant difference in weight loss
4-4
-------
TABLE 4-2. SUMMARY OF THERMOGRAVIMETRIC ANALYTICAL RESULTS
Test
No.
Train
ID
1Aa
lBd
1C
ID
2A
2B
2C
2Da
3A
3B3
3Cd
3Dd
4A
46 a
4Ca
4D
5A
5Bd
5C
5D
6A
6Ba
6C
6D
7A
7Ba
7Cd
7D
Sample
type
M5
M5
M5W
M5W
M5W
M5W
MSB
M5B
M5W
M5W
M5
M5
M5W
M5W
MSB
M5B
M5
M5
M5W
M5W
M5W
M5B
MSB
M5W
MSB
MSB
M5-450
M5-450
Filterable particulate following conditioning at
indicated temperatures, mg
Ambient
Probe
73.7
18.6
-
^_
_
29.0
26.7
_
.
83.9
40.1
—
_
36.1
30.2
57.7
24.5
-
-
24.4
34.7
-
46.5
27.6
20.9
24.9
Filter
136.3
147.3
113.0
142.2
142.7
140.1
108.1
136.0
253.8
261.4
192.0
245.8
152.9
162.6
126.6
148.2
235.0
248.6
204.7
244.2
171.7
181.8
150.2
172.7
247.7
250.8
143.3
226.6
160°C (320°F)
Probe
61.6
11.2
-
-
_
_
16.5
17.4
—
_
_
-
_
_
_
-
_
_
-
-
_
-
-
37.5
22.1
13.8
13.1
Filter
134.6
146.3
-
-
—
«
106.2
133.4
—
_
_
-
_
_
_
-
^
_
-
-
_
-
-
244.5
246.8
142.4
224.6
232°C (450°F)
Probe
55.3
8.2
-
—
_
14.4
14.3
^
_
61.8
20.8
_
_
22.7
16.9
_
-
-
-
_
-
-
35.5
20.9
10.6
13.2
Filter
129.2
137.6
-
_
_
106.5
130.6
_
_
182.8
228.1
_
_
114.0
143.0
_
_
-
-
_
-
-
237.7
239.7
128.2
215.1
316°C (600°F)
Probe
51.1
7.2
-
_
_
12.6
14.4
_
_
61.5
20.1
_
_
22.3
15.3
38.8
13.5
-
-
12.3
18.4
-
33.3
19.9
10.4
12.4
Filter
126.6
136.8
-
_
_
103.0
129.1
_
_
174.6
227.0
_
_
114.0
138.3
200.3
206.7
-
-
168.1
136.3
-
221.2
218.8
126.7
213.1
(continued)
4-5
-------
TABLE 4-2 (continued)
Test
No.
8
9
Train
ID
8Aa u
8Ba'b
8Ca
8Da
9Aa
9Ba
9Ca
9D
Sample
type
MS.
M5b
MSB
MSB
MSB
M5-450
M5B
M5-450
Filterable parti cul ate following conditioning at
indicated temperatures, mg
Ambient
Probe
53.5
339.1
38.5
38.9
90.1
44.2
35.6
43.1
Filter
173.4
249.2
167.0
209.7
305.2
280.2
244.5
293.6
160°C (320°F)
Probe
42.9
325.5
25.5
28.7
™
Filter
170.3
245.2
164.4
205.0
"**
232°C (450°F)
Probe
37.6
319.5
22.4
25.7
*"
Filter
149.6
215.7
162.2
183.4
™
316°C (600°F)
Probe
36.4
318.1
20.8
22.1
66.3
21.3
14.1
26.9
Filter
147.2
212.5
144.6
180.0
286.9
269.2
226.2
275.6
Heated for 24 hours; all others heated for 6 hours.
Vrobe rinse contamination not included in averages.
4-6
-------
TABLE 4-3. DIRECT COMPARISON OF PARTICULATE WEIGHTS
AT 6- AND 24-HOUR HEAT-CONDITIONING PERIODS
Run No.
2C
2D
4C
4D
5A
5B
6B
6C
7C
7D
8C
8D
Heat
time, h
6
24
24
6
6
24
24
6
24
6
6
24
Sample
type
MSB
MSB
MSB
MSB
M5
MS
MSB
MSB
M5-450
M5-450
MSB
MSB
Ambient
weight, mg
137.1
162.7
162.7
178.4
292.7
273.1
206.2
184.9
164.2
251.5
205.5
248.6
Weight after heating .
at indicated temperature, mg
160°C
122.7 (11)
150.8 (7)
NA
NA
NA
NA
NA
NA
156.2 (5)
237.7 (5)
189.9 (8)
233.7 (6)
232°C
120.9 (12)
144.9 (11)
136.7 (16)
159.9 (10)
NA
NA
NA
NA
138.8 (15)
228.3 (9)
184.6 (10)
209.1 (16)
316°C
115.6 (16)
143.5 (12)
136.3 (16)
153.6 (14)
239.1 (18)
220.2 (19)
180.4 (13)
154.7 (16)
137.1 (17)
225.5 (10)
165.4 (20)
202.1 (19)
aAmbient weight (rinse and filter fractions) in milligrams.
Weight after heat treatment at indicated temperature. The numbers in
parentheses represent the relative percent weight loss for each heat interval
compared with the ambient weights.
NA = Not applicable (see Analytical Plan, Section 3, Page 3-10).
4-7
-------
TABLE 4-4. COMPARISON OF WEIGHT LOSSES ABOVE 160°C
Run No.
2C
2D
7C
7D
8C
8D
Sample
type
MSB
MSB
M5-450
M5-450
MSB
MSB
Heating
time, h
6
24
24
6
6
24
Ambien.t
weight, rug
137.1
162.7
164.2
251.5
205.5
248.6
Total weight
loss to 315°C
(600°F), mg
21.5
19.2
27.1
26.0
40.1
46.5
Weight loss
above 160°C
(320°F), mg
7.1
7.3
19.1
12.2
24.5
31.6
4-8
-------
at 160°C between the 6- and 24-hour samples. As presented in
Table 4-4, however, the samples heated for 24 hours showed higher
weight loss above 160°C than those heated for 6 hours regardless
of the initial ambient weight. Also, the total weight loss to
315°C (600°F) is comparable (Table 4-4) for each conditioning
time. Based on these results, a thermogravimetric conditioning
period of at least 6 hours seems appropriate.
Table 4-5 presents a comparison of particulate concentra-
tions after heat conditioning at the indicated temperatures. The
average concentrations and standard deviations are given in
milligrams per dry normal cubic meter for all samples of a simi-
lar type and temperature. The number of data points at each
temperature is also shown. Table 4-6 summarizes the relative
percent weight loss by sample fraction at the indicated tempera-
tures, and Figure 4-1 graphically depicts these data for Run 7.
Table 4-7 summarizes the EPA Method 8* analytical results for
sulfuric acid (H-SO.) and sulfur dioxide (S0«).
Data presented in Tables 4-5 and 4-6 for each sample type
exhibit the same basic characteristics as samples collected at
other FCCU sources and analyzed in a similar manner. The data
suggest, however, that biases in particulate measurements caused
by sulfuric acid (H-SO.) are less significant than observed for
samples collected at other FCCU sources evaluated under this task
assignment. The ambient concentration of the M5 samples averaged
40 CFR 60, Appendix A, Reference Method 8, July 1982.
4-9
-------
TABLE 4-5. COMPARISON OF FILTERABLE PARTICULATE CONCENTRATION
AFTER CONDITIONING AT TEMPERATURES OF 160°, 232°, AND 315°C
Run
No.
1A
IB
3C
3D
5A
SB
8A,
8Bh
Sample
ID
M5
M5
M5
M5
M5
M5
M5
M5
2C
2D
4C
40
6B
6C
7A
7B
8C
8D
9A
9C
MSB
MSB
MSB
MSB
MSB
M5B
MSB
MSB
MSB
MSB
M5E
M5B
7C
7D
9B
9D
M5-450
M5-450
M5-450
M5-450
Ambient
Total
weight ,
mg
210.0
165.9
275.9
285.9
292.7
273.1
226.9
588.3
Concen-
tration,
mg/dNm3
143.6
114.3
112.4
124.0
106.7
98.2
95.2
215.1
Average = 113.5
o^ = 16.5
Nd = 7
137.1
162.7
162.7
178.4
206.2
184.9
294.2
278.4
205.5
248.6
395.3
280.1
110.2
132.6
104.1
115.3
102.7
93.0
103.6
97.0
79.1
91.7
138.2
103.2
Average = 105.9
°A = 16.7
Nd = 12
164.2
251. S
324.4
336.7
75.5
88.8
113.6
117.1
Average = 98.8
c^ = 20.0
Nd = 4
160°C
Total
weight,
mg
196.1
157.5
-
-
213.2
570.7
Concen-
tration
mg/dNm3
134.1
108.5
-
-
89.4
208.6
Average = 110.7
o = 22.4
N = 3
122.7
150.8
-
-
282.0
268.9
189.9
233.7
-
98.7
122.9
-
-
99.3
93.6
73.1
86.2
-
Average = 95.6
o = 16.5
N = 6
156.2
237.7
-
71.8
84.0
-
Average = 77.9
o = 8.6
N = 2
232°C
Total
weight,
mg
184.5
145.8
244.6
248.9
.
187.2
535.2
Concen-
tration,
mg/dNm3
126.1
100.4
99.7
108.0
-
78.5
195.7
Average = 102.5
o = 17.1
N = 5
120.9
144.9
136.7
159.9
-
273.2
260.6
184.6
209.1
-
97.2
118.1
87.4
103.3
-
96.2
90.8
71.1
77.1
-
Average = 92.7
o = 14.8
N = 8
138.8
228.3
-
63.8
80.6
-
Average = 72.2
o = 11.9
N = 2
315°C
Total
weight,
mg
177.7
144.0
236.1
247.1
239.1
220.2
183.6
530.6
Concen-
tration,
mg/dNm3
121.5
99.2
96.2
107.2
87.2
79.2
77.0
194.0
Average =95.4
o = 15.8
N = 7
115.6
143.5
136.3
153.6
180.4
154.7
254.5
238.7
165.4
202.1
354.7
240.7
92.9
117.0
87.2
99.3
89.9
77.8
89.6
83.1
63.7
74.6
124.0
88.7
Average = 90.7
o = 16.8
N = 12
137.1
225.5
290.5
302.5
63.1
79.6
101.7
105.2
Average = 87.4
o = 19.8
N = 4
Total filterable catch (probe and filter).
Probe rinse contamination; value not included in group averages.
Standard deviation with N-l weighting of data.
Number of data points.
4-10
-------
TABLE 4-6. FILTERABLE PARTICULATE RELATIVE PERCENT WEIGHT LOSS
AFTER CONDITIONING AT TEMPERATURES 160°, 232°, AND 315°C
Run
No.
1A
IB
3C
3D
5A
5B
8A
88 a
2Ca
2Da
4C
40
6B
6C
7A
7B
8C
8D
9A
98
7C
70
98
90
Sample
10
M5
M5
M5
M5
M5
M5
M5
MS
MSB
MSB
M5B
MSB
MSB
MSB
MSB
MSB
MSB
MSB
MSB
MSB
M5-450
M5-450
M5-450
M5-450
Ambient
temperature
Cone. , mg
Rinse
73.7
18.6
83.9
40.1
57.7
24.5
53.5
339.1
29.0
26.7
36.1
30.2
24.4
34.7
46.5
27.6
38.5
38.9
90.1
35.6
20.9
24.9
44.2
43.1
Filter
136.3
147.3
192.0
245.8
235.0
248.6
173.4
249.2
108.1
136.0
126.6
148.2
181.8
150.2
247.7
250.8
167.0
209.7
305.2
244.2
143.3
226.6
280.2
293.6
160"C (320°F)
Cone. , mg
Rinse
61.5
11.2
_
-
42.9
325.5
16.5
17.4
-
_
37.5
22.1
25.5
28.7
-
13.8
13.1
-
Filter
134.6
146.3
-
-
170.3
245.2
106.2
133.4
-
-
244.5
246.8
164.4
205.0
_
142.4
224.6
-
Wt. loss, %
Rinse
17
40
-
-
20
43
35
-
-
19
20
34
26
.
34
47
-
Filter
1
1
-
-
2
2
2
2
-
-
1
2
2
2
_
1
1
-
232"C (450°F)
Cone. , mg
Rinse
55.3
8.2
61.8
20.8
-
37.6
319.5
14.4
14.3
22.7
16.9
-
35.5
20.9
22.4
25.7
-
10.6
13.2
-
Filter
129.2
137.6
182.8
228.1
-
149.6
215.7
106.5
130.6
114.0
143.0
-
237.7
239.7
162.2
183.4
-
128.2
215.1
-
Wt. loss, J
Rinse
25
56
26
48
-
30
50
46
37
44
-
24
24
42
34
-
49
47
-
Filter
5
7
5
7
-
14
13
2
4
10
4
-
4
4
3
13
-
11
5
-
315°C (600°F)
Cone. , mg
Rinse
51.1
7.2
61.5
20.1
38.8
13.5
36.4
318.1
12.6
14.4
22.3
15.3
12.3
18.4
33.3
19.9
20.8
22.1
67.8
14.1
10.4
12.4
21.3
26.9
Filter
126.6
136.8
174.6
227.0
200.3
206.7
147.2
212.5
103.0
129.1
114.0
138.3
168.1
136.3
221.2
218.8
144.6
180.0
286.9
226.6
126.7
213.1
269.2
275.6
Wt. loss, X
Rinse
31
61
27
50
33
45
32
57
46
38
49
49
47
28
28
46
43
25
60
50
50
52
38
Filter
7
7
9
8
15
17
15
15
5
5
10
7
8
9
11
13
13
14
6
7
12
6
I '
• "lent "e'1ht
Probe rinse contaminated.
-------
120
100
80
60
o:
-------
TABLE 4-7, SUMMARY OF
AND SC>2 ANALYTICAL DATA
Test
No.
7
f
Train
ID
1A
IB
1C
ID
2A
2B
2C
2D
3A
3B
3C
3D
4A
4B
4C
4D
5A
SB
5C
5D
6A
6B
6C
6D
7A
7B
7C
7D
8A
8B
8C
8D
9A
9B
9C
9D
Sample
type
MS
MS
M5W
MSW
M5W
MSW
MSB
MSB
MSW
MSW
MS
MS
MSW
MSW
MSB
MSB
MS
MS
MSW
MSW
MSW
MSB
MSB
MSW
MSB
MSB
MS -450
M5-450
MS
MS
MSB
MSB
MSB
M5-450
MSB
M5-450
Total HoSO,8
mg
210
39.7
74.8
57.8
45.7
63.0
63.4
72.7
252
253
137
179
80.5
40.2
59.8
68.0
23.2
12.1
-
32.4
22.4
25.2
17.3
27.6
17.3
31.1
34.5
45.5
23.2
24.3
37.4
29.6
320
53.6
26.7
24.2
Ihg/aNm*
143.8
27.4
51.9
39.1
33.6
48.1
51.1
59.1
105.9
95.8
55.9
77.5
50.3
25.3
38.3
43.9
8.5
4.4
-
12.2
11.1
12.5
8.7
13.9
6.1
10.8
15.8
16.1
9.8
8.9
14.4
10.9
111.9
18.7
9.9
8.4
Total SO.,6
mg
23.0
37.1
11.9
13.5
32.0
8.3
9.7
23.2
0.5
2.6
1.8
1.8
10.5
15.4
6.4
10.4
414.8
493.0
392.8
443.0
320.0
320.0
289.3
317.7
525.5
525.2
324.2
501.9
272.2
349.3
271.4
304.1
358.6
540.9
504.6
561.5
mg?dNms
15.8
25.6
8.3
9.1
23.5
6.3
7.8
18.9
<1.0
<1.0
<1.0
<1.0
6.6
9.7
4.1
6.7
151.4
177.3
144.4
167.2
159.2
159.2
145.4
160.5
185.0
183.0
148.7
177.3
114.4
127.5
104.4
112.2
125.4
189.1
186.2
195.0
Total sulfur
mg
80.2
31.6
30.4
25.7
31.0
24.8
25.6
35.4
82.6
84.0
45.7
59.4
31.6
20.9
22.8
27.4
215
251
197C
232
167
168
150
168
269
273
174
266
144
183
148
162
284
288
261
289
mg/dNm3
54.8
21.7
21.2
17.4
22.7
18.9
20.6
28.8
34.8
31.8
18.6
25.8
19.8
13.1
14.6
17.7
78.4
90.1.
72. 2C
87.5
83.1
83.9
75.7
84.7
94.6
95.1
79.8
94.0
60.3
66.8
57.0
59.7
99.3
100.9
96.3
100.4
Total H2SO. (SO-j/H^SO. mist) analyzed by ion chromatography. Runs 1 through 4 are
considered void because of contamination of sample solutions (see Page 4-15).
Total SO- analyzed per Method 8 (40 CFR 60, Appendix A, Reference Method 8,
July 1982).
cBased on S02 values only.
4-13
-------
113.5 mg/dNm3 with a standard deviation of 16.5 mg/dNm3, compared
with 105.9 mg/dNm3 and 16.7 mg/dNm3 for the MSB runs and 98.8
V
mg/dNm3 and 20.0 mg/dNm3 for the M5-450 runs. The average ambi-
ent M5 concentration was approximately 7 percent higher than the
MSB and 15 percent higher than the M5-450 sample concentrations
at ambient conditions. Considering the variability in results
for each sample type as exhibited by the group standard devia-
tions, the ambient weight results are reasonably comparable for
each sample type.
Correspondingly, the M5 concentration after heating to 315°C
averaged 95.4 mg/dNm3 with a standard deviation of 15.8 mg/dNm3,
compared with 90.7 mg/dNm3 and 16.8 mg/dNm3 for the MSB runs and
87.4 mg/dNm3 and 19.8 mg/dNm3 for the M5-450 runs. On a concen-
tration basis, the MS runs exhibited a total weight loss averag-
ing 16 percent when heated to 315°C, compared with 14 percent for
MSB and 11 percent for M5-450.
In the M5 runs for which gravimetric data are available
after heating to 160°C, the maximum relative percent weight loss
of the filter fraction was 2 percent, which is comparable to the
available MSB and M5-450 filter data (Table 4-6). The percent
weight loss of the filter fraction heated to 315°C ranged from 7
to 17 percent for the M5 runs, from 5 to 14 percent for MSB, and
from 4 to 12 percent for M5-450. In general, the greater filter
weight loss occurred at a conditioning temperature of 232°C.
Since the thermogravimetric procedure would be expected to elim-
inate the H.SO. and associated water at 160°C, the data suggest
4-14
-------
that residual H2S04 or other sulfate species not volatilized at
160°C are being eliminated at the higher conditioning tempera-
tures. The probe rinse fraction for each sample type consist-
ently showed higher percent weight losses than the filter frac-
tions, but the total catch in the probe was significantly less
than the filter fraction and resulted in a distortion of the
rinse weight loss data when compared with the filter data. With
few exceptions, the largest percent weight loss from the rinse
fractions for each sample type occurred at 160°C and additional
losses were consistently observed at the higher conditioning
temperatures. Figure 4-1 shows a comparison of the Method 5B
data and Method 5-450 data for Run 7. These data show a rela-
tively consistent difference (18 mg/dNm3) at each conditioning
temperature. The data also indicate little bias due to sulfuric
acid (<5 percent) and little bias due to other condensible matter
(<14 percent total weight loss). The average difference of 18
mg/dNm3 between these two methods for this run could be attribut-
able to particulate stratification in the stack (see Section 3).
The Method 8 analytical results for sulfates as H-SO. and
S02 were shown in Table 4-7. The results of Tests 1 through 4
and 9A are considered void because traces of hydrogen peroxide
(H_0_) found in the isopropanol sample fractions caused an
anomaly in reported results. The IPA fractions were initially
titrated per Method 8, but the titrametric end points were diffi-
cult to distinguish and results were highly variable. These
sample fractions were then reanalyzed by ion chromatography (1C)
4-15
-------
and the 1C results are reported. The reported SO- titrametric
results were also checked by 1C and these results compared favor-
ably. Although the H-SO. and S02 data for Tests 1 through 4 are
considered void, the total sulfur measured in the back half of
the sampling train should not be affected by the presence of H-O-
in the IPA. Note that the total sulfur reported for Tests 1
through 4 is less than that reported for Tests 5 through 9; thus,
the increase in total sulfur is probably due to process variation
or minor upset condition. This increase in total sulfur for Runs
5 through 9 corresponds to an overall increase in the filterable
particulate catch for each sample type in these tests. No sig-
nificant difference was noted in the weight loss for these sam-
ples, however, compared with data from Runs 1 through 4. The
total amount of H SO. measured during Runs 5 through 9 is sig-
nificantly less than the H_SO. measured at similar sources evalu-
ated under this task assignment. Also, no consistent correlation
exists between increased sample temperature and a corresponding
increase in back-half H-SO. content for these sample runs. The
H?SO. catch averaged 23.5 mg during the M5 and M5W runs compared
with 26.3 mg during the MSB runs. During the M5-450 runs, H-SO.
averaged 39.5 mg in the back half, but the variability in re-
ported results (1C and D and 9B and D) and the limited amount of
data preclude making an accurate assessment of sample temperature
and back-half H-SO. content.
Tables 4-8 and 4-9 present precision estimates for the heat
treatments evaluated. In Table 4-8, each group represents two
4-16
-------
TABLE 4-8. STATISTICAL DATA FOR GROUPED RUNS AFTER CONDITIONING AT INDICATED TEMPERATURES
I
M
*J
Run No.
1A-B
3C-D
5A-B
8A-Bd
2C-0
4C-D
6B-C
7A-B
8C-D
9A-C
7C-D
9B-D
Sample
type
M5
M5
MS
M5
MSB
MSB
MSB
MSB
MSB
MSB
M5-450
M5-450
X a
A »
mg/dNm3
129.0
118.2
102.5
-
121.4
109.7
97.9
100.3
85.4
120.7
82.2
115.4
Ambient
b
o,
mg/dNm3
20.7
8.2
6.0
-
15.8
7.9
6.9
4.7
8.9
24.7
9.4
2.5
CV,C
%
16.1
6.9
5.9
-
13.0
7.2
7.0
4.7
10.4
20.5
11.4
2.2
16C
X,
mg/dNm3
121.3
-
-
-
110.8
-
-
96.5
79.7
-
77.9
-
°C (320°F)
o,
mg/dNm3
18.1
-
-
-
17.1
-
-
4.0
9.3
-
8.6
-
cv,
%
14.9
-
-
-
15.4
-
-
4.2
11.7
-
11.1
-
2.
X,
mg/dNm3
113.3
103.9
-
-
107.7
95.4
-
93.5
74.1
-
72.2
-
52°C (450°F)
o,
mg/dNm3
18.2
5.9
-
-
14.8
11.2
-
3.8
4.2
-
11.9
-
cv,
%
16.1
5.7
-
-
13.7
11.7
-
4.1
5.7
-
16.5
-
316°C (600°F)
X,
mg/dNm3
110.4
101.7
83.2
-
105.0
93.3
83.9
86.4
69.2
106.4
71.4
103.5
0.
mg/dNM3
15.8
7.8
5.7
-
17.0
8.6
8.6
4.6
7.7
25.0
11.7
2.5
cv,
%
14.3
7.7
6.9
-
16.2
9.2
10.3
5.3
11.1
23.5
16.4
2.4
Mean filterable concentration.
Within-run standard deviation with N-l weighting for sample data.
Coefficient variance is the standard deviation expressed as a percent of the mean concentration.
Run 8B considered void due to probe rinse contamination. Runs 8A and B excluded from statistical analysis.
-------
TABLE 4-9. SUMMARY OF PRECISION ESTIMATES AFTER CONDITIONING AT INDICATED TEMPERATURES
i
M
OO
Run No.
1,3,5
2,4,6,
7,8,9
7,9
Sample
type
M5
M5B
M5-450
-
mg/dNm3
116.5
105.9
98.8
Ambient
t_
0,
mg/dNm3
11.6
Nd = 6
11.5
N = 12
6.0
N = 4
CV,C
(V
10.0
10.8
6.0
16(
X,
mg/dNm3
121.3
95.6
77.9
)"C (320°F)
o ,
mg/dNm3
18.1
N = 2
10.1
N = 6
8.6
N = 2
CV,
14.9
10.6
11.0
2
X,
mg/dNm3
108.6
92.7
72.2
32°C (450°
o,
mg/dNm3
12.1
N = 4
8.5
N = 8
11.9
N = 2
n
cv,
11.1
9.2
16.5
31
X,
mg/dNm3
98.4
90.7
87.4
6°C (600°F
o,
mg/dNM3
9.8
N = 6
11.9
N = 12
7.1
N = 4
)
cv,
0'
9.9
13.1
8.1
Mean filterable concentration based on grouped run values. Test No. 8 not included in calculations.
Mean standard deviation of grouped runs (-j^-)-
°Mean coefficient variation (percent) calculated using the mean standard deviation and the filterable concentration of grouped runs.
N = Number of data points.
-------
simultaneous runs of the same sample. Presented for each run
group and temperature are the mean filterable concentration, the
standard deviation with N-l weighting for sample data, and the
percent coefficient of variation (CV), which expresses the stan-
dard deviation as a percent of the mean concentration. Table 4-9
summarizes precision estimates for M5, MSB, and M5-450 test data
at each conditioning temperature. The mean filterable concentra-
tions were calculated by averaging the individual run data to
minimize roundoff errors. The mean standard deviations were
calculated by averaging standard deviation values for each set of
grouped runs (Table 4-8) to minimize the effect of temporal
variation in emissions. In this way, the mean standard deviation
of the grouped runs (a in Table 4-9) more accurately reflects
method precision than does the standard deviation of individual
run concentrations (o in Table 4-5). The number of data points
included in each calculation is shown to assist in evaluating the
precision estimates.
The mean standard deviation for three M5 run groups (Runs 8A
and B excluded) was 11.6 mg/dNm3 at ambient conditions, which
corresponds to a mean coefficient of variation (CV) of 10.0
percent. The six MSB run groups had a mean standard deviation of
11.5 mg/dNm3 and a corresponding mean CV of 10.8 percent at
ambient conditions. The two M5-450 run groups had a mean stan-
dard deviation of 6.0 mg/dNm3 and a corresponding mean CV of 6.0
percent at ambient conditions.
4-19
-------
The mean standard deviation for one M5 run group heated to
160°C (320°F) was 18.1 mg/dNm3, and the corresponding CV was 14.9
to-
percent. For the three M5B run groups heated to 160°C, the mean
standard deviation and CV were 10.1 mg/dNm3 and 10.6 percent,
respectively. The one M5-450 run group heated to 160°C showed a
mean standard deviation of 8.6 mg/dNm3 and a CV of 11.0 percent.
Precision data for the two M5 run groups heated to 232°C
(450°F) included a mean standard deviation of 12.1 mg/dNm3 and a
CV of 11.1 percent. For the four MSB run groups, the average
standard deviation and CV were 8.5 mg/dNm3 and 9.2 percent. The
one M5-450 run showed a mean standard deviation of 11.9 mg/dNm3
and a CV of 16.5 percent.
Precision data for the three M5 groups heated to 316°C
(600°F) included a mean standard deviation of 9.8 mg/dNm3 and a
CV of 9.9 percent. Two of the three run groups had a mean stan-
dard deviation of less than 8.0 mg/dNm3 and a corresponding CV of
less than 8 percent. For the MSB samples, the mean standard
deviation and CV were 11.9 mg/dNm3 and 13.1 percent, respec-
tively. Four of the six MSB runs, however, had a mean standard
deviation of less than 9 mg/dNm3 and a corresponding CV of less
than 10 percent. For the M5-450 runs, the mean standard devia-
tion and CV were 7.1 mg/dNm3 and 8.1 percent.
All of these statistical results indicate a variable degree
of precision for the majority of samples. Generally, the within-
run agreement was expected to improve after each stage of heat
treatment due to elimination of sulfate biases. The precision
4-20
-------
estimates for each sample type remained relatively consistent,
however, and actually decreased in some cases, particularly at
the conditioning temperatures of 160° and 232°C. At 315°C, the
precision estimates were comparable to those obtained at ambient
conditions for each sample type. As noted in Section 3, the
scrubber outlet gas stream contained waste droplets and exhibited
cyclonic flow conditions as determined by plume observation and
turbulent flow data obtained by means of procedures described in
EPA Reference Method 2.* As a result of these flow conditions,
spatial and temporal variations in gas velocity and particulate
concentration could have affected the within-run precision and
thus increased the degree of variability in reported results.
4.3 WATER-SOLUBLE SULFATE ANALYTICAL DATA
Table 4-10 summarizes results from the water-soluble sulfate
analysis performed on the indicated samples. Because particulate
cannot be determined gravimetrically in the presence of sulfuric
acid due to the inexact amount of water retained by the acid, the
method is designed to convert the acid to a nonhydroscopic,
nonvolatile product—in this case ammonium sulfate. The acid is
converted to ammonium sulfate, and the weight of ammonium sulfate
formed is subtracted from the total weight. The reported values
represent the particulate concentration corrected for total
water-soluble sulfates determined by ion chromatography. The
TACB method also allows the use of a barium perchlorate titra-
tion, as described in EPA Method 6.* Previous studies have shown
*
40 CFR 60, Appendix A, Reference Methods 2 and 6, July 1983.
4-21
-------
TABLE 4-10. SUMMARY OF WATER-SOLUBLE SULFATE ANALYTICAL RESULTS
Test No.
1C
ID
2A
2B
3A
3B
4A
4B
5C
5D
6A
6D
Sample
ID
M5W
M5W
M5W
M5W
M5W
M5W
M5W
M5W
M5W
M5W
M5W
M5W
Total
NWSSP,
mg
105.2
126.7
131.5
130.8
239.6
230.7
147.5
142.0
162.3
212.2
152.2
154.6
Average =84.9
o = 12.2
N = 12
Concen-
tration,
mg/dNm3
73.3
85.8
96.6
99.9
100.9
87.3
92.3
89.6
59.6
80.0
75.6
78.0
Statistical data
for grouped runs
x,b
mg/dNm3
79.6
98.3
94.1
91.0
69.8
76.8
o,C
mg/dNm3
8.8
2.3
9.6
1.9
14.4
1.7
CV,d %
11.1
2.4
10.2
2.1
20.7
2.2
Total non-water-soluble sulfate particulate matter (probe and filter
fractions) determined by the modified Texas Air Board analytical method.
Mean filterable concentration.
cStandard deviation with N-l weighting of data.
Coefficient of variance is the standard deviation expressed as a percentage
of the mean concentration.
4-22
-------
that titrametric and 1C analyses yield comparable results. '
Section 3 and Appendix D of this report detail the sample prep-
aration and analytical techniques as well as equipment and rea-
gents used to perform this analysis. Appendix C contains ana-
lytical data for this method.
The M5W sample runs showed an average concentration of 84.9
mg/dNm3 and a standard deviation of 12.2 mg/dNm3. Statistical
data for grouped runs exhibited an acceptable degree of preci-
sion, as characterized by a mean standard deviation of 6.5
mg/dNm3 and a corresponding mean coefficient of variation (CV) of
7.6 percent. A comparison of the M5W results with the thermo-
gravimetric results for M5 and MSB at 315°C (concentration basis)
shows that the M5W results averaged 81.2 mg/dNm3 during Runs 1,
3, and 5, whereas the M5 results averaged 98.4 mg/dNm3 during the
same runs. In Runs 2, 4, and 6, the average values were 88.7
mg/dNm3 for M5W and 94.0 mg/dNm3 for MSB. No direct comparison
with M5-450 is possible; however, the overall average M5-450
concentration at 315°C of 87.4 mg/dNm3 compares favorably with
the average M5W concentration of 84.9 mg/dNm3.
These data are consistent with the basic principle of the
thermogravimetric procedure in that only H.SO. and associated
water are removed by heating at 160°C, whereas additional sulfate
species (metal sulfates, ammonium sulfate, and possibly some
res_idual H-SO.) could not be removed. If these other sulfate
species were water-soluble, the expected M5W results would be
4-23
-------
lower because the method is designed to correct for total water-
soluble sulfate, which includes H2SO..
In an effort to characterize this difference, the within-run
heat-conditioned samples were extracted with water, and aliquots
were analyzed by ion chromatography (1C) for residual water-
soluble sulfates as S0.~. Table 4-11 presents the results of the
within-run residual sulfate analysis. As shown, both the rinse
and filter fractions contained residual sulfate ranging from 2.0
to 15.3 mg in the probe fraction and from 5.2 to 28.2 mg in the
filter fraction. All samples exhibited the same basic character-
istics; i.e, residual water-soluble sulfate was found in each
sample fraction.
The particulate concentrations of samples conditioned to
315°C were corrected for residual sulfate and compared against
the M5W results. Table 4-12 summarizes the comparative data.
The M5 samples in Runs 1, 3, and 5, averaged 90.2 mg/dNm3, com-
pared with 81.2 mg/dNm3 for M5W. The MSB samples in Runs 2, 4,
and 6 averaged 83.2 mg/dNm3 compared with 88.7 mg/dNm3 for M5W.
Considering the variability in reported results because of source
conditions, the overall complexity of the M5W sample analysis,
and the number of analytical steps involved in the thermogravi-
metric and 1C procedures, the data presented in Table 4-12 show
that the M5W results are comparable to the M5 and MSB results
heated to 315°C and corrected for residual sulfate. These data
substantiate the general conclusion that the primary difference
in particulate concentrations between samples heat-conditioned to
4-24
-------
TABLE 4-11. SUMMARY OF RESULTS FOR RESIDUAL SULFATE (SO/)
IN WITHIN-RUN SAMPLES CONDITIONED AT 315°C H
Sample
ID
1A
IB
2C
2D
3C
3D
4C
4D
5A
5B
6B
6C
Sample3
type
M5
M5
MSB
MSB
MS
MS
MSB
MSB
MS
MS
MSB
MSB
_b
Residual sulfate as SO/
Probe, mg
15.3
2.0
2.4
3.5
4.0
5.3
4.1
3.2
14.0
3.0
3.0
2.1
Filter, mg
9.6
6.1
5.4
11.4
9.4
11.5
5.2
13.7
11.7
8.9
28.2
25.5
Total , mg
24.9
8.1
7.8
14.9
13.4
16.8
9.3
16.9
25.7
11.9
31.2
27.6
These samples, previously heat conditioned to 3_15°C, were extracted
with water and analyzed for total sulfate (SO. ) with ion chromato-
graphy (1C).
3Total sulfate (SO/) determined by 1C from aliquots of probe rinse
and filter fractiSns.
4-25
-------
TABLE 4-12. COMPARISON OF WITHIN-RUN PARTICULATE CONCENTRATION
AFTER CORRECTION FOR RESIDUAL SULFATE TO THE M5W TEST RESULTS
Sample
ID
1A M5
IB M5
1C M5W
ID M5W
2A M5W
2B M5W
2C M5B
2D M5B
3A M5W
3B M5W
3C M5
3D M5
4A M5W
4B M5W
4C MSB
4D MSB
5A MS
SB MS
5C M5W
5D M5W
6A M5W
6B MSB
6C MSB
6D M5W
Uncorrected parti cul ate
concentration, mg/dNm3
121.5
99.2
NAC
NA
NA
NA
92.9
117.0
NA
NA
96.2
107.2
NA
NA
87.2
99.3
87.2
79.2
NA
NA
NA
89.9
77.8
NA
Corrected partnculate
concentration, mg/dNm3
104.4
93.6
73.3
85.8
96.6
99.9
86.6
104.8
100.9
87.3
90.7
99.9
92.3
89.6
81.2
88.3
77.8
74.9
59.6
80.0
75.6
74.3
63.9
78.0
The uncorrected concentrations for MS and M5B samples are calculated
concentrations after conditioning at 315°C.
5The corrected participate concentrations for the MS and MSB samgles were
calculated by subtracting the weight of residual sulfate as SO, (Table 4-11)
from the total catch at 315°C and dividing by the sample volume.
"Not applicable.
4-26
-------
315°C and the calculated M5W results represents water-soluble
sulfate species not removed by thermal treatment. As atfurther
check, the remaining M5, M5B, and M5-450 sample fractions were
extracted with water and analyzed by 1C for residual sulfate.
These data, which are summarized in Table 4-13, exhibit similar
characteristics to the data presented in Tables 4-10, 4-11, and
4-12. Residual sulfate as SO ~ was found in each sample and
ranged from 1.7 to 23.2 mg in the rinse fraction and 9.6 to 32.8
mg in the filter fraction.
In the characterization of sulfate species other than H-SO.,
extracts of samples from each sample type were analyzed by ICP
for cations; the results of this analysis are presented in Table
4-14.
A 40-element ICP scan was determined. Four nonmetals were
included in this list: boron, phosphorous, silicon, and sulfur.
Sulfur was converted to the equivalent amount of sulfate. The
results are presented in Table 4-15. Twenty elements were below
0.01 mg/liter or not detected, and nine others were less than 0.1
mg/liter. These nine elements were cadmium, cobalt, copper,
lithium, molybdenum, nickel, strontium, titanium, and vanadium.
At 0.1 mg/liter, the contribution of these elements is insignifi-
cant compared with that of sodium at 5 mg/liter. The agreement
of the sulfate values obtained on the 1C and by ICP is good when
one considers the ability of ICP to detect sulfur.
Table 4-16 presents the charge balance for the 16 samples
analyzed by ICP. Calculations of this charge balance are based
4-27
-------
TABLE 4-13. SUMMARY OF RESULTS FOR RESIDUAL SULFATE (SO/)
ON SAMPLES CONDITIONED AT 315°C 4
Run
No.
7A
7B
7C
7D
8A.
8Bd
8C
8D
9A
9B
9C
9D
Sample3
ID
M5B
M5B
M5-450
M5-450
M5
M5
M5B
MSB
M5B
M5-450
MSB
M5-450
Total sul fates
as S0.~, mg
Rinse Filter
13.8 19.0
5.1 14.2
1.7 9.6
3.0 11.1
14.7 10.7
2.4 18.4
2.7 9.6
8.1 13.4
23.2 32.8
5.0 25.8
2.1 28.4
9.7 30.2
Total b
residual
sulfate
(S04 ), mg
32.8
19.3
11.3
14.1
25.4
20.8
12.3
21.5
56.0
30.8
30.5
39.9
Particulate
concentration, mq/dNm3
Uncorrected
89.6
83.1
63.1
79.6
77.0,
194.0°
63.7
74.6
124.0
101.7
88.7
105.2
Corrected
78.1
76.4
57.8
74.6
66.4,
186.3°
58.9
66.6
104.4
90.9
77.4
91.3
These samples, previously heat-trea£ed at 315°C, were extracted with water
and analyzed for total sulfate (S04 ) by 1C.
Total residual sulfate is the summation of the rinse and filter fraction
in mg.
GThe uncorrected concentrations represent rinse and filter catch after heat
conditioning at 315°C. The corrected values represent p§rticulate concen-
tration after adjusting the sample fractions for the SO, values obtained
by 1C analysis.
Probe rinse contaminated.
4-28
-------
TABLE 4-14. CATIONS FOUND IN WATER EXTRACTION BY ICP
1
Sample
ID
5A Probe rinse
5A Filter
5B Probe rinse
5B Filter
5C Probe rinse
5C Filter
50 Probe rinse
5D Filter
6A Probe rinse
6A Filter
6B Probe rinse
6B Filter
6C Probe rinse
6C Filter
6D Probe rinse
6D Filter
Sample
type
M5
M5
M5
M5
M5W
M5W
M5W
M5W
M5W
M5W
MSB
MSB
MSB
MSB
M5W
M5W
Metal , mg '
Al+3
<0.01
0.28
<0.01
0.22
<0.03
0.03
0.05
0.04
0.06
0.78
0.01
0.04
0.01
<0.01
<0.03
1.08
Ca+2
0.14
0.65
0.08
0.85
0.14
0.58
0.24
0.42
0.29
0.48
0.11
0.30
0.06
0.30
0.17
0.45
Fe+3
<0.01
0.07
<0.01
0.03
<0.01
0.01
0.13
0.01
0.04
0.10
0.01
0.04
0.02
<0.01
0.17
0.12
K+
0.05
<0.01
0.02
0.09
0.12
0.58
0.76
0.52
1.12
0.04
0.28
0.14
0.01
0.02
0.18
0.25
Mg+2
<0.01
0.32
0.01
0.04
<0.02
0.22
0.10
0.15
0.13
0.02
0.05
0.06
0.01
0.18
<0.02
0.09
Na+
7.20
2.75
1.42
1.78
1.70
8.50
5.06
9.50
8.94
1.58
1.32
10.75
0.86
10.25
3.44
1.35
Zn+2
0.06
0.03
0.06
0.02
0.13
<0.01
0.11
<0.01
0.15
0.01
0.05
0.01
0.05
<0.01
0.10
0.01
I
NJ
VD
-------
TABLE 4-15. SOLUBLE SULFATE PRESENT IN SAMPLE
ANALYZED BY ICP
Sample ID
5A M5
5B M5
5C M5W
5D M5W
6A M5W
6B M5B
6C MSB
6D M5W
1C sulfate, mg
Probe rinse
14.0
3.0
3.9
14.3
17.2
3.0
2.1
13.1
Filter
11.1
8.9
38.2
39.1
24.8
28.2
25.5
27.2
ICP sulfate, mg
Probe rinse
13.2
2.8
4.4
11.6
14.6
2.8
2.2
11.9
Filter
10.5
6.7
27.7
29.2
21.0
24.7
23.2
25.5
4-30
-------
TABLE 4-16. CHARGE BALANCE RESULTS FOR SAMPLES
ANALYZED BY ICP
Sample ID
5A M5
5B M5
5C M5W
5D M5W
6A M5W
6B MSB
6C MSB
6D M5W
Charge balance
Probe rinse
1.11
1.10
1.09
0.93
1.28
1.25
1.03
0.63
Filter
0.93
0.82
0.55
0.57
0.36
0.85
0.90
0.39
Milliequivalents cations/milliequivalents sulfate.
4-31
-------
on each cation having the charge listed in Table 4-14 and all
_2
sulfate presented as SO. . The values for the M5W filter sam-
ples averaged 0.47 and 0.98 for the probe rinse samples. These
values indicate that possibly 53 percent of the sulfate on the
filter is present as sulfuric acid and less than 2 percent of the
probe rinse sulfate is sulfuric acid. This is in direct contrast
to the MS and MSB sample after heat-conditioning at 315°C, in
which the average charge balance for the probe rinse was 1.12;
this value is close enough to 1.00 at these low levels to lead to
the conclusion that the remaining sulfate is present as metal
salts. The average charge balance for the M5 and MSB filters is
0.88, which means either that 88 percent of the sulfate is pre-
sent as metal sulfates or that some of the metal salts are pre-
sent as bisulfates, which would account for the charge balance
being less than 1.00.
The results of the charge balances for the samples collected
at this site, although similar for the filters, differ consider-
ably for the probe rinse samples from those collected at previous
sites ' that did not have a scrubber as part of the emission
control system. The charge balance for the filter samples aver-
aged 0.44 and 0.54 for the previous sites ' compared with 0.47
for this site. The probe rinse samples averaged 0.11 and 0.09
for the previous sites ' compared with 0.98 for this scrubbed
source. The charge balance for the probe rinse portions of these
runs shows that the sulfates are present as metal salts rather
4-32
-------
than sulfuric acid. This is consistent with the thermogravi-
metric results, which showed a smaller percent weight loss upon
b-
heat conditioning than in previous studies. The residual sulfate
remaining after heat conditioning of these samples was also
significantly higher than in previous studies.
4.4 RECOMMENDATIONS FOR SAMPLING AND ANALYTICAL METHODOLOGY
Data in this study show that emissions from this scrubber-
controlled FCCU unit are considerably different from emissions
from ESP-controlled FCCU units. Compared with the two previous
sites, ' which were ESP-controlled FCC units, the scrubber-
controlled unit showed the following variations:
1) Less condensible H-SO. collection; i.e., lower percent-
age weight loss upon heating.
2) The percent weight losses of M5, MSB, and M5-450 are
comparable rather than widely different as in previous
studies.
3) Considerably lower H2SO. and SO. concentrations were
determined in the back half of the sampling train.
4) Considerably more residual sulfate remained in the
heat-conditioned particulate samples.
5) The M5W results are considerably lower than the thermo-
gravimetric results before correction for residual
sulfate.
6) The cation concentration, especially sodium, found in
the extracts of the heat-conditioned samples is signif-
icantly higher for this site.
After allowances were made for the sampling difficulties
cited in Section 3 of this report, these data indicate that
although the scrubber reduces concentrations of sulfuric acid and
S02, it also increases the relative amount of metal sulfates that
contributes to the particulate catch and is not removed by
4-33
-------
thermal conditioning. The EPA has stated that the intent of the
NSPS for particulate emissions from FCC units is to control
b-
"catalyst fines" or "mineral dust" and not the condensible sul-
fates that are present in the gas phase at the control device
operating temperature. Emissions from fluid catalytic cracking
and thermoform catalytic cracking units are known to contain
sulfur oxides; therefore, sulfuric acid (H.SO.) and/or its metal
and ammonium salts are the most probable forms of water-soluble
sulfates. Water-soluble sulfate exists in many complex chemical
forms, the most common being sulfuric acid. The results of this
study show that H_SO. is a significant sulfate species in the
FCCU emission stream, and metal sulfates (primarily sodium)
constitute another significant water-soluble sulfate species.
Therefore, a particulate sampling method designed to minimize the
collection of residual H_SO. immediately prior to gravimetric
analysis is not sufficient to minimize potentially high biases in
pa-rticulate measurements from this type of source.
It is evident that sample temperature affects the retention
of condensible sulfate (H_SO. in this case) material in the front
half of the standard Method 5 sampling train, but sulfuric acid
is not necessarily the predominant sulfate species in a scrubber-
controlled source.
From an analytical standpoint, the thermogravimetric proce-
dure is the easiest and least expensive technique for reducing
H-SO. bias on rinse and filter samples collected at this source.
The analytical data indicate that H?SO. and its associated water
4-34
-------
are significantly reduced by heating sample fractions to at least
160°C (320°F) prior to gravimetric analysis. Observed weight
b-
losses at higher treatment temperatures (which are significant
for these samples) are primarily attributable to the volatiliza-
tion of residual H_S04 and other water-soluble sulfates not
removed by heating at 160°C. The results of the water-soluble
sulfate analysis of selected samples support this conclusion.
The average particulate concentrations in M5, MSB, and M5-450
samples conditioned to 315°C and corrected for residual water-
soluble sulfate (as determined by ion chromatography) compared
within 10 percent of the M5W results.
Based on the results of this and similar studies, PEDCo
offers the following recommendations relative to the sampling and
analytical methodology for particulate measurement at FCCU
sources.
4.4.1 Texas Air Control Board Method
The results of this study show that the Texas Air Control
Board (TACB) method entitled "Determination of Particulate in
Stack Gases Containing Sulfur Dioxide" is applicable to these
sources. Modifications to procedures detailed in the TACB method
are presented in Section 3 and Appendix D of this report. A copy
of the method as received from the TACB is presented in Appen-
dix D.
Prior to analyzing field samples, PEDCo conducted an exten-
sive laboratory evaluation of the method. The experimental
design and the results of this study are described in a separate
4-35
-------
method evaluation report issued under this task assignment. This
method entails the use of the sampling procedures and temperature
[121°C (250°F)] described in EPA Reference Method 5, except that
deionized, distilled water is used instead of acetone as the
rinse reagent. The method converts any sulfuric acid present to
a suitable form for accurate gravimetric analysis. Ammonium
hydroxide is added to form ammonium sulfate in the aqueous solu-
tions. The procedure allows for the determination of gross
particulate (sulfate as ammonium sulfate and other particulate),
the determination of sulfate as ammonium sulfate from a Method 6
titration or ion chromatography, and subsequently, the determina-
tion of non-water-soluble sulfate particulate by the subtraction
of sulfate (as ammonium sulfate) from the gross particulate. No
direct comparison between this method and the thermogravimetric
procedures is possible because the method corrects for total
water-soluble sulfate (including H_SO.), whereas the thermogravi-
metric procedure corrects primarily for H-SO. and associated
water. The fact that the average heat-treated sample concentra-
tions corrected for residual water-soluble sulfate compare within
10 percent of the calculated M5W concentration explains the con-
sistently lower values obtained by M5W compared with the heat-
treated samples. Obviously, because this method corrects for
total water-soluble sulfate, the particulate results obtained are
expected to be equal to or lower than those in samples collected
at the same or higher temperature and analyzed thermogravimetri-
cally, regardless of treatment temperature. Considering the
4-36
-------
overall complexity of the analytical procedure, the precision and
accuracy of the method are exceptionally good, as characterized
by a standard deviation of 6.5 mg/dNm3 and a relative standard
deviation of 7.6 percent for this set of data.
4. 4". 2 Modified Method 5
Because of the complexity of the TACB analytical procedure,
an alternative methodology incorporating higher sampling tempera-
tures in conjunction with the thermogravimetric analysis seems
appropriate if not totally sufficient to remove the sulfate bias.
Sampling and analytical procedures contained in EPA Reference
Method 5 are recommended, with the following modifications, to
determine particulate emissions from sources of this type:
1. Sample collection temperatures should be maintained at
no less than 160°C (320°F), and the probe and filter
temperature should be monitored directly by thermo-
couple leads located at the exit of the sample probe
and immediately behind the filter frit. Most commer-
cially available stack sampling equipment is capable of
maintaining front-half temperatures of 160°C (320°F);
however, sampling at temperatures above 205°C (400°F)
will probably require equipment modifications to ensure
maintenance of the temperature required and the integ-
rity of sampling train components.
2. Prior to the gravimetric analysis, probe rinse and
filter fractions should be heated in an oven at not
less than 160°C (320°F) (higher temperatures would be
preferable) for 6 hours or more. Prior to weighing,
sample fractions should be allowed to cool in a desic-
cator for approximately 2 hours and weighed according
to the constant weight criteria detailed in Reference
Method 5.
3. These results also indicate that the combination train
(Reference Methods 5 and 8) may not yield accurate
results for particulate and H_S04 unless the probe and
filter temperatures are considerably higher than 160°C
(320°F). The data indicate that the combination Refer-
ence Methods 5 and 8 train can be used to sample for
4-37
-------
particulate and SO- simultaneously, provided the train
is air-purged for at least 15 minutes after testing is
completed to remove the S0~ adsorbed in the IPA.
4-38
-------
SECTION 5
QUALITY ASSURANCE
Because the goal of testing is to produce representative
emission results, quality assurance is one of the main facets of
stack sampling. Quality assurance guidelines provide the de-
tailed procedures and actions necessary for defining and pro-
ducing acceptable data. Four such documents were used in this
test program to ensure the collection of acceptable data and to
provide a definition of unacceptable data. The following docu-
ments comprise this source-specific test plan prepared by PEDCo
and reviewed by the Emission Measurement Branch: the EPA Quality
Assurance Handbook Volume III, EPA-600/4-77-027; the PEDCo Envi-
ronmental Emission Test Quality Assurance Plan; and the PEDCo
Environmental Laboratory Quality Assurance Plan. The last two,
which are PEDCo's general guideline manuals, define the company's
standard operating procedures that are routinely followed by the
emission testing and laboratory groups. In addition, data ob-
tained from similar test programs conducted under this task
assignment were utilized to assess the between-source analytical
trends of the methods employed.
Appendix F provides more detail on the quality assurance
procedures, such as QA objective; data reduction; quality control
5-1
-------
checks; performance and system audits; preventive maintenance;
precision, accuracy, and completeness; corrective action; and
to-
quality assurance reports to management.
The following steps were taken in this specific test program
to-ensure that the testing and analytical procedures produced
quality data.
0 Calibration of all field sampling equipment. (Appendix
E describes calibration guidelines in more detail.)
0 Checks of train configuration and calculations.
0 Onsite quality assurance checks such as leak checks on
the sampling train, pitot tube, and Orsat line and
quality assurance checks of all test equipment prior to
use.
0 Use of designated analytical equipment and sampling
reagents.
Table 5-1 lists the sampling equipment used and the calibra-
tion guidelines and limits. In addition to the pre- and post-
test calibrations, a field audit was performed on the meter
consoles used for sampling. PEDCo constructed critical orifices
for use in this audit. Figures 5-1 through 5-4 show example
audit runs for each dry gas meter used for testing. Onsite
calculations were performed for each test run to ensure isoki-
netic sampling sites. Pertinent test data were compared with
expected values as an additional validation check.
As a check on the reliability of the thermogravimetric
analytical procedure, sets of blank filters and acetone were
resubmitted to the laboratory for blank analysis. Table 5-2
presents example results of the thermogravimetric blank analysis.
5-2
-------
TABLE 5-1. FIELD EQUIPMENT CALIBRATION
u>
Equipment
Meter box
Pi tot tube
Digital
indicator
Thermocouple
Orsat
analyzer
Train
A
B
C
D
A&B
C&D
A&B
C&D
A&B
C&D
All
ID
No.
FB-2
FB-3
FB-5
FB-6
402
401
250
124
125
178
121
206
145
Calibrated
against
Wet test meter
Geometric
specifications
Millivolt
signals
ASTM-2F
Standard gas
Allowable
deviation
Y ±0.05 Y pre-test
AH @ ±0.15
(Y ±0.05 Y post-test)
See Appendix E
0.5*
±1.5%
(±2% saturated)
±0.5%
Actual
deviation
+0.001
-0.02
-0.001
+0.005
-0.04
+0.013
+0.006
-0.07
+0.002
+0.002
-0.03
+0.02
0.0
0.034
0.034
0.37%
0.23%
-0.4
1.2
1.2
-0.25*
0.0%
Within
allowable
limits
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Comments
CP = 0.84
co2
°2
(continued)
-------
TABLE 5-1 (continued)
m
i
Equipment
Implnger
thermometer
Balance
Barometer
Dry gas
thermometer
Probe
nozzle
Train
A
B
C'
D
All
All
A
B
C
D
A
B
C
D
ID
No.
391
452
390
451
196
229
FB-6 inlet
FB-6 outlet
FB-3 inlet
FB-3 outlet
FB-5 inlet
FB-5 outlet
FB-2 inlet
FB-2 outlet
1A
2A
3A-9A
IB
2B
3B-9B
1C
2C
3C-9C
ID
20
3D-9D
Calibrated
against
ASTM-2F
Type S weights
NBS traceable
barometer
ASTM-2F
Caliper
Allowable
deviation
±2°F
±0.5 g
±0.10 in.Hg
(0.20 post-test)
±5°F
Dn ±0.004 in.
Actual
deviation
0°F
-1°F
2°F
-2°F
+0.25 g
0.0
in.Hg
-1°F
•H°F
-3°F
-3°F
+4°F
-3°F
-2°F
-2°F
0.001
0.000
0.001
0.001
0.002
0.001
0.001
0.001
0.000
0.002
0.000
0.002
Within
allowable
limits
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Comments
-------
DATE: Z/2JL/&3
CLIENT:
BAROMETRIC
ORIFICE NO.
PRESSURE (Pbar): 21. 1 1 in
\5
. Hg METER BOX NO. ~l^
PRETEST Y: O.c?77
ORIFICE K FACTOR: 5- V/ 1 \ \ O~i AUDITOR: /ft . Ph»ll'1>$
Orifice
manometer
reading
AH
in H20
a.ir
/
Dry gas
meter
reading
Vvf
ft3
k>/.Voo
8/^.086
Temperatures
Ambient
Tai/Taf
°F
s56
51
Dry gas meter
Inlet Outlet
w w
op op
<5o 53
^3 5+
Duration
of
run
0
min
/
-------
DATE:
CLIENT:
BAROMETRIC PRESSURE (Pfa ):3o.tff in. Hg
ORIFICE NO.
5
ORIFICE K FACTOR: 5M)l\ lo"l
METER BOX N
PRETEST Y:
AUDITOR:
Orifice
manometer
reading
AH
in H20
^o
Dry gas
meter
reading
yvf
•5
ft3
843.0U3
S55."7^>
Temperatures
Ambi ent
Tai/Taf
°F
"74
16?
Dry gas meter
Inlet
VTif
°F
So
So
Outlet
Toi/Tof
°F
"71
7|
Duration
of
run
0
mm
/J.O
Dry gas
meter
volume
Vm
ft3
U.kYI
Average temperatures
Ambient
Ta
°F
1^
Dry gas
meter
Tm
°F
~70
Vm
mstd
ft3
ll.L,^
v_
mact
ft3
ll.^o
Audit
Y
/Oo^
Y
deviation
%
0-S7,;
Vmstd
(17.647)( Vm )(Pbar + AH/13.6)
(Tm + 460)
Audit Y
Vm
ao.t
mstd
\ct
(1203)( 0 )( K )(P )
(T + 460)1/2
a
Y deviation, %
(Y audit - Y pre-test)(100%)
(Y audit)
Audit Y must be in the range, pre-test Y ±0.05 Y
Figure 5-2. Audit report dry gas meter (Meter Box FB-3)
5-6
-------
DATE:
CLIENT:
BAROMETRIC
ORIE-ICE NO.
PRESSURE (P
V
bar^9;77
Dai
in. Hg METER BOX NO. FA£
PRETEST Y: O.9^s3
ORIFICE K FACTOR: S.aSC, Xl 0~? AUDITOR: /^./%////i\5
_
Orifice
manometer
reading
AH
in H20
Z-0
1
Dry gas
meter
reading
Vvf
ft3
6
-------
DATE: Z/2J?/&3
CLIENT:
BAROMETRIC
ORIEICE NO.
ORIFICE K F
Orifice
manometer
reading
AH
in H20
no
PRESSURE (Pbjr):o0.oy in. Hg METER BOX N
7 PRETEST Y:
0. F£6
O.^n
ACTOR: 5.03k x|o~y AUDITOR: Al.fhJliP^
Dry gas
meter
reading
v./vf
ft3
6/0,7/7
b Z-B.o&r
1
Temperatures
Ambient
W
°F
7V
1'L
Dry gas meter
Inlet
°F
76,
19
Outlet
VTof
°F
7/
75^
Duration
of
run
min
/XO
Dry gas
meter
vol ume
Vm
ft3
)2,3L>L>
Average temperatures
Ambient Dry gas
meter
Ta Tm
op op
7s* 1C,
ft3
I*.?-*)
Vt
ft3
II. lot*
Audit
Y
O.fol
Y
deviation
o. Ho
Vmstd
(17.647)( Vm )(Pbar + AH/13.6)
(T + 460)
Audit Y
v"> .
act
V|T1std
-
Xct
(1203)( 0 )( K )(Pbar)
(Ta + 460)1/2
a
Y deviation, %
(Y audit - Y pre-test)(100%)
(Y audit)
Audit Y must be in the range, pre-test Y i.0.05 Y
Figure 5-4. Audit report dry gas meter (Meter Box FB-6)
5-8
-------
TABLE 5-2. EXAMPLE OF A THERMOGRAVIMETRIC ANALYSIS OF FILTER AND ACETONE BLANKS
(milligrams)
Ul
i
VD
Sample
Filter
Acetone3
Lab
ID
CW507
CW504*
CW506D
Tare
weight
324.9
106,912.7
106,546.6
Ambient temp.
Weight
324.9
106,917.8
106,548.6
Net
0.0
+5.1
+2.0
Weight after conditioning at Indicated temperature
160°C
Weight
325.0
106,913.0
106,546.3
Net
+0.1
+0.3
-0.3
232°C
Weight
324.6
106,911.9
106,545.5
Net
-0.3
-0.8
-1.1
316°C
Weight
324.5
106,911.8
106,545.4
Net
-0.4
-0.9
-1.2
alnitial volume, 457 ml.
blnitial volume, 292 ml.
-------
The reported net weights are reasonable (considering the number
of times each sample fraction was handled) and show good analyti-
cal techniques. In addition, several particulate samples were
reanalyzed to preclude weighing error as a source of variability
in-reported results.
Audit solutions prepared by EPA were used to check the
analytical procedures and reagents for SO- sampling analysis.
Figure 5-5 presents the results of this analytical audit. Table
5-3 summarizes the reagent blank analysis performed on the iso-
propyl alcohol (IPA) and hydrogen peroxide used for sampling.
The audit test shows that the analytical techniques were good.
Table 5-4 presents the results of the H-SO. analysis by both
titration and ion chromatography. As discussed in Section 4,
traces of peroxide found in IPA sample solutions caused some SO-
to be absorbed in the IPA and thus increased the reported values
for SO." as H-SO./SO., and decreased the reported SO- values. Ion
chromatography was used to validate titrametric data because the
end points were difficult to distinguish, especially for the IPA
solutions. The data show good correlation between titrametric
and 1C SO- analysis. The 1C analytical results for SO ~ and
H-SO./SO- compared favorably.
As an assessment of the reliability of the procedures used
in determining the non-water-soluble sulfate particulate concen-
tration, blank filter and deionized, distilled water samples were
analyzed along with each set of field samples. Table 5-5 pre-
sents example blank results for this method.
5-10
-------
Plant
S - COT ft
PN Number
Date samples received 3-H-63 Date analyzed 3-c?.i-83 ••'. j-;zj,-fc.
Samples analyzed by C Si/,//
Reviewed by JT, &«?A^JLCL /v'c Date of Review 3-f>^
»* DsTX.
C w/-1 5^; /
it-,c 5^Jt>X
^ B-v/^j
C a.' -"5 ^ 3L
K: v ^c Jfe: -
• V 4 I -\
mg S02/dscm
Determined
/4i«i -ci
i^-^7.6
i'Pi^o.)
Source of
Sample
/ . UJ (XCj NA »'
T.LUx^^r^
1 . lA'iv.fjCi"
Accepted
Value
iHfC-^
asair.q
if!-l5/i
%
Difference
+ O, uS
- /. Z.1
-C',>'i&
Figure 5-5. Audit report S02 analysis,
5-11
-------
Plant
PN Number 5 S".V» ••/*-/
Date samples received i-^-f' Date analyzed ;-;. -•*;_-£ ~>
Samples analyzed by C . OtrJi
Reviewed by
Sample
Number
/ . . _ „ f- 1
' . uovwuxs- /';.<_ Date of Review -5 -3* -'?• "S
Determined
H3..I
Source of
Sample
r.u „,.„,
Accepted
Value
„,,,
Difference
*- /.V;
Figure 5-5. Audit report S02 analysis (continued)
5-12
-------
TABLE 5-3. REAGENT BLANK ANALYSIS FOR IPA AND
Sample type
10% H009
L C.
80% IPA
Sample
ID
CW588
CW589
CW590
CW549
CW550
CW551
mg of SO/
as S02
<0.1
<0.1
<0.2
_
-
-
" as H2S04
—
_
-
<0.3
<0.2
<0.3
5-13
-------
TABLE 5-4. ION CHROMATOGRAPHY CHECKS
H202 Impinger Solutions":
Run No.
1BM5
4AM5W
6AM5W
7DM5450
Lab No.
CW553
CW564
CW572
CW579
SOj. total ma
Titration
37.1
10.5
320
502
1C
39.3
11.7
334
510
1PA Impinger Solutions:
Run No.
1AM5
1BM5
1CM5W
1DM5W
2AM5W
2BM5W
2CM5B
2DM5B
3AM5W
3BM5W
3CM5
3DM5
4AM5W
4BM5W
4CM5B
4DM5B
5AM5
5BM5
5CM5W
5DM5W
6AM5W
6BM5B
6CM5B
6DM5W
Lab No.
CW514
CW515
CW516
CW517
CW518
CW519
CW520
CW521
CW522
CWS23
CW524
CW525
CW526
CW527
CW528
CW529
CW530
CW531
H^SO., total mg
TitrStiSn
135
62.8
76.3
80.1
60.8
79.1
78.2.
78.0
264
278
178
212
107
75.3
89.5
99.1
31.9
21.9
No Sample
CW532
CW533
CW534
CW535
CW536
31.8
32.6
31.0
24.4
27.7
1C
210
39.7
74.8
57.8
45.7
63.0
63.4
72.7
252
253
137
179
80.5
40.2
59.8
68.0
23.2
12.1
32.4
22.4
25.2
17.3
27.6
(continued)
5-14
-------
TABLE 5-4 (continued)
IPA Impinger Solutions (continued):
Run No.
7AM5B
7BM5B
7CM5450
7DK5450
BAMS
8BM5B
8CM5B
8DM5B
9AM5B
9BK5450
9CM5B
9DM5450
IPA Blank No. 1&2
IPA Blank No. 3-6
IPA Blank No. 7-9
HpO Blank
Lab No.
CW537
CW538
CW539
CW540
CW541
CW542
CW543
CW544
CW545
CW546
CW547
CW548
CW549
CW550
CW551
CW513
H^SO,,, total mq
Titr&tidn
29.5
47.3
40.1
58.9
33.6
26.5
31.0
37.3
327
66.6
27.0
24.1
<0.3
<0.2
<0.3
-
1C
17.3
31.1
34.5
45.5
23.2
24.3
37.4
29.6
320
53.6
26.7
24.2
a
a
a
<0.60
aIPA concentration caused erratic results.
5-15
-------
TABLE 5-5. NON-WATER-SOLUBLE SULFATE BLANK ANALYTICAL DATA
Sample
type
Filter
Probe rinse
Filter
Probe rinse
Filter
Probe rinse
Lab ID
CW508
CW511
CW510
CW512
CW509
CW513
Net participate
weight includ-
ing ammonium
sulfate, mg
0.0
2.7
5.7
1.7
1.2
1.8
Cso4,
mg/nter
2.22
<1.00
2.19
<1.00
2.11
1.61
Volume
evapo-
rated, ml
235
593
235
622
235
572
NWSSP,
mg
-0.7
2.7
5.0
1.7
0.5
0.5
15 ml were removed for 1C analysis.
5-16
-------
Table 5-6 summarizes blank analytical data for the ion
chromatography analyses performed on rinse and filter samples.
b-
The 1C was calibrated daily with standard solutions of 1.0,
2.5, 5.0, 10.0, and 15.0 mg/liter. A standard reference solution
(SRS) at 8.0 ppm was analyzed at the beginning, the end, and
after every 10 samples during an analysis day. The measured SRS
value had to be within ±5 percent of the theoretical value as the
previous 10 samples were reanalyzed after the instrument was
recalibrated. Ten percent of the samples were analyzed in dupli-
cate. These samples agreed within ±5 percent.
5-17
-------
TABLE 5-6. ION CHROMATOGRAPHY BLANK ANALYTICAL DATA
Sample
type
Filter
Rinse
(acetone)
Rinse
(acetone)
Lab ID
CW509
CW503
CW505
^cn
iU4'
mg/ liter
3.55
<1.0
<1.0
Volume,
ml
250
250
250
S04", mg
0.9
<0.2
<0.2
5-18
-------
REFERENCES
1. Mitchell, W. J., and M. R. Midgett. A Means to Evaluate the
Performance of Stationary Source Test Methods. Environ-
mental Science and Technology, 10:85-88, 1976.
2. Oldaker, G. B. Condensible Particulate and Its Impacts on
Particulate Measurements. Draft Report. Prepared under EPA
Contract No. 68-01-4148, Task No. 69. May 1980.
3. Peters, E. T., and J. W. Adams. Sulfur Dioxide Interaction
With Filters Used for Method 5 Stack Sampling. In: Work-
shop Proceedings on Primary Sulfate Emissions From Combus-
tion Sources, Volume I - Measurement Technology. EPA-
600/9/78-020a, 1978. pp. 199-202.
4. Gushing, K. W. Particulate Sampling in Process Streams in
the Presence of Sulfur Dioxide. In: Workshop Proceedings
on Primary Sulfate Emissions From Combustion Sources, Volume
I - Measurement Technology. EPA-600/9-78-020a, 1978. pp.
202-227.
5. PEDCo Environmental, Inc. Comparative Evaluation of EPA
Methods 5 and 17. Draft Report. Prepared under EPA Con-
tract No. 68-02-3431, Task Nos. 88, 103, and 163. February
1983.
6. PEDCo Environmental, Inc. Emission Test Report - Method
Development and Testing for FCCU Regenerators, Arco Petrole-
um Products Company. EPA Contract No. 68-02-3546, Task Nos.
14 and 20. March 1984.
7. PEDCo Environmental, Inc. Emission Test Report - Method
Development and Testing for FCCU Regenerators, Phillips
Petroleum Company. EPA Contract No. 68-02-3546, Task Nos.
14 and 20. March 1984.
R-l
------- |