&ER&
United States
Environmental Protection
Agency
Environmental Sciences Research
Laboratory
Research Triangle Park NC 27711
EPA 600 2-79 141
August 1979
Research and Development
Evaluation of
Stationary Source
Particulate
Measurement
Methods
Volume IV.
Basic Oxygen
Furnaces
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology required for the control and treatment
of pollution sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/2-79-141
August 1979
EVALUATION OF STATIONARY SOURCE PARTICULATE
MEASUREMENT METHODS
Volume IV. Basic Oxygen Furnaces
by
J. E. Howes, Jr., W. M. Henry, and R. N. Pesut
Battelle, Columbus Laboratories
505 King Avenue
Columbus, Ohio 43201
Contract No. 68-02-0609
Project Officer
Kenneth T. Knapp
Emissions Measurement and Characterization Division
Environmental Sciences Research Laboratory
Research Triangle Park, North Carolina 27711
ENVIRONMENTAL SCIENCES RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
-------
DISCLAIMER
This report has been reviewed by the Environmental Sciences Research
Laboratory, U. S. Environmental Protection Agency, and approved for publica-
tion. Approval does not signify that the contents necessarily reflect the
views and policies of the U. S. Environmental Protection Agency, nor does
mention of trade names or commercial products constitute endorsement or
recommendation for use.
ii
-------
ABSTRACT
The work described in this report is part of a study to evaluate
the EPA Method 5 procedure for sampling a variety of stationary sources.
Specifically, this study addresses the reliability of this reference method
for measurement of particulate emissions from basic oxygen furnaces (EOF).
The facilities at which sampling was performed are equipped with emission
controls which are representative of the two principal systems used to abate
BOF emissions, namely, high-energy wet scrubbers and electrostatic precipi-
tators.
Experiments were conducted to study the effects of sampling
system temperature, filter material, and anisokinetic sampling on mass
results obtained with Method 5. In-stack sampling was used as a comparative
technique to assess the Method 5 procedure. Chemical and limited physical
characterizations were performed to evaluate the representativeness of the
collected particulates and to identify variations which lead or may lead
to mass differences introduced by various sampling parameters. Gaseous
emissions were analyzed to identify species which may interact with the
particulate sampling process.
Operationally, the Method 5 procedure performed satisfactorily
when used to sample particulate emissions from the wet scrubber-equipped BOF.
Although some fractionation of certain species in the sampling train was
observed, the general chemical composition of the Method 5 collections were
representative of the stack emissions. Experiments in which sampling was
performed at anisokinetic rates (0.7 and 1.3 times isokinetic), and sampling
system temperature was varied from 84 to 191°C, did not show observable dif-
ferences in mass loading results.
Problems were encountered in the use of an in-stack method for
sampling the wet scrubber emissions. The moisture-ladened stack gas with
entrained droplets saturated the filter with water causing a high pressure
drop and rupture of the filter. Isolation and heating of the filterholder
were required to keep the filter dry. Although the results of the in-stack
Method 5 comparison were scattered, the data indicate that in-stack sampling
may give higher mass loading measurements.
Experiments at the BOF equipped with electrostatic precipitators
indicate that Method 5 gives reliable results and provides collections which
are representative of the stack emissions. Differences in mass loading
measurements were not observed when sampling was performed with different
filter materials (glass fiber and quartz) or when the sampling system tempera-
ture was increased to stack gas temperature (M_77°C). A limited number of
experiments indicates that in-stack sampling may yield lower mass results
than Method 5.
This report was submitted in fulfillment of Contract No. 68-02-0609
by Battelle's Columbus Laboratories under the sponsorship of the U.S. Environ-
mental Protection Agency.
iii
-------
CONTENTS
Abstract iii
Figures vi
Tables vii
Acknowledgments viii
1. Introduction 1
2. Conclusions 3
EOF with wet scrubber 3
EOF with electrostatic precipitator 3
3. Recommendations 5
4. Experimental Work and Results 6
Experimental approach 6
Process and sampling site descriptions 7
Sampling equipment 13
Sample collection and analysis procedures ... 19
Test descriptions and results - 22
Characterization of particulate collections
and gaseous emissions 30
5. Discussion 44
References 46
Appendices
A. Method 5 - Determination of particulate emissions
from stationary sources 47
E. Stack gas measurement data 51
C. Sampling system operation data 55
-------
FIGURES
Number
1 Wet scrubber gas cleaning system 9
2 Velocity pressure and gas temperature profile
of EOF/wet scrubber stack at sampling location 11
3 Basic oxygen furnace with electrostatic precipitator
emission control 12
4 Velocity pressure and gas temperature profile of
BOF/ESP stack quadrant at sampling location 14
5 Schematic diagram of Method 5 train 15
6 Thermocouple assembly used to measure gas
temperature at probe outlet 17
7 Thermometer assembly used to measure gas
temperature at the box filter outlet 17
8 In-stack filterholder for glass fiber thimbles 18
9 Temperature measurement points 20
10 NOX and S02 concentrations in EOF emissions during
the oxygen blow - EOF with ESP emission control 39
11 Particle size distribution of EOF/wet scrubber
emissions collected on Method 5 filter - run 5B 41
12 Particle size distribution of EOF/wet scrubber
emissions collected on in-stack filter - run 5A 42
13 Particle size distribution of emissions from ESP
equipped basic oxygen furnace 43
vi
-------
TABLES
Number Page
1 Operational Steps in EOF Process 7
2 Velocity Pressure and Temperature Profile During
Blow - EOF with Wet Scrubber 10
3 Summary of Experiments - EOF with Wet Scrubber 23
4 Sample Weight Data 24
5 Sample Weight and Mass Loading Data - EOF with
Wet Scrubber 26
6 Randomized Test Pattern for Study of Filter Media,
Temperature, and Their Interaction - EOF with
ESP Control 27
7 Sample Weight Data - Temperature/Filter Experiments -
EOF with ESP Emission Control 28
8 Analyses of Variance-Temperature/Filter Experiments -
EOF with ESP Emission Control 29
9 Sample Weight and Distribution Data - Method 5 and
In-Stack Comparison Tests - EOF with ESP Emission
Control 31
10 Analysis of Method 5 and Scrubber Samples - EOF with
Wet Scrubber Emission Control 32
11 Chemical Compositions of Particulate Collections
From EOF with Wet Scrubber 33
12 Percentage Distributions of Major Cations and Anions
Within the Sampling Train Components 34
13 Analysis of Particulate Emissions from EOF Equipped
with ESP Control - 36
14 Chemical Analysis of Particulate Emissions from EOF
with ESP Control 37
15 Gas Analysis of EOF Emissions with Wet Scrubber Control ...... 38
16 Gas Chromatographic and Mass Spectrometric Analysis of
Gaseous 'Emissions from EOF with ESP Control 38
17 Results of Analysis of Impinger Collections 40
vii
-------
ACKNOWLEDGEMENTS
The authors wish to acknowledge Dr. Kenneth Knapp for his
direction of this program as the EPA Project Officer.
Messrs. D. L. Sgontz, D. F. Kohler, R. L. Livingston, W. C.
Baytos, S. E. Miller, D. Talbert, and E. J. Schulz of Battelle-Columbus
participated in the field sampling. Their dedicated efforts are appre-
ciated and gratefully acknowledged.
viii
-------
SECTION 1
INTRODUCTION
The Clean Air Act as amended in 1970 provides the impetus for
programs to improve the air quality in the U.S. through research to broaden
the understanding of the effects of air pollutants, research and develop-
ment of techniques to control emissions, and the enactment of air quality
regulations to protect the public welfare. Pursuant to Section 111 of the
Act, the Environmental Protection Agency (EPA) on December 23, 1971, pro-
mulgated Standards of Performance for New Stationary Sources (amended) for
fossil fuel-fired steam generators, incinerators, Portland cement plants,
and nitric and sulfuric acid plants.^) On March 8, 1974, similar perfor-
mance standards were issued for asphalt concrete plants, petroleum refiner-
ies, storage vessels for petroleum liquids, secondary lead smelters,
secondary brass and bronze ingot production plants, iron and steel plants,
and sewage treatment plants. (2) All new and modified sources in the pre-
ceding categories are required to demonstrate compliance with the standards
of performance.
The performance standards are intended to reflect "the degree of
emission limitation achievable through the application of the best system
of emission reduction which (taking into account the cost of achieving such
reduction) the Administrator determines has been adequately demonstrated".
Compliance with required performance is determined by testing pro-
cedures specified with the standards. The use of a procedure called "Method
5 Determination of Particulate Emissions from Stationary Sources"™' is
specified in all instances where particulate mass emission measurements must
be made. The Method 5 procedure consists of isokinetic extraction of a
sample from the emission stream with a heated probe and collection of the
particulates on a heated filter. With the recent exception of fossil fuel-
fired power plants (5) } the same sampling system operating parameters have
been adopted for all stationary sources.
The source categories subject to Method 5 particulate measurements
include diverse processes which encompass a wide range of the following emis-
sion characteristics; moisture content, gas temperature , gas composition,
particulate concentration and composition, and flow dynamics. Interaction
of these emission properties with the Method 5 sampling technique can produce
significant variations in the results of particulate emission measurements.
The following are examples of some of the reactions which may affect particu-
late measurements:
(1) 803 or I^SOi^ in emissions can condense to form sulfates
which increase the mass of collected "particulates".
The SOs-H^SO^ dew point is dependent on 803 concentra-
tion and moisture content of the emissions.
-------
(2) The filter particulate catch may present a surface for
reactions with gaseous emission components such as SOX
and NOX- Reactivity would be dependent on particulate
loading and composition and on gas composition of the
emissions.
(3) Changes in gas temperature in the sampling system may alter
the apparent particulate concentration through condensation
or evaporation.
Such interactions with the sampling process must be recognized
and controlled if Method 5 is expected to yield reliable particulate measure-
ments for individual source categories.
The work presented in this report was performed as part of an EPA
program to study the applicability of the Method 5 procedure to measurement
of particulate emissions from a variety of stationary sources. Specifically,
this work addresses the question of whether Method 5 provides an accurate,
reliable measurement of particulate emissions from basic oxygen furnaces (BOF),
The study included BOF facilities with the two types of emission controls in
common use, wet scrubbers and electrostatic precipitators.
Volume I in this series covers a similar study of Portland Cement
Plants(6) and Volume II Oil Fired Power Plants^7).
-------
SECTION 2
CONCLUSIONS
The results of this study lead to the following conclusions re-
garding methodology for measuring mass emissions from basic oxygen steel-
making furnaces (EOF).
EOF WITH WET SCRUBBER
• Method 5 as promulgated in the Federal Register, December 23, 1971,
appears to be a reliable procedure for particulate mass emission measure-
ments. The precision (repeatability) of mass measurements made by con-
current sampling with two systems is estimated to be about 2.8 percent.
Chemical analyses show that although fractionation of some species may
occur between the probe and filter collections, the composition of the
total system catch is representative of the stack emissions.
• Mass emission results obtained with Method 5 are not significantly
affected by sampling variations at 0.7 and 1.3 times isokinetic rate.
Particle size measurements of the emissions indicate a mass mean
diameter of about 0.2 ym. Consequently, anisokinetic sampling would
be expected to have little effect on the mass measurements.
• Variation in sampling system operating temperature was found to have
no effect on mass emission measurements. Experiments in which sampling
was performed with Method 5 probe outlet gas and box filter temperatures
of about 84, 149, and 191°C gave essentially the same results as obtained
with a Method 5 train operated at the minimum specified temperature,
Operational problems were encountered in use of in-stack sampling on the
wet scrubber. Isolation and external heating of the filterholder are
required to prevent saturation of the filter with water. Even with these
measures, it was difficult to regulate the in-stack filter temperature
to prevent collection of moisture.
Comparisons of in-stack and Method 5 show a considerable amount of
scatter; however, generally results indicate that in-stack sampling may
give higher mass emission data.
EOF WITH ELECTROSTATIC PRECIPITATOR
• Method 5 as promulgated in the Federal Register, December 23, 1971,
appears to be a reliable procedure for measurement of particulate mass
emissions. The repeatability of measurements made by concurrent sampling
with two trains is estimated to be about 3.4 percent. Chemical analysis
-------
confirm that samples collected with the Method 5 procedure are represen-
tative of the in-stack particulate composition.
Statistically designed experiments show that there is not a
difference between mass results obtained with Method 5 when two
filter materials are used and when the sampling train is °?era;
stack gas temperature. There were no statistically significant
ence in results obtained with MSA 1106BH and ADL quartz filter
Results from experiments in which sampling trains were operated at stacK
temperature, -vl77°C, were not statistically different from ™*^s
obtained with a sampling system temperature of 121°C, the minimum speci-
fied in Method 5.
In-stack sampling appears to yield mass emission measurements which are
lower than those obtained with Method 5. Experiments with use ot MbA
1106BH flat filters and thimbles marketed by Carborundum Company for
in-stack particulate collection gave results which were, on the average,
about 25 and 16 percent lower, respectively, than Method 5 values.
Particle size measurements indicate a mass mean diameter of 4 urn as
compared to 0.2 ym for the , wet scrubber. These results are consistent
with these types of control devices.
-------
SECTION 3
RECOMMENDATIONS
The study demonstrates that particulate mass emissions from
basic oxygen furnaces both with wet scrubber and electrostatic precipi-
tators can be determined with acceptable precision and representativeness
with Method 5 as currently promulgated. Therefore, major revisions in
Method 5 when applied to EOF facilities are neither necessary nor recom-
mended .
The use of in-stack sampling for wet-scrubber installations
is not recommended. When used without external heating, the in-stack
filter becomes saturated with water leading to restricted flow rates.
When external heating is used, considerable difficulty is encountered
in controlling the filterholder temperature.
Operationally, there are no apparent problems in the use of
in-stack sampling techniques for ESP equipped EOF facilities. However,
additional studies are recommended to determine if in-stack techniques
give results equivalent to Method 5.
-------
SECTION 4
EXPERIMENTAL WORK AND RESULTS
EXPERIMENTAL APPROACH
In line with the program objectives, the experimental plan was
formulated (1) to study parameters of Method 5 which might affect particu-
late mass measurements, (2) to characterize the emissions, both particulate
and gaseous, to identify reactive species which might affect sampling re-
sults, and (3) to compare Method 5 with in-stack sampling techniques.
Sampling parameters which were studied included filter material,
sampling system temperature, and deviation from isokinetic sampling rate.
These studies were intended to reveal the sensitivity of particulate measure-
ments to the sampling variables ^and to determine if current Method 5 operating
parameters are within a range which will produce accurate, reliable results.
Various in-stack filter configurations were compared with Method 5.
Particulates collected under in-stack conditions might be considered less
subject to compositional alterations, especially by condensation products
and reactions which may occur upon cooling below stack temperature assuming
stack temperature is greater than 121°C (250°F) . Accordingly, comparison of
the in-stack and Method 5 particulate collections provides an approach to
study of sample alterations which may be introduced by the Method 5 sampling
procedure. In addition, in-stack collection essentially eliminates deposi-
tion of the particulates in the probe and facilitates the study of reactions
which may occur in the probe, e.g., condensation of H^SO^ or organic materials.
Another consideration behind the methods comparison is that EPA is
considering adoption of the in-stack technique as an optional performance
test method. Equivalency of Method 5 and the in-stack method must be demon-
strated to maintain consistency with established performance standards.
The approach selected to conduct the experimental study consisted
of concurrent sampling at approximately the same point in the EOF emissions
stack with two sampling systems operated in various sampling configurations
and under the various conditions under study. Filter and probe collections
were analyzed gravimetrically and chemically to detect differences resulting
from various sampling parameters. Analyses of the gas composition of the
emissions were performed to identify components which might interact with the
sampling process. Where possible, the mass emission data from the experi-
ments were analyzed statistically to determine the significance of observed
differences.
-------
PROCESS AND SAMPLING SITE DESCRIPTIONS
Basic Oxygen Steel Process
In 1974, the production of steel in the U.S. totaled 145,720,000
tons. Production over the past 4 years, a relatively weak period economi-
cally, has increased at a rate of about 2 percent per year. In 1974, carbon
steel accounted for 86.9 percent of the total production with alloy and
stainless steels accounting for 11.6 and 1.5 percent respectively.
The principal steelmaking process used in the U.S. is the basic
oxygen furnace (EOF) process. Of the total 1974 production, 56 percent of
the steel was made by this method. Open hearth and electric furnace pro-
cesses produced 24 and 20 percent, respectively. In the past 3 to 4 years,
steelmaking by the open hearth process has leveled off, while basic oxygen
steelmaking has continued to increase by 29 percent.
The basic oxygen process or "EOF process" as it is commonly called
is a batch reactor process wherein hot metal and scrap are charged into the
furnace, and lime, fluorspar, plus other fluxes are added and these are
reacted with oxygen which oxidizes out the major impurities—principally
carbon, phosphorus, silicon, and magnesium—to form a low-carbon steel. Under
ideal conditions, the entire operation cycle takes 35 minutes and includes
the sequence of operations listed in Table 1.
TABLE 1. OPERATIONAL STEPS IN EOF PROCESS
(1) Scrap metal charge is loaded in vessel.
(2) Hot metal (pig iron) from the blast furnace
is added to complete the charge.
(3) The lance is lowered near the bath and the
oxygen blow is initiated.
(4) Flux is added.
(5) Blow is terminated after about 20 minutes.
(6) Temperature of the steel is measured and a
sample is taken for chemical analysis.
(7) Reblowing may be necessary to achieve desired
chemical composition and/or temperature.
(8) Steel is teemed into ladle.
(9) Slag is discharged.
-------
Particulate emissions from the EOF process includes metal, slag,
and impurities swept from the vessel by the supersonic oxygen stream and
the violent exothermic reaction of the oxygen with the impurities in the
pig iron. Gaseous emissions are high in CO and C02 resulting from oxidation
of the carbon in the pig iron. Some form of emission control device is
required to bring particulate emissions into compliance with EPA Performance
Standards. Wet scrubbers or electrostatic precipitators are the types of
control systems which are commonly used.
Sampling Site—Wet Scrubber Control
Sampling experiments involving a wet scrubber emission control
system were performed at a modern EOF factory. The facility has two 200-
ton vessels, only one of which is in operation at any one time. The charge
to the furnace consists of approximately 50 tons of scrap and 150 tons of
hot metal. Scrap preheating is not used.
The emission control system utilizes a high energy wet scrubber
gas cleaning process illustrated in the simplified flow diagram, Figure 1.
The system includes a movable skirt which seals over the vessel during the
blow to conduct particulates and hot gases to a quench section where they
are sprayed with water causing the coarse particulates to drop out. Then
the gases go through a secondary venturi where a large pressure drop provides
additional particulate removal. The gas, ^tien cooled to a temperature of
about 65°C (150°F), goes through the fan and up the stack where a pilot
burner ignites the CO present in combustible quantities. The wet scrubbing
process is designed to operate at about 99.98 percent efficiency.
At the initiation of the blow and near its termination, the head
space above the vessel is purged with nitrogen to sweep oxygen from the
emission control system.
The emissions to the stack are heavily moisture-laden and have the
following characteristics.
Composition
CO - 70 to 75 percent
C02 - 10 to 20 percent
02 - 0.1 to 0.3 percent
NOX - vLOOO ppm max during N2 purge
S02 - negligible
particulates - <50
moisture - saturated with entrained droplets
Velocity
8 to 13 m/sec
Temperature
49 to 65°C
-------
Burner-
Main
cooler
Upper
hood
/—Primary
venturi
Secon-
dary
venturi
Fan
Figure 1. Wet scrubber gas cleaning system
-------
Sampling was performed in 7.62-cm (3-in.) ports on a platform
which is located at the 18.3-m (60-ft) level of the 1.83-m (6-ft) diameter
by 70.1-m (230-ft) stack. Access to the sampling area is by way of a
ladder on the stack.
'. .-JS1-'
The velocity pressure and temperature profile of the stack at
the sampling location are given in Figure 2. The velocity pressure changed
significantly during the course of the 20 minute oxygen blow as may be noted
by a typical pattern given in Table 2.
TABLE 2. VELOCITY PRESSURE AND TEMPERATURE
PROFILE DURING BLOW - EOF WITH WET
SCRUBBER
Time into Blow, Stack Gas Temp.,
minutes ^C AP, cm H20
0
2
4
6
8
10
12 .
14
16
18
20
54
54
54
56
60
60
58
59
57
59
59
0.43
0.89
0.43
0.43
1.14
1.22
1.14
1.14
0.97
0.89
1.78
Sampling Site—ESP Control
Sampling at the site equipped with electrostatic precipitators
(ESP) was performed at a shop with two 180-ton furnaces. During sampling
only one furnace was in operation; however, the furnaces are frequently
operated in tandem. The vessels are charged with about 45 tons of scrap,
without preheating, and 135 tons of hot metzl. Several low alloy steels
are made at the facility through additions to the ladle during teeming.
Only molybdenum additions were made directly to the vessel.
During the blow, emissions are collected by an open hood above
the vessel and conducted through a cooling section (water spray), then into
electrostatic precipitators to remove particulates and finally out through
a 5.79-m (19-ft) diameter by 50-m (165-ft) high stack. A diagram of the
system is shown in Figure 3.
The general characteristics of the stack emissions from the BOF
with ESP are as follows:
10
-------
pitot located
at center of
Stack diameter - 1.83 m (6 ft)
Reading taken at ^0.3 m (1 ft)
spacings
Upper values are velocity pres-
sure, cm H20
Lower values are temperature, °C
Port
B
Figure 2. Velocity pressure and gas temperature profile
of BOF/wet scrubber stack at sampling location
-------
Stack-
/—Sampling ports
Fan
'Electrostatic
precipitator
•Oxygen lance
Steam
injection
\
-Vessel
Figure 3. Basic oxygen furnace with electrostatic
precipitator emission control
-------
Composition
CO - 0.5 percent
C02 - 13 percent
02 - 13 percent
NOx - 176 ppm max
S0£ - 4 ppm max
Moisture - 16 to 25 percent
Particulates - 30 to 180 mg/m
Velocity
^8 m/s
Temperature
130 to 160°C
Sampling was performed at the 46-m (150-ft) level of the stack.
The velocity head (AP) and temperature profile in the quadrant of the stack
in which sampling was performed is shown in Figure 4. Slight variations in
AP were noted during the first three to five minutes of each oxygen blow and
then the AP remained steady. The gas temperature changed steadily during
the blow, usually increasing by about 25 degrees centigrade from the starting
temperature.
SAMPLING EQUIPMENT
Particulates
The particulate sampling was performed with two identical, commer-
cially built, Method 5 trains comprised of components assembled as shown
schematically in Figure 5. The sampling probes, which were constructed at
Battelle, were 2.06 m (6.8 ft) in length and were constructed from approxi-
mately 11-mm-I.D. x 16-mm-O.D. glass tubing. Heating was provided by a 3.05-m
(10 ft) glass fiber-insulated heating tape wrapped around the glass probe.
A thermocouple junction was taped to the outer surface of the glass tubing
at a point between the heater windings and at the midpoint of the probe.
The glass probe and heating tape assembly was insulated by a wrapping of
asbestos tape. A stainless steel sheath was used to protect the glass probe
and to hold the fitting to attach the nozzle. The seal between the glass
probe and the 1.6-cm (5/18-in.) Swagelol© nozzle connection was made with a
silicone 0-ring.
A type "S" pitot tube attached to one of the sampling probes was
used for velocity head readings. The tube was constructed at Battelle from
approximately 7.5-mm-I.D. x 9.5-mm-O.D. stainless steel tubing. A stainless
steel sheathed thermocouple was attached to the pitot tube to provide stack
gas temperature measurements. The type "S" pitot tube was calibrated
against a standard pitot tube over the velocity range of 10.7 to 29.6 m/sec
(35 to 96 ft/sec) in the Battelle wind tunnel facility.
13
-------
Measurement Points-Distance
from Port, meters
.22
.49
.78
1.13
1.55
2.15
Port 3
Upper values are
velocity pressure, cm of H20
Lower values are gas
temperature, C
Port 2
Port
Figure 4. Velocity pressure and gas temperature profile of BOF/ESP
stack quadrant at sampling location
-------
II 12
1. Stainless steel nozzle 7.
2. Glass-lined probe
3. 7.62 cm (3 in.) filter 8.
4. Heated box for filter 9.
5. Ice bath for impinger 10.
6. Modified G-S impinger 11.
with 100 ml water 12.
Greenburg-Smith impinger
with 100 ml water
Modified G-S impinger
Silica gel trap
Thermometers or thermocouples
Vacuum gauge
Flow control valve
13.
14.
15.
16.
17.
18.
Pump
Flow control valve
Dry test meter
Calibrated orifice
Manometer (AH)
"S" type pitot tube
19. Manometer (AP)
Figure 5. Schematic diagram of Method 5 train
-------
Two modifications were made in the Method 5 sampling trains to
permit additional gas temperature measurements at various points and to
control the gas outlet probe temperature. The glass connectors from the
probe outlet to the filter and the filter outlet to the first impinger
were modified for some experiments as shown in Figures 6 and 7. The probe-
to-filter connector contained a thin-wall thermocouple well which extended
about 5.1 cm (2 in.) into the outlet end of the probe. The modified
filter-to-first impinger connector was fitted with a metal thermometer,
the tip of which was positioned about 1.3 cm (0.5 in.) from the frit in
the filterholder.
In-Stack Filters—
In-stack sampling was performed at the BOF/wet scrubber using a
Gelman #2220, 47-mm filterholder. Due to the entrained moisture in the
emissions, it was necessary to heat the filterholder to maintain a dry
filter during sampling. Heating was provided by a silicone rubber-insulated
heating tape wrapped around the filterholder. A metal case was fitted
around the entire assembly to protect it from the moisture. A thermocouple
was attached to the outside of the filter case to measure and regulate
temperature.
Two types of in-stack filter configurations were used in sampling
at the ESP equipped EOF facility. The one type was a 6.25-cm (2.5 in.)
filterholder marketed by Sierra Instrument Company Model 8145. The other
type of in-stack filterholder, shown in Figure 8, was constructed at Battelle
and was designed to use Munktell glass fiber thimbles* which are marketed
by Carborundum Company. The thimble was sealed to the inlet with the spring
loaded, stainless steel collar. A TefIons' gasket was used to seal the filter-
holder body.
The in-stack filterholders were inserted between the nozzle and
the glass-lined probe. All other components of the sampling train were
assembled as shown in Figure 5.
Gas Sampling
Continuous monitoring for SOg and NOX was performed at both EOF
facilities with an Environmetrics Model NS 300AC Faristor unit. The gas
sample was extracted from the stacks through a 0.63-cm (0.25-in.) diameter
stainless steel tube and passed through a moisture trap prior to the Faristor
unit. Teflon® lines were used to connect the monitor to the sampling probe.
Evaculated three liter flasks were used at the EOF/ESP facility to
collect samples for NOX and gas mass spectrometric analysis. CO, C0£, and
02 were determined with Orsat and Fyrite equipment.
*Munktell's Swedish glass fiber thimbles are made by Grychksbo Papersbruk,
AB, Sweden.
16
-------
Figure 6. Thermocouple assembly used to measure gas
temperature at probe outlet
Figure 7. Thermometer assembly used to measure gas
temperature at the box filter outlet
17
-------
Figure 8. In-stack filterholder for glass
fiber thimbles
18
-------
SAMPLE COLLECTION AND ANALYSIS PROCEDURES
Particulate Sampling
In all tests, particulate sampling was performed concurrently
with two systems (designated A and B), each with a separate operator. The
sampling was performed at a fixed-point in the stack in an area of nearly
uniform velocity. Sampling probes of the two systems were inserted into the
duct through two adjacent ports so that the pitot tube attached to one of
the probes was positioned equidistance between the sampling nozzles. The
separation between the pitot tube and each nozzle was about 2.5 cm (1 in.).
The relative nozzle-pitot tube positions and the point of sampling within
the duct are indicated in Figures 2 and 4.
At the start of each test day, the laboratory calibration of the
gas metering components of both sampling systems was checked by setting the
orifice manometer (AH) to the meter box calibration factor (AH/g) and measur-
ing the flow rate through the dry gas meter over a 5-minute period. A flow
rate of 0.21 m3/min (0.75 cfm) confirmed that the gas metering system
remained in calibration.
The preparation of the particulate collection trains for all tests
was performed as specified in Paragraph 4.1.2 of Method 5 with the following
modifications:
(1) The entire sampling train was leak checked by plugging the
sampling nozzle inlet and evacuating to 38.1 cm (15 in.)
of Hg. Leak rates did not exceed 566 cm3/min (0.02 cfm).
This procedure was used for both the regular Method 5 train
and the in-stack filter train. Prior to leak testing, the
in-stack filter assembly was heated to stack temperature
with a heat gun. Heating was continued until the probe was
inserted into the duct to initiate sampling.
(2) The probe was heated until the thermocouple at the outlet
indicated that the desired operating temperature was achieved
prior to initiation of sampling.
(3) Heat guns (260-399°C) were necessary on some occasions for
supplemental heating when the box filter was operated at
temperatures above 121°C (250°F).
In performance of the experiments, sampling trains were operated
as prescribed in Paragraph 4.1.3 of Method 5. The AP, AH, and system
temperatures were read each minute during the first 5 minutes of the blow
and at 5-minute intervals thereafter. Frequent adjustment of sampling rate
was required to maintain isokinetic conditions during the initial minutes of
sampling. The stack velocity was determined by only one pitot tube. However,
two nomographs were used to obtain the proper sampling rate (AH). Tempera-
ture measurements were obtained at the points shown in Figure 9.
19
-------
Probe
Mid-point
Probe outlet
Filter outlet
Filter box
• Impinger
Figure 9. Tempera.ture measurement points
20
-------
Sampling periods were initiated after the hot metal charge and
continued until the end of the oxygen blow. Each test consisted of sam-
pling during three or four blows.
After completion of the tests, the trains were again leak checked,
sealed to prevent contamination, and transferred to the sample recovery area.
Sample Recovery and Analysis
Filters were removed from holders, sealed in Petri dishes or glass
jars (thimbles), and immediately placed in a desiccator. In Method 5-type
tests, the probe and nozzle were disassembled and washed separately. The
probe was first rinsed with acetone without brushing, then rinsed with acetone
while slowly inserting and removing a Nyloi$) brush in a rotating fashion.
The acetone wash and brushing were continued until visual inspection indi-
cated that all particulates were removed. The brush was thoroughly flushed
with acetone prior tb removal from the probe. The probe wash (usually about
100 to 150 ml) was collected in an Erlenmeyer flask sealed onto the probe
outlet ball joint. Particulates were recovered from the nozzle and the inlet
half of the filter holder by alternately brushing and rinsing with acetone.
The wash solutions from all three components (probe, nozzle, and filterholder)
were combined for analysis.
In all tests with in-stack filters, the in-stack filter was removed
from the probe and particulates were recovered from the probe as described
in the previous paragraph. The outlet side of the in-stack filterholder and
the inlet side of the back-up filterholder (box filter) were separately
brushed and washed with acetone and the solutions were combined with the
probe wash. The nozzle and inlet side of the in-stack filterholder, as one
unit, was alternately brushed and washed with acetone; This wash solution
was not combined with the probe wash, but analyzed separately.
At least one 200-ml acetone blank was obtained each day from the
wash bottle dispenser. All acetone wash solutions and blanks were stored in
glass bottles with Teflorflw-lined caps for transfer to the laboratory for
analysis.
The MSA 1106 BH and ADL quartz filters and particulate catches
were desiccated at least 24 hours (usually longer) prior to weighing. It
was found necessary to desiccate the Munktell '(Carborundum) thimbles at least
72 hours prior to weighing (both tare and final) to achieve a constant weight.
The acetone wash solutions were evaporated to dryness in a reverse
airflow, clean hood and the residues were desiccated to a constant weight to
the nearest 0.1 mg.
Calculations were performed as described in Section 6 of Method 5.
Gas Analyses
Continuous analyses for S02 and NOX were performed concurrently
with all tests in the preliminary series of particulate sampling experiments.
21
-------
A gas sample was withdrawn from the EOF stack emissions at the rate of
about 1.5 liters/minute passed through an ice-cooled moisture trap and
analyzed with an Environmetrics NS300 AC monitor. Calibrations were per-
formed at the start and end of each test day by passing standard gas
mixtures through the sample inlet system. Grab samples for mass spectro-
metric analysis were taken with evacuated 2-liter glass flasks. Orsat
analysis were made on integrated bag samples (Method 3) collected over the
oxygen blowing period.
TEST DESCRIPTIONS AND RESULTS
BOF with Wet Scrubber Emission Control
Experiments were performed to compare in-stack sampling with
Method 5 and to evaluate the sensitivity of Method 5 to the operating
variables—system temperature and deviation from isokinetic sampling rate.
A summary of the experimental conditions is given in Table 3. Runs 1
through IS^wire conducted to study the characteristics of in-stack sample
collections and to compare this technique with Method 5. Two different
filter^materials, Pallflex Tissuquartz 2500 QAO and MSA 1106BH, were used
in the Method 5 trains to study filter/emission interactions.
The results of the BOF/wet scrubber experiments involving com-
parison
-------
TABLE 3. SUMMARY OF EXPERIMENTS — EOF WITH WET SCRUBBER
U)
Run #
1A
B
2A
B
3A
B
4A
B
5A
B
6A
7A
B
8A
B
9A
B
Filter
Configuration
In-stack
Method 5
In-stack
Method 5
In-stack
Method 5
In-stack
Method 5
In-stack
Method 5
In-stack
Method 5
In-stack
Method 5
In-stack
Method 5
In-stack
Method 5
Filter
Material
Tissuquartz
Tissuquartz
Tissuquartz
Tissuquartz
Tissuquartz
Tissuquartz
Tissuquartz
MSA 1106
Tissuquartz
MSA 1106
Tissuquartz
Tissuquartz
Tissuquartz
Tissuquartz
Tissuquartz
Tissuquartz
Tissuquartz
Tissuquartz
Run #
Other variables
10A
B
11A
B
12A
B
13A
B
14A
B
ISA
B
16A
B
17A
B
ISA
B
19A
B
Filter
Configuration
In-stack
Method 5
In-stack
Method 5
In-stack
Method 5
In-stack
Method 5
Method 5
Method 5
Method 5
Method 5
Method 5
Method 5
Method 5
Method 5
Method 5
Method 5
Method 5
Method 5
Filter
Material
Tissuquartz
Tissuquartz
Tissuquartz
MSA 1106
Tissuquartz
MSA 1106
Tissuquartz
MSA 1106
MSA 1106
MSA 1106
MSA 1106
MSA 1106
MSA 1106
MSA 1106
MSA 1106
MSA 1106
MSA 1106
MSA 1106
MSA 1106
MSA 1106
Other variables
System Temperature "V121C
System Temperature ^191C
System Temperature ^12 1C
System Temperature
-------
TABLE 4. SAMPLE WEIGHT DATA
Run No.
1A
IB
2A
2B
3A
3B
4A
4B
5A
5B
6A
6B
7A
7B
8A
SB
9A
9B
10A
10B
HA
11B
12A
12B
13A
13B
In-Stack
Filter
46.3
93.4 (b)
64.9
12.9
42.4
38.1
14.8(C>
29.8
39.7
54.5
64.9
41.0
32.6
Sample
Box
Filter
0.0
31.1
33.4
100.0
39.8(b)
119.7
32.3
57.4
0.0
61.7
0.1
40.5
34.5
50.4
0.0
15.1
1.4
24.9
0.0
26.5
0.0
29.6
0.8
32.6
0.0
22.9
Weights,
Probe
8.6
181.7
87.5
10.6
4.5
8.0
5.6
4.4
5.8
2.2
6.2
4.1
2.8
mg
m H.....JJL - .J .—.!-.- .-...-I
Nozzle
8.7
31.3
16.9
8.5
4.6
4.4
1.8
9.5
6.2
11.6
4.7
8.0
8.3
, n . (•=,) Matribution in Sampling System. % Total Catch
Probe/
Nozzle
18.9
66.3
34.1
10.0
4.7
4.4
8.1
32.2
18.0
29.0
37.3
53.2
35.1
Total
63.6 (55.0)
50.0
339.8
166.3
209.1
153.8
64.3
67.4
51.5 (47.0)
66.5
50.6 (42.5)
44.9
56.7
58.5
43.7 (39.3)
47.3
53.1 (45.9)
42.9
68.3 (66.1)
55.5
75.8 (69.6)
66.9
53.9 (49.0)
85.8
43.7 (40.9)
58.0
rag/Nm3 ng/H« -B
62.0 (53.6) 1.56
39.8
130.4 — (b)
34.7
100.5 — *b*
33.3
36.1 — (c)
28.5
24.9 (22.7) 1.00 (0.91)
24.9
30.5 (25.6) 1.18 (0.99)
25.8
33.1 — fc)
32.8
69.4 (62.4) 1.26 (1.13)
55.1
37.2 (32.2) 1.30 (1.12)
28.7
43.7 (42.3) 1.27 (1.23)
34.4
38.4 (35.3) 1.20 (1.10)
32.1
32.1 (29.2) 0.64 (0.58)
50.0
27.6 (25.8) 0.78 (0.73)
35.4
. System A
Averages - „
6 System B
In-Stack
Filter
72.8
—
—
—
82.4
75.4
—
68.1
74.7
79.8
85.6
76.1
74.6
76.7
Box
Filter
0.0
62.2
60.1
77.8
85.2
0.0
92.9
0.2
90.2
86.2
0.0
31.8
2.7
58.0
0.0
47.7
0.0
44.2
1.5
38.0
0.0
39.5
0.4
62.6
Probe
13.5
—
—
—
8.7
15.7
—
10.1
10.9
3.2
8.2
7.5
6.5
9.4
Nozzle
13.7
—
—
—
8.9
8.6
—
21.8
11.6
17.0
6.2
14.9
18.9
13.5
Probe/
Nozzle
37.8
39.1
22.2
14.8
7.1
9.8
13.8
68.2
42.0
52.3
55.8
62.0
60.5
37.3
(a) Data based on total system catch. Values in parenthesis are calculated from filter and nozzle collections
(b) In-stack filter appeared that moisture had condensed on it sometime during test.
(c) In-stack filter ruptured.
only.
-------
The operating temperature of the in-stack filter was difficult
to maintain at a constant level. Average reading for Runs 5 through 13
ranged from 113°F to 164°F. However, the comparative mass loading results
(A/B) show no correlation with in-stack temperature variations.
The results of experiments with Method 5 to evaluate the effects
of anisokinetic sampling and system temperature are given in Table 5. Stack
gas and sampling data are given in the appendices.
In Runs 17 and 18 the sampling was performed by a pair of Method
5 trains, one of which was operated under anisokinetic conditions. The
comparison of the results indicate that deviations of 1.3 and 0.7 from iso-
kinetic sampling rate do not introduce appreciable error in the Method 5
mass measurement.
In the other experiments sampling was performed with a pair of
Method 5 trains, one of which was operated at the minimum specified gas
temperature at the probe outlet, i.e., 121°C. The filter box on this system
was also held at 121°C (250°F). The gas temperature at the probe outlet
and the filter box temperature of the other Method 5 trains was maintained
at 191°C (375°F) for Run 14, 149°C (300°F) for Run 15, and at about stack
temperature, M34°C (183°F) for Run 19. The mass results obtained with the
systems in which temperature was varied above and below 121°C (250°F) exhibit
excellent agreement with the "normal" Method 5 data. The variations are not
significantly different from those observed in the two concurrent Method 5
measurements (Run 16).
If it is assumed that there is no interaction of system tempera-
ture with the mass results, Runs 14, 15, 16, and 19 may be used to obtain
an estimate of Method 5 precision. Based on these data, repeatability
(within-laboratory precision) of Method 5 for sampling the wet scrubber
equipped BOF is estimated to be 2.8 percent.
BOF with ESP Emission Control
The experiments at the BOF/ESP facility included a series of
experiments of confounded factorial design^ to study the effects of filter
media and sampling system temperature on the mass results and a test series
to compare Method 5 and two types of in-stack filter configurations.
The effect of Method 5 sampling temperature (filter box and gas
at probe outlet) was studied at two levels: the minimum specified tempera-
ture, 121°C (250°F) and at approximately the temperature of the BOF stack
gas, 177°C (350°F). Two filter materials, MSA 1106BH and an ADL-developed
high purity quartz fiber, were included in the experimental design. The
pattern for the filter media—sampling system temperature experiments is pre-
sented in Table 6 and results of the experiments are given in Table 7.
Statistical analysis of the data was performed by analyses of variance tech-
niques (8) yielding the statistical data and conclusions given in Table 8.
Based on the criteria of a 95 percent confidence level, it is concluded that
differences in results obtained with sampling system temperatures of 121°C
25
-------
TABLE 5. SAMPLE WEIGHT AND MASS LOADING
DATA - EOF WITH WET SCRUBBER
Sample Weights, mg
Run No.
14A
14B
15A
15B
16A
16B
17A
17B
ISA
18B
!9A(a)
19B(a)
System M.21°C
System ^191°C
System M.21°C
System vL49°C
System VL21°C
System vL21°C
M.00% Isokinetic
VL30% Isokinetic
^70% Isokinetic
^100% Isokinetic
System VL21°C
System - stack temp.
Filter
85.3
85.5
43.7
43.1
79.2
76.8
42.6
52.0
51.8
69.7
18.1
17.4
Probe
34.3
24.0
14.6
13.3
37.4
33.3
13.7
25.1
12.0
14.7
6.6
6.2
Total
119.6
110.5
58.3
56.4
116.6
110.1
56.3
77.1
63.8
84.4
24.7
23.6
mg/m^
56.3
53.1
33.4
34.0
61.9
59.9
33.3
36.2
46.8
46.1
54.2
55.4
(a) Run included only one blow. Process interruption precluded
additional sampling.
26
-------
to
TABLE 6. RANDOMIZED TEST PATTERN FOR STUDY OF FILTER MEDIA, TEMPERATURE,
AND THEIR INTERACTION - (BOF WITH ESP EMISSION CONTROL)
Rep
1
2
3
Block
1
2
1
2
1
2
Test
Number
1
2
3
4
5
6
Temp.,°C
VL77
121
121
^177
121
vL77
System A
Filter
ADL
ADL
ADL
MSA
MSA
ADL
Operator
1
2
2
2
1
1
Temp. , °C
121
-------
00
TABLE 7. SAMPLE WEIGHT DATA - TEMPERATURE/FILTER EXPERIMENTS -
BOF WITH ESP EMISSION CONTROL
(„•)
System Temperalture, °CV '
Run No.
1A
IB
2A
2B
3A
3B
4A
4B
5A
5B
6A
6B
Box Filer
176
122
121
177
123
177
177
120
126
122
179
177
Probe Outlet
179
122
129
178
123
179
176
128
127
127
179
185
Filter
Material
ADL
MSA
ADL
MSA
ADL
ADL
MSA
MSA
MSA
ADL
ADL
MSA
Sample weights,
Filter
43.2
44.2
36.8
50.7
199.1
229.4
123.7
93.5
154.2
137.0
71.7
70.8
Probe
18.9
17.7
21.2
20.4
108.4
94.0
34.5
64.6
47.4
55.2
34.3
27.1
mgs
Total
62.1
61.9
58.0
71.7
307.5
323.4
158.2
158.1
201.6
192.2
106.0
97.9
mg/Nm3
33.7
34.1
32.8
40.5
165.4
177.4
118.2
120.4
139.1
138.4
92.6
85.4
(a) Actual average system temperatures.
-------
TABLE 8. ANALYSES OF VARIANCE-TEMPERATURE/FILTER EXPERIMENTS--
EOF WITH ESP EMISSION CONTROL
t-o
Source
Filter (F)
Temp. (T)
FxT
Reps
Block/Reps
Remainder
Total
(a) Critical
a
0.100
0.050
0.025
0.010
Degrees of Sum of
Freedom Squares
1
1
1
2
3
3
11
0.32
36.55
61.05
25,713.52
5,197.04
32.39
31,040.87
Mean .. ..
Square F-Ratio W
0.32 0.03
36.55 3.39
61.05 5.66
12,856.76
1,732.35
10.80
2,821.90
Blocks with
Conclusion Information
Not significant 1, 2,
Not significant 1, 2,
Not significant for 3, 4,
a < 0.05
Significant for
a = 0.10
All
All
All
All
5, 6
3, 4
5, 6
values of F-Ratio.
Critical Value
5.54
10.13
17.44
34.12
-------
(250°F) and 177 (350°F) and with the MSA 1106BH and the ADL developed
quartz fiber filter materials are not statistically significant. Further-
more, there is no statistically significant interaction effect between
sampling system temperature and filter media.
Assuming no filter or temperature effects, results of Runs 1
through 6 may be used to estimate Method 5 precision. Based on these data,
the repeatability (within-laboratory precision) expressed as the coefficient
of variation is 3.4 percent.
Four experiments (Runs 7 through 10) were performed to compare
results of Method 5 and in-stack sampling techniques. In two experiments,
in-stack glass fiber thimbles marketed by Carborundum Company were used and
in the others, 6.25 cm (2.5 in.) MSA 1106BH in an in-stack flat filterholder
were employed. MSA 1106BH filter material was used in the Method 5 trains.
The results of the experiments are presented in Table 9- Both the in-stack
thimble and flat filter based on the filter and nozzle catches gave lower
mass loading results than the corresponding Method 5 values. With exception
of Run 7, mass loadings were also lower when calculated from the total
in-stack system particulate collection. Based on nozzle and filter catches,
the in-stack flat and thimble filters gave results which were an average of
25 and 16 percent lower, respectively, than Method 5 results.
Statistical analysis of the test pairs does not indicate that the
differences between the in-stack and Method 5 results are significant. How-
ever, the power of the statistical analysis is greatly diminished by the
experimental design and number of replications. Statistical conclusions
aside, the consistency of the in-stack and Method 5 differences would seem
to argue that the two methods do not give equivalent results.
CHARACTERIZATION OF PARTICULATE COLLECTIONS
AND GASEOUS EMISSIONS
Chemical Composition of Particulates
Extensive chemical analyses of particulate and gas samples from
both EOF facilities were conducted to investigate possible interactions in
the sampling process and to determine if the particulate collections were
representative of the stack emissions. Table 10 presents the results of
optical emission spectroscopy analysis of Method 5 filter and probe catches
and samples taken from the wet scrubber emission control system. In general,
the compositions of the Method 5 filter collections are essentially the same
as particulates removed by the wet scrubber (clarifier sludge). Comparison
of the filter and probe collections shows a disproportionate distribution of
several elements, e.g., Fe, K, and Na. The relatively higher Na and K con-
centrations in the probe may arise from entrained liquid droplets in the
stack emissions. Analysis indicates that these elements comprise the major
fraction of the dissolved solids in the scrubber liquid.
Compositional analysis of particulates collected from the BOF/wet
scrubber facility by Method 5 and in-stack sampling are given in Table 11.
30
-------
TABLE 9. SAMPLE WEIGHT AND DISTRIBUTION DATA - METHOD 5 AND IN-STACK
COMPARISON TESTS - EOF WITH ESP EMISSION CONTROL
Run No.
7A
7B
8A
8B
9A
9B
10A
10B
Run No.
7A
78
8A
8B
9A
9B
10A
10B
Sampling (a)
Method
Method 5
In-stack thimble
Method 5
In-stack thimble
Method 5
In-stack flat
Method 5
In-stack flat
r B a>
Method
Method 5
In-stack thimble
Method 5
In-stack thimble
Method 5
In-stack flat
Method 5
In-stack flat
Sample Weights, rags
Box Filter
115.1
0.9
74.6
0.0
60.5
0.9
67.3
1.2
In-stack
Filter
__
138.7
— —
77.5
—
69.1
—
73.6
Probe
46.7
16.3
37.4
15.5
41.o(c)
17.3
Sample Distribution,
Box Filter
71.1
0.5
66.6
0.0
57.2
1.0
62.1
1.2
In-Stack
Filter
__
83.6
—
78.8
74.3
73.2
Probe
28. 9 (c)
9.8
33.4
12.1
42.8
16.7
37.9
17.2
Nozzle
.._
10.1
__
9.0
—
7.5
—
8.5
Percent
Nozzle
__
6.1
__
9.1
—
8.1
__
8.4
Total
161.8
166.0
112.0
98.4
105.7
93.0
108.3
100.6
of Total Collection
In-Stack Filter
and Nozzle
__
89.6
87.9
__
82.4
-.—
81.6
m8/Nm3(b)
131.6
121.1 (135.2)
93.4
71.6 (81.5)
93.1
68.3 (82.9)
86.4
66.4 (81.4)
Behind
In-Stack Filter
__
10.4
__
12.1
— _
17.6
— _
18.4
(a) MSA 1106BH filters were used for Method 5 and in-stack flat filter experiments. In-stack thimble was
Munktell glass fiber filter.
(b) Based on nozzle and in-stack filter catch. Values in parenthesis are calculated from total system catch.
(c) Probe and nozzle rinse combined.
-------
TABLE 10. ANALYSIS OF METHOD 5 AND SCRUBBER SAMPLES
EOF WITH WET SCRUBBER EMISSION CONTROL
Weight Percent in Sample
Sample Fe Si Na K Ca Zn Mg Mn Al Pb B Tl Sn Hi Co Cr V Mo Cu Ag
Method 5 Filter 30-50 1 1 2-3 0.5 1 0.1 0.5 0.02 0.5 0.01 <0.005 — 0.003 0.003 0.01 0.005 0.003 0.01 <0.001
Run 14A
Method 5 Filter 30-50 1 1 2-32 5-10 0.3 0.5 0.02 1 0.01 0.005 — 0.003 0.003 0.01 0.01 0.003 0.03 0.001
Run 16A
Method 5 Probe 15-30 1 4-7 10-20 1 2-3 0.3 0.5 0.3 0.5 0.02 0.01 — 0.03 <0.003 0.03 <0.005 0.003 0.03 0.005
Residue
Run 14A
Method 5 Probe 15-30 1 4-7 10-20 1 2-3 0.3 0.5 0.3 0.5 0.03 0.02 — 0.03 <0.003 0.01 <0.005 0.003 0.03 0.003
Residue
Run 16A
^ Clarifier 40-60 0.5 <0.1 <0.1 2 2-3 0.3 0.6 0.03 0.3 <0.01 0.01 <0.01 0.003 <0.01 0.01 0.005 0.003 0.03 0.001
Suspended 40-60 0.5 0.1 <0.1 2 0.2 0.6 2 0.03 0.03 <0.01 0.005 <0.01 0.005 <0.01 0.01
solids 0>)
Dissolved 0.01 0.1 30-50 30-50 0.3 <1.0 0.2 <0.1 0.01 <0.01 0.01 <0.005 <0.01 <0.005 <0.01 <0.01
solids 0>)
(a) Analyses performed by optical emission spectroscopy.
(b) Samples of stack particulates taken from scrubber cleanup system.
-------
u>
TABLE 11. CHEMICAL COMPOSITIONS OF PARTICIPATE COLLECTIONS
FROM EOF WITH WET SCRUBBER
(Results in weight percent)
Run
8A
9A
11A
12A
13A
Average
8B
9B
11B
12B
13B
Average
9A
11A
8A, 12A,
Average
9A
11A
Average
8B
9B
"11B
12B
13B
Average
System/
Component
In-stack/
Filter
Method 5/
Filter
In-stack/
Nozzle
13A
In-stack/
Probe
Method 5/
Nozzle and
Probe
Fe
27.5
31.5
25.0
38.5
40.5
32.2
47.5
55.0
59.0
41.0
60.0
51.5
•v.15.
M.9.
26.0
20.0
6.6
2.4
4.5
13.2
•v.11.
8.7
7.2
7.4
9.5
Si Na K
•v.3 9.0 16.0
•v.3 5.0 6.3
••v-3 6.1 14.0
•v.3 5.1 10.5
•v.3 5.5 16.0
3 6.1 12.6
•v.3 6.0 13.1
•v-3 3.0 9.0
•v.3 1.0 8.0
•v.3 6.5 4.5
•v.3 7.0 6.5
3 4.7 8.2
•v.3 "v.1 . *2 .
•v.3 'U) . 5 *v.2 .
VJ 4.0 5.0
•v.3 1.8 3.0
•v.3 ^2 . -v.2
•v.3 "HO. "v.10.0
•v.3 -\,6. -v4>.
— 7.9 18.5
•v.3 "v.10. MO
•v.3 8.0 22.9
— 10.0 20.0
— 8.7 16.3
3. 8.9 18.0
Mn
1.0
1.2
ITT
1.7
2.1
—
1.9
•v.1.
•v.2.
3.0
2.0
•vfl.7
M).4
0.5
•v.1
0.85
—
1.
1.
Ca
0.9
0.5
377
1.6
1.0
1.9
1.5
•v.1.
•v.2.
—
1.5
•v.1.
•vfl.9
1.
__
•v-1
0.6
—
1.4
1.0
Al
MI!?
MM
•vfl.5
•vfl.5
—
•v.0.5
VI.
•vfl.3
•vfl.2
0.5
•v.1.
•vfl.5
0.8
„.
•v.1.
0.5
—
-vfl.25
075-
Mg
•vfl.5
•v.0.5
•v.0.5
•\5~5"
•vO.2
•v.0.2
—
M).2
•v«.3
M).5
—
0.4
•V0.8
•vO.5
0.7
•vfl.5
•v.0.2
—
1.
0.5
Zn
0.3
1.2
0.6
0.7
0.5
1.8
—
1.2
0.5
0.5
—
0.5
•v.1.
•vfl.3
0.7
•v.0.2
•v.1.
—
•v.0.2
•vfl.4
Other
Pb Metals
0.5
0.4
0.3 ~
—
ifl.3 <0.1
•v.0.3
^•0.5
— —
0.4 <0.1
•vO.2
•vfl.2
0.2 <0.1
M).5
•vfl.3
—
_ _
0.4 -
C
Uncombined
1.0
0.75
0.4
3.0
0.2
0.4
1.2
—
12
12
__
1.0
1.0
10.
—
0.3
3.0
4.0
co3-
15.2
10.0
16.0
13.7
10.0
7.3
6.5
7.9
—
--
31.5
_,
__
20.0
20.0
27.9
—
15.0
20.0
25.0
22.0
so"
4.0
4.0
6.0
3.5
4.4
8.0
2.1
4.2
3.0
5.3
4.5
—
6.5
6.5
__
13.0
13.0
12.9
—
29.0
18.8
13.8
18.6
F~
2.1
1.1
1.8
2.7
1.8
1.9
3.4
2.9
1.5
3.5
4.0
3.1
—
—
2.5
2.5
„
2.5
2.5
2.8
—
2.0
3.2
2.6
2.7
Total of Average
Individual . .
Cl~ Determinations
4.5
9.0
3.0
5.3
2.3
2.3
4.5
5.0
3.9
—
2.0
2.0
2.0
2.0
6.2
—
9.0
9.3
8.0
8.1
•vlOOZ
117. 4Z
99. 5%
67. 2Z
106.2%
(a) Fe and Si calculated as oxides.
-------
In general, the analyses were obtained by use of an optical emission spectro-
graphic method to show the approximate cation content of the samples. These
analyses were followed by use of the more precise atomic absorption method
for the major cations, Fe, Na, K, and for confirmatory determinations on
important minor elements, Mn, Zn, Pb, Ca. Anion contents were determined by
classical wet chemical, and ion-selective electrode methods. The principal
anions found present were C0|, SO^, F~, and Cl~. Total N analyses indicated
little or no NH^ or NO3 ions could be present. Limited X-ray diffraction
examinations showed iron oxide as the probable form of iron in the collec-
tions.
Comparison of the collections within the Method 5 and in-stack
system components show a disproportionate distribution of several anions and
cations. Method 5 filter samples contain a higher percentage of iron than
the probe, while the probe samples show higher fractions of Na, K, COf, and
SO^. The in-stack filter catch contains higher percentages of Fe, Na, and K
and a lower percentage of CO^ than detected in the nozzle particulates. Dis-
similarities are also exhibited among the fractions of Fe, Na, K, CO^, F~,
and Cl~ in the Method 5 and in-stack filter collections. With exception of
the in-stack probe data, the sum of the average cation and anion values (Fe
and Si calculated as oxides) provides essentially a quantitative account of
the composition of the various samples.
Although collections of the two trains (in-stack and Method 5) show
fractionation, the differences in composition of major cations and anions in
the total system catches when adjusted for weight distribution in the sampling
system and normalized to 100 percent (last column of Table 11) are less
evident as illustrated in Table 12 below.
TABLE 12. PERCENTAGE DISTRIBUTIONS OF MAJOR CATIONS AND
ANIONS WITHIN THE SAMPLING TRAIN COMPONENTS
Fe
Na
K
COf
ci-
In-Stack System
In-stack filter 24.50 4.65
Nozzle 2.80 2.52
Probe .67 .82
Totals: 27.97 7.99
9.60
.42.
.82
10.40
4.41
2.70
3.35
.90
2.00
1.45
.35
.34
10.84
17.51
6.25
2.14
4.05
.28
.27
4.50
Method 5
Box filter 23.8
Nozzle and probe 4.1
Totals: 27.9
3.8 3.65 2.1
7.8 9.40 8.0
11.6 13.05 10.1
1.78
3.50
5.28
34
-------
Caution must be taken in drawing firm conclusions from the
above tabulations because of the overall averaging and the possible indi-
vidual analytical errors. Within the cautionary limitations, it appears
indicated that, with the exception of SO^, the overall composition of
loadings of the two train systems are similar, i.e., the train configura-
tion and mode of collection does not materially alter the chemical form of
the particulate collections. If alterations do occur, it appears indicated
that there are somewhat higher contents of Na+ and COf in the in-stack
collections and somewhat higher K+, SO^, Cl~, and F~ in the Method 5 collec-
tions.
The general chemical composition of Method 5 samples and a grab
sample taken from the BOF/ESP emissions is shown in Table 13. The data,
which were obtained by optical emission spectroscopy, do not show .any signi-
ficant compositional differences between Method 5 filter and probe collections
and between the Method 5 grab samples.
More quantitative data for several cations and anions in the EOF/
ESP samples are presented in Table 14. Cations analyses were performed by
atomic absorption spectroscopy, and the anions were determined by classical
wet chemistry and ion-selective electrode procedures. The Method 5 filter
catches and the grab sample show similarity in composition except for Zn,
SO^, F , and C which were found in higher concentrations in the grab sample.
Iron predominates the sample contents, accounting for 91+ percent of the
weight when assumed to be present as the oxide. The higher carbon value in
the probe samples indicates the presence of organics probably introduced in
the Method 5 acetone probe washing procedure.
Gaseous Emissions Analysis
The typical composition of the stack gas at the BOF/wet scrubber
facility a£ determined by Orsat and an Environmetrics Faristor monitor are
given in Table 15 . The gas contains a high concentration of CO which is
flared at the top of the stack. Rather high NOX levels probably result from
nitrogen introduced during the blow to purge oxygen from the emission control
system. S02, if present in the furnace emissions, was probably efficiently
removed by the scrubber system and consequently was not detected in the stack
gas.
The concentrations of S02 and N0x.in the stack gas emission during
a typical blow at the BOF/ESP facility are shown in Figure 10. The NOX
level increases steadily during the blow to a maximum level in the range of
about 90 to 150 ppm at the termination. The S02 level remains fairly constant
at a level of about 4 ppm.
Analysis for other gaseous species and organics was performed on
samples withdrawn with evacuated glass flasks. The analytical results are
reported in Table 16. The primary gas components are C02, CO, 02, and N.
S02 was detected at approximately the same levels measured with the Environ-
metrics Faristor. Other sulfur compounds such as H2S, COS, and CS2 were not
detected (minimum detectable level is 2 ppm). The only organic present at a
detectable level was methane.
35
-------
TABLE 13. ANALYSIS OF PARTICULATE EMISSIONS FROM EOF
EQUIPPED WITH ESP CONTROL^a'b^
Element
Fe
Si
Na
K
Mn
Ca
Mg
Zn
Pb
Ni
V
Cr
Co
Cu
B
Al
Mo
Sn
Ti
Filter (3A)
40-60
1
0.2
0.5
1 -
1
0.3
0.5
0.2
0.005
0.01
0.01
<0.001
0.03
0.001
0.02
0.01
0.003
0.005
Filter (3B)
40-60
1
0.2
0.5
1
2
0.3
0.5
0.2
0.005
0.0
0.01
<0.001
0.03
0.001
0.02
0.01
0.005
0.005
Probe (3A)
40-60
1
0.2
0.5
1
3-4
0.3
0.5
0.2
0.01
0.01
0.02
<0.001
0.03
0.001
0.1
0.01
0.01
0.03
Probe (3B)
40-60
1
0.2
0.5
1
3-4
0.3
0.5
0.2
0.02
0.01
0.02
<0.001
0.03
0.001
0.3
0.01
0.02
0.03
Grab
40-60
1
0.2
0.5
1
3-4
0.3
0.5
0.2
0.01
0.01
0.01
<0.001
0.03
0.001
0.03
0.01
0.1
0.02
(a) Analysis performed by optical emission spectroscopy.
(b) Results in weight percent.
-------
TABLE 14. CHEMICAL ANALYSIS OF PARTICULATE EMISSIONS
FROM EOF WITH ESP CONTROL
Weight Percent in Sample
Sample (Run) Fe Mn Ni V Zn S04~ Cl F COg C H N P
Method 5 Filter (5A) 70.5 1.27 0.011 0.010 0.43 0.68 0.89 0.28 ND 0.1 <0.1 <0.1 0.03
Method 5 Filter (5B) 71.6 1.28 0.011 0.010 0.44 0.66 0.96 0.64 -- 0.1 <0.1 0.1 0.06
Method 5 Probe
u> Residue (5A) 53.8 1.01 0.015 0.008 0.81 0.92 1.15 0.50 ND 7.3 1.0 0.1 0.13
•^j
Method 5 Probe
Residue (5B) 56.4 1.03 0.017 0.009 1.47 0.84 0.92 0.83 — 5.7 0.8 0.1 0.05
Grab Sample 71.1 1.19 0.010 0.013 0.87 0.98 0.88 1.10 ND 0.4 0.1 <0.1 0.08
-------
TABLE 15. GAS ANALYSIS OF EOF EMISSIONS
WITH WET SCRUBBER CONTROL
Gas
Concentration in Stack Emissions
(a-)
CO
C02
02
NOX
S02
63 - 81%
9 - 20%
0.1 - 0.3%
1034 - 1050 ppm
Not detected, >5 ppm
(a) Concentration ranges observed over
four heats.
TABLE 16. GAS CHROMATOGRAPHIC AND MASS SPECTROMETRIC
ANALYSIS OF GASEOUS EMISSIONS FROM EOF
WITH ESP CONTROL
Volume Percent
C02 12.6 to 13.4
02 12.2 to 13.0
CO 0.37 to 0.87
N2 72.3 to 73.0
A 0.93
H2 <0.1
S02 <0.0001 to 0.0004
C2H2 <0.1
C2Ht^ ^O.l
CgHg ^0.1
CijHg <0.1
ppm
HC1 <2
CS2 <2
H2S <2
COS <2
NOX 3 to 21
38
-------
u>
MD
Range - NOX 200 ppm full scale
S02 200 pp full scale
Figure 10. NOX and S02 concentrations in EOF emissions during
the oxygen blow - EOF with ESP emission control
-------
Impinger Solution Analysis
Impinger solutions from sampling at both EOF facilities were
analyzed according to the Method 5 procedure, August 18, 1977. The results
shown in Table 17 were obtained.
TABLE 17. RESULTS OF ANALYSIS OF IMPINGER COLLECTIONS
Sample Residue Weight, mgs(a)
Facility
BOF with
Scrubber
Run No.
16A
16B
Water (extracted)
2.8
4.1
Chloroform/Ether
1.4
1.4
Acetone
0.3
0.0
18B 1.8 1.0 0.7
BOF with ESP
3A
3B
4.9
5.1
2.0
6.9
1.9
0.2
(a) Weights corrected for blanks.
Both the aqueous and organic extract residues gave very low weights. Total
impinger catches are approximately 4 to 5 percent of the front end collection
for the BOF/wet scrubber and about 4 percent of the front end catch for the
BOF/ESP.
Particle Size Measurements
Particle size distribution measurements of the Method 5 and in-stack
filter collections from the BOF/wet scrubber emissions are presented in Figures
11 and 12, respectively. The measurements were made by electron microscopic
examination of the particulate catches. The data show that the emissions con-
sist of particulates which are primarily submicron in size. Collections by
both techniques show essentially the same size distribution with a mean mass
diameter of about 0.2 microns.
Particle size distribution measurements of the BOF/ESP facility were
made with an Anderson cascade impactor. The mass distribution plotted versus
particle diameter is given in Figure 13. Particulate emissions from the ESP
are larger in size than from the wet scrubber, the mass mean diameter being
about 4 microns.
40
-------
ffi
s
I
I
E
3
u
99.99
99.9
99
96
95
90
80
TO
60
50
40
30
20
10
5
2
0.5
0.2
0.1
0.05
:
•
m
•
•
:
•
;
£
m
;
=
E
;
7
•
-
:
•
*
•
:
-
X
°'00.02 0.03 0.04
X
*
• i
X
X
X
X
X
X
X
* —
*
X
X
x
*-
l^»
0.06 0.08 O.I 02 0.3 0.4 0.6 0.8 1 2
Equivalent Particle Diameter, micrometers
Figure 11. Particle size distribution of BOF/wet
scrubber emissions collected on
Method 5 filter - run 5B,
41
-------
a>
m
c
>
>
X
X
X
I 1 i 1 1 1 1 l l
X
X
X
X
X
X
X
X
0.02 0.03 0.04 0.06 0.06 O.I 02 0.3 0.4 0.6 0.8
Equivalent Particle Diameter, micrometers
Figure 12. Particle size distribution of BOF/wet
scrubber emissions collected on in-stack
filter - run 5A
42
-------
99.99
99.9
99
98
95
90
'§" 80
>» 70
03
IE 60
L- 50
°~ 40
>
•£ so
0 JVI
"5
C Oft
3
m
rt *?
O.Z
O.I
O.O5
n ni
T
|-
:
.
~
i
r
1
;
I
•
;
f
•
X
X
'
II 1
~T~| X
1 IHI 1 1 i i i i i I I
0.2 0.3 0.4 OS 06 0.8 I 2 3 4 5 6 7 8 9 10
Equivalent Particle Diameter, micrometers
20
Figure 13. Particle size distribution of emissions
from ESP equipped basic oxygen furnace
43
-------
SECTION 5
DISCUSSION
The study does not indicate any problems associated with the
use of Method 5 to determine particulate emissions from basic oxygen
furnaces equipped with wet scrubber or electrostatic precipitator con-
trols. Potentially, wet scrubber emissions could present the most diffi-
culty since the sampling system must handle a moisture-laden gas stream
with entrained water droplets. If water accumulates on the filter, a
high pressure drop will result and limit sample flow. In addition, a wet
filter would be more prone to reaction with gas phase species. However,
experiments in which the Method 5 sampling train temperature was held at
121°C, the minimum specified temperature, and below, ^84°C, showed no
accumulation of moisture on the box filter.
On the other hand, in-stack sampling presents a formidable problem
in maintaining a dry filter during sampling. Without heating, moisture
plugs the filter and rupture frequently occurs. External heating may be
used with limited success, but adds additional complexity to the sampling
process.
The experiments with the BOF/wet scrubber demonstrate that results
obtained with Method 5 are not affected by rather large variations in sam-
pling system temperature or deviations from isokinetic sampling rate. Sam-
pling with system temperatures over the range of 84 to 191°C gave results
which were in good agreement with sampling performed at the normal Method 5
operating temperature, 121°C.
Variation of the sampling rate at 0.7 and 1.3 times isokinetic
also did not significantly affect the accuracy of the mass measurements.
The particulate emissions from the scrubber are shown to be very small in
size (MMD ^0.2y), consequently, deviations from isokinetic sampling would
be expected to have a negligible influence on mass measurements.
The chemical characterization work indicates that Method 5 collec-
tions are generally representative of the stack particulates. The Method 5
filter collections show the same composition, within the limits of accuracy
of the analytical method used, as the particulates removed from the stack by
the wet-scrubber system.
In sampling emissions from an ESP equipped EOF, Method 5 was also
found to give reliable, reproducible results. Use of a higher purity, lower
pH filter medium (ADL quartz) gave mass results which were not statistically
different from those obtained with MSA 1106BH, the commonly used filter
material. Operation of the sampling system at stack temperature, which would
be expected to reduce errors due to condensation reactions, gave mass loading
results which were statistically indistinguishable from the values obtained
with the train operated at 121°C.
44
-------
The chemical analyses confirms that the Method 5 procedure ex-
tracts a representative sample of the BOF/ESP stack particulates. The same
general chemical composition was found in Method 5 filters and grab samples
removed from the stack at the sampling point.
Method 5, when compared to two in-stack sampling configurations,
appears to give higher mass results. Additional experimentation is necessary
to confirm this observation and to identify the source(s) of the discrepancy.
45
-------
REFERENCES
1. Federal Register, 36(247):24876-24895, December 23, 1971.
2. Federal Register, 39(47):9308-9323, March 8, 1974.
3. Federal Register, 39(177):32852, September 11, 1974*
4. Federal Register, 36(247):24888-24890, December 23, 1971.
5. Federal Register, 39(177):32856, September 11, 1974.
6. Howes, J. E., Jr., Pesut, R. N., and Henry, W. M. Evaluation of Sta-
tionary Source Particulate Measurement Methods. Volume I, Portland
Cement Plants. EPA-650/2-75-051a, U.S. Environmental Protection Agency,
Research Triangle Park, N.C., June 1975.
7. Peters, E. T., and Adams, J. W. Evaluation of Stationary Source Parti-
culate Measurement Methods. Volume II, Oil-Fired Steam Generators.
EPA-600/2-77-026, U.S. Environmental Protection Agency, Research Triangle
Part, N.C., February 1977.
8. Davies, Owen L., Editor. The Design and Analysis of Industrial Experi-
ments. Hafner Publishing Company, New York (1960), Chapter 9.
46
-------
APPENDIX A
EPA Method 5
Federal Register, December, 1971
47
-------
CUES AND SEGULATiONS
2.1.4 nter Holder—Pyrei' glass with
bsMlOg system CMMible Crf TTtRJptftt n 1 j^g mini-
mum temperature of 326' P.
U.JS Implngers / Condenser—Pour imp In-
gtra connected in series with glaae boll Joint
OttliicjB. Tne first, third, and fourth isnpln-
g«r» *re of the Greenburg-Smlth design,
modified by replacing the tip with ft % -incb
ID gfct1"? tutoe extending to one-half inch
from i..i- bottom of the fia&k. The second 1m-
pioger J* of the Greenburg-Smlth design
With the- standard tip. A condenser may be
used In place of the impingers provided that
Hie moisture content of the stack gas oan
•dU be determined.
3.1 JS Metering system—Vacuum gauge.
leak-free pump, thermometers capable of
measuring temperature to within 5* P., dry
g«a meter with 2% accuracy, and related
equipment, or equivalent, as required to
maintain an leoklnetic sampling rate and to
determine sample volume.
2.1.7 Barometer—To measure atmospheric
pressure to ±0.1 Inches Hg,
2.2 Sample recovery.
2.2.1 Probe brush—At least as long as
probe.
3.2.2 Glass wash bottles—T> .
2.2-J Glass sample storage i./amlners,
25.4 Graduated cylinder—250. ml.
2.3 Analysis.
2.3.1 Glass weighing dishes.
2.3.2 Desiccator.
2,3,3 Analytical balance—To measure to
±0.1 mg.
2.3.4 Trip bali.nce—300 g. capacity, to
measure to ± 0,05 g.
3. Reagents,
S.I Sampling,
3.1.1 Filters—Glass fiber, MSA 1106 BH *,
or equivalent, numbered for Identification
and preweighed.
3.1.2 Silica p.fl—Indicating type. &~I£
mesh, dried at 175' C. (350* P.) for 2 hours.
3-1.3 Water.
-. 3.1.4 Crushed ice.
"a.2 Sample recovery.
3,2.1 Acewme—«eag*nt grade.
3.3 Analysis.
3.3.1 Water.
IMP1NGEF TRAIN OPTIONAL MAY BE REPLACED
BY AN EQUIVALENT CONDENSER
HEATED AREA JILTED HOLDER / THERMOMETER CHECK
^VALVE
STACK
|—WALL
METHOD 5—DRKEUINATION or
EMISSIONS PROM STATK>NA*T SOUKCES
1. Principle and applicability.
l.l Principle. ParttcuJate matter is wlth-
dravn tookinetically from the source and Its
weight ie determined gravlmetrlcally after re-
moval of uacombined water.
1.2 Applicability. This method Is applica-
ble for the determination of paniculate emis-
sions from stationary sources only when
specified by the test procedures lor determin-
ing compliance with New Source Perform-
ance Standards.
2. Apparatus.
2.1 Sampling train. The design specifica-
tions of the paniculate sampling train used
by EPA {Figure 5-1) are described In AFTD-
0581. Commercial models at this train are
available,
2.1.1 Noezte—Stainless steel (316) with
sharp, tapered leading edge.
3.12 Probe—Pyrex1 glass with a heating
cybtein capable of maintaining a minimum
gas temperature of 260* P. at the exit end
during sampling to prevent condensation
from occurring. When length limitations
(greater thar. about 8 ft.) are encountered at
temperatures leafi than 600* P., Incoloy 825',
or equivalent, may be used. Probe* for sam-
pling gas streams at temperatures In excess
of 600' P. must have been approved by the
Adznl nis&rator.
2.:.3 Pilot tube—Type 8, or equivalent,
attached to probe to monitor stack gas
Telocity.
REVERSE-TYPE
PITOT TUBE
^
PITOT MANOMETER
GAUGE
DWTESTWET1:R
Figure 5-1. Parliculate-sampling train.
> Trade &*ma.
3.3.2 Deslccact—Drierite,1 indicating.
4. Procedure.
4.1 Sampling
4.1.1 After selecting the sampling site and
the Tnittiwiiim number of sampling points,
determine the stack pressure, temperature,
moisture, and range of velocity head.
4.1.2 Preparation of collection train.
Weigh to the nearest gram approximately 300
g, of silica gel. Label a filter tit proper diam-
eter, desiccate* for at least 24 hours and
•weigh to the nearest O.S mg. in a room where
the relative humidity Is less than &0%. Place
100 ml. of water In each of the first two
Impingers, leave the third Implnger empty,
and place approximately 200 g. of preweighed
silica gel in the fourth impinger. Set up the
train without the probe as In Figure 5-1.
Leak check the sampling train ai Lite sam-
pling site by plugging up the inlet to the ni-
ter bolder and pulling a 15 in. Hg vacuum. A
leakage rate not In excess of 0.02 cfan. at a
vacuum of 15 In. Hg is acceptable. Attach
the probe and adjust the heater to provide a
gas temperature ol about 250* p. at the probe
outlet. Turn on the filter heating system.
place crushed Ice around the Impingers. Add
» Trade name.
•Dry using Driertte1 at 70" P.±10' P.
more ice during the run to keep the temper-
ature of the gasio leaving the last Impinger
as low as pubaib'.i; and preferably at 'TO" F.,
or less. Temperatures above 70C F. may result.
In damage to the dry gas meter from either
moisture condensation or exccb&ivc heat,
4,1.3 Purticulate train operation. For each
run, record the dr.la required on the example
sheet shown i:i Figure 5-2. Take readings at
each sampling poiM, at least every 5 minutes.
and when signliicu.nt change;; In stuck con-
ditions necessitate addition..' ftdjxittnieuts
in flow rate. To begin samp!):..:. pObUlou tht-
11022!e at mo first traverse point w:th U:e
tip pointing directly Into the gas stream.
Immediate!)' start the pump and adjjst the
flow to isc.Sv.i^'.ii: conditions. Sample for a:
least 5 ininu:^.: a: u.ch uwcr-'-e point, sam-
pling UiJiC u:u,; "oe the suiuc lor eak.h pah;..
Muititain isok:i,e:ic sampling tliroughout the
sampling p'-ri^d. Nomographs are available
which aid :i; :rje rupid adjiii'meiii- of the
»ampliiii» r^c without other cumputatlons.
APTD-0570 derails The procedure for using
these noniLgTv.pUs. 7-.irn off the puR-.p at the
Conclusion o' each run and record the final
readings. Hei'i.Me the probe and uczzle from
the stack ai.u handle In accordance with th*1
sample recovery process described in secuou
4.2.
FEDERAL REGISTER. VOL. 36, NO. 247—THURSDAY, DECEMBER 23, 1971
48
-------
HUB AND K6MADONS
T_i- Average dry (is meter «*•••['— »*"•-
*V,— Barometric pressure at j» orifice
meter. Inches Hg.
AH- Average pressure drop across the
orifice meter, inrTioi Sjo,
13.6 -Specific gravity at mercury.
P.,,— Absolute pressure at standard e«n-
drUone, » JZ aches at
«* Volume of water vapor.
etpnttonS-*
V».,4—Volume of
•™ ^ -^.
CO. tt.
fi.-Total voluma of liquid eoUaotod la
imfto&t* and iUic« fti («•* Fig-
ure 3-3). mL
Ma,o->4ol«-.'.ar we«lrt of, •atat,
Sample ncov
y. Ettrdae care In mov-
ing til* collection train Horn tto« tot ttt* to
ti» Euspic leiujwuy am to mlnlmlm tta
lax at collected uunplr or tb* gKla «(
aBMMOM pansmute aittMr. an wM* •
wtftton of OM aoctoiw HSM] lo ttv auiiidA
raeovcry u • bluk tor uxlfmlx. Ucoure tto
voimnc of wuter tnn tb» flnrt ttme 1m-
ptngeis. tbcii ^fayaTM Fiaovtbe Hmpioft-ln
containers u fallen:
CoRtoiiur Mo. 1. Bemote tie fltur Iroitt
tu tiaioti, plsai IB out cfint»tmir. and ml.
ContaUUET Wo. Z. — ----- " ------
OontBtner »o.». W«lgh th. qMct tfilc* ga
•Mnpart to tiie i
ctBttcca prior to tba niter
In this container and leal. HBB a racer blade,
brush, or rubber policeman to lo» adhering:
particles.
Container Wo. 3. Transfer tbc silica gel
from tfie fourth Unptager to tbe original con-
tainer and >eat. Use a robber policeman aa
an aid in removing silica get from tuv
4J Analysis. Record t&e datft required on.
example «he« shown ' —
BMMhejBeW *pajB«i«fce> tBBamwMewe* %Mp»ie»^^^^ ^» • • ~~
Container no, 1. Transfer toe Otter ami
any loose parttenlate matter (tarn the sample
container to a land glaee weighing dttb.
deaiceate, and dry to a constant weight. Be*
pott results to the nearest OJS mg.
Container ML 2. Transfer the acetone
washings to a tared beaker and evaporate to
dryneea a* ambient temperature and pres-
sure Desiccate and dry to a constant weight
Report results to the nearest Of mg.
eqnmtionS-1
»»^-Volume of gae aample through the
dry net meter (standard eandi-
tune), en. ft
V_-Votmne of a-c aample through the
dry gae meter (meter condl-
tlon«).cu.ft.
TM-Abaolute Ton—i*"- at standard
*" oondltJona.SMCa.
B-W««1 '(at ooMUnt, ujs. mebat
Hg— cu. ft./lb.-mole-'R_
T...- Absolute tcmpcmtm at rtandard
ooodltfanaa. 630* B.
*„»— Absoluie pnuuzc at standard con-
ations. 39.82 inches Hg.
0.4 Uolttun content.
equation 5-3
at waiar vapor Ir. the p^4
ka.
of water In the tat sample CRtrndoid
•,cu.R.
(XauaudwuiltKw). euro.
Total paniculate weight Determine
tbc total pamoalate-cateb from the nrm of
the waaghta OB the analysis data sheet
(Figure Ml.
04 Concentration.
equation 5-4
eXwicoaraUoe at pBrtlctilalc matter la Back
Vulum« ot eas sunrle tirouch dir H» "»«•»
(atudanl torallitoni). cu. It.
VOl. 34, MO. J47—IHUMBAY, DCOMBEt tl. I»71
49
-------
JO90
MILES AND REGULATIONS
PLANT_
DATE_
RUN NO..
CONTAINER
NUMBER
WEIGHT Of PARTICIPATE COLLECTED,
FINAL WEIGHT '.TARE MIGHT WEIGHT GAIN
FINAL
INITIAL
LIQUID COLLECTED
TOTAL VOLUME COLLECTED
VOLUME OF LIQUID
WATER COLLECTED
IMPINGER
VOLUME.
ml
n.
SILICA GEL
WEIGHT.
9
9*1 ml
CONVERT WEIGHT OF WATER TO VOLUME BY DIVIDING TOTAL WEIGHT
INCREASE BY DENSITY OF WATER. (1 0 ml):
I—l'oirenl of isokiiietic sampling.
\\«=Tulul volume of liquid collected In bnptogen
and silica gel £5ee Fig. 6-3), mL
pBjO-Uousitj' of water, 1 g./QiL
R™ Ideal gas constant, 21.83 Inches Hg-co. ft./lb.
juolt*~*R..
MH,o=Mo!«:Liiar weight of water, 18 Ib./lb.-molfc
VB •» Volant* ol gas sample through the dry gat mater
i uii-tet conditions), cu- ft.
Tu.=AbsuluiR average dry gas meter temperature
(se« Figaro 5-2), °R.
Fb«r* Barometric pressure at templing site, tochv
Hg.
AUK A vertigo pressure tiiup across the oilier (sw
KV. 5-2), inches JljO.
T,™Absolute ftvtsrape stack gas temperature (sw
Fig. !>-2),0R.
ff=Totaj sampUnc time, min.
V.-= Stack gB.<: vefocity calculated by Method 2,
Krjuali'-n 2-'J. ft.^«.
P.°Al)s-,(Lili >uck pas pressure, iocbw Eg.
An^Cioss-BOctluimiiiritaof oozzle, stj. ft.
6.8 Acceptable results. The following
range sets the limit on acceptable Uotineilc
sampling results;
ZT90%<1< 110%, the re^uJtc are acceptable,
otherwise, reject the results and repeat
the test.
7. Reference.
Addendum to Specifications for Incinerator
Testing at Federal Facilities, PHS, KCAPC,
Dec. 6, 1967.
Martin, Robert M., Construction, Details of
IsoJclnettc Source Sampling Equipment, En-
vironmental Protection Agency, AFTD-0581.
Bom, Jerome J.. Maintenance, Calibration,
and Opei'&Uon of laokinetic Source Sam-
pling Equipment, Environmental Protection
Agency, APTD-0576.
Smith, W. S-. R. T. Sblgehara, and W. F.
Todd, A Method of Interpreting Stec^ f^Jn-
pllng Data, Paper presented at the 63d An-
nual Meeting of the Air Pollution Control
Association, St. Loulfl. Mo,, June 14-19, 1970.
Smith, W. &., et al.. Stack Gas Sampling
Improved and Simplified with New Equip-
ment, AFCA paper No. 67-119, 1967.
Specifications for Incinerator Testing at
Federal Facilities, PHS. NCAPC, 1967.
INCREASE,
(1 9/ml)
VOLUME WATER, ml
FI0ur«5-3. Analytical data,
Concentration in Ib/ou, It.
I
Ib.
v,
equation 5-5
M»~Ttotal amount of parttcolate nutter collected,
Vmpl4-Votame of gu lamplt through dry BUS meter
(Mandard eondlttons), cu. ft.
«.7 laoktnetlc variation, i",':
Equation 6-8
MCISTBI, VOL J4, NO. »47—THURSDAY, OiCEMtM J3, 1971
50
-------
APPENDIX B
-STACK GAS MEASUREMENT DATA
51
-------
TABLE B-l. STACK GAS DATA - BOF WITH WET SCRUBBER
EMISSION CONTROL
V
Run No. (
1A
B
2A
B
3A
B
4A
B
5A
B
6A
B
7A
B
8A
B
9A
B
10A
B
11A
B
12A
B
13A
B
r3i> (avg)
:mH20 S
0.69
0.97
1.13
0.83
1.00
0.73
0.92
0.86
0.91
0.86
0.94
0.99
0.92
Ts (avg)
C
56
54
54
61
56
64
66
54
53
55
51
58
53
mm Hg
745.5
744.7
744.0
749.3
749.3
748.8
749.3
739.4
739.4
747.8
747.8
756.7
751.1
02, %
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
C02, %
14
14
14
14
14
14
14
14
14
14
14
14
14
CO, %
75
75
75
75
75
75
75
75
75
75
75
75
75
BWO> %
20.6
21.4
19.4
16.2
25.5
18.9
30.8
20.8
19.0
20.6
13.1
13.4
17.9
17.4
22.0
17.4
16.0
15.0
16.8
15.4
14.3
13.6
18.2
17.2
20.0
19.5
Md
Ib/lb-mole
30.2
30.2
30.2
30.2
30.2
30.2
30.2
30.2
30.2
30.2
30.2
30.2
30.2
Vs (avg)
m/s
8.0
11.1
13.2
9.7
11.4
8.2
10.7
9.9
10.2
9.8
10.4
11.2
10.4
52
-------
Ui
TABLE B-2. STACK GAS DATA - BOF WITH WET SCRUBBER
EMISSION CONTROL
Run No.
14A
B
ISA
B
16A
B
17A
B
ISA
B
19A
B
AP (avg),
cm H205
1.00
0.89
1.00
0.89
1.02
0.94
C
58
56
59
59
60
84
mm Hg
739.6
745.0
746.0
748.8
748.3
752.9
02, %
0.1
0.1
0.1
0.1
0.1
0.1
C02, %
14.0
14.0
14.0
14.0
14.0
14.0
CO, %
75.0
75.0
75.0
75.0
75.0
75.0
Bwo, %
16.4
15.9
14.9
15.1
19.3
19.2
16.6
16.0
18.2
17.9
15.2
16.1
Md,
Ib/lb-mole
30.2
30.2
30.2
30.2
30.2
30.2
Vfl (avg),
tn/s
11.54
10.20
11.60
10.32
11.79
11.15
-------
TABLE B-3. STACK GAS DATA - EOF WITH ESP
EMISSION CONTROL
Run No.
1A
B
2A
B
3A
B
4A
B
5A
B
6A
B
7A
B
8A
B
9A
B
10A
B
/?(avg)
cm HjO f
0.64
0.65
0.64
0.59
0.75
0.56
0.57
0.59
0.59
0.60
T (avg)
C
133
155
127
144
144
149
153
154
164
147
v
mm Hg
742.7
734.6
737.1
739.4
734.1
727.5
742.2
745.7
741.2
730.8
%2
17.9
17.9
17.9
17.9
17.9
17.9
17.9
17.9
17.9
17.9
C?2
5.5
5.5
5.5
5.5
5.5
5.5
5.5
5.5
5.5
5.5
V'
10
22.61
22.78
25.19
25.76
18.91
18.7
22.25
22.51
16.39
15.14
24.93
25.16
21.66
22.08
21.22
21.36
23.06
22.70
21.28
21.24
Mi,
Ib/lb mole
29.6
29.6
29.6
29.6
29.6
29.6
29.6
29.6
.
29.6
29.6
Vs (avg)
m/s
8.2
8.8
8.2
7.7
9.6
7.5
7.6
7.7
7.9
7.9
54
-------
APPENDIX C
SAMPLING SYSTEM OPERATION DATA
55
-------
TABLE C-l. SAMPLING DATA - BOF WITH SCRUBBER
Ui
Run No.
1A
B
2A
B
3A
B
4A
B
5A
B
6A
B
7A
B
8A
B
9A
B
10A
B
11A
B
12A
B
13A
B
Meter Volume
-------
TABLE C-2. SAMPLING DATA - EOF WITH WET SCRUBBER EMISSION CONTROL
. Average System Tenroera Cures. C
Run No.
14A
B
15A
B
16A
B
17A
B
ISA
. B
19A
B
Meter Volume
(Vm), m3
2.17
2.04
1.71
1.59
1.88
1.78
1.66
2.04
1.36
1.79
0.45
0.41
Barometer,
mm Hg
739.6
745.0
746.0
748.8
748.3
752.9
AH,
ran H20
62.0
62.5
49.0
48.3
59.2
58.9
49.0
80.5
30.0
61.5
55.6
Avg. Meter
Temp. (Tm), C
22
11
11
6
17
8
13
7
18
12
14
10
Dry Gas
Std. Cond.
2.
2.
1.
1.
1.
1.
1.
2.
1.
1.
0.
0.
Sampled
(Wd>. ™3
12
08
74
65
88
83
69
12
36
83
45
42
Percent
Isoklnetic
104
101
105
100
105
103
100
125
74
99
105
99
Filter Box
129
191
123
149
124
124
122
132
125
132
125
171
Gas at Filter
> Outlet
107
180
109
148
109
111
114
131
114
133
116
66
Gas at Probe
Outlet
130
183
126
154
128
122
131
134
127
133
131
95
Probe
Mid-point
142
126
124
127
126
128
-------
TABLE C-3. SAMPLING DATA - EOF WITH ESP EMISSION CONTROL
Average Svstem Temperatures . C
Run No
1A
B
2A
3A
B
4A
B
5A
Ln „
00 B
6A
B
7A
B
3A
B
9A
B
10A *
B
Meter Volume,
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
! REPORT NO. 2.
MEPA-600/2-79-141
4. TITLE AND SUBTITLE
EVALUATION OF STATIONARY SOURCE PARTI CULATE MEASUREMENT
METHODS
1 Volume IV. Basic Oxygen Furnaces
j^AUTHORtS)
|J. E. Howes, Jr., W. M. Henry, and R. N. Pesut
J9, PERFORMING ORGANIZATION NAME AND ADDRESS
Battelle, Columbus Laboratories
505 King Avenue
Columbus, Ohio 43201
[l2. SPONSORING AGENCY NAME AND ADDRESS
Environmental Sciences Research Laboratory - RTP, NC
Office of Research and Development
U.S. Environmental Protection Agency
(Research Triangle Park, N. C. 27711
3. RECIPIENT'S ACCESSION NO.
5. REPORT DATE
August 1979
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO.
10. PROGRAM ELEMENT NO.
1AD605 Task DA-002 (FY-77)
11. CONTRACT/GRANT NO.
68-02-0609
13. TYPE OF REPORT AND PERIOD COVERED
Interim 10/73 - 6/77
14. SPONSORING AGENCY CODE
EPA/600/09
hs. SUPPLEMENTARY NOTES y^ ^ j ^ .^^ ^ ^ 650/2_75_051 3 , June ^ 975 . Vol ume JJ 35
EPA 600/2-77-026, February 1977; Volume III as EPA 600/2-79-115, July 1979; Volume V as
Ippa finn/?-7Q_Tifi .ii,iy 1070
HeTABSTRACT ' y
JA procedure, EPA Method 5, for sampling and determining particulate concentrations in
emissions from stationary sources was specifically evaluated at basic oxygen furnaces
(BOF) equipped with wet-scrubbers or electrostatic precipitator (ESP) controls.
Although wet-scrubber emissions could potentially present the most difficulty since the
sampling system must handle a moisture-laden gas stream with entrained water droplets,
no problems were found when using Method 5. Variation of the sampling rate at 0.7 and
1.3 times isokinetic also did not significantly affect the accuracy of the mass measure-
ments. Use of a higher purity, lower pH filter medium (ADL quartz), gave mass results
that were not statistically different than those obtained with MSA 1106 BH, the
commonly used filter material. ,;,
Chemical analyses confirmed that the Method 5 procedure extracts a representative
sample of the BOF/ESP stack particulate emissions. The same general chemical composi-
tion was found on Method 5 filters and in grab samples removed from the stack at the
sampling point.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b-IDENTIFIERS/OPEN ENDED TERMS
COSX
Air pollution
Particles
Collection methods
Evaluation
Basic converters
['«• DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
> Form 2220-1 (Rev. 4-77) PREVIOUS EDITION is OBSOLETE
EPA Method 5
~g SECURITY CLASS (ThisReport)
UNCLASSIFIED
13B
14B
11F
21. NO. OF PAGES
67
20 SECURITY CLASS (Thispage)
UNCLASSIFIED
59
------- |